article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
substantial progress has been made over the past few decades in understanding the nature of energy flow in highly excited molecules@xcite . at higher levels of excitation
the usual rigid rotor - normal mode description breaks down and the modes get strongly coupled .
the resulting spectral complexity in terms of the splittings and intensities encode the flow of energy through the molecule .
this phenomenon of intramolecular vibrational energy redistribution ( ivr ) can lead to subtle but important deviations from the predictions of statistical theories like rrkm@xcite and thus profound consequences for the rates of chemical reactions@xcite .
the mode mixings in general have classical as well as quantum origins .
much of the classical effects arise from vibrational anharmonicities , coriolis and centrifugal distortion effects .
possible quantum contributions@xcite manifest as indirect state - to - state explorations involving a sequence of intermediate , off - resonance virtual states ( `` vibrational superexchange '' ) . in molecules with symmetry
the phenomenon of dynamical tunneling provides a quantum route to mode mixing@xcite .
theoretically it is crucial to establish and understand ivr pathways from time dependent as well as time independent perspectives .
rrkm theory has been quite successful@xcite but evidence is mounting for a number of reactions where the intramolecular dynamics is inherently nonstatistical@xcite . accounting for such nonstatistical dynamics is theoretically challenging and crucial for control of reactions . in this work
our aim is to understand ivr in the deuterated formyl radical dco .
our choice of the system is primarily due to the considerable experimental@xcite and theoretical@xcite work on ivr and mode specificity of the unimolecular decay of highly vibrationally excited dco .
in contrast to hco the deuterated analog involves a strong accidental @xmath5 resonance between the vibrational frequencies . the experiments and analysis of renth _ et al
_ using a effective hamiltonian clearly shows that the unimolecular decay in dco is ivr limited@xcite . although dco has three vibrational modes , @xmath6 ( dc stretch ) , @xmath7 ( co stretch ) , @xmath8 ( bend ) , the coupling structure of the effective hamiltonian@xcite yields a conserved quantity called as the polyad number @xmath9 .
experimentally the existence of @xmath10 is revealed by bunching of vibrational levels as evident in the sep spectrum of dco@xcite . as a small molecule exhibiting nonstatistical intramolecular dynamics dco provides an ideal system to test the various theoretical approaches for studying highly excited molecules .
impressive advances have been made towards fully quantum investigation of the ivr dynamics in large systems@xcite .
there is little doubt that dco can be subjected to such exact treatments which are important in their own right .
however there are strong reasons@xcite to believe that powerful insights into ivr can be gained by approaches based on classical - quantum correspondence .
for example , jung _ et al _ have provided dynamical assignments of the highly excited states of dco based on certain key periodic orbits in phase space@xcite .
a key to the success of the many classical - quantum correspondence studies@xcite is that the understanding of ivr is firmly rooted in the invariant structures inhabiting the underlying phase space .
this provides a very clear dynamical picture of ivr in molecules and can serve as a baseline for estimating the important quantum routes to mode mixings .
currently one of the main obstacles in extending such useful classical - quantum correspondence studies is the issue of dimesionality . despite the existence of potentially @xmath11 coupled vibrational modes it is now well understood that not all of the modes are coupled strongly@xcite .
although such hierarchicality of the couplings leads to much smaller dimensionalities of the ` ivr space ' it is possible that for large enough molecules or smaller molecules at higher levels of excitation the effective dimesionality of the system can be larger than three . in such instances
the dimensionality of the system poses both technical as well as conceptual problems in the context of classical - quantum correspondence studies .
the technical problems are well known and have to do with the inability to visualize poincar surface of sections or the quantum states and , computation of the periodic orbits or higher dimensional tori .
thus the knowledge of the global phase space structure along with its influence on the quantum states is at present difficult to obtain for systems with three or more degrees of freedom . for more detailed discussions on the dimensionality issues ,
technical and conceptual , we refer the reader to the earlier works@xcite .
one of the central goals of our research is to come up with techniques that would allow a detailed classical - quantum correspondence study of ivr in molecules with three or greater degrees of freedom and in this context two approaches will be of particular interest .
the first one is the promising technique of local frequency analysis ( lfa ) put forward nearly two decades ago by martens , davis and ezra@xcite . in their pioneering work@xcite on energy flow in the ocs molecule
it was shown that lfa was capable of highlighting the role of various phase space structures including partial barriers and resonance zones .
the important point is that although motion may be irregular when viewed over long time scales ( several picoseconds ) , quantities such as frequencies can be constant over times corresponding to many vibrational periods .
this gives rise to long time correlations and the notion of dynamically significant phase space regions@xcite .
since then very little work has been done in utilizing this technique for studying energy transport in multimode molecules .
interestingly much more efforts have been made in celestial mechanics@xcite with fewer applications in atomic physics@xcite and molecular systems@xcite .
recently arevalo and wiggins@xcite have proposed a technique for performing time - frequency analysis ( lfa ) which include strongly chaotic trajectories as well by using wavelet analysis .
applications to various systems demonstrated the advantages of a wavelet based lfa in understanding phase space transport@xcite .
a second technique@xcite for analyzing the quantum states , based on parametric variations , was recently proposed by us and applied to a 2-mode model hamiltonian .
it was shown that variations of the eigenvalues with appropriate parameters _
i.e. , _ level velocities were strongly correlated to the important classical phase space structures organizing the quantum states . in the case of mixed and near - integrable systems the magnitudes and relative signs of the level velocties clearly indicated@xcite the dynamical region of phase space associated with a given eigenstate thereby providing a dynamical assignment of the states .
the technique based on parametric variations is dimensionality independent , like the local frequency analysis mentioned before , and more importantly basis invariance is guaranteed .
moreover parametric variations can be used to identify localization , resulting in deviations from random matrix theory ( rmt ) predictions , even in strongly chaotic systems@xcite .
the full utility of parametric techniques is yet to be explored , particularly in the context of ivr@xcite , and there do exist detailed semiclassical theories in the ( near)integrable@xcite and chaotic situations@xcite .
our system of choice dco is effectively two dimensional due to the existence of the polyad @xmath10 .
thus the intramolecular dynamics can be investigated in great detail using the existing methods .
however dco provides a stringent test for the dimensionality independent techniques mentioned above since there is significant chaos in the classical phase space for the lowest energies in @xmath1 and @xmath12 .
this paper is organized as follows : in section ii the main features of the effective hamiltonian and its classical limit are presented .
this is followed , in section iii , by a dynamical assignment of states in @xmath13 using the level velocities .
it will be shown that a dynamical assignment based on level velocities , in conjunction with the inverse participation ratios@xcite ( ipr ) , can be given despite the presence of significant chaos . in section
iv the time dependent dynamics of dco are analyzed using local frequency analysis .
this reveals the global picture of phase space transport and highlights chaotic trajectories undergoing rapid frequency jumps mainly between the stretch - stretch @xmath14 resonance and the dc stretch - bend @xmath15 fermi resonance .
moreover it is seen that chaotic trajectories can be trapped in resonances for many vibrational periods .
this is related@xcite to the stickiness of the chaotic trajectories and is of importance to the ivr dynamics in the molecule .
the local frequency analysis also highlights the nature of the co stretch - bend @xmath15 fermi resonance which is seen to play a role in the ivr for short times .
section v concludes .
the spectroscopic hamiltonian , proposed recently by trllsch and temps@xcite , was obtained through a fit to the experimental df and sep spectra .
the relevant matrix elements are @xmath16 \nonumber \\
h_{mnb;m-2,n+2,b } & = & \frac{\lambda_{2}}{2 } [ m(m-1)(n+1)(n+2)]^{1/2 } \\
h_{mnb;m-1,n , b+2 } & = & \frac{k_{mbb}}{2 } \left[\frac{1}{2}m(b+1)(b+2)\right]^{1/2 } \\
h_{mnb;m , n-1,b+2 } & = & \frac{k_{nbb}}{2 } \left[\frac{1}{2}n(b+1)(b+2)\right]^{1/2}\end{aligned}\ ] ] with @xmath17 denoting the zeroth order dc stretch , the co stretch , and the bend modes respectively .
the coupling terms @xmath18 and @xmath19 represent a @xmath14 and a darling - dennison @xmath20 resonance between the stretch modes respectively .
the fermi resonant couplings @xmath21 correspond to a @xmath22 resonance between the dc ( co ) stretch and the bend modes .
the effective hamiltonian @xmath23 commutes with the polyad @xmath24 and hence block diagonal in @xmath25 .
a given @xmath25 block has @xmath26 states for integer polyads and the states are labelled as @xmath27 in order of decreasing energy .
note that the above hamiltonian is able to approximately reproduce the experimental df spectra for polyad values upto @xmath28 since @xmath29 was the highest polyad considered in the fitting procedure .
consequently the current work focuses on the polyads @xmath30 as examples .
the techniques used for the analysis in this work are , however , applicable to higher polyads as well . in what follows we will adopt a shorthand notation for the two main resonances , @xmath14 dc stretch - co stretch and @xmath22 dc stretch - dco bend , denoting them as @xmath31 and @xmath32 respectively .
the classical limit hamiltonian can be obtained using the heisenberg correspondence @xmath33 between creation - annhilation operators and the action - angle variables .
the resulting classical hamiltonian is @xmath34 \nonumber\end{aligned}\ ] ] with @xmath35 $ ] .
as expected the zeroth order hamiltonian @xmath36 is a function of the actions alone and the various couplings manifest themselves as anharmonic resonances .
the quantum zero point energy @xmath37 has been subtracted from the classical hamiltonian for consistency .
the classical system is effectively a two degree of freedom system due to the existence of the conserved quantity @xmath38 which is the analog of the quantum polyad . as a result
the classical dynamics can be studied in a four dimensional reduced phase space .
the reduction can be obtained by performing a canonical transformation via the generating function : @xmath39 the new action - angle variables are related to the old ones , @xmath40 , and @xmath41 .
the resulting hamiltonian@xcite is independent of the angle variable @xmath42 which is conjugate to @xmath43 and hence an effectively two - dimensional system .
it is possible to perform a chirikov@xcite analysis of the classical hamiltonian and obtain a resonance template in the state space@xcite .
however it is seen that in this case the chirikov analysis , which is based on the zeroth order @xmath36 , is not very suitable .
this is mainly due to the reduced bend anharmonicity of dco . despite this limitation useful qualitative information
can be obtained regarding the approximate location of the resonance zones and overlaps .
for instance , such an analysis predicts that the co stretch - bend fermi resonance ( @xmath44 ) does not show up in the state space for most of the polyads .
thus one anticipates the ivr dynamics in dco to be dominated by the stretch - stretch @xmath14 and dc stretch - bend @xmath15 fermi resonances .
this observation is in agreement with earlier works@xcite . in what follows
we will show the resonance template as a guideline and the local frequency analysis of the dynamics establishes the significant deviations from the zeroth order picture due to strong couplings .
considerable work has been done in the last few years to understand and possibly assign highly excited states by invoking the classical - quantum correspondence principle@xcite . every single method , without exception ,
is based on identifying a dominant classical phase space structure that organizes the eigenstates in a given energy range . indeed as dco is effectively two dimensional due to the existence of the polyad it is possible to use any one of the several techniques proposed in the literature .
however our objective in this work is to use techniques that do not explicitly rely on determining the exact periodic orbits and/or visualizations of the phase space and wavefunctions .
this is especially important for analyzing systems with more than two degrees of freedom .
one such method was recently proposed by us@xcite and it is based on the method of parametric variations . in this approach
the `` level velocities '' @xmath45 associated with an eigenstate @xmath46 are determined and analyzed .
thus if the effective , spectroscopic , hamiltonian has the general form @xmath47 with @xmath48 and @xmath49 denoting the various resonances and the corresponding coupling strengths , then the level velocity of a specific eigenstate associated with a coupling @xmath48 is obtained as : @xmath50 in the above equation @xmath51 is the zeroth - order basis and @xmath52 .
the level velocity of a state @xmath53 with respect to @xmath48 can be written as a sum over the polyads @xmath54 , corresponding to the resonant perturbation @xmath48 , of the partial level velocities @xmath55 . in the integrable limit _
i.e. , _ with only @xmath48 present it is easy to see that @xmath56 . in the presence of other perturbations
@xmath57 the level velocity @xmath58 can have contributions from many different @xmath54 indicating extensive mixing .
thus the level velocity essentially measures the response of the quantum eigenvalue to a specific perturbation .
it was established@xcite in the earlier work that @xmath58 is correlated to the phase space nature of the eigenstate .
for integrable , single resonance cases the various properties of the level velocities are well known@xcite . in the case of mixed phase space systems
it is still possible@xcite to dynamically assign states based on the level velocities since the regular regions strongly influence the eigenstates . in case of near integrable systems and , to some extent , for mixed systems the correlation of the level velocities with the phase space can be understood semiclassically since it is possible to write@xcite @xmath59 \\ & \approx & \frac{1}{(2 \pi \hbar)^{n } } \int d{\bf i } d{\bm \theta } v_{j}^{w}({\bf i } , { \bm \theta } ) w_{\alpha}({\bf i } , { \bm \theta})\end{aligned}\ ] ] where @xmath60 is the weyl symbol of the operator @xmath61 and @xmath62 is the wigner function associated with the eigenstate @xmath53 . for near integrable systems
the wigner function condenses on to the invariant torus@xcite thus explaining the observed correlation .
however for systems with considerable chaos it is not _ a priori _ clear wether the correlation continues to hold .
the @xmath13 cases for dco considered herein show extensive chaos at low energies and hence provide a critical test for the level velocity approach . in the extreme limit of an ergodic system the berry - voros hypothesis states
that@xcite @xmath63\ ] ] in this limit it is easy to see that @xmath64 is essentially the parametric derivative of the smooth part of the spectral staircase function@xcite .
this term is independent of the nature of the dynamics and in particular refers to contributions from short time orbits .
nevertheless it is well known that even in strongly chaotic systems the periodic orbits ( unstable ) play a crucial role@xcite .
thus it is expected that in such systems the level velocities are influenced by the periodic orbits .
a more general expectation is that , independent of the nature of the dynamics , closed orbits will influence the level velocities .
as the purpose of the current work is not to provide a semiclassical expression for computing the level velocities , we will briefly outline the reasons for such an expectation .
the arguments provided are not new and appear in an early work by eckhardt _
_ where semiclassical matrix elements were determined from periodic orbits@xcite .
to begin with define a quantity @xmath65 with @xmath66 being the energy green s function and @xmath61 is the perturbation of interest . using the formal definition of @xmath66 it is straightforward to show that @xmath67 thus @xmath68 has poles at the eigenvalues @xmath69 and the residues are precisely the level velocities @xmath58 .
the quantity @xmath70 can be written in the semiclassical limit as : @xmath71 with @xmath72 being the wigner transform of the green s function .
the above integral can be evaluated via stationary phase approximation and , apart from the smooth short time contribution , gives rise to an oscillating term .
the oscillating part can be evaluated for integrable ( berry - tabor)@xcite , near - integrable ( almeida - ullmo - grinberg - tomsovic)@xcite , and chaotic ( gutzwiller)@xcite cases . in every case the oscillating part depends on the stabilities of the closed orbits and involves the classical time average of @xmath48 over the period of the closed orbits@xcite .
since the level velocities are extracted as residues of the oscillating part it is reasonable to expect a strong influence of the closed orbits and their stabilities .
note that extracting the velocities from this approach is rather involved .
since it is difficult to obtain bounds on the velocities using semiclassical techniques in the mixed phase space case and the effect of various bifurcations on the level velocities is not known as of yet we will explore the utility of the technique in dynamical assignments from a qualitative perspective .
the basis for the assignments will be explicitly discussed for the case @xmath1 since the arguments are fairly general .
.dynamical assignments for states in @xmath1 [ cols="^,^,^,^,^,^,^,^ " , ] the states for @xmath2 can be assigned in a similar fashion and the assignments are shown in table [ table 2 ] .
in particular , all the states except 4@xmath73 and 4@xmath74 can be assigned in analogy with the states in @xmath1 .
thus the states 4@xmath73 and 4@xmath74 are classified as mixed . in fig.[fig 2 ] states with large positive ( 4@xmath75 ) and large negative ( 4@xmath76 )
@xmath32 velocities are shown .
note the localization of the husimis in the stable and unstable regions of the phase space .
the state space , husimi and partial velocity data for the mixed state 4@xmath73 are shown in fig.[fig
2 ] as an example .
extensive delocalization is seen for this state in both the state space and the phase space .
the state 4@xmath73 also happens to have the lowest ipr among the states in @xmath2 .
we note that for @xmath2 the maximum expected velocities for the @xmath31 and the @xmath32 are @xmath77 and @xmath78 respectively .
a general observation regarding the assignments is that the lower energy states of the polyads are dominated by @xmath32 whereas the higher end states are influenced more by the @xmath31 resonance .
states in the middle of the energy regions exhibit transitional character .
similar observations were made in the dynamical assignments of the states in @xmath79 by jung _
et al_@xcite .
note that the assignments made in this work are essentially based on the level velocities and the iprs .
thus our assignment does not invoke any explicit quantum numbers based on periodic orbits .
however it is clear that attaching an approximate polyad and noting the signs on the level velocities is providing dynamical information .
for instance state 4@xmath80 having a large @xmath81 implies that the state is influenced by a stable periodic orbit of the @xmath31 type . in particular this state would be considered as a @xmath82 class state with the assignment @xmath83 in the notation of the previous work@xcite .
the approximate polyad @xmath84 can be identified with the longitudinal quantum number @xmath85 .
a crucial point regarding the assignments using level velocities is worth reiterating at this juncture .
the advantage of using the level velocities is based on strong correlation of the velocities with the phase space nature of the eigenstates .
such dynamical informations can not be obtained by using the iprs alone or the various representations of the eigenstates .
for example , in @xmath1 the iprs @xmath86 and one would correctly infer that both states are of @xmath31 type .
however with @xmath87 it is possible to provide a finer distinction between the two states ; 3@xmath88 is associated with a stable region of the phase space whereas 3@xmath89 is located in the unstable region of the phase space .
as mentioned before the chirikov analysis of dco resonances based on the zeroth order classical @xmath90 is not very accurate due to strong resonant couplings of the various modes .
the surface of sections indicate that for polyads @xmath13 there is considerable stochasticity even at the lowest energy .
this implies that the various resonances overlap strongly for these polyads and most trajectories are chaotic . despite the significant amount of chaos in the phase space the husimis of many eigenstates
were seen to be rather localized .
such observations have been made earlier in other studies and it is well known that a largely chaotic phase space can still contain structures that act as partial barriers to phase space transport .
the partial barriers can arise due to cantori as well as broken seperatrices and existence of such structures can lead to eigenstate localization .
moreover even in the case of strongly chaotic systems eigenstates can get scarred by periodic orbits leading to localization . in recent years
it has been understood that such partial barriers and periodic orbit scarring lead to statistically significant deviations from the predictions of rmt .
infact the level velocities , studied in the previous section , already encode such localization information . from the ivr perspective
, several studies@xcite have now established that the energy flow process is hierarchical and that a hierarchical local random matrix ( hlrm ) approach is more appropriate to understand ivr .
the important factor is the local density of coupled states and not the total density of states at a given energy which leads , amongst other observations , to a power law decay of the survival probability on intermediate timescales@xcite . and ( b ) @xmath91
. solid line is the exact hamiltonian and dashed lines are with @xmath92.,width=192,height=288 ] based on the detailed work on classical phase space transport@xcite it is tempting to conjecture that the hierarchical nature of ivr is intimately connected to the various partial barriers in the phase space .
the important work@xcite by davis and martens _ et al _ on ocs seems to support the conjecture .
however , except for earlier works on ocs , there have not been many efforts to identify partial barriers for transport in the context of ivr in molecules .
the importance of such an undertaking lies in the fact that it might prove crucial in controlling ivr as well . in dco
it has been established that the unimolecular decay is ivr limited@xcite . moreover , previous studies@xcite on dco suggest that the fermi resonance between the co stretch and dco bend is unimportant .
the approximate chirikov analysis also indicates the muted role of the @xmath93 fermi resonance . in fig.[fig 3 ] the survival probabilities for the zeroth order bright states @xmath94 and @xmath91 in the absence of the @xmath93 fermi resonance are shown and compared to the full system .
it is clear from the figure that although the initial decay from the bright states is controlled by the @xmath14 ( for about 100 fs ) the @xmath93 fermi resonance starts to play a role around 200 fs .
is it the case that the @xmath93 resonance does play a role , albeit in a secondary fashion , in the short time tier structure
? why is the initial decay from @xmath91 apparently complete by about 150 fs as opposed to those of @xmath94 and @xmath95 ?
in other words what makes the @xmath14 so effective for the bright state @xmath91 ?
the answers to these questions are not straightforward and in this work we will try to provide some approximate answers .
we believe that the lfa technique is capable of addressing some of these crucial issues . ahead
a brief introduction is given to the technique of lfa and we refer the reader to the work by martens _ et al_@xcite and arevalo _
et al_@xcite for more details .
the classical limit of the spectroscopic hamiltonians are nonlinear , multiresonant hamiltonians which are functions of the zeroth order actions @xmath96 and the angles @xmath97 .
the classical limit hamiltonians have a typical form @xmath98 where the various resonant perturbations are denoted by @xmath48 . in the absence of the perturbations
the system is described by @xmath36 which is classically integrable . in such cases
the various nonlinear frequencies @xmath99 are time - independent and one can define a frequency map @xmath100 which , if invertible , can be used to characterize the phase space .
the frequency map can be used to characterize the phase space , thanks to the kam theorem , for near - integrable systems as well . for a system like dco
there are three frequencies @xmath101 and it is useful to visualize the dynamics in the frequency ratio space .
however for non - quasiperiodic trajectories , associated with the full nonintegrable hamiltonian , the various frequencies @xmath102 vary with time since increasing perturbation strengths lead to the breakdown of kam tori and widespread chaos . despite the time dependence of the frequencies
it has been observed in many systems that the frequencies can remain approximately constant over times corresponding to many vibrational periods .
thus even non - quasiperiodic trajectories can spend an appreciable amount of time in one part of the phase space before passing on to another region of the phase space .
this in turn implies existence of long time correlations and gives rise to the concept of dynamically significant regions of phase space .
the concept of local frequencies _
i.e. , _ time - varying frequencies allows one to identify such dynamically significant regions of phase space .
two main approaches are available to extract the local frequencies from a classical trajectory .
the first one proposed by martens _
et al_@xcite , and a similar one by laskar@xcite , uses fourier transforms of short time segments of long trajectories .
we are interested in the more recent work@xcite by arevalo and wiggins wherein the local frequencies are obtained by a continuous wavelet transform of the trajectories . in this approach a general time
dependent function @xmath103 is expressed in terms of the basis functions constructed as translations and dilations of a mother wavelet@xcite @xmath104 the coefficients of this expansion are given by the wavelet transform of @xmath103 , defined as @xmath105 for @xmath106 and @xmath8 real .
the wavelet transform gives the local frequency of @xmath103 over a small interval of time around @xmath107 and inverse of the scale factor @xmath108 is proportional to the frequency .
a distinct advantage of the wavelet approach is that the time window automatically narrows for high frequency and widens for low frequency .
the value of the local frequency at a given time @xmath107 is extracted@xcite by determining the scale @xmath108 which maximizes the modulus of the wavelet transform . throughout this work , as in arevalo and wiggins , we will use the morlet - grossman mother wavelet @xmath109 with @xmath110 and @xmath111 .
the parameters @xmath112 can be tuned to improve the resolution . within the wavelet approach
the local frequencies can be obtained even for chaotic trajectories .
for a strongly chaotic trajectory the frequencies will vary considerably and one can measure the `` diffusion '' of the @xmath113 frequency by evaluating @xmath114 where @xmath115 is the mean frequency and @xmath116 is the total time to which the trajectory is propagated .
this measure of diffusion characterizes the unstable regions of phase space and thus low diffusion implies either quasiperiodic motion or extensive trapping near resonance zones ( see below ) . for our application to dco we choose initial conditions for the trajectories in the following fashion .
a uniform grid is chosen in the state ( zeroth order action ) space @xmath117 with @xmath118 being fixed by the polyad conservation . a particular choice of the angles @xmath119
then provides initial conditions corresponding to a specific slice of the phase space .
the numerically integrated trajectories over certain time interval @xmath120 $ ] are expressed as @xmath121 for @xmath122 .
the three local frequencies @xmath123 are extracted by computing the maxima of the modulus of the wavelet transforms of @xmath124 _ i.e. , _
note that the maxima are obtained with a fractional precision of @xmath126 .
details regarding the numerical procedure to extract the maxima can be found in the work of arevalo _
et al_@xcite .
total time of propagation @xmath127 ps .
@xmath128 is shown in blue , @xmath129 is shown in red and @xmath130 is in green .
see text for details on each trajectory.,width=288,height=288 ] @xmath4 .
the variables are @xmath131 and @xmath132 .
selected trajectories ( cf .
fig.[fig 4 ] ) are highlighted with filled circles ( a ) , triangles ( d ) , and squares ( e ) .
the highlighted trajectories are plotted for about 5 ps.,width=288,height=288 ] a large number of trajectories for both @xmath1 and @xmath2 at various energies were analysed and many distinct dynamical behaviours were repeatedly seen . in fig.[fig 4 ] we show , as examples of such dynamical patterns , six trajectories whose frequency ratios were computed to about @xmath127 ps .
note that 10 ps is a fairly long time and corresponds to about 500 co vibrational periods .
all of the six trajectories are at an energy @xmath133 @xmath4 which is energetically close to the middle of the polyad @xmath1 .
the phase space structure is shown in fig.[fig 5 ] for reference and one can clearly see large regions of stochasticity .
figure ( a ) shows a typical quasiperiodic , nonresonant trajectory which in this case represents a distorted local mode . the corresponding trajectory is shown in the surface of section ( circles in fig.[fig 5 ] ) as well . in fig.[fig 4]b
a @xmath14 resonant regular trajectory corresponding to the small island on the surface of section is shown .
notice the locking of the relevant frequency ratio _ i.e. , _
@xmath134 for the entire duration . in fig.[fig 4]c
a trajectory with initial conditions very close to a periodic orbit of the full system is shown .
all the frequency ratios are locked perfectly for about 7 ps and later show large diffusion .
interestingly in this case the @xmath93 locking @xmath135 is clearly seen and thus any other initial condition close to this would certainly involve the @xmath93 resonance to some extent . in fig.[fig 4]d
we show an example of a trajectory exhibiting resonance trapping . this particular trajectory is trapped in the @xmath14 resonance for nearly 5 ps and then drifts away .
the corresponding trajectory is plotted in the surface of section ( triangles in fig.[fig 5 ] ) and it is seen that this trapping is due to the stickiness of the small @xmath14 island .
the trajectory is chaotic for longer times but the observed stickiness suggests that even tiny regions of regularity in the phase space can sufficiently influence the dynamics .
another example of a trapped trajectory is shown in fig.[fig 4]e with the corresponding behaviour on the surface of section shown in fig.[fig 5 ] ( squares ) . in this case
the trajectory starts out with frequency locked in the @xmath14 but drifts away quickly and gets trapped into the @xmath136 ratio for almost 6 ps . finally in fig.[fig 4]f
we show a typical chaotic trajectory with large diffusion and no apparent long time trapping in any of the fundamental resonances .
two points are worth reiterating in the context of the examples shown in fig.[fig 4 ] and fig.[fig 5 ] .
the first is that all these various , and other , dynamical behaviors are found at the same energy indicating the rich structure inherent to the classical phase space .
this energy is by no means special and indeed similar observations hold for other energies in the polyad @xmath1 as well as different polyads .
secondly , the overall phase space might look chaotic on long time scales but the regular regions , even those with phase space area @xmath137 , lead to significant stickiness@xcite and hence long time trappings in one or the other significant resonance zone .
this observation , we emphasize again , is not new and is well known in the nonlinear dynamics community . however in this work we are demonstrating this effect for a realistic spectroscopic hamiltonian with important consequences for ivr in the molecule . exhibiting rapid frequency jumps . for clarity the ratios @xmath129 ( solid line with points ) and @xmath130 ( dashed )
are shown . note
that the jumps ( indicated by arrows ) take the trajectory from being near @xmath32 resonance to being near the @xmath31 resonance .
only a 2.5 ps portion of the 15 ps trajectory is shown .
the jumps occur over the entire time span.,width=288,height=288 ] a further dynamical effect that occurs in dco has to do with rapid jumps in the frequency ratio space .
an example for @xmath2 is shown in the fig.[fig 6 ] where the ratio @xmath129 is undergoing rapid jumps between otherwise smooth variations .
note that the ratio @xmath130 is fairly smooth and does not exhibit the rapid jumps .
the jumps in the frequency ratio happen from the trajectory being near the @xmath32 resonance to the trajectory being near the @xmath31 resonance .
such rapid jumps have been observed by arevalo in ocs as well as the planar 3-body problem@xcite . in this instance
the jumps seem to correspond to heteroclinic orbits that connect the @xmath32 and the @xmath31 resonance zones . however confirming this involves computing invariant manifolds and their intersections which is far from easy .
nevertheless , such rapid resonance transitions could have important implications for the nature of ivr in dco . to indicate the strongly coupled nature of the dco dynamics we show the total diffusion for @xmath2 over the state space in fig.[fig 7 ] .
in fig.[fig 7]a the diffusion is shown for the phase space slice @xmath138 . for reference figs.[fig
7]c , d show the diffusion data for integrable @xmath31 and @xmath32 subsystems .
it is easy to see the @xmath31 and @xmath32 resonance zones in the integrable single resonance cases . comparing to fig.[fig 7]a
it is clear that the resonances have overlapped strongly .
further in fig.[fig 7]b we show the diffusion for the phase space slice @xmath139 .
note that the diffusions for the two slices of phase space are dramatically different .
this implies that it is not possible to understand the intramolecular dynamics of dco based on @xmath36 alone .
a key observation here is that the diffusions in state space show significant heterogeneity .
for instance in fig.[fig 7]a , b we have regions of similar diffusion strengths seperated by a region of larger diffusion strengths .
thus if one imagines going along a specific @xmath31 or @xmath32 polyad in state space then alternating diffusion strengths are encountered .
similarly along a specific energy contour the diffusion values can vary considerably .
this leads to trappings of the classical trajectory and hence bottlenecks to energy flow .
for instance in fig.[fig 7]b the region of low diffusion around @xmath140 is due to the existence of a periodic orbit of the full system . in the semiclassical limit
the heterogeneous nature of the state space diffusion can have interesting consequences for eigenstate localizations .
temps _ et al .
_ discuss@xcite the differences between choosing a co - stretch bright state ( experimental ) versus the dco bend bright states .
it was shown that the dco bend states undergo much faster unimolecular decay . in order to gain dynamical insights into their findings
the various bright states were analyzed from the lfa perspective .
several initial conditions corresponding to the bright states _
@xmath141 were generated with the actions @xmath142 corresponding to the bright state of interest and randomly choosing the angle @xmath119 variables .
the various dynamical quantities of interest are then averaged over the angles .
this is done for two main reasons .
first , although the correspondence between actions and quantum numbers is well known a similar correspondence for angles is not very clear . in the deep semiclassical limit one
expects a correspondence but the problem at hand is far from such a limit .
second observation is that the quantum dynamics associated with a particular zobs , semiclassically speaking , would naturally involve a family of classical trajectories which are modeled here with the random angle distributions .
the fact that the diffusion of the frequency ratios are strongly dependent on the choice of the phase space slice further support the procedure of angle averaging .
the quantities of interest are number of crossings per unit time of the two main resonances ( denoted @xmath143 ) , total time spent near the resonances ( denoted @xmath144 ) , longest locking time in a given resonance ( denoted @xmath145 ) , and the diffusions @xmath146 . in this study
we have averaged over 1000 trajectories with the angles chosen such that @xmath147 @xmath4 and each trajectory was propagated to @xmath148 ps . , and @xmath149 respectively .
averaging is done over 1000 trajectories and @xmath148 ps . in ( a ) ,
( b ) , and ( c ) the filled circles and triangles correspond to the @xmath31 and @xmath32 resonances respectively while the open symbols are for the case with @xmath150 .
see text for details.,width=288,height=288 ] in fig.[fig 8 ] we show the various dynamical quantities of interest for three of the experimental co stretch bright states and three of the bend only bright states corresponding to the polyads @xmath151 . in fig.[fig 8]a
the number of resonance crossings per unit time are shown .
it is immediately clear that the @xmath31 resonance plays a dominant role for the co stretch bright states and that the @xmath32 resonance is key for the bend bright states .
notice that for the @xmath152 and the @xmath149 bright states the flux across @xmath31 is almost zero . on the other hand despite the dominance of @xmath31 flux for the co stetch zobs , the @xmath32 flux is not negligible .
this already provides a hint to the drastically different ivr dynamics from the two classes of zobs with @xmath153 being an exception .
fig.[fig 8]b shows the longest locking time , on average , for the different bright states .
interestingly even though the @xmath31 flux is larger than the @xmath32 flux for the co stretch states the @xmath154 s show the opposite trend .
the total time spent near a particular resonance as shown in fig.[fig 8]c also indicates that the ivr from co stretch states is strongly effected by the @xmath32 resonance .
the diffusions in fig.[fig 8]d show marginal differences .
a clear picture emerges for the ivr dynamics associated with the co stretch bright states .
initially the @xmath31 resonance transfers energy to the dc stretch ( dissociation mode ) but prior to a substantial buildup of excitation in the dc stretch the energy is `` siphoned off '' into the bend mode via the @xmath32 resonance .
this process happens for a fairly long time leading to a highly nonstatistical nature of the dynamics and unimolecular decay of dco . on the other hand for the bend bright states corresponding to @xmath155 the initial ivr dynamics
is dominated by the @xmath32 resonance .
however the @xmath31 resonance is unable to compete and this results in unhindered transfer of energy to the dissociative mode via the @xmath32 coupling .
exception to the nature of bend zobs is clearly seen in the case of @xmath156 wherein the @xmath31 resonance does compete and leads to transfer of energy from the dissociative mode to the co stretch mode .
an important observation is that the diffusion of frequencies computed here is the lowest for the zobs @xmath149 which undergoes the fastest unimolecular decay .
this is not a contradiction since large difusion corresponds to many hops and consequent locking leading to incomplete flow of energy into the dissociative mode .
for example the largest diffusion is predicted for the zobs @xmath156 which is consistent with the fact that the times spent near the two main resonances are similar as compared to the other states .
thus @xmath156 should decay the slowest which is confirmed by computing the associated survival probability ( cf .
figure 3(b ) in ref . ) .
it is interesting to observe that the diffusions for the co stretch bright states are not very different whereas in the case of the bend bright states there is a large variation .
this demonstrates the mode - specificity of the intramolecular dynamics of dco . in fig.[fig 8 ] the effect of setting @xmath150 is shown .
it is clear that the largest effect of the @xmath93 fermi resonance is for the co stretch bright states @xmath94 and @xmath91 . in the absence of @xmath93 the @xmath31 flux
is decreased and the @xmath32 flux is increased with larger time being spent , on the average , near the @xmath32 fermi resonance .
this indicates that the co stretch bright states would lead to smaller rates of unimolecular decay in the absence of the @xmath93 resonance .
this is true for the bend bright states as well although the effect is very small .
thus the role of @xmath93 fermi resonance is not negligible for the lower polyads . as a final observation we point out that amongst the co stretch bright states
there is a sharp increase , on going from @xmath94 to the @xmath91 , in the average residence time and the average longest locking time for the @xmath31 resonance .
this perhaps explains the efficiency of the @xmath31 resonance in @xmath2 but at this stage it is conjectural .
a long term goal of our research is to come up with techniques that would allow fairly detailed classical - quantum correspondence studies of highly excited states in systems with three degrees of freedom or more . to this end
two dimensionality independent techiniques , parametric variations and local frequency analysis , have been employed in this study .
we have demonstrated the utility of both the techniques for understanding the intramolecular dynamics of dco .
the successes of both the techniques , in dynamical assignments of states and establishing the nonstatistical nature of the dynamics , given the strongly coupled nature of the system ( coupling strengths comparable to the mean level spacing ) is quite encouraging .
we conclude with brief comments on the techniques and future outlook .
the level velocity approach , used to dynamically assign the eigenstates , is certainly capable of identifying strongly localized states .
the close connections between the level velocity of an eigenstate and its corresponding phase space nature makes this approach attractive .
previous application@xcite to a mixed 2-mode system demonstrated the utility of the level velocities for dynamical assignments of the highly excited states . in this work
we have shown that the method is useful even in the presence of significant chaos .
although avoided crossings lead to difficulties , and this is not surprising , it is debatable wether the resulting mixed states can be or should be assigned . demixing the states invovled in avoided crossings and then analyzing them
can be done within this approach as in any other approach .
however the procedure of demixing , especially in multistate or broad avoided crossings , should not be considered as an assignment .
this work shows that analyzing the level velocities in conjunction with measures like iprs can be a fairly useful way of dynamically assigning highly excited states . at this stage
the effects of various bifurcations on the level velocity spectrum is not known .
interestingly in the case of dco , despite substantial change in the phase space structure with changing energies , the level velocity corresponding to a state localized about stable regions is very close to the classical estimate .
this has been observed before@xcite in the case of hcp as well .
this suggests a very detailed classical - quantum correspondence and needs to be studied further . in order to make this approach more useful
it is important to obtain some bounds on the level velocities in the mixed and strongly chaotic situations .
thus a semiclassical theory for the velocities in the chaotic regimes needs to be developed and work is in progress in our group along these lines .
note that the distribution of the velocities has been found to be gaussian in the strong chaos limit@xcite .
the general technique of parametric variations has a wide range of applications and level curvatures have also been studied@xcite .
it remains to be seen if the level curvatures have any utility in understanding highly excited states and ivr .
for example conductance of disordered systems have been related to level curvatures@xcite and given the tantalizing similarities between the example mentioned and ivr it seems worthwhile to investigate the utility of level curvatures in molecular systems .
the lfa technique is able to provide detailed phase space information in a strongly coupled system like dco .
in particular the time - frequency analysis revealed the existence of trajectories trapped near important resonance zones . in an effective spectroscopic hamiltonian the key resonances are easy to expect since the fit already builds in the main resonant perturbations but lfa
is able to provide information on the competing roles of these various resonances .
an example is the clear differences between the co - stretch bright state and dco bend bright state dynamics for dco .
lfa is also able to highlight the role of weak resonances like the @xmath93 fermi resonance .
as shown in this work and in the previous works the wavelet based approach to lfa is useful in analyzing effective hamiltonians and various dynamical quantities can be determined to understand ivr .
lfa clearly shows that despite significant chaos the dco phase space is far from the rmt regime . in this paper
it is shown that even for a strongly coupled system like dco the state space diffusion has a lot of structure with regions of high and low diffusions interspersed thorughout .
the diffusion of frequencies provides clear insights into the dynamics at a particular polyad and/or a specific energy .
this information can then be used to determine the nature of the random walk in action space for strongly coupled systems and thus provide more insights into certain recent conjectures@xcite regarding the intermediate time behavior of survival probabilities . although not shown here most of the trajectories exhibit across - resonance diffusion which is quite fast and implicates higher order secondary resonances as well@xcite .
the rapid frequency jumps which perhaps correspond to heteroclinic orbits joining unstable periodic orbits have not been investigated in detail .
surely the rapid jumps from one resonance to another have important consequences for ivr and needs further study .
we also note that there is not much of an arnold web for dco for @xmath13 in contrast@xcite to h@xmath157o .
dco being effectively two dimensional implies that there is no transport along resonance lines and thus many of the interesting conjectures put forward@xcite by martens , davis , and ezra remain to be tested .
currently work is in progress for true three degree of freedom systems . as a final note
we mention that both techniques used in the current study probe localiztion characteristics
. the parametric response is probing eigenstate localization and the lfa is probing localization of phase space transport due to partial barriers and broken seperatrices .
localization explicitly implies nonstatistical behavior and thus deviations from rmt and incomplete ivr in the system .
evidence exists@xcite that the quantized versions of the partial barriers can be more restrictive than the classical analogs . a interesting link between these two approaches
is provided by the recently proposed intensity - velocity correlator@xcite .
detailed studies of true three degrees of freedom system via the intensity - velocity correlator can lead to a better understanding of the ivr dynamics .
it is a pleasure to acknowledge useful discussions with dr .
l. v. arevalo regarding the wavelet based approach to time - frequency analysis .
we are also grateful to prof .
greg ezra for sending us a copy of the unpublished work ref .. this work is supported by funds from the department of science and technology , india . | intramolecular dynamics of highly excited dco ( @xmath0 ) is studied from a classical - quantum correspondence perspective using the effective spectroscopic hamiltonian proposed recently by trllsch and temps ( z. phys .
chem . * 215 * , 207 ( 2001 ) ) .
this work focuses on the polyads @xmath1 and @xmath2 corresponding to excitation energies @xmath3 5100 @xmath4 and 7000 @xmath4 respectively .
the majority of states belonging to these polyads are dynamically assigned , despite extensive stochasticity in the classical phase space , using the recently proposed technique of level velocities .
a wavelet based time - frequency analysis is used to reveal the nature of phase space transport and the relevant dynamical bottlenecks .
the local frequency analysis clearly illustrates the existence of mode - specific ivr dynamics _
i.e. , _ differing nature of the ivr dynamics ensuing from co stretch and the dco bend bright states .
in addition the role of the weak fermi resonance involving the co stretch and dco bend modes is investigated .
a key feature of the present work is that the techniques utilized for the analysis _
i.e. , _ parametric variations and local frequency analysis are not limited by the dimensionality of the system .
this study , thus , explores the potential for understanding ivr in larger molecules from both time domain and frequency domain perspectives . |
there are numerous phenomena , where first - order phase transitions occur between phases of different densities . in nuclear physics various first - order phase transitions may take place in the early universe @xcite , in heavy ion collisions @xcite and in neutron stars ( e.g. , pion , kaon condensation and deconfinment or chiral phase transition @xcite ) . at low collision energies
the nuclear gas - liquid ( ngl ) first - order phase transition is possible @xcite .
the experimental status of the latter is discussed in refs . @xcite . at high collision energies the hadron
quark gluon plasma ( hqgp ) first - order phase transition may occur ( see e.g. ref .
@xcite ) . within the hydrodynamical approach dynamical aspects of the ngl and hqgp transitions have been recently studied in refs .
. an important role of effects of non - ideal hydrodynamics has been emphasized in ref .
however in these works density fluctuations were carefully analyzed , whereas thermal transport effects were only roughly estimated .
dynamical trajectories of the system in heavy - ion collisions can be approximately determined by constant initial values of the entropy per baryon ( i.e. entropy density per net baryon density ) .
this conclusion follows from the analysis of heavy ion collisions at energies less than several gev@xmath2 performed in an expanding fireball model ( e.g. , see ref .
@xcite ) , from ideal hydrodynamical calculations in a broad energy range ( see e.g. refs .
@xcite ) and also from transport calculations ( e.g. , see ref .
@xcite ) .
it is well known , see refs .
@xcite , that , at least in the mean field approximation , the isothermal spinodal ( its ) line and adiabatic spinodal ( as ) line are different . from the exact thermodynamic relation one
obtains @xmath3 ^ 2 .
\label{uc } \]]here @xmath4 is the temperature , @xmath5 is the pressure , @xmath6 is the density of the conserving charge ( here the baryon density ) , @xmath7 is the baryon mass , @xmath8 is the entropy per baryon , @xmath9 is the specific heat density at the fixed volume @xmath10 , @xmath11 and @xmath12 .
the value @xmath13 has the meaning of the adiabatic sound velocity ( at @xmath14const ) .
namely , this quantity characterizes propagation of sound waves in ideal hydrodynamics .
as we show below , in non - ideal hydrodynamics at a finite value of the thermal conductivity , @xmath15 , the propagation of sound waves is defined by the interplay between @xmath16 ( at @xmath17const ) and @xmath13 .
conditions @xmath18 and @xmath19 define @xmath20-curves : the its line ( at fixed temperature @xmath4 ) and as line ( at fixed @xmath21 ) , respectively .
the maximum temperature points on these lines are the critical temperature @xmath22 in case of @xmath4=const and the adiabatic maximum temperature @xmath23 in case of @xmath24const . in the mean field approximation ,
the specific heat density @xmath9 has finite non - negative values .
thus one has @xmath25 on the as line , where @xmath26 .
calculations performed in mean field models show that @xmath27 . in refs .
@xcite and numerous subsequent works , crossing of the as line by effective trajectories @xmath21=const was considered as the starting point for the multifragmentation following an adiabatic expansion of the system .
in contrast , in ref .
@xcite and later refs .
@xcite it was assumed that the multifragmentation may appear already at higher temperatures , when the its curve is crossed by the @xmath21=const lines .
the argumentation of the former group of works can be traced back to the fact that in ideal hydrodynamics instabilities indeed arise at the as boundary ( see extensive discussions in ref .
isothermal instabilities at first order phase transitions were studied in details in refs .
@xcite .
in mean field theories , difference between values of @xmath22 and @xmath23 is usually rather large .
for example , within the njl model one obtains @xmath28 mev ( see ref .
@xcite ) for the chiral restoration transition . in the model used in ref .
@xcite @xmath29 mev for hqgp phase transition .
note that simultaneously with the low value @xmath30 mev a sufficiently high value of @xmath31 should be reached .
this looks unlikely to be achieved in heavy - ion collisions . hence ,
if instability occurs for @xmath32 , one could face difficulties in exploring the hqgp spinodal instabilities in actual experiments with heavy ions .
it was argued in ref .
@xcite that above mentioned difference of @xmath22 and @xmath23 is probably an artifact of the mean field approximation and both values of the critical temperature should coincide provided thermal fluctuation effects are included .
indeed , it is known ( e.g. , see ref .
@xcite ) , that inclusion of fluctuations may result in a weak divergence of the specific heat .
for example , for the three - dimensional ising universality class @xmath33^{-\alpha}$ ] , where @xmath34 . it can be concluded from eq .
( [ uc ] ) , that for divergent @xmath9 both temperatures @xmath35 and @xmath36 should coincide .
although this observation is fully true , it seems legitimate to rise some questions .
* first , thermal fluctuation effects are strong enough only in a critical region , where the mean field theory breaks down and system properties are governed by non - trivial critical exponents .
for many systems the critical region is a rather narrow region near the critical point , while contrary examples are known as well @xcite .
although it is not known a priori how large is the hqgp critical region , there is an indication in favor of the hqgp critical region being small @xcite . only if the system trajectory passes trough the critical region , one can hope that fluctuation effects appearing during evolution of the system may lead to @xmath37 . *
second , even if the system trajectory passes through the critical region and @xmath35 coincide with @xmath36 , the its and the as lines may not collapse into one line even within the critical region because the fluctuation contribution to the specific heat diverges only at the critical point and then sharply decreases .
thus we should understand crossing of which of two lines ( the its or as line ) results in the spinodal instability . *
third , the divergence of @xmath9 at the critical point is due to long - scale fluctuations , which need a long time to develop . therefore the time , which the system spends in a vicinity of the critical point in the course of the heavy ion collision , might be insufficient for the developing of such fluctuations , see discussions in refs .
@xcite .
moreover , if the initial collision energy is such that the system trajectory passes outside the critical region , i.e. entirely in the mean field regime , the thermal fluctuation effects do not contribute at all .
then the question arises how one can treat the problem purely within the mean field approximation .
this is actually an important question , since most of equations of state used in the description of heavy ion collisions disregard the fluctuation effects . in refs .
@xcite we focused on the study of dynamical effects of the viscosity and the surface tension .
the aim of the given paper is to consider the influence of thermal conductivity effects on the dynamics of first - order phase transitions and to answer above risen questions .
below we find out that in non - ideal hydrodynamics ( with non - zero thermal conductivity , @xmath38 ) instabilities show up at the crossing of the its line and only for @xmath1 , e.g. in ideal hydrodynamics , they appear at the as line .
phase diagram of the van der waals equation of state , @xmath20-plane .
bold solid , dashed and dash - dotted curves demonstrate boundaries of the maxwell construction , the spinodal region at @xmath17const and @xmath24const , respectively .
the short dash lines show adiabatic trajectories of the system evolution , see explanation in the text.,height=226 ] phase diagram for the van der waals equation of state in @xmath39 plane , see explanation in the text . ,
to simplify further considerations we describe the dynamics of a first - order phase transition by means of the standard system of equations of non - ideal non - relativistic hydrodynamics : the navier - stokes equation , the continuity equation , and the general equation for the heat transport : @xmath40= -\nabla_i p
\\ \label{contin } & & + \nabla_k \left [ \eta \left ( \nabla_k u_i + \nabla_i u_k -\frac{2}{d } \delta_{ik } \mbox{div } \mathbf{u } \right ) + \zeta \delta_{ik } \mbox{div } \mathbf{u } \right ] , \nonumber\end{aligned}\]]@xmath41@xmath42=\mbox{div}(\kappa \nabla t ) \\ & & + \eta \left(\nabla_k u_i + \nabla_i u_k -\frac{2}{d } \delta_{ik } \mbox{div } \mathbf{u } \right)^2 + \zeta ( \mbox{div } \mathbf{u})^2.\nonumber\end{aligned}\ ] ] here @xmath43 and @xmath44 are the shear and bulk viscosities ; @xmath45 is the velocity of the fluid element ; @xmath46 is the entropy density ; @xmath15 is the thermal conductivity ; @xmath47 is the dimensionality of space .
the short - scale fluctuations can be taken into account by adding noise terms on the right sides of eqs .
( [ navier]-[therm ] ) . in what follows we neglect these contributions .
before we proceed further we aware the reader , that eqs .
( [ navier]-[therm ] ) are derived in the first gradient order approximation , see ref .
@xcite . in the kinetic regime
the typical inverse relaxation time of the system @xmath48 and the typical inverse spatial scale @xmath49 characterizing the collective processes fulfil the inequalities @xmath50 and @xmath51 , see @xcite , where @xmath52 and @xmath53 characterize microscopic processes . for quasi - equilibrium nucleon system with @xmath54 ( @xmath55 is the nucleon fermi energy ) one has @xmath56 and @xmath57 .
the hydrodynamic regime is reached if the typical time scale and the typical length scale characterizing the collective motion are larger than the corresponding kinetic values , @xmath58 and @xmath59 . in what follows
we assume that these inequalities are fulfilled .
an approximate validity of the adiabatic expansion ( i.e. conservation of specific entropy density , @xmath21 , along the fluid flow ) in heavy ion collisions @xcite means that ( i ) @xmath15 is rather small , or the system is spatially homogeneous and ( ii ) @xmath43 and @xmath44 are small , or the velocity of the expansion of the system is very low ( with the guaranty for absence of shock waves ) .
note that in the general case solutions of ref .
@xcite , which we will exploit below , are obtained at a non - zero viscosity therefore we will not additionally assume smallness of the viscosity coefficients , rather we exploit smallness of the velocity @xmath60 .
the best known example to illustrate principal features of a first order phase transition in the mean field approximation is the van der waals fluid .
expressions for thermodynamical quantities in this particular case are presented in appendix a. trajectories for the expansion of the uniform system , obtained at the assumption @xmath61 const , are shown in fig . 1 on the plot @xmath62 .
the upper convex curve ( mc , bold solid line ) is the boundary of the maxwell construction , the next bold dashed line ( its ) shows the boundary of the its region and the lower bold dash - dotted curve ( as ) , the boundary of the as region .
the short dashed lines crossing the regions under these three curves are calculated at two values of the entropy per baryon : @xmath63 and @xmath64 .
different scenarios of the dynamics for first - order phase transitions in expanding system at @xmath61 const have been considered previously in ref .
@xcite . the super - cooled vapor ( sv ) and the overheated liquid ( ol ) regions are between the mc ( bold solid curve ) and the its ( bold dash ) curves , on the left and on the right respectively . for @xmath65 , where @xmath63 corresponds to the critical point and the line with @xmath66 passes through the point @xmath67 at @xmath68 , the system traverses the ol state ( region ol in fig . 1 ) and then the its region ( below its line ) and the as region ( below as line ) . for @xmath69
the system trajectory passes through the sv state ( the region sv in fig .
1 ) and then the its region . in fig .
2 we show the same regions as in fig.1 , but in the plane @xmath70 for @xmath71 .
the line @xmath72 corresponds to the maximum pressure at the as curve .
the critical point is determined by @xmath73 , @xmath74 . with the help of eq .
( [ p ] ) of appendix a from these conditions
we find @xmath75 , @xmath76 .
the maximal pressure point at as line is defined by @xmath77 , @xmath78 . thus using eq .
( [ ps ] ) we obtain : @xmath79 , @xmath80 , see fig .
2 . the maximal temperature point on the as curve corresponds to @xmath81 , @xmath82 ,
another relevant point on each line @xmath83 is the turning point , where @xmath84 has a minimum as a function of the density , @xmath6 . in this point
the pressure @xmath85 equals zero and the fireball expansion is slowing down .
the lines labeled @xmath86 and @xmath87 cross respectively the as and its lines at @xmath88 .
note , that with realistic equation of state , @xmath88 might be reached at the ngl first - order phase transition , see ref .
@xcite , but not at the hqgp transition .
we stress that we use the van der waals equation of state only for the illustrative purpose .
our main conclusions do not depend on the used equation of state provided there is a first order phase transition .
we may expand all quantities near an arbitrary reference point @xmath89 , or @xmath90 .
analytical solutions can be found , provided the system is in the vicinity of the critical point @xmath91 , or near the maximum point @xmath92 . therefore considering the problem in @xmath93 variables
it is convenient to take @xmath91 as the reference point and considering the problem in @xmath94 variables , to take @xmath95 as the reference point .
the following remark is in order .
as we have mentioned , in this paper we exploit the mean field approximation . thus using @xmath91 as the reference point
we actually assume that the fluctuation region ( @xmath96 , @xmath97 ) is very narrow and we are able to chose the reference point outside this region ( but nevertheless very close to the critical point ) . also , mentioned problem is avoided provided one considers the system at the time scale , being shorter than that necessary for a development of long - scale critical fluctuations .
we construct the generating functional in ( @xmath98 , @xmath99 ) or in ( @xmath100 , @xmath101 ) variables , the landau free energy , such that @xmath102 : @xmath103 ^2}{2}+\frac{\lambda m^3 ( \delta n)^4}{4}-\frac{\lambda v^2 m(\delta n ) ^2}{2}-\epsilon \delta n \right\},\end{aligned}\ ] ] where @xmath104 is expressed through the pressure on the maxwell construction .
the maximum of the quantity @xmath105 is @xmath106 , therefore we may count @xmath107 from this value supposing further that @xmath108 with @xmath109 .
the first term in eq . ( [ fren ] ) is due to the surface tension .
it produces a surface pressure @xmath110 ( see eq . ( [ v - t ] ) below ) . in case of the van der waals form of the equation of state ,
see appendix a , expanding the pressure of the uniform matter in @xmath111 , @xmath99 , for @xmath112 , @xmath113 , i.e. in the vicinity of the critical point @xmath114 , we obtain @xmath115=f_0 \left[\frac{9\delta { { t}}\delta
n } { 4}+\frac{9t_{cr } ( \delta n)^3}{16n_{cr}^2 } + \frac{3}{2 } n_{cr } \delta t + ... \right].\ ] ] the last explicitly presented term in eq .
( [ modif ] ) is unimportant if @xmath116 const , since the addition of any constant does not change equation of motion ( [ navier ] ) .
comparing eq .
( [ modif ] ) with the value , which follows from eq .
( [ fren ] ) , we find relations between coefficients : @xmath117 expanding the pressure with respect to @xmath118 , @xmath119 in the vicinity of the maximum of @xmath39 at fixed @xmath120 we obtain @xmath121=\frac{15}{512}t_{p , max } \left [ \delta \tilde{s } \delta n + \frac{f_0 ( \delta n)^3 } { 2 n_{p , max}^2}+\frac{114}{15}n_{p , max}\delta\tilde{s}+ ...
\right].\end{aligned}\ ] ] comparing this expression with the value , which follows from eq .
( [ fren ] ) , we find relations @xmath122 further , following refs .
@xcite , we use smallness of the velocity @xmath123 of the growth / damping of the density and temperature fluctuations .
we linearize all terms in hydrodynamical equations in the velocity @xmath124 , the entropy density @xmath125 and the temperature @xmath126 variables . here
index `` r '' stays for an arbitrary reference point .
excluding a dependence on the velocity , @xmath45 , from eqs .
( [ navier ] ) , ( [ contin ] ) we obtain @xcite @xmath127 , \ ] ] where @xmath110 is the surface pressure ; @xmath128 , @xmath129 for spherical geometry , @xmath130 .
we count @xmath5 from its value in the final equilibrium state ( reached at @xmath131 ) .
namely this difference , @xmath132 , has the meaning of the thermodynamical force driving the system to the final equilibrium state .
( [ v - t ] ) should be supplemented by the equation for the heat transport ( [ therm ] ) , which after linearization owing to eq .
( [ contin ] ) simplifies as @xmath133=\kappa_{\rm r}\delta \delta t.\ ] ] the variation of the temperature is related to the variation of the entropy density @xmath134 $ ] by @xmath135 where , as in eq .
( [ uc ] ) , @xmath136 is the specific heat density .
since in this section the `` r''-reference point can be taken arbitrary , we suppress index `` r '' . to find solutions of hydrodynamical equations we present , cf .
@xcite , @xmath137 , \nonumber \\
\delta s= \delta s_0 \mbox{exp}[\gamma t + i \mathbf{p}\mathbf{r } ] , \\
t = t_{>}+\delta t_0 \nonumber \mbox{exp}[\gamma t + i \mathbf{p}\mathbf{r}].\end{aligned}\ ] ] where @xmath138 is the temperature of the uniform matter . for finite thermal conductivity , @xmath0 , from eqs .
( [ v - ts ] ) , ( [ t - s ] ) we obtain @xmath139 } \left[1+\frac { n \kappa p^2(\partial s/\partial n)_{t } } { \gamma s c_{v } } \right].\end{aligned}\ ] ] the increment , @xmath140 , is determined by eq .
( [ v - t ] ) , where we put @xmath141 thus we obtain @xmath142 .
\end{aligned}\ ] ] this equation differs from that has been derived in ref .
@xcite by presence of extra surface tension term ( @xmath143 ) .
( [ g2 ] ) has three solutions . expanding the solutions for small momenta ( long - wave limit )
we find @xmath144\frac{p^2}{2},\label{solg1 } \\ \gamma_3 & = & - \frac{\kappa u_t^2 p^2}{u_{\tilde{s}}^2c_{v } } \left [ 1 - \frac{u_t^2-u_{\tilde{s}}^2}{u_{\tilde{s}}^2u_t^2}\left ( c + \frac{\kappa u_t^4 } { u_{\tilde{s}}^2 c_{v } ^2 } - \frac{(\tilde{d}\eta + \zeta ) \kappa u_t^2}{mn c_{v } u_{\tilde{s}}^2 } \right ) p^2 \right ] .
\label{solg}\end{aligned}\ ] ] the solutions @xmath145 correspond to the sound mode in the long wavelength limit .
the third solution , @xmath146 , is responsible for the thermal transport mode .
below the its line ( and above the as line ) @xmath147 , @xmath148 , solutions @xmath145 correspond to an oscillation and damping , whereas @xmath146 describes an unstable growing solution .
below the as line , since @xmath149 and @xmath150 , the modes exchange their roles : the sound modes become unstable , while the thermal mode is damped .
we solved the system of non - relativistic hydrodynamical equations .
solutions generalizing eqs .
( [ solg1 ] ) and ( [ solg ] ) to the relativistic case are presented in appendix b. further we will consider two limiting cases of a large and small thermal conductivity , @xmath15 . for @xmath151 from eq .
( [ g2 ] ) we arrive at @xmath152,\ ] ] that we have derived in refs .
note that dependence on @xmath15 disappeared in this limiting case .
the solutions are @xmath153 here @xmath154 is the dimensionless parameter . for @xmath155 , and @xmath156 for the particular case of the van der waals fluid in the vicinity of the critical point @xmath157 ( taken as the reference point , cf
( [ modif ] ) ) .
an instability arises within the its region , where @xmath147 .
the most rapidly growing mode corresponds to @xmath158 , and @xmath159 : @xmath160 using these expressions we rewrite the condition @xmath161 as @xmath162 then also we have @xmath163 and following ( [ 123 ] ) : @xmath164}{c_{v } n \left[1+{\kappa } /(\sqrt{c}c_{v } ) \right]}.\ ] ] we see that an assumption on the spatial uniformity of the system fails right after the its region is reached .
an aerosol ( mist ) of bubbles and droplets is formed for a typical time @xmath165 .
for the van der waals equation of state using eq .
( [ s ] ) of appendix a we find that @xmath166 , i.e. the temperature is larger in denser regions .
however the amplitude of the temperature modulation is rather small for @xmath167 in the limiting case , one can neglect at all small deviations of @xmath168 and put @xmath169 in eq .
( [ t0 ] ) .
formally this limit corresponds to setting @xmath170 in eq .
( [ t0 ] ) . in order to avoid misunderstanding
we stress that we continue to consider only slow motions staying in the framework of the validity of the non - relativistic hydrodynamics @xcite . as before
, we assume that at a slow expansion with @xmath171 const the initially uniform system reaches its region ( see corresponding curves in fig .
2 ) . solution of eq .
( [ therm ] ) for @xmath172 corresponds to @xmath173 , and to the approximate conservation of the entropy ( provided the velocity @xmath60 or the viscosities are small ) , as it follows from the l.h.s of eq .
( [ therm ] ) .
we may also see it from eq .
( [ v - ts ] ) . for that
let us divide this equation on @xmath15 .
then we can put zero the l.h.s . of thus obtained equation ( since @xmath172 ) .
any solution @xmath173 fulfils this equation . using eq .
( [ delns ] ) we arrive at @xmath174 ( not dependent on @xmath175 and @xmath176 ) .
the latter approximation was used in refs .
@xcite , where we solved eq .
( [ v - t ] ) for @xmath17const and derived eq .
( [ g ] ) and eq .
( [ gsol ] ) . on the other hand , from eq .
( [ t - s ] ) at @xmath177 we find @xmath178 , that follows from eq .
( [ 123 ] ) as well . using that @xmath177 in eq .
( [ varp ] ) we again arrive at eq .
( [ g ] ) and eq .
( [ gsol ] ) . for the sake of simplicity
, here we consider the case , when both shear and bulk viscosities are zero , but thermal conductivity is small but non - zero . dropping @xmath179 term in the l.h.s .
( [ g2 ] ) ( we justify this approximation below ) we immediately find the solution @xmath180 expanding this value in small @xmath181 we see that it , indeed , corresponds to the solution @xmath146 given by eq .
( [ solg ] ) .
the most rapidly growing mode corresponds to @xmath182 slightly below the its line ( for @xmath183 ) , @xmath184 now we are able to check the validity of our solution in this case . comparing different terms in eq .
( [ g2 ] ) we see that it is legitimate to drop the term @xmath179 in the l.h.s . of equation ,
provided @xmath185 comparing eq .
( [ gam3 m ] ) with eq .
( [ gmax ] ) for @xmath186 ( zero viscosity ) we see that at @xmath187 , when both solutions @xmath188 and @xmath189 are valid , the following inequality is satisfied @xmath190 .
thus we found that in all considered above cases the instability occurs at the crossing of the its line .
however , for @xmath191 the most rapidly growing mode corresponds to @xmath188 and in the opposite limit @xmath192 , to @xmath189 . in this specific limit
there are significant differences compared to the case of any non - zero @xmath193 .
indeed , for @xmath1 there is no solution with @xmath194 . below the its line and above the as line the thermal mode , that drives the system towards equilibrium for non - zero thermal conductivity , does not exist for @xmath1 .
thus the evolution in the spinodal region is entirely governed by approximately adiabatic sound excitations . from eq .
( [ v - ts ] ) we find @xmath195 and from eqs .
( [ t - s ] ) and ( [ s0 ] ) @xmath196}{c_{v } n } .\ ] ] thereby the temperature is modulated , similar to the entropy density .
for the van der waals equation of state using eq .
( [ s ] ) of appendix a we find that @xmath197 .
the amplitude of the temperature modulation is here larger than in case of large @xmath198 , see eq .
( [ t0 ] ) . the increment @xmath199 is given by eq .
( [ g ] ) with @xmath13 replaced by @xmath16 : @xmath200.\ ] ] thus , contrary to the case @xmath0 , instability appears , when the system trajectory crosses the as line rather than the its line .
the value @xmath201 for the particular case of the van der waals fluid in the vicinity of the point @xmath202 ( taken as the reference point , cf .
( [ modifs ] ) ) .
this result holds also in case of the ideal hydrodynamics , where in addition to @xmath1 , the viscosity coefficients ( @xmath43 , @xmath44 ) are zero .
note that for any small but finite value @xmath15 the solution @xmath146 results in instability already for @xmath203 ( i.e. below the its line rather than below the as line ) .
since in reality always @xmath204 , we conclude that the instability condition @xmath149 , as one usually derives it within the framework of ideal hydrodynamics , should be replaced by the condition @xmath203 .
during its evolution a dynamical system crosses a metastable state region ( binodal ) before entering to the region of the instability ( spinodal ) , see figs . 1 and 2 .
if a system evolves rather slowly a phase transition occurs already in a metastable region .
therefore let us consider the system in the metastable state .
small seeds of the stable phase , being produced in fluctuations , converge , while seeds of overcritical sizes start to nucleate , see ref .
@xcite . in nuclear physics
these processes have been considered in many works , e.g. see ref .
@xcite and references therein . in refs .
@xcite we described the seeds dynamics using equations of non - ideal hydrodynamics . here
we continue this description . an analytical description of metastable states is possible to perform in the limiting cases of infinitely large thermal conductivity ( when one can put @xmath17const , see refs .
@xcite ) and of zero thermal conductivity .
therefore , below we study the limiting cases at first and then we perform estimates for an arbitrary value of thermal conductivity , @xmath15 .
let us consider the situation , when at very slow expansion with @xmath171 const spatially uniform spherical system of a large radius @xmath206 achieves either the ol- state or the sv- state ( see corresponding curves in fig .
1 ) . during subsequent slow expansion , after a while a bubble of an overcritical size @xmath207 of stable gas phase , or respectively a droplet of liquid phase may appear in a fluctuation and begin to grow . in case @xmath208 , corresponding solutions ( [ v - t ] ) describing dynamics of the density in the fluctuation were searched in ref .
@xcite in the form : @xmath209+(\delta n)_{cor},\ ] ] where @xmath210 , the upper sign corresponds to the growth of the bubble and the lower one , to the growth of the droplet , and the solution is valid for @xmath211 , i.e. for @xmath212 const @xmath213 and for the initial bubble / droplet size @xmath214 . in refs .
@xcite we neglected a small correction @xmath215 in eq .
( [ delr ] ) . with this small correction one is able to recover exact baryon number conservation , see below .
one can easily see that eq .
( [ delr ] ) fulfils eq .
( [ v - t ] ) provided the temperature , @xmath216 , does not depend on time and space .
as we argued above , this result follows from eq .
( [ v - ts ] ) , if one formally sets @xmath172 . in case of expanding system
we also assume that typical time for the formation and growth of a fluctuation of our interest ( @xmath217 ) is much smaller than the typical expansion time @xmath218 . substituting eq .
( [ delr ] ) to eq .
( [ v - t ] ) and considering @xmath219 in the vicinity of the bubble / droplet boundary we obtain equation describing growth of the bubble / droplet size @xcite : @xmath220 where @xmath221 $ ] is the time scale .
it follows from the solution of this equation that @xmath222 .
first the bubble / droplet size @xmath223 grows with an acceleration and then it reaches a steady grow regime with a constant velocity @xmath224 . in the bubble / droplet interior @xmath225 , and this region grows with time .
thus @xmath226 produces compensation of the baryon number . note that this correction is very small for @xmath227 and
does not affect solution at @xmath228 .
substituting eq .
( [ delr ] ) to eq .
( [ t - s ] ) for @xmath17 const we obtain @xmath229+(\delta n)_{cor}\right\}. \end{aligned}\ ] ] thus the temperature stays constant in the bubble / droplet interior and exterior and the entropy and the density differ in the interior and exterior regions . in spite of that
, approximation of a quasi - adiabatic expansion of the system as a whole might be used even , when the system reaches metastable region , provided the gas of bubbles / droplets is rare , or the value @xmath230 is small .
let us determine , when in the case @xmath0 we are still able to exploit solutions found above for @xmath231 . estimating from eq .
( [ t - s ] ) @xmath232 ( for @xmath233 ) with @xmath234 from eq .
( [ delr ] ) and setting this estimate in eq .
( [ v - ts ] ) we find the typical time for the heat transport @xmath235 . for @xmath236 one
has @xmath237 and the droplet / bubble dynamics is determined by eq .
( [ dim ] ) for the change of the density , cf .
@xcite . at @xmath172 for all times the droplet / bubble dynamics is determined by eq .
( [ dim ] ) . at @xmath204 for @xmath238 ,
i.e. for @xmath239 , the dynamics is controlled by the heat transport .
limit @xmath1 is again specific . in case
latexmath:[$|n -n_{p , max } @xmath92 , corresponding solutions ( [ v - t ] ) describing dynamics of the density in the fluctuation can be searched in the form ( [ delr ] ) with the only difference that @xmath242 should be replaced by @xmath243 , @xmath244 by @xmath245 and @xmath230 by @xmath246 .
thus @xmath247+(\delta n)_{cor},\ ] ] where @xmath248}$ ] , the upper sign corresponds to the growth of the bubble and the lower one , to the growth of the droplet , and the solution is valid for @xmath249|\ll 1 $ ] , i.e. for @xmath212 const @xmath250 and for @xmath251 .
deriving eq .
( [ delrs ] ) we used that @xmath252 does not depend on @xmath176 and @xmath175 .
dynamics of @xmath253 is determined by eq .
( [ dim ] ) , where one should replace values calculated at @xmath254 to the corresponding values at @xmath92 . from eq .
( [ v - ts ] ) we obtain @xmath255 with @xmath234 given by eq .
( [ delrs ] ) . knowing @xmath256 and @xmath234 we find @xmath242 from eq .
( [ t - s ] ) .
thus not only density , @xmath6 , but also the entropy density , @xmath46 , and the temperature , @xmath4 , change in the surface layer and get different values inside and outside the bubble / droplet .
contrary , the value @xmath21 remains constant .
let @xmath257_{\tilde{s}}$ ] is the typical time for the expansion of the system ( at approximately constant entropy ) prepared in heavy ion collision and the system trajectory passes through the its region . in case , if @xmath258 _ { \tilde{s}}$ ] the aerosol has no time to develop in the corresponding density interval .
then the system ( being covered by fine ripples ) continues a quasi - uniform expansion traversing the its region .
further it may cross the as region ( note that it also can be that the expanding system freezes out before the trajectory crossed the as line ) . as is seen from fig .
2 , the turning line can be further reached , where @xmath259 . for @xmath260
the turning points are lying inside the as region .
for @xmath261 the turning points are outside the as region , but within the its region . in the vicinity of the turning point , the expansion time @xmath262 can be significantly increased yielding @xmath263 .
note that , as follows from the analysis with the realistic equations of state the turning points can be reached only for low values of the entropy @xmath21 .
such low values are reachable in case of low energy heavy ion collisions , where there arises ngl transition . in case of violent heavy - ion collisions at conditions of the first - order hqgp
phase transition values of @xmath21 are much larger and @xmath5 is always positive .
in this work we found various solutions describing dynamics of the density , temperature , and entropy density in the spinodal and metastable regions at first - order phase transitions .
we tried to answer the question how one should treat the fact that critical temperatures at the isothermal and adiabatic lines are different .
in particular , we demonstrated that correct region of instability is not appropriately found within standard ideal hydrodynamical calculations ( where instability manifests itself at the crossing of the adiabatic line ) .
this might be an important finding since ideal hydrodynamics is often used in actual simulations of heavy ion collisions . since in reality @xmath0 , spinodal instabilities take place , when the system phase trajectory crosses the isothermal spinodal line in the course of the heavy ion collision .
the value of the critical temperature , @xmath22 , at the pressure isotherm [ @xmath39 for fixed @xmath4 ] is significantly larger than the value of the maximum @xmath23 at the pressure adiabat [ @xmath39 for fixed @xmath21 ] .
this means that according to our findings the prospects of observing signatures of spinodal instabilities in experiments with heavy ions become more promising .
we are grateful to yu.b .
ivanov , b. friman and e.e .
kolomeitsev for fruitful discussions .
we are grateful to l. grigorenko for the reading of the manuscript and making numerous useful comments .
v. s. acknowledges the financial support by the frankfurt institute for advanced studies ( fias ) .
this work was supported in part by the dfg grants wa 431/8 - 1 and 436 rus 113/558/0 - 3 .
all thermodynamical quantities can be found from expression for the free energy @xmath266 = nf_0 \left\ { t\left [ \ln ( \lambda^{3}n)-1\right]- t\ln ( 1-bn)-an\right\},\ ] ] where @xmath267 is the thermal wave - length , @xmath268 is the baryon number .
parameter @xmath269 governs the strength of a mean field attraction and @xmath270 controls a short - range repulsion , @xmath271 is an extra constant , which can be fitted to obtain more realistic values of thermodynamic quantities for the specific phase transition ( e.g. , for ngl or hqgp ) . from eq .
( [ f ] ) we find the pressure in @xmath93 variables : @xmath272=f_0\left [ { nt}/{(1-bn)}-n^2 a \right],\ ] ] and the entropy per baryon @xmath273=f_0\left [ -\ln ( \lambda^3/b)+\ln ( 1/(bn ) -1)+5/2\right]+c_1.\ ] ] since entropy in classical systems is defined up to a constant , an arbitrary constant @xmath274 is added . from eqs .
( [ p ] ) and ( [ s ] ) we also find the energy @xmath275=nf_0
( 3b x^{2/3}-an),\ ] ] the pressure @xmath276=f_0 ( 2b x^{5/3}- an^2 ) , \ ] ] and the temperature @xmath277=2bx^{2/3},\ ] ] expressed in terms of @xmath94 variables .
here @xmath278 $ ] , @xmath279 .
in numerical calculations , which results are illustrated by figs .
1,2 , being performed just with a demonstration purpose we put @xmath280 and @xmath281 , as for the case of the standard van der waals equation of state . in this appendix
we generalize results obtained in sect .
[ instabilities ] to the relativistic case .
we use the landau approach @xcite .
it is well - known that equations of relativistic first - order hydrodynamics are parabolic and formally suffer of the causality problem .
however the causality problem may arise only , when one tries to describe the flow with a fast velocity and for short wavelength phenomena . to avoid this problem following ref .
@xcite we assume that the velocity of the collective motion is much slower than the mean particle velocity in the system ( the latter can be relativistic ) and we are interested in description of a long wavelength phenomena .
let the equilibrium state be the fluid rest frame . then @xmath282 .
as in non - relativistic case we linearize the hydrodynamical equations with respect to the variations @xmath234 , @xmath256 , @xmath242 and @xmath283 . with the help of the standard thermodynamic relations
the landau equations are reduced to @xcite : @xmath284 = 0 , \\ & & \partial_t \delta s - \frac{s}{n } \partial_t \delta n + \frac{\kappa}{w}\nabla^2\left ( \delta p - c n \nabla^2\delta n - \frac{w}{t } \delta t \right ) = 0 , \label{rel_hydro3}\end{aligned}\ ] ] where @xmath285 and @xmath46 are the enthalpy density and the entropy density , respectively . here
we additionally included surface tension effects . the pressure and the entropy density variations are expressed through the density and the temperature variations ( see e.g. eq . ( [ varp ] ) for the pressure variation ) . substituting eqs .
( [ delns ] ) in eqs .
( [ rel_hydro1]-[rel_hydro3 ] ) we find the dispersion law . in the long wavelength limit ( up to @xmath286-terms )
we obtain @xmath287\frac{p^2}{2}- \frac{c_{\tilde{s}}^2t}{w } ( 2\alpha-1)\kappa\frac{p^2}{2 } , \\ & & \gamma_3^{\rm rel } = - \frac{\kappa u_t^2}{u_{\tilde{s}}^2 c_v } p^2 \label{rel_disp}.\end{aligned}\ ] ] here @xmath288 , @xmath289 .
similar to the non - relativistic case , the solutions @xmath290 describe sound excitations in the long wavelength limit , while @xmath291 corresponds to the thermal mode .
the last term in expression for @xmath290 is a small correction @xmath292 for @xmath293 of our interest .
dropping it we arrive at eqs .
( [ solg1 ] ) , ( [ solg ] ) , which we have derived in the non - relativistic limit in the paper body .
thus we come to precisely the same conclusions about instability region , as in the non - relativistic case .
e. witten , phys .
d30 * , 272 ( 1984 ) ; w. greiner , h. stoecker and p. koch , in hirscheg proc .
1987 , p. 16 .
p. chomaz , m. colonna , and j. randrup , phys .
rep . * 389 * , 263 ( 2004 ) .
e. shuryak , arxiv:0807.3033 [ hep - ph ] .
g. bertsch and p. siemens , phys .
b126 , 9 ( 1983 ) . h. schulz , d. n. voskresensky , and j. bondorf , phys . lett . *
b133 * , 141 ( 1983 ) .
a. d. panagiotou , m. w. curtin , h. toki , d. k. scott , and p. j. siemens , phys .
lett . * 52 * , 496 ( 1984 ) ; j. a. lopez and p. j. siemens , nucl . phys . *
a431 * , 728 ( 1984 ) ; a. vicentini , g. jacucci and v. r. phandharipande , phys . rev . , * c31 * , 1783 ( 1986 ) .
m. dagostino _ et al .
_ , phys .
b * 473 * , 219 ( 2000 ) .
m. schmidt , r. kusche , t. hippler , j. donges , w. kronmuller , b. von issendorff and h. haberland , phys .
* 86 * , 1191 ( 2001 ) .
m. dagostino _ et al .
_ , nucl .
a * 749 * , 55 ( 2005 ) , [ arxiv : nucl - ex/9906004 ] .
i. c. arsene _ et al .
_ , phys .
c * 75 * , 034902 ( 2007 ) , [ arxiv : nucl - th/0609042 ] . g. papp and w. norenberg , heavy ion phys .
* 1 * , 241 ( 1995 ) . c. sasaki , b. friman , and k. redlich , phys
lett . * 99 * , 232301 ( 2007 ) ; phys . rev . *
d77 * , 034024 ( 2008 ) . c. pethick and d. g. ravenhall , nucl
a471 * , 19c .
j. a. lopez and g. lbeck , phys .
b219 * , 215 ( 1989 ) . v. l. ginzburg and a. a. sobyanin , sov .
* 19 * 773 ( 1976 ) .
y. hatta and t. ikeda , phys .
d * 67 * , 014028 ( 2003 ) , [ arxiv : hep - ph/0210284 ] . ya .
b. zeldovich and a. s. mikhailov , ufn ( sov . ) , * 187 * , 467 ( 1987 ) .
kadanoff and p.c .
martin , ann . phys . * 24 * , 419 ( 1963 ) .
l.p . kadanoff and g. baym , _ quantum statistical mechanics _ ( benjamin , 1962 ) ; yu.b .
ivanov , j. knoll and d.n .
voskresensky , nucl . phys . * a672*,313 ( 2000 ) .
d. n. voskresensky , phys .
scripta * 47 * , 333 ( 1993 ) .
e. m. lifshitz and l.p .
pitaevskii , physical kinetics , ( pergamon , oxford , 1981 ) .
l. d. landau and e. m. lifshitz , fluid mechanics , 2nd ed .
( pergamon , oxford , 1987 ) . | effects of thermal conductivity on the dynamics of first - order phase transitions are studied .
important consequences of a difference of the isothermal and adiabatic spinodal regions are discussed .
we demonstrate that in hydrodynamical calculations at non - zero thermal conductivity , @xmath0 , onset of the spinodal instability occurs , when the system trajectory crosses the isothermal spinodal line . only for @xmath1
it occurs at a cross of the adiabatic spinodal line .
therefore ideal hydrodynamics is not suited for an appropriate description of first order phase transitions . and |
low - dimensional ( one - dimensional ( 1d ) and quasi - one dimensional ) gapped spin systems such as spin - peierls , haldane and spin ladder systems have drawn attention from both theorists and experimentalists in the literature after the prediction by haldane @xcite that a 1d heisenberg antiferromagnet should have an energy gap between the singlet ground state and the first excited triplet states in the case of an integer spin quantum number @xmath0 , while the energy levels are gapless in the case of a half - odd integer values of @xmath0 .
the most fascinating characteristic of these systems is to show magnetic plateaus i.e. topological quantization of the magnetization at the ground state of the system due to magnetic excitations .
a general condition for the quantization of the magnetization in low - dimensional magnetic systems is derived from the lieb - schultz - mattis theorem @xcite .
oshikawa , yamanaka and affleck ( oya ) @xcite discussed this plateau problem and derived a condition @xmath1 , necessary for the appearance of the plateau in the magnetization curve of one - dimensional spin system , where @xmath0 is the magnitude of spin , @xmath2 is the magnetization per site and @xmath3 is the spatial period of the ground state , respectively .
this condition emphasize that the @xmath4 magnetization plateaus ( contained the saturated magnetization @xmath5 ) can appear when the magnetic field increases from zero to its saturation value @xmath6 , however , it does not directly prove its existence @xcite . in many theoretical studies , it has been found that low - dimensional gapped spin systems have magnetic plateaus at low temperatures or ground state as providing oya criterion .
for example , chen et al .
@xcite employed the classical monte carlo technique to investigate the magnetization plateaus of 1d classic spin-@xmath7 and spin-@xmath8 af ising chain with a single - ion anisotropy under the external field at low temperatures , and they showed that the systems have @xmath4 magnetic plateaus .
tonegawa et al .
@xcite observed the plateau at @xmath9 , and @xmath10 in the ground state of spin-@xmath7 af heisenberg chain with bond alternating and uniaxial single - ion anisotropy .
sato and kindo obtained a magnetic plateau at @xmath9 , and @xmath10 for spin-@xmath7 bond alternating heisenberg chain by using renormalization group technique @xcite .
hida observed that an @xmath11 antiferromagnetic chain with period-3 exchange coupling shows a plateau in the magnetization curve at magnetization per site @xmath12 @xcite .
totsuka found a plateau at @xmath13 for the familiar @xmath11 antiferromagnetic heisenberg chain with the next - nearest - neighbor and alternating nearest - neighbour interactions using the bosonization technique @xcite .
sakai and takahashi found that a magnetization plateau appears at @xmath14 for the @xmath15 antiferromagnetic heisenberg chain by the exact diagonalization of finite clusters and finite - size scaling analysis @xcite .
obtained half - saturated plateau of @xmath13 in the @xmath11 xxz chain in @xmath16 phase exposed to a weak period-4 field @xcite .
_ numerically obtained the magnetic plateau @xmath14 in the @xmath15 antiferromagnetic heisenberg chain with single ion anisotropy @xmath17 @xcite .
aydiner and akyz obtained magnetic plateaus for spin-1 using classical transfer matrix technique @xcite .
okamoto also found out the magnetization plateau in alternating spin chains @xcite . on the other hand , magnetic plateaus
have been predicted not only in theoretical study but also have been observed in experimental studies .
for example , narumi et al .
@xcite observed the magnetic plateaus in the magnetization curve for both [ ni@xmath18(medpt)@xmath18(@xmath19-ox)(h@xmath18o)@xmath18](clo@xmath21)@xmath182h@xmath18o and [ ni@xmath18(medpt)@xmath18(@xmath19-ox)(@xmath19-n@xmath22)](clo@xmath21)0.5 h@xmath18o .
goto et al .
@xcite reported the existence of the magnetization plateau at @xmath23 in spin-@xmath7 3,3,5,5-tetrakis ( n - tert - butylaminxyl ) biphenyl ( bip - teno ) . in the majority of plateau mechanisms which have been proposed up to now the purely quantum phenomena play crucial role .
the concept of magnetic quasi - particles and strong quantum fluctuation are regarded to be first importance of these processes .
particularly , for a number of systems it was shown that the plateau at @xmath24 are caused by the presence of the spin gap in the spectrum of magnetic excitation in the external magnetic field @xcite .
another mechanism lies in so - called crystallization of the magnetic particles @xcite . meanwhile , even though the phenomenon of magnetization plateaus are often regarded to have purely quantum origin , it is obviously not clear since one - dimensional @xcite and two dimensional @xcite quasi - classical spin models which exhibit a magnetization plateaus at zero temperature or near the ground state .
in fact , these studies show to us that heisenberg and ising spins leads to the qualitatively same structures of the magnetization profiles , particularly to the formation of magnetization plateaus . according to our knowledge , magnetic plateaus
can appear depend on dimerization , frustration , single - ion anisotropy or periodic field for one - dimensional af spin chain independent of whether the classical or not quantum models .
therefore , we say that one of these ingredients is required for a plateau mechanism . in this study , we shall investigate the one - dimensional antiferromagnetic spin-3/2 and spin-2 ising chains with single - ion anisotropy to obtain magnetic plateaus and other thermodynamical quantities taking into account the analogy of the magnetization profiles between heisenberg and ising spin systems .
the approach is based on very simple idea to use the ising spin instead of the heisenberg operators and to develop the classical transfer matrix technique .
it is worth to emphasize that in contrast to the majority of existing approaches to the problem of magnetization plateau , this technique is entirely based on analytical calculations and allows to obtain the magnetization profiles for arbitrary finite temperature and arbitrary values of @xmath25 , @xmath26 and @xmath27 . on the other hand
, we must remark that the transfer matrix approach to the 1d classical system is rather simple as compared with other techniques , however , the implementation of this technique leads to remarkable simplicity and clarity and produces considerably exact results for 1d infinite spin system .
the organization of this paper is as follows . in the next section ,
we present classical hamiltonian for spin-@xmath0 system with single - ion anisotropy and briefly introduce transfer matrix technique . in section 3 and 4 we give numerical results for spin-3/2 and spin-2 , respectively .
the last section gives our conclusions .
in point of classical view , hamiltonian for the 1d spin-@xmath0 system with single - ion anisotropy can be written as follows : @xmath28 where @xmath25 denotes the exchange coupling of antiferromagnetic type ( i.e @xmath29 ) , @xmath26 is the single - ion anisotropy , @xmath27 is the external field , and _
n _ is the total number of site in the system .
herein , in order to investigate the magnetic and thermal properties of 1d af spin-@xmath8 and spin-@xmath30 ising chain , we are interested in the three thermodynamical expressions ( i ) ferromagnetic order @xmath2 , ( ii ) the specific heat @xmath31 , and ( iii ) the magnetic susceptibility @xmath32 .
these quantities are calculated in terms of free energy as @xmath33 @xmath34 @xmath35 then the free energy for the system can be found exactly in terms of the partition function with the help of the transfer matrix technique .
hence , all thermodynamic functions in eq . ( 2 ) are respectively calculated .
the free energy per spin of this system is given by @xmath36 with @xmath37 where @xmath38 represents transfer matrix of the spin system , which we give below .
it is known that a conventional way to calculate the free energy is to write the partition function ( 4 ) in terms of eigenvalues of the transfer matrix @xmath38 as follows : @xmath39 according to the standard assumptions of this technique , partition function ( 5 ) is represented as , @xmath40,\ ] ] where @xmath41 , @xmath42 , @xmath43 , ... which are sorted from the biggest to smallest .
it is clearly seen that the second and the third terms on the rhs of eq .
( 6 ) go to zero as @xmath44 due to @xmath45 , @xmath46 , and so on .
hence the free energy eq .
( 3 ) is practically reduced to @xmath47 this result implied that the magnetic and thermodynamical quantities in eq .
( 2 ) for the present system can be evaluated depend on the biggest eigenvalue @xmath48 . for spin-3/2
, @xmath49 takes on @xmath50 , and @xmath51 values , on the other hand , for spin-2 , @xmath49 takes on @xmath52 , @xmath53 , and @xmath54 values .
considering the fact that , for example , we can write the transfer matrix @xmath38 for spin-3/2 as @xmath55 where @xmath56 , @xmath57 , @xmath58 , and @xmath59 where @xmath60 is the boltzmann constant and @xmath61 is the temperature . the transfer matrix @xmath38 for spin-2 can be constructed in a similar way .
the numerical results for spin-3/2 are as follow : in fig .
1(a ) and ( b ) , the magnetization @xmath2 were respectively plotted as a function of external field @xmath27 for fixed values of @xmath62 , and @xmath63 at @xmath64 ( unit by @xmath25 ) .
it is seen from fig .
1(a ) and ( b ) that @xmath65 plateaus appear , located at @xmath66 , and @xmath63 , which satisfies the oya s criterion for all values of positive single - ion anisotropy @xmath26 at the low temperature near the ground state under the external field @xmath27 .
appearance of the magnetic plateaus indicate that the 1d af spin-3/2 ising chain has a gap mechanism owing to single ion - anisotropy .
previous theoretical studies @xcite show that single ion - anisotropy play a crucial role to appear magnetic plateaus for classical ising chains .
our result also confirms previous theoretical predictions @xcite .
the transitions between magnetic plateaus occur at critical field values .
the nonzero magnetization at @xmath67 begins from a finite external field which called as initial critical field @xmath68 .
second plateau at @xmath69 appears at second critical field value @xmath70 .
the last plateau corresponds to the value of saturation magnetization of the system , that we called it as saturated field @xmath71 . on the other hand ,
we have no observed a magnetization plateau for @xmath72 .
this observation indicate that single ion - anisotropy is one of the necessary ingredient to plateau mechanism for the present system .
in fact , @xmath26 values change width of plateaus as seen in fig .
1(a ) and ( b ) .
in order to see this effect , magnetic phase diagram @xmath73 was plotted in fig .
this phase diagram displays the behavior of the system depend on single ion - anisotropy near the ground temperature i.e. , @xmath64 .
however , this phase picture does not clearly show the effect of exchange interaction @xmath25 .
three lines in fig .
2(a ) correspond to critical field values which are responsible of plateau transition .
four plateau areas i.e. , @xmath74 , @xmath75 , @xmath10 , and @xmath63 are divided by the initial critical field @xmath68 , second critical field @xmath70 and the saturated field @xmath71 lines for values of @xmath76 as shown in fig.2(a ) .
areas between these lines give information how width of the plateaus can change depend on single ion - anisotropy .
the longitudinal coordinate of the square - dot is the beginning point of the field for appearance of the plateau @xmath67 which corresponds to the first critical field @xmath68 values , the plateau @xmath69 is denoted by the circular - dot which corresponds to the second critical field @xmath70 values , and its ending point is signed by the triangle - dot which correspond to the saturated field @xmath6 values .
it is clearly seen that the initial field line and second critical field line have a critical behavior at the @xmath77 while the saturated field increasing monotonously with the rise of @xmath26 .
the first critical field decreases in the regime @xmath78 with the increase of @xmath26 , however , it lies longitudinal for @xmath79 . on the other hand , the second critical field increasing monotonously with the rise of @xmath26 for @xmath79 , while it lies longitudinal for @xmath80 . as seen in fig .
2(a ) , value of @xmath77 behaves as a critical point for both first and second critical field values .
reason of this behavior is chosen as @xmath81 .
the magnetization phase diagram in fig.2(a ) is compatible with the monte carlo prediction @xcite .
the width of the plateau @xmath67 increases with @xmath26 increasing for the present system , whereas it remains in heisenberg chain with single - ion anisotropy @xmath82 .
furthermore , the width of the plateaus @xmath74 and @xmath69 are independent of the single - ion anisotropy @xmath26 , which is similar to that of @xmath67 in one - dimensional @xmath83 antiferromagnetic ising chain with single - ion anisotropy @xcite .
though the magnetic phase diagram in fig .
2(a ) of the hamiltonian ( 1 ) gives information about depending single ion - anisotropy , however , it does not display how the spin exchange strength @xmath25 contribute to plateau mechanism . in order to understand effect of @xmath25 to the plateau ,
we have studied in the limit of @xmath84 . for the simplicity we have chosen an extreme values of @xmath85 and @xmath86 and magnetization
were plotted versus external field for these values in fig.2(b ) .
it is seen that the only intermediate plateau @xmath67 expands to depress the range of the other plateaus @xmath74 and @xmath69 in this limit .
this result also indicate that spin exchange parameter @xmath25 plays a significant role on the plateau mechanism as well as single ion - anisotropy @xmath26 . at the extreme case of @xmath85
, we can easily understand that it is equal to single - particle system with two plateaus @xmath67 and @xmath63 in the ground magnetization . in order to obtain the behavior of the specific heat @xmath31 depending on the external field @xmath27 and the single - ion anisotropy @xmath26 at low temperatures , they are plotted as a function of temperature @xmath61 , for the fixed value of the single - ion anisotropy , @xmath87 , at various values of the external field @xmath88 , and @xmath89 in fig.3(a ) and for the fixed value of @xmath90 at various values of @xmath91 , and @xmath92 in fig.3(b ) , respectively .
the specific heat curves in fig.3(a ) and fig.3(b ) have a relatively board maximum like the schottky peak near @xmath93 and @xmath94 respectively . as can be seen from fig.3(a ) , the height of the specific heat peak for the fixed value of @xmath26 decreases and moves towards zero temperature when the external field @xmath27 increases . on the other hand ,
the height of the specific heat peak for the fixed value of @xmath27 decreases and moves towards zero temperature when the value of single - ion anisotropy increases in fig.3(b ) .
we note that the schottky - like round hump in the specific heat probably reflects the fact that the energy of the system depends on the external field @xmath27 , and the single - ion anisotropy @xmath26 .
however , they do not indicate that a second - order phase transition occurs in the one - dimensional system .
all of them may be thought as a schottky - like peak resulting from the antiferromagnetic short - range order .
another way of the confirmation of multi - plateau magnetization curves by numerical calculation in the 1d af ising chain is to examine the field and temperature dependence of susceptibility @xmath32 in fig.4(a ) and ( b ) , respectively .
it is expected that the temperature dependence of the susceptibility gives information that whether the low - dimensional system has an energy gap or not , while field dependence of it verify formation of plateaus .
therefore , the susceptibility is plotted in fig.4(a ) as a function of external field @xmath27 .
it can be seen that there are three peaks in the susceptibility for @xmath87 and @xmath64 .
the first peak occurs at the initial critical field value , the second peak occurs at the second critical field value and the third peak occurs at the saturated field value for the fixed value of @xmath26 .
each peak indicates the transitions from one magnetic plateau to another one .
furthermore , we have also plotted the magnetic susceptibility as a function of temperature for fixed value of @xmath87 and for three selected external fields @xmath95 , and @xmath96 in fig.4(b ) .
the curve for @xmath97 marked by triangular - dot line show a relatively sharp peak at low temperature , while the other curves for @xmath98 , and @xmath96 have a broad maximum which are denoted by circular - dot and square - dot line , respectively .
we roughly say that the susceptibility curves decay exponentially with the increasing temperature by the relation @xmath99 .
the peaks or round hump behavior in the magnetic susceptibility curves in fig.4(b ) may be interpreted as evidence the formation of the short - range correlations within the chains . in that case
, these correlations may play a role for gap mechanism just as in heisenberg af systems .
this characteristic behavior in the susceptibility clearly compatible with the experimental results for spin-@xmath8 compound cs@[email protected]@xmath18o @xcite .
we start our discussion once again with investigating the properties of one - dimensional af spin-@xmath30 ising chain with single - ion anisotropy .
now , we consider the above hamiltonian ( 1 ) for the case of spin-@xmath30 . in the case of spin-3/2 the physical quantities such as the magnetization ,
the magnetization phase diagram , the specific heat , and the magnetic susceptibility were evaluated numerically to keep in this system of spin-2 .
we have plotted the magnetization @xmath2 as a function of external field @xmath27 for only fixed values of @xmath86 and @xmath64 ( unit by @xmath25 ) as shown in fig .
the number of the plateaus obey to oya criterion for spin-2 chain in the case of the spin-@xmath8 chain .
the plateaus appear at @xmath74 , @xmath75 , @xmath10 , @xmath63 , and @xmath96 for all values of positive single - ion anisotropy @xmath26 .
it is seen from fig .
5(a ) that there are three critical fields as @xmath68 , @xmath70 and @xmath101 in the spin-@xmath30 chain . on the other hand ,
we have plotted the magnetization @xmath2 as a function of the external field @xmath27 for only fixed values of @xmath86 and for the extreme case of @xmath85 in fig .
5(b ) to see behavior of the present system in the limit of @xmath84 . in this limit
, the spin-2 ising chain has the three plateaus at @xmath74 , @xmath10 , and @xmath96 . at the extreme case of @xmath85
, one can easily suggest that it is equal to two - particle system with three plateaus at @xmath74 , @xmath10 , and @xmath96 in the ground magnetization . the single ion - anisotropy dependence of the magnetization plateaus of 1d af spin-2 ising chain near the ground state were presented in @xmath73 space in fig .
the characteristic behavior of the phase diagram of the spin-@xmath30 chain is similar to that of the spin-@xmath8 chain .
the saturated field increasing monotonously with the rise of @xmath26 , however , all critical field values i.e. , @xmath68 , @xmath70 , and @xmath101 have a critical behavior at the @xmath77 due to choosing of @xmath102 .
the first and second critical fields decrease in the regime @xmath78 with increase of @xmath26 , but they increasing monotonously with the rise of @xmath26 .
the third critical field increase slowly with rise of @xmath80 , however , it increase monotonously with the rise of @xmath26 . on the other hand
, the width of the plateau @xmath69 increases with @xmath26 increasing for the present system , however , the width of the plateaus @xmath67 and @xmath69 remain constant for values of @xmath79 , while they increase with @xmath26 increasing up to @xmath77 as seen from fig .
the magnetization phase diagram of the ground state of af spin-@xmath30 ising chain under finite magnetic field .
the square - dot line , the circular - dot line , the upper triangle - dot line , and the lower triangle - dot line represent the initial critical field , the second critical field , the third critical field , and the saturated field , respectively.,width=283,height=245 ] the temperature dependence of the specific heat for different field values and fixed value of @xmath26 is shown in fig.7(a ) and for different single ion - anisotropy values and fixed field value of @xmath27 is shown in fig.7(b ) , respectively .
one can suggested that the physical behavior of specific heat of spin-2 chain has the same characteristic with spin-3/2 .
in fact , as can be seen from fig.7(a ) , the height of the specific heat peak for the fixed value of @xmath26 decreases and moves towards zero temperature when the external field @xmath27 increases . on the other hand ,
the height of the specific heat peak for the fixed value of @xmath27 decreases and moves towards zero temperature when the value of single - ion anisotropy increases in fig.7(b ) .
finally , we have plotted the magnetic susceptibility versus external field for fixed @xmath26 in fig.8(a ) and versus the temperature for different values of @xmath27 and fixed value of @xmath26 in fig.8(b ) , respectively .
the peaks in fig.8(a ) indicate the critical field values transition from one plateau to another one for the spin-2 chain . on the other hand ,
the behavior of the susceptibility curves in fig.8(b ) clearly provide that 1d af spin-2 ising chain has an energy gap which leads to the magnetic plateaus .
in this study , using by quasi - classical hamiltonian(1 ) , we have evaluated the magnetic and thermal properties of 1d af spin-@xmath8 and spin-@xmath30 ising chain with a single - ion anisotropy in the presence of an external magnetic field at low temperature in the framework of the transfer matrix formalism .
the most important result stemming from this study is the confirmation of a multi step magnetization process by a numerical calculation for both in the case of a half - odd integer , and an integer spin quantum @xmath0 .
indeed , we found out that the @xmath4 step - like magnetization plateaus of the system appear for all values of positive single - ion anisotropy @xmath103 in both spin system , when the external field varied from zero to the saturated field .
in addition , the magnetic plateau does not appear for @xmath72 in these spin systems , however , at extreme value of @xmath85 we observed that the spin-@xmath8 ising chain has two plateaus at @xmath67 and @xmath63 and then spin-@xmath30 ising chain has three plateaus at @xmath74 , @xmath10 , and @xmath96 .
these evidences confirm that single ion - anisotropy and the nearest - neighbor interaction contribute to be formed the magnetic plateaus .
therefore , plateau forms in the 1d af spin-@xmath0 ising chains can be explain owing to the co - existence of antiferromagnetic interaction of @xmath104 and positive single - ion anisotropy @xmath105 .
the plateau predictions in this study are compatible with previous theoretical studies @xcite .
therefore , the present results provide that there is a class of one - dimensional spin systems which in terms of heisenberg and ising spins lead to the qualitatively same structures of the magnetization profiles , particularly to the formation of magnetization plateaus . hence , we suggest that the quasi - classical approach can be used to examine the magnetic and thermal behaviors of 1d af spin-@xmath0 ising chains as well as the heisenberg hamiltonian
. unfortunately , we have not compare our thermodynamical results with the experimental results since there are no enough experimental studies for these systems in the literature .
however , it is understand from present theoretical evidences that the magnetic and thermal characteristic of the 1d af spin-@xmath0 ising chains are the same .
therefore , we expected that our results will be agreement with the experimental results of the ideal spin-@xmath8 and spin-@xmath30 compounds .
finally , we stated that one does not expect any spin - peierls transition in these models , because there is no physical mechanism for it within a model constructed purely from @xmath106 operator , without any coupling to the lattice .
a spin - peierls transition may occurs only due to a small amount of magnetic impurities and derivations from stoichiometry of the compound synthesized experimentally @xcite . | in this study , we have employed the classical transfer matrix technique to investigate the magnetization plateaus , phase diagrams and other thermodynamical properties of the one - dimensional antiferromagnetic spin-3/2 and spin-2 ising chains with single - ion anisotropy in the presence of an external magnetic field at very low temperature .
we have showed that single ion - anisotropy is one of the indispensable ingredients for an energy gap which leads to magnetic plateau mechanism in one - dimensional antiferromagnetic ising spin chains .
other thermodynamical predictions seem to be provide this argument .
* keywords : * antiferromagnetic ising chain ; classical transfer matrix technique ; magnetic plateau ; single - ion anisotropy . |
most theoretical protostellar disk models have relied on the @xmath0parametrization of the anomalous turbulent viscosity proposed by shakura & sunyaev ( 1973 ) . in this context , the kinematic turbulent viscosity @xmath1 is written @xmath2 , where @xmath3 is the sound speed , @xmath4 is the disk scale height and @xmath0 is a dimensional parameter ( @xmath5 ) .
so far , only mhd instabilities ( balbus & hawley 1991 ) have been shown to be able to produce and sustain turbulence in accretion disks . however , because they develop only in adequately ionized fluid , the parameter @xmath0 is probably not constant through protostellar disks , and it may even be that only parts of these disks can be described using this @xmath0 prescription ( balbus & hawley 1998 and this volume ) .
however , for the purpose of considering planet formation , as we are interested in here , we will use such models .
since they have already been described in previous papers , we will not present them in detail here again .
instead we refer the reader to papaloizou & terquem ( 1999 ) for references and a detailed description .
below we shall make use of the results of some of the numerical calculations they performed .
+ _ vertical structure : _ we consider thin disks which are in keplerian rotation around a star of mass @xmath6 m@xmath7 .
the opacity in the disk , taken from bell & lin ( 1994 ) , has contributions from dust grains , molecules , atoms and ions
. we are free to choose two parameters to construct @xmath0disk models .
we take @xmath0 and the local mass flow rate through the disk which is defined as @xmath8 , where @xmath9 is the vertically averaged kinematic viscosity and @xmath10 is the surface mass density of gas .
if the disk were in a steady state , @xmath11 would not vary with radius and would be the constant accretion rate through the disk .
we note that @xmath11 is related to @xmath0 through @xmath1 , but it also depends on the distribution of mass in the disk . at a given radius @xmath12 and for given values of @xmath11 and @xmath0
, we solve the differential equations describing the disk vertical structure ( equations of vertical hydrostatic equilibrium , energy conservation and radiative transport ) with appropriate boundary conditions to find the dependence of the temperature , pressure , mass density and radiative flux on the vertical coordinate . an important point to note
is that as well as finding the disk structure , we also determine @xmath10 for a given @xmath11 , so that a relation between @xmath13 and @xmath10 is derived . in order to investigate the solutions of the diffusion equation which governs the disk evolution ( see below ) and for other purposes , we found it convenient to derive analytic piece wise power law fits to this relation .
details of these fits , which can be used for a very wide range of disk parameters , are given in papaloizou & terquem ( 1999 ) .
+ _ time dependent evolution and quasi
steady states : _ in general , an accretion disk is not in a steady state but evolves diffusely according to the following equation ( see lynden bell & pringle 1974 , papaloizou & lin 1995 and references therein ) : @xmath14 \label{aa4 } .\ ] ] from this we see that the characteristic diffusion timescale at radius @xmath12 is @xmath15 , where @xmath4 is the disk semithickness . for disks with approximately constant aspect ratio @xmath16
as applies to the models considered here , @xmath17 scales as the local orbital period .
one thus expects that the inner regions relax relatively quickly to a quasi steady state which adjusts its accretion rate according to the more slowly evolving outer parts ( see lynden bell & pringle 1974 and lin & papaloizou 1985 ) . for estimated sizes of protostellar disks of about 50 au ( beckwith & sargent 1996 )
, the evolutionary timescale associated with the outer parts is about 30 times longer than that associated with the inner parts with @xmath18 au , which we consider here in the context of planetary formation .
thus these inner regions are expected to be in a quasi steady state through most of the disk lifetime . + _ steady state models : _ in the models we present here , we assume that the disk is immersed in a medium with background temperature @xmath19 k. in figures [ fig1]a b we plot @xmath20 ( where @xmath4 is defined as the vertical height at approximately unit optical depth above which the atmosphere is isothermal ) and @xmath10 versus @xmath12 for @xmath11 between @xmath21 and @xmath22 m@xmath7 yr@xmath23 ( assuming this quantity is the same at all radii , i.e. the disk is in a steady state ) and for illustrative purposes we have adopted @xmath24 . in figures [ fig2]a
b we plot the midplane temperature @xmath25 versus @xmath12 and @xmath25 versus the midplane pressure @xmath26 for the same parameters .
the surface mass density of planetesimals can be derived from @xmath10 by noting that in protostellar disks the gas to dust mass ration is about 100 . in these models ,
reprocessing of the stellar radiation by the disk has not been included .
figure [ fig1]a indicates that the outer parts of the disk beyond some critical radius ( @xmath27 to a few au , depending on @xmath0 and @xmath11 ) are shielded from the radiation of the central star by the inner parts ( the very outer parts may not be shielded but since they are optically thin they do not absorb the stellar radiation anyway ) . since reprocessing of the stellar radiation by the disk is not an important heating factor below these radii
, this indicates that this process can be self - consistently ignored .
however , it is possible that a multiplicity of solutions exists when reprocessing is taken into account ( chiang & goldreich 1997 ) , with it being important for cases in which the disk is flared , as may be the case with hh30 ( burrows et al .
1996 ) , and unimportant when it is not , as may be the case with hk tau ( stapelfeldt et al .
1998 , koresko 1998 ) . in any case ,
reprocessing , when present , does not affect significantly the structure of the disk at @xmath285 au , where planets are supposed to form .
below we shall consider the migration of protoplanetary cores from @xmath295 au , where they are supposed to form under conditions where ice exists , down to the disk inner radii .
it is therefore of interest to estimate the mass of planetesimals @xmath30 contained inside the orbit of a core located at @xmath12 , since this can potentially be accreted by the core during its migration .
we assume a gas to dust mass ratio of 100 and list the values of @xmath31 corresponding to @xmath24 and @xmath32 and @xmath33 and @xmath34 m@xmath7 yr@xmath23 ( characteristic of the early stages of disk evolution ) in table [ tab1 ] .
.estimate for the mass of planetesimals @xmath30 contained within a radius @xmath12 for different values of @xmath0 and gas accretion rate @xmath35 through the disk .
+ [ cols="^,^,^,^ " , ] [ tab2 ] the previous section indicates that migration is relatively fast . in the context of the disk models presented above ,
a core which builds up at a location between 1 and 5 au will migrate inwards once it reaches a mass of about 0.1 m@xmath36 on a timescale shorter than the time it takes to form a giant planet , according to current theories ( pollack et al .
1994 ) . therefore , unless there is a mechanism to stop the migration of this core , it will never become a giant planet but will plunge onto the star .
i migration could be avoided if the interaction could become nonlinear and a gap open up around the core .
this would require small values of @xmath20 and @xmath0 ( see eq .
[ [ crit1 ] ] and [ [ crit2 ] ] ) and then type ii migration would be very slow .
if a gap opens up around a core before it has accreted a gaseous envelope , further accretion may take place ( e.g. , kley 1999 ; bryden et al .
1999 ) so that it is conceivable it could become a giant planet .
however , observations show that in some cases migration has to take place to push the protoplanet down to intermediate radii ( @xmath37 au ) , and has to stop there .
if the viscosity were controlling the migration rate , unless there is a fortuitous disk dispersal , this would require @xmath0 to decrease as @xmath12 decreases , in contrast to what is proposed in current disk models ( see balbus & hawley in this volume ) . in the context of the disk models presented in
[ sec : disks ] , only a low surface mass density of gas would give long enough migration timescales to enable planet formation to take place before protoplanets with masses @xmath38 fall onto their parent star .
in this context , ward ( 1997a ) has suggested that several small mass ( @xmath39 m@xmath36 ) cores could be built up and maintained in isolation until the disk is partially depleted , at which point they would assemble into a larger core .
however , it seems that such cores would interact with each other on a rather short timescale ( chambers , wetherill & boss 1996 ) .
also , even if the core could assemble only in the late stages of the disk evolution , there may no longer be enough gas left in the disk at this point to form a giant planet , let alone a system of three giant planets , as observed around ups and ( butler et al .
1999 ) .
therefore either giant planet formation occurs faster than currently thought , on a timescale shorter than @xmath40 , or type
i migration can be halted somehow .
we note that migration would cease if the disk were terminated by a magnetospheric cavity and the core were sufficiently far inside it ( lin et al .
however , gas accretion is likely to be very much reduced in this case .
the core migration should be halted before the disk is terminated for a giant planet to be formed .
as noted above , ward ( 1986 , 1997b ) finds the direction of type
i migration could reverse from inwards to outwards if the disk midplane temperature decreased inwards fast enough .
such a condition would not be expected in the disk models considered here . however ,
if the inner disk is terminated through interaction with a stellar magnetic field , physical conditions may start to differ in the interaction zone where magnetic field lines penetrate the disk .
additional energy and angular momentum transport mechanisms due to a wind for example may start to become important .
as a result , an inward midplane temperature decrease might be produced .
it may then be possible that migration could be halted such that the protoplanet retains contact with disk gas .
in addition , work in progress ( the results of which will be published elsewhere ) shows that type
i migration is affected by the presence of a magnetic field in the disk .
as described above , the torque exerted by the protoplanet on the disk depends mainly on the location of the radii where the perturbation is in resonance with the free oscillations of the disk . when a magnetic field is present
, the perturbation can resonate not only with the epicyclic motions of the fluid but also with the alfven and magnetoacoustic waves which propagate through the fluid .
therefore there are additional locations in the disk where a torque can be exerted .
this additional contribution to the torque has to be considered .
even if it were not important , the results derived by ward ( 1986 , 1997b ) and described above would be affected . whether the outer lindblad resonances are closer or not to the protoplanet than the inner lindblad resonances indeed depends on the gradient of the lorentz force in the vicinity of the protoplanet . also
, preliminary estimates of the effect of a finite eccentricity @xmath41 of the planet s orbit on type
i migration indicate that it may be reversed for reasonable disk models once @xmath42 .
details will be published elsewhere .
giant planets are believed to form out of protostellar disks by either gravitational instability ( kuiper 1951 ; cameron 1978 ; boss 1998 ) or by a process of growth through planetesimal accumulation followed by gas accretion ( safronov 1969 ; wetherill & stewart 1989 ; perri & cameron 1974 ; mizuno 1980 ; bodenheimer & pollack 1986 ) .
the first mechanism is expected to produce preferentially massive objects in the outer parts of the disk , if anything . here
we study planetary formation within the context of the second mechanism , which is commonly accepted as the most likely process by which planets form in at least the inner ten astronomical units of protostellar disks .
we note however that important issues related to this model still remain to be resolved ( see lissauer 1993 for a review ) .
the build up of the atmosphere of giant planets has first been considered in the context of the so called core instability model by perri & cameron ( 1974 ) and mizuno ( 1980 ; see also stevenson 1982 and wuchterl 1995 ) . in this model ,
the solid core grows in mass along with the atmosphere in quasi
static and thermal equilibrium until the core reaches the so called critical core mass @xmath43 above which no equilibrium solution can be found for the atmosphere .
as long as the core mass @xmath44 is smaller than @xmath43 , the energy radiated from the envelope into the surrounding nebula is compensated for by the gravitational energy which the planetesimals entering the atmosphere release when they collide with the surface of the core . during this phase of the evolution ,
both the core and the atmosphere grow in mass relatively slowly . by the time @xmath44
reaches @xmath43 , the atmosphere has grown massive enough so that its energy losses can no longer be compensated for by the accretion of planetesimals alone . at that point , the envelope has to contract gravitationally to supply more energy .
this is a runaway process , leading to the very rapid accretion of gas onto the protoplanet and to the formation of giant planets such as jupiter .
further time
dependent numerical calculations of protoplanetary evolution by bodenheimer & pollack ( 1986 ) support this model , although they show that the core mass beyond which runaway gas accretion occurs , which is referred to as the crossover mass ( because it is reached when the mass of the atmosphere is comparable to that of the core ) , is slightly larger than @xmath43 .
the similarity between the critical and crossover masses is due to the fact that , although there is some liberation of gravitational energy as the atmosphere grows in mass together with the core , the effect is small as long as the atmospheric mass is small compared to that of the core .
consequently , the hydrostatic and thermal equilibrium approximation for the atmosphere is a good one for core masses smaller than @xmath43 .
therefore we use this approximation here and investigate how @xmath43 varies with location and physical conditions in the protoplanetary disk . to calculate the structure of an atmosphere in hydrostatic and thermal equilibrium around a core of mass @xmath44 , we need to solve the equations of hydrostatic equilibrium , mass conservation and radiative transport ( mizuno 1980 ) . both radiative and convective transport
are taken into account here , and it is assumed that the only energy source comes from the accretion of planetesimals onto the core .
the atmosphere is confined between the radius of the core and that of the protoplanet s roche lobe .
details of the calculations presented here are given in papaloizou & terquem ( 1999 ) , who used the equation of state given by chabrier et al .
( 1992 ) and the same opacity law as in [ sec : disks ] . in figure [ fig3 ] ,
we plot curves of total protoplanet mass @xmath45 against @xmath44 at different radii in a disk with @xmath46 and @xmath47 m@xmath7 yr@xmath23 ( using the models described in [ sec : disks ] ) . in each frame , the different curves correspond to planetesimal accretion rates @xmath48 in the range @xmath49@xmath22 m@xmath36 yr@xmath23 .
the critical core mass @xmath43 , above which the equations have no solution , is attained at the point where the curves start to loop backwards . for masses below @xmath43 , there are ( at least ) two solutions , corresponding to a low mass and a high mass envelope , respectively . when the core first begins to gravitationally bind some gas ,
the protoplanet is on the left on the lower branch of these curves .
assuming @xmath48 to be constant , as the core and the atmosphere grow in mass , the protoplanet moves along the lower branch up to the right , until the core reaches @xmath43 .
at that point the atmosphere begins to undergo very rapid contraction . since the atmosphere in complete equilibrium
is supported by the energy released by the planetesimals accreted onto the protoplanet , @xmath43 decreases as @xmath48 is reduced . for @xmath24 and @xmath50 m@xmath7 yr@xmath23 , @xmath43 at 5 au varies between 16.2 and 1 m@xmath36 as @xmath48 varies between the largest and smallest value .
the former result is in good agreement with that of bodenheimer & pollack ( 1986 ) .
note that there is a tendency for @xmath43 to increase as @xmath12 decreases , the effect being most marked at small radii . at 0.05au ,
@xmath43 varies from 42 to 9 m@xmath36 as @xmath48 varies between the largest and smallest value .
similar calculations for different disk models ( see papaloizou & terquem 1999 ) indicate a relatively weak dependence of @xmath43 on disk conditions except when rather high midplane temperatures @xmath51 k are attained , as in the inner regions .
these results are consistent with the fact that @xmath43 depends on the boundary conditions only when a significant part of the envelope is convective ( wuchterl 1993 ) , being larger for larger convective envelopes ( perri & cameron 1974 ) .
we note that it is unlikely there are planetesimals at radii @xmath52 au . therefore , although critical core masses for the same planetesimal accretion rates may be higher there , a lack of planetesimals may result in a fall in the core luminosity , making @xmath43 relatively small at these radii . as noted in
[ sec : migration ] , cores of several earth masses form at about 5 au and migrate inwards to small radii before they can become a giant planet . in doing so , they continue to grow . to know whether they can reach the critical mass on their way , and thus begin to accrete significant amount of gas , we have to evaluate @xmath48 , i.e. the mass of planetesimals they accrete as they migrate in .
papaloizou & terquem ( 1999 ) performed analytical estimates and numerical calculations to evaluate the fraction of planetesimals accreted by a core of mass @xmath5310 m@xmath36 migrating inward from @xmath29 au on a timescale @xmath54@xmath55 yr .
they found that the migrating protoplanet accretes about 25% of the planetesimals inside its orbit , as if it were in a homogeneous medium without a gap forming in the planetesimal distribution .
even more planetesimals are accreted when the migration is slower .
given that the disk models typically contain at least about 8 m@xmath36 interior to 5 au ( see [ sec : disks ] ) , this means that the accretion of planetesimals is high enough to maintain the core luminosity such that attainment of @xmath43 ( and therefore gas accretion ) does not occur as long as planetesimals are present , i.e. above @xmath560.1 au .
runaway gas accretion onto a small core can then occur at these radii .
however , if the core is too small , the gas accretion phase may be longer than the disk lifetime . even for core masses in the range
1520 m@xmath36 , the build up of a massive atmosphere may take a time @xmath57 yr ( bodenheimer et al . 1998 ) .
the reason for this is that once the core starts to accrete a significant atmosphere , energy production occurs through its gravitational contraction .
the luminosity produced then slows down the evolution .
an estimate of the evolutionary ( kelvin helmholtz ) timescale @xmath58 at this stage has been calculated by papaloizou & terquem ( 1999 ) .
typically , they found @xmath59 yr for core masses in the range 1020 m@xmath36 for radii larger than 0.075 au .
the core masses required to get such a characteristic timescale increase rapidly interior to 0.06 au .
however , we note that they decrease as the mass transfer rate through the disk does .
the fact that fairly large core masses are required to give evolutionary timescales comparable or less than the expected disk lifetime means that mergers of additional incoming cores may be required in order to produce a core of sufficient mass that real runaway gas accretion may begin .
table [ tab1 ] indicates that , in a disk with @xmath60 and @xmath61 or @xmath34 m@xmath7 yr@xmath23 , 40 m@xmath36 of planetesimals are contained within 1 or 5 au .
therefore , the timescale for building up a core with a mass between 20 and 40 m@xmath36 at small radii is typically the timescale it takes for planetesimals to migrate from 1 or 5 au down to these small radii . according to ward ( 1997b ; and see [ sec : migration ] ) , the migration timescale of cores of @xmath62 m@xmath36 located at 1 or 5 au is at most @xmath63 yr in such a disk , and it decreases with increasing core mass . if the disk has @xmath24 , 40 m@xmath36 of planetesimals are contained within 5 or 11 au . in this case again , the migration timescale of cores of @xmath62 m@xmath36 at 5 or 11 au is about @xmath63 yr . therefore ,
if planetesimals can be assembled into cores of at least a few tenths of an earth mass at these radii on a reasonably short timescale , a massive core could be obtained at small radii ( if at least three cores are present , they can not stay in isolation but interact with each other , as shown by chambers , wetherill & boss 1996 ) .
a massive core can be built up through the merger of additional incoming cores either after having stopped at small radii or on its way down to small radii ( where it would still be expected to be stopped ) .
the former process resembles that discussed by ward ( 1997a ) .
the latter scenario would occur if more massive cores , which migrate faster , overtake less massive cores on their way down . in any case , several cores would necessarily interact with each other at some point so that core isolation , which would increase planet formation very much if it occurred ( pollack et al . 1994 ) , would be avoided .
supposing that a protoplanetary core massive enough can be built up on its way down to small radii and that it continues to rapidly move inward until it gets interior to the disk inner boundary , it can only accrete the gas which is in its vicinity , i.e. typically the amount of gas contained within @xmath640.1 au .
since the core is expected to reach these radii early in the life of the protoplanetary disk , there may still be an adequate amount of gas there ( see [ sec : disks ] ) for it to build up a large envelope and become a giant planet .
however the conditions for that to happen are rather marginal .
if the protoplanetary core is stopped at some small radius before the disk is terminated ( see [ sec : migration ] ) , it might be able to retain contact with disk gas . in that case
it might be able to accrete enough gas supplied from the outer disk by viscous evolution to build up a massive atmosphere .
the scenario discussed in this section might be able to produce short period planets in the early stages of the disk evolution .
it would more likely result in a single planet at @xmath640.1 au on a timescale significantly shorter than for _ in situ _
formation if the core were in isolation in this latter process .
in this paper we have described briefly standard @xmath0disk models and presented some numerical calculations for @xmath24 and gas accretion rates through the disk varying from @xmath21 to @xmath22 m@xmath7 yr@xmath23 .
then we have discussed planet migration and formation in the context of these models .
we have reviewed the current theories of protoplanet migration , focusing on the gaseous disk protoplanet tidal interaction . according to these theories ,
protoplanets migrate from the location where they begin to form , which is supposed to be @xmath535 au , down to the disk inner parts , on a timescale significantly shorter than the disk lifetime or even the planetary formation timescale .
clearly the theory has to be developed to explain how at least some of the planets halt their migration before they plunge onto their parent star .
however , these results suggest that planets may not form _ in situ _ but more likely form at the same time as they migrate . accordingly ,
we have considered both the growth of the core and the built up of the envelope of a giant planet in the course of its migration .
our calculations indicate that accretion of planetesimals during the migration is likely to be high enough to maintain the core luminosity such that attainment of the critical core mass ( and therefore significant gas accretion ) does not occur as long as planetesimals are present , i.e. above @xmath640.1 au . although runaway gas accretion can then begin onto small mass cores at these small radii , the timescale for building up a massive envelope becomes longer than the disk lifetime if the core is too small .
however , cores massive enough can be built up through mergers of additional incoming cores on a timescale shorter than for _ in situ _
formation if the core were in isolation in this latter process .
the above considerations can lead to the preferential formation of short period planets on a relatively short timescale in the early stages of the disk evolution .
we thank s. balbus for useful comments on an early draft of this paper .
ct is supported by the center for star formation studies at nasa / ames research center and the university of california at berkeley and santa cruz , and in part by nsf grant ast9618548 .
bodenheimer , p. : 1998
, formation of substellar objects orbiting stars , in _ brown dwarfs and extrasolar planets _ , eds r. rebolo , e. l. martin , & m. r. zapatero osorio ( asp conference series , vol .
134 ) , p. 115 .
bryden , g. , chen , x. , lin , d. n. c. , nelson , r. p. , & papaloizou , j. c. b. : 1999 , tidally induced gap formation in protostellar disks : gap clearing and suppression of protoplanetary growth , _ ap .
j. _ * 514 * , 344367 .
lin , d. n. c. , & papaloizou , j. 1985 : on the dynamical origin of the solar system , in _ protostars and planets ii
d. c. black , & m. s. mathews ( tucson : university of arizona press ) , p. 9811072
lin , d. n. c. , & papaloizou , j. c. b. : 1993 , on the tidal interaction between protostellar disks and companions , in _ protostars and planets iii _ , eds .
e. h. levy , & j. i. lunine ( tucson : university of arizona press ) , p. 749835 .
papaloizou , j. c. b. , & lin , d. n. c. : 1984 , on the tidal interaction between protoplanets and the primordial solar nebula .
i linear calculation of the role of angular momentum exchange , _ ap .
j. _ * 285 * , 818834 . | we review results about protoplanetary disk models , protoplanet migration and formation of giant planets with migrating cores .
we first model the protoplanetary nebula as an @xmath0accretion disk and present steady state calculations for different values of @xmath0 and gas accretion rate through the disk .
we then review the current theories of protoplanet migration in the context of these models , focusing on the gaseous disk protoplanet tidal interaction . according to these theories
, the migration timescale may be shorter than the planetary formation timescale .
therefore we investigate planet formation in the context of a migrating core , considering both the growth of the core and the build up of the envelope in the course of the migration . |
optical beams are widely used in many areas of modern science such as microscopy , lithography , optical tweezers , and communications .
the extent of these applications is essentially determined by the attributes of the employed beams such as their polarization features , diffractive properties , and orbital angular momentum ( oam ) . for instance , oam - carrying beams beams possessing a helical phase structure and a donut - shaped transverse intensity profile can be used to rotate micro - particles @xcite , enhance optical resolution in stimulated emission depletion ( sted ) microscopy @xcite , and increase a communication channel s capacity @xcite .
likewise , shape - invariant optical beams modulated by airy @xcite and bessel @xcite functions have potential applications in optical trapping @xcite .
finally , space - varying polarized light beams beams with certain polarization topologies have been shown to produce needle - shaped longitudinal electric and magnetic dipoles @xcite and exotic 3d polarization structures @xcite , hence their usefulness for exploring novel physical and optical phenomena .
the above applications require individually specific beams characterized by well - defined intensity , phase , and polarization attributes , thus motivating the development of novel beam shaping methods and devices .
spatial light modulators ( slms ) pixelated liquid crystal devices that are capable of introducing a specific phase modulation onto an incident beam have been widely used in different configurations to tailor certain optical phase , intensity , and polarization distributions @xcite .
however , slms have a few drawbacks bounding them to laboratory - scale usage .
to begin with , they are pixelated and limited to a certain phase modulation resolution ( usually associated with 8-bit , 256 grayscale encryption ) .
this restrains slms to diffractive configurations which makes them effectively bulky and lowers their efficiency .
in addition , their operation is characterized by fairly low phase modulation rates ( @xmath1 hz ) , which affect their dynamical capabilities .
similar _ diffraction _
devices , such as deformable mirror arrays ( dma ) and digital micro - mirror devices ( dmd ) have similar problems @xcite .
a few optical devices , such as axicons @xcite , mirror - cones @xcite , @xmath2-plates @xcite , astigmatic mode converters @xcite , and achromatic - oam generators @xcite have been introduced to generate specific optical beams .
however , the most efficient and compact of these devices are pancharatnam - berry phase optical elements ( pboes ) @xcite which have the ability to tailor optical beams through the addition of a geometric phase .
in essence , they add a transverse phase profile to a beam upon flipping its circular polarization state .
this phase alteration depends on the device s geometry .
only a few methods have so far been proposed to manufacture pboes @xcite .
these techniques are mainly capable of generating azimuthally structured devices @xcite . here
, we present a novel configuration capable of manufacturing arbitrary pboes to demonstrate their versatility .
we use them to generate oam - carrying , bessel , airy , and ince - gauss optical beams .
we also exploit our pboes spin - to - orbit coupling properties through the generation of space - varying polarized beams by providing linearly polarized input beams to our devices .
many physical entities can acquire a global phase upon being subjected to cyclic adiabatic processes .
such phases are commonly known as pancharatnam - berry ( geometric ) phases @xcite .
the latter of these terms is associated with the geometric transformation of the parameters defining the corresponding cyclic process . among the many systems that can acquire a geometric phase
are polarized optical beams through the adiabatic variation of their polarization states .
these states can be expanded in terms of a two dimensional vector space and can thus be effectively mapped onto the surface of a unit sphere known as the poincar sphere .
an arbitrary unitary transformation of an optical beam s polarization state can be performed by a sequence of ( half- and quarter- ) wave plates . upon a cyclic polarization transformation along a closed path on the poincar sphere
, a phase corresponding to half of the solid angle enclosed by the path will be added to the beam @xcite .
this effect can be clearly illustrated through the example of a half - wave plate ( hwp ) acting upon light whose polarization is expressed in the circular polarization basis @xmath3 where @xmath4 and @xmath5 are respectively the left- and right - handed circular polarization vectors .
namely , when the half - wave plate is oriented at an angle of @xmath6 with respect to its optical axis , the solid angle enclosed by the path of the beam s polarization vector on the poincar sphere is @xmath7 . as a result ,
the beam s left - handed component becomes right - handed and acquires a phase of @xmath8 while the beam s right - handed component becomes left - handed and acquires a phase of @xmath9 .
this action is described by @xmath10 besides half - wave plates , which are defined by a uniform optical axis and add a constant phase to a beam , there are other types of devices , such as pboes , designed to add a space - varying geometric phase to an incident beam . due to the identical nature of the phase added by both of these optical elements , one can readily establish an analogy between the optical properties of a half - wave plate and those of a pboe .
in essence , much like how the uniformity of the phase added by a half - wave plate arises from its uniform optical axis , the spatial dependence of the phase added by a pboe can be seen as stemming from a non - uniform optical axis distribution .
more specifically , this distribution is given by @xmath11 , where @xmath12 and @xmath13 are coordinates defining a position on the beam s transverse profile . as in the case of a half - wave plate , the corresponding added geometric phase is given by @xmath14
however , in most cases , pboes are not perfect half - wave retarders , i.e. they are not defined by an optical retardation of @xmath15 , but rather by a parameter @xmath16 .
hence , it follows that a pboe s unitary action @xmath17 on an incident light beam in the circular basis , @xmath4 and @xmath5 , for a given optical retardation @xmath16 and a general optical axis distribution @xmath11 is provided by @xcite @xmath18 therefore , any beam that may be generated by imparting a space - varying phase to its transverse profile can be readily produced by pboes characterized by an appropriate optical axis distribution .
some distributions can in certain cases be more conveniently expressed in specific coordinate systems ( @xmath19 ) such as cartesian ( @xmath20 ) , cylindrical ( @xmath21 ) , and elliptical ( @xmath22 ) systems . in the following ,
we provide several examples of existing and potential pboes that generate beams whose mathematical formulations are more easily expressed with a certain coordinate system .
these beams include oam - carrying and bessel beams ( cylindrical ) , ince - gauss beams ( elliptical ) , and airy beams ( cartesian ) .
pboes have so far been mostly designed to make devices known as @xmath2-plates @xcite .
@xmath2-plates , a special class of pboes , possess an azimuthally symmetric optical axis given by @xmath23 , where @xmath2 is either a half or a full integer , @xmath24 is an arbitrary constant , and @xmath25 and @xmath26 are the radial and azimuthal cylindrical coordinates , respectively @xcite .
these devices are of interest because of their ability to generate beams carrying a phase of @xmath27 , where @xmath28 is an integer , thus causing the beam to carry oam @xcite .
it has also been demonstrated that @xmath2-plates can be used to generate space - varying polarized light beams by either feeding linearly polarized light to the device or by adequately tuning its optical retardation @xcite .
adding a linear radial phase dependence , e.g. @xmath29 , to these oam carrying beams will simulate passing the beam through an axicon .
the beam s resulting total phase , @xmath30 , corresponds to that of a bessel beam .
the latter s transverse field is proportional to a @xmath31 term , where @xmath28 is an integer , @xmath32 is the @xmath33 order bessel function of the first kind , and @xmath34 is the radial component of the beam s wavevector @xcite .
a potential design for a pboe that generates such a beam could be defined by an optical axis satisfying @xmath35 .
similar to the previous case , different circular polarization components will acquire opposite values of oam .
however , one circular polarization component will be attributed with a phase of @xmath36 while the other with a phase of @xmath37 .
effects related to these radial phase components will be more thoroughly discussed later on in the following sections .
bessel and oam beams can be readily generated by imposing a cylindrically symmetric phase onto an incident gaussian beam .
other types of beams require the impartment of phase patterns that are more conveniently expressed in other coordinate systems .
such beams include the elliptically symmetric ince - gauss ( ig ) beams .
ig beams form an additional complete set of orthonormal solutions to the paraxial wave equation expressed in elliptic coordinates ( @xmath22 ) @xcite .
they include an ince polynomial in their formulation defined by two integer indices @xmath38 and @xmath39 , along with an additional parameter @xmath40 referred to as the beam s ellipticity .
these polynomials are categorized either as even or odd ince polynomials and correspondingly yield the so called even , @xmath41 , and odd , @xmath42 , ince - gauss modes respectively .
the number of hyperbolic and elliptic nodal lines in the transverse intensity profile of both sets of modes is determined by their @xmath39 and @xmath38 indices .
beams resulting from the superposition of even and odd ig modes with a relative phase difference of @xmath43 , i.e. , @xmath44 ( @xmath45 sign denotes the direction along which the beam s phase varies ) , are called helical ig beams and exhibit optical vortices and an elliptic phase dependence @xcite .
helical ig modes were used to investigate several quantum features including two- and three - dimensional entanglement between their constituent photons @xcite and hence have potential applications in quantum information protocols @xcite .
pboes characterized by an optical axis satisfying the relation @xmath46 , where @xmath47 is the phase of a helical ig beam , would be able to generate beams defined by interesting polarization properties . for instance , consider the case where @xmath47 is associated with an @xmath48 beam . according to eq .
( [ eq : unitaryaction ] ) , the beam resulting from feeding an arbitrarily polarized gaussian beam into a half - wave retarding ( @xmath49 ) ig pboe will consist of two components , namely @xmath50 and @xmath51 .
as previously mentioned , the sign in the formulation of helical ig modes defines the direction along which the beam s phase elliptically varies . in our case , adding a phase of @xmath52 would be equivalent to generating the @xmath53 helical ig mode . with these factors brought into consideration
, we can equivalently write the two components of the device s output beam as @xmath54 and @xmath55 . therefore , by feeding linearly polarized light into the device , the relative amplitude of both components should be equal from which it follows that the resulting beam can be decomposed into two orthogonally polarized linear components , one of which being the helical ig mode s _ even _ mode and the other being the corresponding _ odd _ mode .
other types of beams , such as airy beams , require the addition of a phase that can most readily be expressed in cartesian coordinates .
like bessel beams , airy beams are diffraction - free solutions to the paraxial wave equation @xcite .
one of their distinguishing characteristics is the fact that their main intensity maxima seem to freely accelerate along a parabolic trajectory . to produce these beams
, one can exploit the fact that the fourier transform of an airy function @xmath56 is proportional to a term with a phase varying with the cube of the frequency @xcite .
thus , adding a phase with a cubic spatial dependence on an incident gaussian beam and then taking its fourier transform using an appropriate lens leads to a lab - scale realization of the corresponding airy beam @xcite within a certain propagation range .
for instance , the following space - varying phase pattern can be used to generate a 2d airy beam @xmath57 ^ 3+\left[y_0\left(\frac{2\pi}{\lambda f}y\right)\right]^3\right\ } , \end{aligned}\ ] ] where @xmath58 is the focal length of the spherical converging lens used to take the fourier transform of the beam , @xmath59 is the beam s wavelength , and @xmath60 and @xmath61 are parameters that describe the beam s deflection along the @xmath62 and @xmath63 coordinates , respectively , upon propagation . as in the case of the aforementioned pboes , an airy beam generated by a pboe is defined by certain polarization features .
in essence , the left - handed component of an incident beam will acquire a phase of @xmath64 while the right - handed component a phase of @xmath65 .
an appealing feature of this particular device is the fact that its added phase is an odd function satisfying @xmath66 .
this odd symmetry causes the left handed component of the converted beam to look inverted with respect to the right handed component , which leads to an `` acceleration '' along the opposite direction .
and @xmath67 ) .
the beam then interacts with the dmd programmed to reflect a specific light intensity pattern .
the reflected beam goes through another 4-f system ( @xmath68 and @xmath69 ) and an iris in order to select only its brightest diffraction orders as a mean to clearly image the pattern onto the sample .
the polarization of the beam is then adjusted by a half - wave plate to the required orientation .
after this rotation , the resulting beam aligns a specific region of the sample s photo - alignment layer . the device s structural frame is depicted in the inset in the lower - left region of the figure . ]
cylindrically symmetric pboes can be fabricated by a variety of methods such as through the use of sub - wavelength gratings @xcite , total internal reflection @xcite , plasmonic metasurfaces @xcite , or an inhomogeneous and anisotropic layer of patterned nematic liquid crystals @xcite . here
, we employ the last of these methods to produce our pboes designed to add an arbitrary geometric phase to optical beams .
we first describe the method to construct a pboe with a desired optical - axis pattern imprinted into its liquid crystal layer .
such a construction can be achieved through the photoalignment of an azobenze - based dye . to perform this alignment
, the azodye molecules must be exposed to specific linearly polarized light intensity patterns whose wavelength lies within a peak of the dye s absorption spectrum , thus aligning the molecules with the pattern s polarization @xcite .
when such an aligned azodye molecule is in contact with the pboe s liquid crystals , an electrostatic interaction between the two types of molecules will align the liquid crystal molecules along the axis of the azodye . in practice
, the contact between the two molecules is achieved by spin coating an azodye layer onto the inner region of two substrates enclosing the device s liquid crystals .
thus , by exposing well - defined areas of the coated substrates to specifically polarized intensity patterns , one is able to construct the pboe s space dependent optical axis . this design can also be modified to allow one to control the pboe s optical retardation .
more specifically , when substrates coated with a layer of a transparent conductive material such as indium tin oxide ( ito ) are used , electrical wires can be soldered onto these substrates .
this design modification allows one to apply an electric field to the pboe s liquid crystals , thus giving the liberty to control its optical properties including its birefringence . as a result ,
the element s optical retardation becomes tunable , hence allowing one to use the device to generate a wider selection of optical beams and providing the ability to use the device on beams with different wavelengths @xcite .
we produce the device s dye - coated substrates by spin - coating the azobenzene - based aligning material paad-22 provided by beam co onto the conductive side of an ito - coated substrate using the procedure described in @xcite .
more specifically , the sample is spin coated for 60 s at a rotational speed of 3000 rpm and then soft baked at 120@xmath70c for 5 minutes .
this procedure typically results in the formation of a 50-nm - thick alignment layer , as we verified by observation using a profilometer . before being exposed to light patterns
, the two coated substrates are attached together with the coated region facing towards the inside of the device using an epoxy glue .
we glued them with a slight lateral offset in order to ensure that wires can be soldered onto the inner conductive sides of the substrates for tuning .
spacer - grade silica microspheres with diameters below 4.8 m placed between the two plates ensure the presence of a uniformly spaced cavity in which liquid crystals can later be added .
prior to permanently gluing the plates together , the cavity s uniformity is visually optimized by minimizing the number of fringes produced by thin film interference resulting from the space between the plates . at this stage of the fabrication process ,
the device consists of the pboe s structural frame and is depicted in the inset of fig .
[ fig : apparatus ] . the current state - of - the - art method used to fabricate liquid - crystal - based pboes
can only be used to fabricate devices with an optical axis that solely depends on the azimuthal coordinate .
it specifically relies on rotating the sample under an angular light intensity pattern of a distinct polarization rotating at a different frequency than that of the sample @xcite .
in essence , the resulting alignment will depend on the rotation frequency ratio of these two elements .
to generate more complex optical beams , we use a photoalignment method similar to @xcite able to fabricate pboes characterized by an arbitrarily distributed optical axis .
our method relies on discretizing a pboe s optical axis pattern into @xmath71 segments each of which is associated with a single orientation .
therefore , we reconstruct this discretized version of the pboe by sequentially exposing linearly polarized light patterns to specific regions of the alignment layer defined by the corresponding optical axis orientation .
this method can be implemented via a dmd photoalignment method where the dmd is used to generate the required optical patterns . as displayed in fig .
[ fig : apparatus ] , the rate at which the dmd produces varying light patterns is then synchronized to the rate at which a half - wave plate rotates the polarization of a linearly polarized light beam .
doing so allows the reconstruction of the pboe s desired optical axis onto its photoalignment layer .
additional systems of lenses are added to the setup in order to refine the quality of the alignment .
the first system consists of a set of two cylindrical lenses effectively rotating the beam by @xmath72 and enlarging its size by a factor of three in order to make it uniformly cover the dmd s display .
the second set of lenses consists of a 4-f system of two spherical lenses imaging the pattern reflected by the dmd onto the sample s alignment layer while enlarging it by a factor of 2.5 thus exposing the pboe s frame to a rectangular beam with a length of 0.9 cm and a width of 1.6 cm .
moreover , using an iris , only the brightest diffraction orders of the pattern reflected by the dmd are selected in order to increase the resolution of the image projected on the sample .
after the azodye layers are aligned , liquid crystals are inserted into the structural frame s cavity which is later sealed using epoxy glue .
finally , wires are soldered on the edge of each plate using an alloy composed of 97% indium and 3% silver . in order to fully characterize the beams generated by the manufactured pboes ( transverse intensity , phase , and polarization profiles ) , we investigate the modulation of a beam with a gaussian profile and a wavelength of 633 nm with the setup shown in fig .
[ fig : characsetup ] . prior
to modulating the beam with the device , we modify its polarization by using a polarizing beam splitter ( pbs ) and a quarter - wave plate ( qwp , @xmath73 ) .
this pbs - qwp combination allows us to send either horizontally or circularly polarized light through the pboe .
hence , we are able to generate several sorts of beams along with their associated superposition combinations . after the desired beam
is generated , it goes through a qwp , hwp ( @xmath74 ) , and pbs combination to investigate the various polarization components of the generated transverse beam s profile @xcite .
finally , a reference gaussian beam , which was split from the beam by a non - polarizing beam splitter before the first qwp , is made to interfere with one of the output beam s polarization components , thus making its phase profile become observable in the resulting interference pattern . the reference beam can be blocked when the setup needs to be used only to investigate the intensity profile of the beams generated by the pboes . the resulting intensity or interference profiles
are then recorded with a ccd camera . ) , and a pboe ; the group of elements that select a single polarization includes a quarter - wave plate , a half - wave plate ( @xmath74 ) , and a pbs .
the beam resulting from this interferometer is then observed using a ccd camera . ]
the devices transmission efficiencies were also evaluated using this apparatus .
these efficiencies usually go up to 72% .
however , most losses are associated with an incident beam s reflections on the device s glass substrates and can be reduced by applying an anti - reflection coating on the optical element .
doing so would allow the pboe s transmission efficiency to reach a value closer to its conversion efficiency , which can reach values as high as 97% .
to test the proposed method , we fabricate a standard @xmath75-plate , which is able to generate a vortex beam with an oam value of @xmath76 .
the device was fabricated using @xmath77 pattern method and was designed to imprint a transverse phase of @xmath78 onto an incoming gaussian beam . the device s optical axis distribution and its corresponding added geometric phase profile are shown in fig .
[ fig : q1fig ] .
the device s expected transmission profile when placed between two linear polarizers is also displayed along with an image of the actual sample placed between such polarizers . of the @xmath75-plate with the corresponding geometric phase @xmath79 , shown in hue colors , added onto an incoming beam ; ( ii ) expected transmission when the sample is placed between two polarizers ; ( iii ) observed transmission pattern . *
b * far - field intensity and polarization profiles of the beams generated using the fabricated plate .
the input polarization and tuning of the @xmath2-plate are indicated for each case .
( i ) uniformly right - handed circularly polarized @xmath80 beam ; ( ii ) `` dipole '' vector vortex beam ; ( iii ) `` spiral '' poincar beam .
* c * intensities and interference patterns with a plane wave of beams generated using ( i ) left - handed circularly , ( ii ) right - handed circularly , and ( iii ) horizontally polarized input light beams . ] as a first test , we electrically tuned the @xmath75-plate to an optical retardation of @xmath49 , which , according to eq .
( [ eq : unitaryaction ] ) , results in the entire input beam obtaining a phase profile of @xmath78 .
this leads to a phase singularity in the beam center imprinted onto the incident gaussian input beam .
such beams are known as hypergeometric gaussian beams @xcite .
the intensity and polarization profile of a beam resulting from illuminating the tuned @xmath75-plate with circularly polarized light is shown in fig .
[ fig : q1fig ] * b*-(i ) . as expected , the beam is circularly polarized and possesses a donut - like intensity profile . as investigated previously ,
@xmath2-plates can be used to generate space - varying polarized light beams , namely vector vortex @xcite and poincar beams @xcite . when linearly polarized light is sent through a tuned @xmath2-plate , a so - called vector vortex beam is produced . in the case of a horizontally polarized input and a @xmath75-plate ,
one obtains a vector beam defined by a linear polarization with an orientation given by twice the azimuthal coordinate .
this specific beam is commonly referred to as a `` dipole''-polarization beam as its polarization pattern resembles the field pattern of a dipole ( see fig .
[ fig : q1fig ] * b*-(ii ) ) . if the @xmath2-plate is tuned to an optical retardation of @xmath81 , the resulting beam consists of an equally weighted superposition of a converted and a non - converted component as seen in eq .
( [ eq : unitaryaction ] ) . such a _ detuned _
@xmath2-plate s output beam is depicted in fig .
[ fig : q1fig ] * b*-(iii ) . here
, the beam consists of an equally weighted superposition of a gaussian beam and an @xmath80 vortex beam .
the equal composition is clearly seen through the beam s relatively flat - top central intensity that quickly decreases as it reaches the radius where the @xmath82 vortex beam s intensity profile starts to drop .
this output also corresponds to the poincar beam that can be generated using a @xmath75-plate @xcite . in our case
, the resulting poincar beam is defined by a polarization profile resembling a spiral , a result which is indeed experimentally observed and depicted in fig .
[ fig : q1fig ] * b*-(iii ) . to observe the generated beam s transverse phase structure , the reference arm in fig .
[ fig : characsetup ] was used to perform interferometric measurements . a right - handed circularly polarized beam characterized by the index @xmath80
was generated at the output of the first arm . when it was interfered with an intentionally slightly _ misaligned _ gaussian reference beam , the interference pattern displayed in fig .
[ fig : q1fig ] * c*-(i ) was obtained .
the misaligned interference pattern reveals a pitchfork with three tines .
given that the interference between a plane wave and a beam with a phase - front defined by @xmath28 intertwined helices results in a pitchfork interference pattern consisting of @xmath83 tines @xcite , the recorded structure confirms that the beam has an @xmath28 value of 2 . when a left - handed circularly polarized beam having an index of @xmath84 was generated and interfered with the gaussian reference beam , we obtained the pattern in fig .
[ fig : q1fig ] * c*-(ii ) which exhibits a similar pitchfork of opposite orientation as it carries oam of opposite sign .
the qwp before the @xmath2-plate was then oriented to provide horizontally polarized light to the @xmath75-plate , thus again generating the vector vortex beam seen in fig .
[ fig : q1fig ] * b*-(ii ) . after projecting onto the horizontal component of the resulting beam by a second pbs
, we find the expected four lobes defined by two possible distinct phases in the vicinity of either 0 or @xmath15 . as a result , the interference fringes observed in the region separating the lobes are out of phase as observed in fig . [ fig : q1fig ] * c*-(iii ) .
all tests examining the intensity , phase , and polarization profiles of beams generated using the manufactured @xmath75 plate reveal that the properties of these beams are in excellent agreement with their corresponding expected and well - established results .
we further demonstrate the method s ability to fabricate highly complex @xmath2-plates by manufacturing a device defined by a high topological charge , namely a @xmath86-plate able to generate optical beams characterized by @xmath0 units of oam .
devices able to generate extreme high - order modes are good demonstrations of the ability to explore the unbounded nature of oam .
they can also be used to produce photonic gears which can potentially be useful in optical / quantum metrology @xcite . due to the relatively high resolution needed to fabricate the device associated with the high frequency at which its optical axis varies
, the @xmath2-plate was made using a lower reconstruction resolution consisting of a four step discretization method .
the high degree at which the @xmath2-plate s optical axis orientation varies results in the formation of a moir pattern in the plate s added phase .
thus , for practical reasons , the device s center is refrained from being exposed to an incident light beam .
this technique is often employed in many areas of beam shaping involving the fabrication of elements defined by abrupt variations in the phase aimed to be added to a beam @xcite we demonstrate the @xmath2-plate s performance by feeding horizontally polarized light to it and only selecting the generated beam s horizontal polarization component , which should consist of @xmath87 lobes . as depicted in fig .
[ fig : photonicgear ] , all 400 lobes of the generated beam are present , clearly defined , and uniformly shaped across the beam s profile .
-plate fabricated using the procedure described in the text . ]
after verifying that our method works , we continue using it to fabricate an extensive range of pboes in a way that has not been possible before .
we first fabricate a pboe generating a bessel beam defined by parameters of @xmath88 and @xmath89 using a @xmath90 pattern method .
similarly to the @xmath2-plate described above , we show our device s optical axis distribution along with the corresponding geometric phase modulation as well as its observed appearance between two crossed polarizers in fig .
[ fig : besselfig ] * a*. of a pboe generating a bessel beam defined by parameters of @xmath88 and @xmath89 along with the corresponding geometric phase @xmath79 , which is shown in hue colors ; ( ii ) its observed transmission between two polarizers .
* b * experimentally observed intensity and interference pattern of the ( i ) right- and ( ii ) left - handed circular components of the generated beam at each of their foci .
* c * reconstructed polarization profile of a generated beam using the bessel pboe . measured intensities for @xmath4 , @xmath91 , and @xmath5
components are depicted on the right side . ] as bessel beams exhibit their defining characteristics at their focus , we modified the characterization apparatus ( depicted in fig .
[ fig : characsetup ] ) to first enlarge the beam and illuminate a wider region of the device to then weakly focus it using a 1000 mm lens .
thus , we are able to better resolve the characteristics in its focal region
. the right - handed and left - handed circular components of this beam have a phase with a radial dependence given by @xmath29 and @xmath37 , thus it will exhibit converging and diverging lensing behaviors , respectively . as a result , the bessel beam attributed to the right - handed component will be seen to focus before the one attributed to the left - handed component .
figure [ fig : besselfig ] * b * shows the intensity profile of both circular polarizations at their respective foci . since @xmath5 focuses closer to the plate , it forms a tighter bessel beam than the one formed by @xmath4 . to examine the phase of the generated beams , we recorded interference patterns resulting from the interference between the converted circular components of the beams and a reference beam with a uniform phase front ( see fig .
[ fig : besselfig ] * b * ) .
the interference patterns exhibit circular discontinuities , which are attributed to the fact that each ring appearing in a bessel beam s intensity profile is out of phase with its adjacent rings .
moreover , these patterns are defined by a pitchfork consisting of two tines , which is a sign that the beam carries an absolute oam value of 1 . in our experimental images
these features are clearly observed for both components of the generated bessel beam .
in addition , the different orientation of the pitchfork for each component demonstrates the opposite sign of their oam quanta .
finally , we explore the ability to use the fabricated device as a mean to generate space - varying polarized light beams .
in particular , we examine the polarization of an emitted beam consisting of equally weighted circular polarization components in a region where both components perfectly overlap .
the resulting beam along with its reconstructed complex polarization profile is shown in fig .
[ fig : besselfig ] * c*. in addition to azimuthal variations in the polarization pattern attributed to the beam s azimuthally dependent phase , one can also see the effects of the radial dependence of the beam s phase through radial variations in the beam s transverse polarization profile .
this feature can be made more visible if the beam is projected onto different polarization components .
namely , both circular components ( @xmath4 and @xmath5 ) have a uniform appearance over the beam s cross - section , yet the linear component @xmath91 is found to form intertwined spirals .
using our custom alignment method , we were able to fabricate a device producing an @xmath92 beam .
again , we used an @xmath90 pattern method in the fabrication procedure to discretize the required continuous modulation . in fig .
[ fig : ig53plate ] * a * the pboe s desired optical axis along with the corresponding added geometric phase are shown . the device s expected appearance when placed between two crossed linear polarizers is also displayed together with an image of the actual sample . , and the corresponding geometric phase @xmath93 shown in hue colors ; ( ii ) expected device appearance between two crossed polarizers ; ( iii ) image of the @xmath92 pboe .
* b * expected ( top ) and experimental ( bottom ) : ( i ) intensity patterns of the completely converted component of the beam produced by the fabricated @xmath92 pboe ; ( ii ) corresponding fringes resulting from the interference of this beam with a plane wave .
theoretical intensity profiles of the ( iii ) even and ( iv ) odd ig modes and their recovered experimental intensity profiles . ]
the ig - pboe s characterization was performed similarly to the previous ones .
the expected and experimentally recorded intensities and interference patterns with a plane wave were examined and are shown in fig .
[ fig : ig53plate ] * b*. to verify the polarization properties of the generated beams , we feed linearly polarized light into the perfectly tuned device and recover the resulting beam s even ig mode and its odd ig mode . as each of these modes
is attributed to a linear polarization component , we were able to recover them by accordingly modifying the orientation of the setup s half - wave plate .
these recovered beams along with their corresponding theoretical transverse intensity profiles are shown in fig .
[ fig : ig53plate ] * b*. as a final demonstration of the flexibility of our fabrication method , we reproduce the fabrication of pboes producing airy beams as performed in @xcite .
more specifically , we manufactured a device designed to add a phase of @xmath64 defined in eq .
( [ eq : airyphase ] ) .
the employed parameters were @xmath94 , @xmath95 , @xmath96 .
similar to the earlier findings , we display in fig .
[ fig : airyplate ] the pboe s optical axis distribution and the corresponding geometric phase along with its expected and observed appearance between two crossed polarizers . and
the corresponding geometric phase @xmath97 , which is shown in hue colors ; ( ii ) expected appearance between two crossed polarizers ; ( iii ) image of the device generating a 2d airy beam . *
b * expected ( top ) and observed ( bottom ) intensity profiles of the ( i ) right - handed circular polarization @xmath5 , ( ii ) left - handed circular polarization @xmath4 , and ( iii ) a superposition of both components of the generated airy beam . ] each polarization component generated by such a device will appear to be inverted with respect to the other .
this inversion is clearly depicted in the intensity patterns of the generated beams shown in fig .
[ fig : airyplate ] * b * along with their corresponding expected appearance , which are in excellent agreement .
in conclusion , we have devised and implemented a dmd - based photoalignment method to fabricate liquid crystal pancharatnam - berry phase optical elements defined by arbitrary optical axis spatial distributions .
these pboes are more compact , cost - effective , and responsive to phase inversions than other beam shaping devices restricted to laboratory - scale usage .
in particular , we manufactured devices capable of generating oam - carrying , bessel , ince - gauss , and airy beams .
moreover , our generation of complex patterns with high resolution displays our ability to produce arbitrary beam patterns .
these complex beams are characterized by distinct transverse intensity , phase , and polarization profiles along with the high conversion efficiency required for many applications such as quantum information and optical tweezers .
for instance , cascading these pboes along with suitably fast pockels cells may be used for ultra - fast ( at gigahertz rate ) and super - dense telecommunications .
canada research chairs ; canada excellence research chairs ( cerc ) ; government of canada ; natural sciences and engineering research council of canada ( nserc ) .
the authors thank sergei slussarenko and svetlana lukishova for useful advice and discussions .
52ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) in @noop _ _ ( , ) pp . @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) in @noop _ _ , vol .
( , ) pp . in @noop _ _ , vol .
( , ) pp . @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | converting spin angular momentum to orbital angular momentum has been shown to be a practical and efficient method for generating optical beams carrying orbital angular momentum and possessing a space - varying polarized field . here , we present novel liquid crystal devices for tailoring the wavefront of optical beams through the pancharatnam - berry phase concept .
we demonstrate the versatility of these devices by generating an extensive range of optical beams such as beams carrying @xmath0 units of orbital angular momentum along with bessel , airy and ince - gauss beams .
we characterize both the phase and the polarization properties of the generated beams , confirming our devices performance . |
it is known that many bubbles levitated in a liquid form a stable structure named `` bubble grapes '' when a weak standing sound wave is applied @xcite . in this stable structure , bubbles or clusters of them maintain a distance each other and tend not to collide , contrary to predictions given by the classical theory by bjerknes which allows either only attraction or repulsion .
the distance between bubbles or clusters in the structure is comparable to their sizes .
a similar phenomenon can be observed in a strong acoustic field @xcite in which bubbles form a complex filamentary structure varying slowly , whose scale is much smaller than the wavelength of the acoustic field . in order to understand these phenomena
, one should know thoroughly how the mutual interaction of the bubbles influences their acoustic properties . in ref .
@xcite , zabolotskaya showed theoretically that the effective resonance frequencies of two interacting bubbles rise or fall as bubbles approach each other and that the variation in the resonance frequencies may sometimes causes the change of the sign of the secondary bjerknes force , which is an interaction force acting between pulsating bubbles , at a certain distance . in ref .
@xcite , doinikov and zavtrak introduced a similar result by taking the multiple scattering of sound between two bubbles into account .
they explain the origin of the change in the sign of the force as follows : when two bubbles oscillate in phase , their surfaces move against each other resulting in a stiffening of the bubbles .
this may increase the effective resonance frequencies ( or eigenfrequencies ) of both bubbles and may sometimes cause the change in the sign of the force from attractive to repulsive .
influences of mutual interaction have also been studied in regard to sonoluminescence @xcite . in ref .
@xcite , mettin et al .
examined numerically the sign and magnitude of the secondary bjerknes force acting between two small air bubbles in a strong acoustic field .
they determined that even when the distance between two bubbles is much larger than their equilibrium sizes the magnitude of the force becomes much larger than that predicted by a linear theory , and the sign of the force changes near the dynamical brake threshold or around the _ nonlinear _ resonance frequency . in ref .
@xcite , doinikov showed that a neighboring bubble drastically varies the magnitude and sign of the primary bjerknes force , which an incident sound wave exerts on a bubble , in a strong acoustic field . in this paper , following zabolotskaya s approach ,
the influences of the mutual interaction on the effective resonance frequencies of bubbles are theoretically examined in detail . as was pointed out and was mentioned previously
, the variations in eigenfrequencies strongly affect the primary and secondary bjerknes forces .
this means that full understanding of the eigenfrequencies surely must be of help in knowing how the forces are changed by mutual interaction . in sec .
[ sec2a ] , we show that the variation in the eigenfrequencies with respect to the distance between two bubbles is more complicated than those expected by zabolotskaya @xcite and doinikov and zavtrak @xcite . in sec .
[ sec2b ] , additionally , we discuss theoretically the influence of the mutual interaction on the amplitude of the bubble pulsation . in sec .
[ sec3 ] , we show some numerical results of the primary bjerknes force , acting on two interacting bubbles , by using a nonlinear model in order to verify the linear theory for the eigenfrequencies . when the frequency of an external sound is equivalent to an eigenfrequency of a bubble , the force vanishes because of the phase difference of @xmath0 between the sound and the bubble pulsation . namely , by observing the points where the force vanishes , one can know how the eigenfrequencies changed by the interaction . in the present study , an external sound field of weak or moderate amplitude with a frequency comparable to the partial resonance frequencies of bubbles
is assumed .
a gas bubble levitated in a liquid pulsates when a sound wave is applied .
the sound pressure at the bubble position drives the pulsation . when other bubbles ( named `` bubble 2 '' @xmath1 `` bubble @xmath2 '' , where @xmath2 is the number of existing bubbles ) exist near the bubble ( bubble 1 ) , the pulsation of bubble 1 is also driven by the sound wave scattered by the other bubbles .
namely , the driving pressure acting on bubble 1 , @xmath3 , is expressed as @xmath4 where @xmath5 and @xmath6 are the sound pressures of the external sound field and the scattered wave emitted by bubble @xmath7 , respectively , at the position of bubble 1 .
the pressure of the external sound can be considered uniform when the wavelength of the standing wave is much larger than the radii of the bubbles and the distance between bubbles . by assuming that those bubbles keep being spherically symmetric and the surrounding liquid is incompressible , the scattered pressure can be estimated with @xcite @xmath8 where @xmath9 is the density of the liquid
, @xmath10 is the distance between the centers of bubble 1 and bubble @xmath7 , @xmath11 is the radius of bubble @xmath7 , and the dot denotes the time derivative .
when the compressibility of the material surrounding the bubbles is completely negligible , one can use the rpnnp equation @xcite : @xmath12 with @xmath13 for modeling the radial pulsation of bubbles , where @xmath14 is the static pressure , @xmath15 is the surface tension at the bubble surface , @xmath16 is the viscosity in the liquid , @xmath17 is the polytropic exponent of the gas inside the bubbles , and @xmath18 is the equilibrium radius of bubble 1 .
substitution of eqs .
( [ eq1 ] ) and ( [ eq2 ] ) into eq .
( [ eq3 ] ) yields @xmath19.\ ] ] when the number of bubbles is two , this equation is reduced to @xmath20,\ ] ] where @xmath21 ( @xmath22 ) .
exchanging 1 and 2 in the subscripts yields the model equation for bubble 2 : @xmath23\ ] ] with @xmath24 in the following sections , we present theoretical and numerical discussions regarding the above systems of equations ( [ eq4 ] ) and ( [ eq5 ] ) .
in this section , some basic properties of eqs .
( [ eq4 ] ) and ( [ eq5 ] ) are discussed by means of linear theory . by assuming that the bubble pulsation can be represented as @xmath25 , @xmath26 , and @xmath27 , @xmath28 , eqs .
( [ eq4 ] ) and ( [ eq5 ] ) are reduced to the following linear formulae : @xmath29 where @xmath30}\ ] ] and @xmath31}\ ] ] are the partial resonance ( angular ) frequencies of bubble 1 and bubble 2 , respectively , @xmath32 in the case where @xmath33 , the system of equations ( [ eq8 ] ) and ( [ eq9 ] ) corresponds to that derived by zabolotskaya @xcite on the basis of lagrangian formalism .
let us analyze in detail the eigenfrequencies of the linear coupled system , although a brief discussion is shown in ref .
when only one bubble exists in an acoustic field and the frequency of the external sound is equivalent to the resonance frequency of the bubble , the phase difference between the bubble pulsation and the external sound becomes @xmath0 , as is well known . according to this analogy
, we call the frequency of an external sound an eigenfrequency of the linear coupled system when the phase difference between @xmath34 and @xmath35 ( or @xmath34 and @xmath36 ) becomes @xmath0 .
we assume that the external sound pressure at the bubble position is written in a form of @xmath37 . a harmonic steady - state solution of the linear coupled system is @xmath38 where @xmath39 @xmath40 with @xmath41 @xmath42 here the solution for only bubble 1 is shown .
the phase difference @xmath43 becomes @xmath0 when @xmath44 it should be noted that a case in which both @xmath45 and @xmath46 become zero does not exist . from eqs .
( [ eq12 ] ) and ( [ eq13 ] ) , one obtains @xmath47 the numerator of this equation always has a nonzero value since @xmath48 ; this result denies the existence of the case where both @xmath49 and @xmath50 are true .
it should be noted also that @xmath51 appearing in the dominator of eq .
( [ eq12 ] ) always has a nonzero value .
when @xmath52 , @xmath53 is reduced to @xmath54 this has a nonzero , negative value because of @xmath55 , @xmath56 and @xmath57 .
this result means that no case exists where both @xmath58 and @xmath52 are true . in a consequence , eq .
( [ eq14 ] ) is reduced to @xmath59 in the following , we analyze this equation . when the viscous terms in eq .
( [ eq15 ] ) are negligible ( but exist ) , one can easily obtain the solutions of this equation . by assuming that @xmath60 and @xmath61 , one obtains @xmath62 or @xmath63 and @xmath64 equation ( [ eq16 ] ) corresponds to that given in ref .
@xcite , and is symmetric ; namely , exchanging 1 and 2 ( or 10 and 20 ) on the subscripts of variables in this equation yields the same equation .
this means that two bubbles have the same eigenfrequencies .
the solutions of eq .
( [ eq16 ] ) are @xmath65 those solutions can be rewritten approximately as follows in the case where the distance @xmath66 is large enough and @xmath67 : @xmath68\omega _ { 10}^2\ ] ] and @xmath69\omega _ { 20}^2 .\ ] ] for @xmath70 , the former and the latter one converge to @xmath71 and @xmath72 , respectively .
as the bubbles approach each other , the former increases and the latter decreases ; namely , the higher eigenfrequency becomes higher and the lower one becomes lower . in the case of @xmath73 and large @xmath66 , the solutions ( [ eq18 ] ) are reduced to @xmath74\omega _
{ 20}^2\ ] ] and @xmath75\omega _ { 10}^2.\ ] ] same as the result for @xmath67 , the higher eigenfrequency becomes higher and the lower one becomes lower as @xmath66 decreases .
the solution of eq .
( [ eq17 ] ) given for the first time is @xmath76 this converges to @xmath72 for an infinite @xmath66 and increases with decreasing @xmath66 .
in contrast with eq .
( [ eq16 ] ) , eq .
( [ eq17 ] ) is asymmetric ; namely this serves to break the symmetry of the zabolotskaya s result mentioned above .
the present results show that , when the viscous effect of the surrounding liquid is negligible and the radii of the bubbles are not equivalent , the bubbles have three asymmetric eigenfrequencies ; one is the `` fundamental eigenfrequency '' that converges to the partial resonance frequency of a corresponding bubble for an infinite @xmath66 , and the remaining two are the `` subeigenfrequencies '' that converge to the resonance frequency of a neighboring bubble .
one of the subeigenfrequencies always increases as bubbles approach each other .
the other subeigenfrequency decreases ( increases ) and the fundamental eigenfrequency increases ( decreases ) when the partial resonance frequency of the bubble is higher ( lower ) than that of the neighboring bubble .
now we discuss briefly the case of identical bubbles .
when two bubbles have the same radii and pulsate in phase , eq .
( [ eq8 ] ) is reduced to @xmath77 this equation has only one eigenfrequency @xcite of @xmath78 which converges to @xmath71 for infinite @xmath66 and decreases with decreasing @xmath66 . using this example
, we here carry out quantitative comparison between the model used here and previous ones . in ref .
@xcite , kobelev and ostrovskii discussed the eigenfrequency of two interacting bubbles by means of a mathematical model in which the surface tension and viscosity are neglected and the potential field forming between the bubbles is accurately described by the method of images .
the model predicts only two eigenfrequencies when @xmath79 , as in zabolotskaya s model , and gives @xmath80 when @xmath81 and both bubbles make contact with each other , i.e. , @xmath82 . in ref .
@xcite , zavtrak also discussed the eigenfrequency of the present case by means of a mathematical model in which multiple scattering between two bubbles is taken into account by the legendre polynomial expansion . in the case of @xmath82 ,
zavtrak s model gives @xmath83 , which is in close agreement with the result obtained by kobelev and ostrovskii .
however , eq . ( [ eq24 ] ) gives @xmath84 which is lower by about 2 % than these values .
this quantitative discrepancy is caused from the difference of functions for describing the scattered sound wave by a bubble .
the amplitude of the scattered wave is assumed to be inversely proportional to @xmath66 in the model used here ( see eq .
( [ eq2 ] ) ) , while those in the models in refs .
@xcite are approximated by an expansion function .
namely , the amplitude of the scattered wave at the position of a neighboring bubble is estimated differently .
the influence of this difference becomes stronger as @xmath66 decreases .
however , all models predict the same result in a qualitative sense , which reveals the downward shift of the eigenfrequency .
figure [ fig1 ] shows the eigenfrequencies as a function of @xmath85 , in the case where the viscous effect is negligible . here ,
the parameters are set to @xmath86 kg / m@xmath87 , @xmath88 atm , @xmath89 n / m , @xmath90 , and @xmath91 . equations ( [ eq18 ] ) and ( [ eq23 ] ) are used in plotting those graphs since eqs .
( [ eq19])@xmath92([eq22 ] ) provide a less - accurate result for small @xmath66 .
the radius of bubble 1 is fixed to 10 @xmath16 m and that of bubble 2 varies from 5 @xmath16 m to 15 @xmath16 m .
the eigenfrequencies shown in those graphs are normalized with the partial resonance frequency of bubble 1 , @xmath93 .
as discussed above , three eigenfrequencies changing with the distance between bubbles appear in each graph except for the case of @xmath81 where only one decreasing eigenfrequency appears .
it should be noted that the larger bubble , whose partial resonance frequency is lower than that of the smaller bubble , has the highest eigenfrequency among those of a pair of bubbles .
this result is inconsistent with the opinion represented in refs .
@xcite in which it is assumed that the effective resonance frequency of the smaller bubble rises faster than that of the larger bubble when two bubbles approach each other .
now we present numerical solutions of eq .
( [ eq15 ] ) for examining the influences of viscosity on eigenfrequency .
figures [ fig2]@xmath94[fig4 ] show the results obtained by using @xmath95 kg/(m s ) that corresponds to the viscosity of water at room temperature . from those figures
, we can observe that , as the viscous effect grows strong , i.e. , the mean radii of the two bubbles become smaller , the subeigenfrequencies vanish gradually from the large - distance region ( and sometimes from the small - distance region ) and only the fundamental eigenfrequency remains . in the case of @xmath96 @xmath97 m , the subeigenfrequencies of a larger bubble disappear , and when the sizes of two bubbles are close each other the fundamental eigenfrequency and the higher one of subeigenfrequencies of a smaller bubble vanish in the small - distance region .
in the case of @xmath98 @xmath97 m , it is difficult to distinguish the subeigenfrequencies from the fundamental one since only a smooth curve , which decreases monotonically with decreasing @xmath99 , appears in the graphs .
( the eigenfrequency of the smaller bubble remaining in the small - distance region may be the lower one of the subeigenfrequencies . )
this result means that in a high - viscosity case the effective resonance frequencies of both bubbles decrease monotonically as @xmath99 decreases .
these results show that the influence of the mutual interaction on the eigenfrequencies depends on the viscosity of the surrounding material and is weakened as the viscous effect becomes stronger , i.e. , the threshold of the distance for the appearance of the subeigenfrequencies is shortened . from the results shown in fig .
[ fig2 ] , we can expect that when @xmath100 @xmath97 m , @xmath101 @xmath97 m , and @xmath102 , for example , the signs of the primary bjerknes force acting on bubbles 1 and 2 change twice as the two bubbles approach each other since the curves shown in the figure for this case close with the line @xmath103 ( @xmath104 ) at two points .
we can expect also from the results shown in fig .
[ fig3 ] that when @xmath96 @xmath97 m , @xmath105 @xmath97 m , and @xmath106 , the sign of the force acting on bubble 1 changes once during the approach since the curves for this case close with the line @xmath107 at one point .
such an influence of the mutual interaction on the primary bjerknes force is discussed in sec .
[ sec3 ] .
next , we briefly discuss the amplitude of the bubble pulsation . by substituting eqs .
( [ eq12 ] ) and ( [ eq13 ] ) into eq .
( [ eq11 ] ) and rearranging it , one obtains @xmath108 when @xmath109 is far apart from the partial resonance frequencies of both bubbles and the viscous term is negligible , eq .
( [ eq25 ] ) is reduced to @xmath110 furthermore , neglecting the second order term of @xmath111 reduces this to @xmath112 ^ 2 \left ( { \frac{1}{{l_1 } } } \right)^2 \left ( { \frac{{p_a}}{{r_{10 } \rho } } } \right ) ^2 .\end{aligned}\ ] ] as @xmath66 decreases , the amplitude @xmath113 increases when @xmath114 , and decreases when @xmath115 .
physically , this result can be understood as follows : when @xmath116 is sufficiently larger than @xmath117 , the scattered wave due to bubble 2 oscillates in phase with the external sound . in this case , the amplitude of the driving pressure acting on bubble 1 increases ( see eq .
( [ eq8 ] ) ) ; thus , the pulsation amplitude of bubble 1 is increased .
in contrast , when @xmath118 is sufficiently smaller than @xmath109 , the scattered wave due to bubble 2 oscillates out of phase with the external sound ; this results in the decrease of the pulsation amplitude .
now we consider the case where two bubbles have the same radii and the frequency of the external sound is equivalent to their partial resonance frequencies , i.e. , @xmath119 and @xmath120 .
this setting reduces eq .
( [ eq25 ] ) to @xmath121 this equation shows that the amplitude of the bubble pulsation decreases with decreasing @xmath66 .
this result can be explained as follows : the decrease of the eigenfrequency represented in eq .
( [ eq24 ] ) causes the difference between the sound frequency and the effective resonance frequency ; resultantly , the amplitude of pulsation decreases .
such a decrease in the pulsation amplitude of identical two small near - resonance bubbles ( @xmath122 @xmath97 m ) in a viscous liquid is shown in ref .
@xcite by employing a direct - numerical - simulation technique .
the present theory may explain the numerical result .
in this section , we show numerical results obtained by using the full system of nonlinear equations [ ( [ eq4 ] ) and ( [ eq5 ] ) ] and sufficiently small amplitude of sound in order to verify the linear theory given in the last section .
we investigate the variations in eigenfrequencies by observing the primary bjerknes force which is a time - averaged force acting on a bubble when the external sound field has a spatial gradient .
the primary bjerknes force @xcite is expressed as @xmath123 where @xmath124 denotes the time average over a period of the external sound . as was mentioned previously
, the phase difference between the external sound and the bubble pulsation becomes @xmath0 when the sound frequency is equivalent to the resonance frequency of a bubble .
in this situation , furthermore , the primary bjerknes force vanishes as the result of the inner product by eq .
( [ eq27 ] ) .
this means that we can observe the eigenfrequencies of interacting bubbles by examining the conditions under which the magnitude of the force becomes zero . here
we assume that the external standing wave field is in a form of @xmath125 with @xmath126 , where @xmath127 is the spatial coordinate vertical to the pressure nodes , @xmath128 is the amplitude of the standing wave field at @xmath127 , and @xmath129 is the wave number of the standing wave field .
this assumption yields @xmath130 where @xmath131 is the unit vector in the @xmath127 direction .
substitution of eq .
( [ eq28 ] ) into eq .
( [ eq27 ] ) results in @xmath132 with @xmath133 where @xmath134 is the position of bubble @xmath7 in the @xmath127 coordinate .
the method for computing the force is the same as what used in refs .
the full system of nonlinear equations of ( [ eq4 ] ) and ( [ eq5 ] ) is solved numerically by the 4th - order runge - kutta method , and @xmath135 is then computed with @xmath136 given by the numerical computation .
the time average in eq .
( [ eq29 ] ) is done over a sufficiently large period after the transition has decayed . here
we assume that the wavelength of the external sound is much larger than the distance between the bubbles , and thus , @xmath137 .
2 figures [ fig5]@xmath94[fig7 ] show @xmath138 curves for the cases of @xmath139 @xmath97 m , @xmath140 @xmath97 m and @xmath141 @xmath97 m , respectively . here
the system of eqs .
( [ eq4 ] ) and ( [ eq5 ] ) is used ; namely , the compressibility of the surrounding material is neglected .
the amplitude of the external sound at the bubble position is assumed as @xmath142 , @xmath143 and @xmath144 for @xmath139 @xmath97 m , @xmath145 @xmath97 m and @xmath146 @xmath97 m cases , respectively .
the frequency @xmath109 of the external sound is set to @xmath147 and @xmath148 for @xmath139 @xmath97 m , @xmath149 and @xmath150 for @xmath140 @xmath97 m , and @xmath151 for @xmath152 @xmath97 m . in those graphs
, we can observe some points where the primary bjerknes force vanishes .
the normalized distances at the points , given theoretically by solving eq .
( [ eq15 ] ) as a function of @xmath66 with the given @xmath109 , are listed in the captions of those figures .
these theoretical values are in good agreement with the numerical ones represented in those graphs .
this result confirms our prediction that the sign of the primary bjerknes force changes at distances where one of the eigenfrequencies becomes equivalent to the sound frequency .
in the present study , the dependency of the eigenfrequencies of two mutually interacting bubbles on the distance between the bubbles was theoretically determined concretely and in detail .
the dependency is more intricate than those predicted or assumed previously by several authors @xcite .
it was shown also that the eigenfrequencies depend also on the viscosity of the bubbles surrounding material , unlike the partial natural frequency . moreover ,
the present theory was verified numerically by observing the primary bjerknes force acting on bubbles .
these present results may be reasonable for understanding the reversal of sign of the secondary bjerknes force @xcite and the pattern formations of bubbles or clusters of them in an acoustic field @xcite . as was pointed out in sec .
[ sec2a ] , zabolotskaya s theory predicts that both bubbles interacting with each other have the same eigenfrequencies .
this result may give rise to simultaneous phase changes of the bubbles , e.g. , from in - phase to out - of - phase with the external sound , and , as a result , may deny the reversal of the sign .
the asymmetric eigenfrequencies derived in the present paper may be able to describe the reversal of the sign more accurately .
( this subject will be discussed in a future paper . ) furthermore , these present results will also affect understandings on some related subjects such as acoustic localization @xcite and superresonances @xcite , which were investigated by employing a system containing only identical bubbles .
y. a. kobelev , l. a. ostrovskii , and a. m. sutin , jetp lett .
* 30 * , 395 ( 1979 ) .
e. a. zabolotskaya , sov .
. acoust . * 30 * , 365 ( 1984 ) .
a. a. doinikov and s. t. zavtrak , phys .
fluids * 7 * , 1923 ( 1995 ) .
a. a. doinikov and s. t. zavtrak , j. acoust .
99 * , 3849 ( 1996 ) .
i. akhatov , u. parlitz , and w. lauterborn , j. acoust .
96 * , 3627 ( 1994 ) ; w. lauterborn , t. kurz , r. mettin , and c. d. ohl , adv .
phys . * 110 * , 295 ( 1999 ) .
r. mettin , i. akhatov , u. parlitz , c. d. ohl , and w. lauterborn , phys .
e * 56 * , 2924 ( 1997 ) .
a. a. doinikov , phys .
e * 62 * , 7516 ( 2000 ) .
w. lauterborn , j. acoust .
59 * , 283 ( 1976 ) .
j. b. keller and m. miksis , j. acoust .
68 * , 628 ( 1980 ) .
a. kobelev and l. a. ostrovskii , sov .
* 30 * , 427 ( 1984 )
. s. t. zavtrak , sov .
. acoust . * 33 * , 145 ( 1987 ) .
m. ida and y. yamakoshi , jpn .
. phys . * 40 * , 3846 ( 2001 ) .
a. i. eller , j. acoust .
43 * , 170 ( 1968 ) ; l. a. crum and a. i. eller , j. acoust .
48 * , 181 ( 1969 ) .
z. ye and a. alvarez , phys .
* 80 * , 3503 ( 1998 ) . c. feuillade , j. acoust .
98 * , 1178 ( 1995 ) . | eigenfrequencies of two mutually interacting gas bubbles in an acoustic field are discussed theoretically and numerically .
it is shown by a linear theory that a bubble interacting with a neighboring bubble has three eigenfrequencies that change with the distance between two bubbles , and the sign and magnitude of the primary bjerknes force acting on bubbles also change according to the change in the eigenfrequencies . 2 |
one of the first , and serendipitous , discoveries of the infrared astronomical satellite ( _ iras _ ) was excess infrared emission from circumstellar dust around main sequence stars @xcite .
these disks are present well after the t tauri phase of stellar evolution and its associated massive accretion disk .
the lifetime of dust grains around such stars is short compared to the ages of these stars and their detection is thus indirect evidence for the presence of larger source bodies that resist poynting - robertson drag , perhaps analogous to asteroids or comets .
our solar system contains a dust cloud ( the zodiacal cloud ) that reprocesses @xmath1 of the sun s radiation @xcite , re - emitting it with a spectral energy distribution similar to that of a 260 k black - body @xcite .
the dust originates in collisions between asteroids and the disintegration of short - period comets .
the early solar system may have contained larger numbers of minor bodies and hence elevated levels of dust .
ancient cratered surfaces in the inner solar system record intense bombardment that occured within 800 million years ( myr ) of planetary accretion @xcite .
such impacts influenced the evolution of inner planet surfaces , atmospheres and volatile inventories @xcite .
these impactors may have appeared suddenly after the breakup of a larger body , or belonged to a population with a long dynamical lifetime , perhaps in the outer solar system @xcite .
gaidos ( 1999 ) used a model of comet - like impactors scaled to the lunar cratering record to show that an early zodiacal dust cloud with @xmath2 was possible if poynting - robertson drag was the dominant removal mechanism .
circumstellar dust is thus a potential probe of the dynamics of extrasolar planetary systems and their impactor history
. the scattered or reprocessed light from residual dust disks around young main sequence stars can be ( and have been ) detected , usually at mid - infrared wavelengths where the contrast ratio with the stellar photosphere is greatest .
_ iras _ detected excess 12 - 60 @xmath0 m emission from dust around many main - sequence a - k stars @xcite .
searches around solar - type ( g and early k ) stars were hampered by the sensitivity limits of the _ iras _ telescope .
an analysis of _ iras _ observations of a sample of single , nearby ( @xmath3 pc ) young ( @xmath4 myr ) solar analogs selected on the basis of their high coronal x - ray emission and chromospheric activity @xcite found excess emission from only one star , however this is now known to be the result of source confusion with a background giant star ( c. beichmann , private communication ) .
_ iras _ detection limits around these stars are typically @xmath5 , or @xmath6 `` zodis '' . a slightly more sensitive search was carried out at 25 - 170 @xmath0 m with the infrared space observatory ( _ iso _ ) ( habing _ et al .
that study found that detectable ( @xmath7 ) levels of dust are present around half of stars younger than 400 myr but only 10% of older stars .
three of five young ( @xmath4 myr ) , solar - type stars in that survey have significant excess emission , including one star ( hd 30495 ) in the gaidos ( 1998 ) sample .
this suggests that yet more circumstellar disks will be found around young main sequence stars at lower detection threshholds .
ground - based detection of dust disks in the infrared is very difficult . at a wavelength of 10 microns ,
the resolution of even the largest telescopes is equivalent to a semimajor axis of 1.3 au at 10 pc , limiting imaging searches for dust to the nearest few bright stars @xcite .
ground - based infrared photometry is notoriously inaccurate and the absolute mid - ir fluxes of stars can not be precisely predicted , rendering such an approach inutile in improving the observational status quo .
the flux density from a hypothetical 260 k dust cloud with @xmath8 around a g5 star at 10 pc is 30 mjy at 11.7 @xmath0 m , representing only 6% of the photospheric signal at this wavelength .
a 3@xmath9 detection would require 2% photometric accuracy , extremely challenging in the mid - infrared .
alternatively , one can use mid - infrared to search for the 10 @xmath0 m emission feature produced by the si - o stretch vibration in crystalline silicate dust .
silicate emission features comparable to the continuum emission have been detected from @xmath10 pictoris and young early - type stars with circumstellar disks @xcite and resemble those observed in comets @xcite . in principle , spectroscopy is more precise than photometry because it is sensitive to the wavelength dependence of the emission within the band - pass and insenstive to the absolute total value .
the marked features associated with crystalline silicate emission are more easily discernable than continuum emission with a planck or power - law spectral energy distribution ( sed ) ( fig .
[ fig.halebopp ] ) .
in addition , the intensity and shape of the emission feature can provide information about the mineralogy of the dust grains @xcite .
we obtained n - band ( 8 - 13 @xmath0 m ) spectroscopy of several nearby young stars using the keck 1 telescope and the long wavelength spectrometer ( lws ) to search for silicate emission from circumstellar dust .
our approach was to correct for the effects of the atmosphere , telescope , and detector combination by comparing target and calibrator stars .
the ideal calibrator star would , of course , be completely dust free .
however , the presence or absence of dust around most nearby stars has not been established at a precision comparable to that of our current observations .
thus , while we observed older calibration stars in the expectation that they would be less likely to harbor significant dust clouds , our conclusion will be limited to statements about the amount of silicate emission flux from pairs of stars .
most of the targets ( table [ table.observations ] ) are young solar analogs from the updated , volume - limited version of the gaidos ( 1998 ) catalog of nearby young solar analogs .
these single or wide - binary stars are all within 25 pc , have spectral types g0-k2 , and coronal x - ray luminosities suggesting ages less than 800 myr .
followup spectroscopy and photometery of these objects have confirmed marked chromospheric activity , high lithium abundance , and rapid rotation , all indicative of stellar youth @xcite .
two other stars in the sample , hd 17925 and hd 77407 were identified as active solar - type stars by montes et al .
. we also observed vega ( @xmath11 lyra ) and the calibrator a stars altair ( @xmath11 cygnus ) and castor ( @xmath11 gem ) .
vega is the archetypal _
iras_-detected infrared excess star and its fluxes at 25 and 60 @xmath0 m exceed those predicted for its photosphere .
circumstellar dust may also have been resolved at sub - millimeter wavelengths @xcite .
no excess has been detected at 12 @xmath0 m to the _ iras _ calibration accuracy of 5% @xcite .
however , ciardi et al . (
2001 ) have reported resolving the stellar disk of vega with interferometry at 2.2 @xmath0 m and suggest that 3 - 6% of the flux is emanating as scattered or reprocessed starlight from circumstellar dust within 4 au of the star .
as equilibrium black - body temperatures at those distances exceed 380 k , this material would be expected to produce a signal at mid - infrared wavelengths as well .
observations were carried out with keck 1 and long wavelength spectrometer ( lws ) in low - resolution spectroscopic mode during two runs ; 2 september 2002 , and 18 - 19 march 2003 ( table [ table.observations ] ) .
the gold - coated secondary has a speed of f/25 and the lws was mounted on the forward cassegrain module of the telescope .
the lws is described in detail by jones & puetter ( 1993 ) .
briefly , the lws detector is a boeing 128@xmath12 pixel si : as array with 75 micron pixels read out by amplifiers with a co - adder inverse gain of 360 e@xmath13 dn@xmath14 at 12 bit accuracy .
the low - resolution grating was used ( dispersion of 0.037 @xmath0 m pixel@xmath14 at the detector ) with a 6-pixel ( 0.5 arc - second ) slit and an n - wide ( 8.1 - 13.0 @xmath0 m ) filter .
the detector is oriented such that rows are approximately parallel to the direction of dispersion .
multiple frames for each star were taken in the usual four telescope states associated with ground - based observations in the mid - infrared , with the chopping secondary and telescope nod set to their on - source and off - source positions . typical chop amplitude and frequency were 10 arc - sec and 2 hz . preprocessing of the spectral frames consisted of subtraction of the average of the ( chop off , nod on ) frames in each file from the average of the ( chop on , nod on ) frames to produce a chop - subtracted nod - on frame .
similarly , the average of the ( chop off , nod off ) frames in each file was subtracted from the average of the ( chop on , nod off ) frames to produce a chop - subtracted nod - off frame . subtracting the nod - off frame from the nod - on frame , dividing by the flatfield , and fixing
the bad pixels produced a stellar spectrum frame suitable for spectral extraction .
a dark current image was constructed by averaging multiple frames with the same exposure time and subtracted from each image .
`` hot '' pixels were identified in spectra of a warm stop that was placed in the beam and are removed from further analysis .
care was taken during the extraction of the stellar spectra to account for the distortion of the image plane , which causes the wavelength corresponding to each column of pixels to differ from row to row , and the file - to - file shifts of the spectra caused by imperfect setting of the grating and pointing of the telescope .
it was important that these effects be corrected before dividing each target spectrum by a calibrator spectrum to avoid producing spurious spectral features , especial in the region of deep atmospheric absorption features .
the wavelength calibration was done using bright telluric emission features in mid - infrared .
an airglow frame was constructed by averaging all the frames in a file ( including both chop and nod positions ) , subtracting the averaged dark frame , dividing by the flatfield frame , and fixing bad pixels .
( the sky emission is much brighter than the stars so the airglow frame was not affected by the presence of the stellar spectrum ) .
the telluric lines are visible as gently curving vertical stripes in the airglow frame .
a narrow horizontal strip across the middle of one of the airglow frames provided a template spectrum for the airglow .
this was stretched vertically to produce an airglow template frame with the same dimensions as the frames from which the stellar spectra were to be extracted .
the airglow lines are nearly vertical in the airglow template frame .
the airglow frame derived from each file was numerically undistorted to fit the airglow template frame , and the same undistortion parameters were applied to the calibrated stellar spectrum frame for that file .
this produced a set of calibrated stellar spectrum frames with consistent wavelength scaling .
because the alignment between the target and calibrator spectra uses sky emission , spectral misalignment can still occur if the centroids of the target and calibrator stars are offset within the slit .
the slit width ( 6 pixels ) is equivalent to 0.22 @xmath0 m and thus significant misalignment is possible .
we mitigate this effect by performing a secondary alignment between the target and calibrator spectra using a weighted cross - correlation in a 0.8 @xmath0m - wide window centered on the 9.6 @xmath0 m ozone feature .
( this portion of the spectrum is not used for subsequent spectral analysis ) .
typical offsets were 0.5 pixels ( 0.019 @xmath0 m ) ( uncertainties in offset are accounted for in [ sec.analysis ] ) a spectrum was extracted from a narrow strip containing the starlight that was projected along columns .
each spectrum was normalized by the average of all columns ( channels ) .
a final spectrum was generated by averaging over the spectra for each star .
an error spectrum was constructing by calculating , channel by channel , the variance of the summed spectra .
finally , a calibrated science spectrum was generated by taking the ratio of the normalized target and calibration spectra .
the errors in each spectral channel was calculated from the target and calibration spectrum errors added in quadrature .
representative spectra are shown in fig .
[ fig.hd166 ]
. a dust - free calibrated spectrum should be essentially flat .
differences in photosphere temperature equivalent to an entire spectral class would induce a slope of less than 1% across the entire n - band .
in none of the spectra is silicate emission obvious , i.e. , at the level of 20% of the photosphere or larger .
we used a matched filter / least - squares fit technique to search for lower levels of silicate emission from a target star . each calibrated spectrum @xmath15 ( where unity has been subtracted ) was compared with a template spectrum of silicate emission @xmath16 .
we derived the template from the mid - infrared spectra of comet hale - bopp near perihelion reported by hayward et al .
the comet exhibited pronounced double - peaked silicate emission with respect to a continuum that can be described by a 390 k blackbody ( fig .
[ fig.halebopp ] ) .
crystalline olivine , glassy olivine , and crystalline pyroxene may all contribute to the complex structure of the emission .
the relatively narrow , intense peak and the hot color temperature with respect to the blackbody prediction ( `` superheat '' ) may reflect a significant contribution by submicron particles to the mid - infrared flux @xcite . to carry out a linear least - squares fit , we construct the template spectrum from the comet spectrum by multiplying the latter by a small number @xmath17 ( here 0.01 ) at 10 @xmath0 m .
the comet spectrum is added to and divided by a 5600 k blackbody spectrum ( representing typical target and calibrator star ) , the resulting template normalized by its average , and unity subtracted .
the least - squares fit amplitude of the silicate emission signal at 10 @xmath0 m ( @xmath18 ) in a calibrated spectrum is given by the ratios ; @xmath19 where the sums are over all channels .
the factor of 0.75 in the left - hand denominator is the average of the silicate emission spectrum relative to its 10 @xmath0 m value .
a @xmath20 value is calculated as follows ; @xmath21 ^ 2\left[\sigma_i t_i^2/\sigma_i^2\right]^{-1}\ ] ] the accuracy and robustness of our analysis were tested by taking one star as both target and calibrator and artificially adding a hale - bopp signal with a specified amplitude to the former .
we then compare the original and recovered amplitudes .
[ fig.recover ] shows that the signal recovered from actual spectra is proportional to and close to the actual signal .
of course , any excess flux with an sed different than that of the stellar spectrum will also produce a signal .
the recovery efficiency ( recovered signal / actual signal ) for excess emission with a blackbody spectrum is plotted in fig . [ fig.bb ] using the star hd 20630 as both target and calibrator .
the matched filter is most sensitive to 170 - 190 k blackbody , although the @xmath20 will be relatively high .
sensitivity to hotter blackbodies is reduced as the spectrum becomes more raleigh - jeans - like and resembles that of the star .
below 150 k the exponential shape of the sed makes any template fit meaningless and regardless , we do not expect a significant signal in the n bandpass from dust at these temperatures .
a detection limit was calculated for each calibrated spectrum by repeating the least - squares fit with @xmath22 artificial spectra generated by adding gaussian - distributed noise ( with errors as calculated above ) to a perfect spectrum ( unity ratio ) .
systematic errors created by the imperfect wavelength alignment between the source and calibration spectra must also be considered .
this error will be proportional to the local slope of the observed spectrum and will thus be larger in regions where there are strong features , e.g. , the ozone absorption line .
we add these errors to the artificial target spectra by introducing a gaussian - distributed misalignment with mean of zero and @xmath23 pixels before dividing the relevant stellar spectrum by itself .
the least - squares fit amplitudes generated from the artificial spectra are sorted and the 99.97% confidence level ( 3@xmath9 for gaussian - distributed errors ) found .
where appropriate , we estimated or put limits on the excess flux in the _ iras _ 12 @xmath0 m channel by fitting a theoretical sed to visual and near - infrared photometry and comparing the predicted mid - ir fluxes to _ iras _ photometry .
visual ( 0.55 @xmath0 m ) magnitudes were obtained from the _ hipparcos _ catalog @xcite .
near - infrared magnitudes in the @xmath24,@xmath25 , and @xmath26 passbands ( 1.24 , 1.66 , and 2.16 @xmath0 m ) measured by the two micron all sky survey ( 2mass ) were extracted from the infrared science archive ( irsa ) using the gator query engine @xcite .
additional near - infrared measurements were obtained from the catalog of infrared observations , version 4.1 @xcite .
most stars also have 12 @xmath0 m fluxes reported in source catalogs generated from observations by the infrared astronomical satellite ( _ iras _ ) @xcite .
the _ iras _ fluxes reported in the fsc are calculated based on a @xmath27 sed and must be color - corrected for stars . as in gaidos ( 1999 ) and others , we correct all fluxes to a 5600 k blackbody : the color corrections for 12 and 25 @xmath0 m are 0.687 and 0.714 , respectively .
analysis was as described in gaidos ( 1999 ) .
kurucz ( 1992 ) spectra for solar metallicity and temperatures of 4750 , 5000 , 5250 , 5500 , 5750 , and 6000 k were linearly interpolated with 100 mesh points between each original spectrum .
weighted least - squares fitting to the visible and near - infrared photometry was performed and the best - fit model used to generate predicted fluxes at _ iso _ and _ iras _ wavelengths .
errors in these predictions were calculated by a standard monte carlo approach using the errors in the photometry .
the significance of an excess was assessed by comparing it with the independent errors in the prediction and photometry added in quadrature .
our lws detections and upper limits were expressed in terms of the equivalent flux in the 12 @xmath0 m _ iras _ channel . the conversion factor ( 0.433 ) between the fractional amplitude of a silicate emission feature at 10 @xmath0 m to the fractional _ iras _ flux at 12 @xmath0 m ( corrected to a 5600 k blackbody )
was found by direct integration of the hale - bopp sed over the _
iras _ response curve .
for example , an infrared excess with the spectrum of hale - bopp and a peak amplitude of 10% of the photosphere at 10 @xmath0 m would create an excess color - corrected flux of 4.3% of the photosphere in the _ iras _
12-micron channel .
one star ( hd 17925 ) in the observed sample has an n - band spectrum that significantly deviates from that of the calibrator star ( hd 20603 ) .
the calibrated spectrum and the silicate emission template scaled to the detection amplitude ( 7.7% of the photosphere at 10 @xmath0 m ) are shown in fig .
[ fig.hd17925 ] .
hd 17925 is a nearby ( 10.4 pc ) active k1 dwarf star identified as a young main - sequence star by montes et al .
habing et al .
( 2001 ) report a marginally significant ( 3.1@xmath9 ) excess at 60 @xmath0 m from _ iso _ observations but no excess was detected at 25 @xmath0 m ( laureijs et al .
the 2mass jhk images of this star were saturated and we used instead 4 measurements from the catalog of infrared observations version 5.1 ( http://ircatalog.gsfc.nasa.gov ) .
photometric errors are not given in the catalog ; we assume 3% precision .
least - squares fitting of an interpolated ( 5090 k ) kurucz model to the vjhk photometry and comparison with _ iras _ and _ iso _ photometry closely reproduces the _ iso _ 60 @xmath0 m excess reported by habing et al .
( @xmath28 mjy ) .
while there is a marginal ( 2.2@xmath9 ) excess of 50 mjy at 25 @xmath0 m , there is no significant excess at 12 @xmath0 m .
the _ iras _ measurement is still insufficiently precise at this wavelength ( @xmath29 mjy ) compared to the expected excess produced by the spectroscopic feature ( 23 mjy ) .
we note that hd 20630 ( the calibrator star ) is also relatively young but does not have _
iras_-detected excess emission . as is customary
, we convert these limits into an optical depth . assuming the template silicate spectrum describes the actual sed of the infrared excess the ratio of the integrated excess signal to the photosphere within the n - band is about 0.057 .
a k1 star emits about 0.13% of its energy in the n band , therefore this detected excess represents @xmath30 . of course ,
if this excess is dust hotter than 300 k , it will contribute a similar amount ( relative to the photosphere ) at longer wavelengths , but inclusion of this effect increases @xmath31 only to @xmath32 , or about 1000 times that of our solar system s zodiacal cloud . for other stars we can place @xmath33 limits on the amount of silicate emission of about 3% of the photosphere at 12 @xmath0 m , a factor of ten below that of _ iras _ and equivalent ( per the calculations above ) to @xmath34 .
of course , these limits are only valid for a dust sed like that of of hale - bopp : our investigation is less sensitive to continuum emission but nonetheless could detect dust at the level of @xmath35 .
our lack of detections suggest that young main - sequence stars possessing warm ( @xmath36200 k ) circumstellar dust with @xmath7 are the exception rather than the rule .
the launch of the _ space infrared telescope _ ( _ sirtf _ ) , currently scheduled for 23 august 2003 and the advent of nulling imaging at the keck interferometer will allow us to probe with still greater sensitivity the dust contents of the circumstellar environments of nearby solar - type stars .
in two of the three observations , high signal - to - noise spectra of vega calibrated by those of altair produced a calibrated spectrum with a negative slope of about -3% across the entire n bandpass .
while the spectrum of castor calibrated by altair also has a similar slope , the spectrum of vega calibrated by castor appears flat .
one possible explanation is a difference in the sed of the photosphere of altair , an a7 star , compared to those of the other two stars ( which have spectral types a0 and a2 ) .
normalization of a 9650 k blackbody ( vega ) by a 8200 k blackbody ( altair ) generates a slope of only 0.8% across the 8 - 13 @xmath0 m range , however stars do deviate from the ideal black - body even in the mid - infrared .
another , less likely , explanation is that altair has a circumstellar infrared excess undetected by _
iras_. the report of a circumstellar signal of 3 - 6% of vega at 2.2 @xmath0 m @xcite must be reconciled with the absence of significant excess emission at n - band wavelengths .
a nominal isothermal blackbody dust model predicts that for a dust albedo of 0.1 ( backman + & paresce 1993 ) and a temperature of 1500 k , the excess flux at 12 @xmath0 m will be @xmath367 times that at k - band .
the _ iras _ limit of 5% thus constrains the k - band excess to @xmath371% .
cooler dust temperatures make the constraint even tighter
. our detection limit of about 0.5% for silicate emission from vega crudely translates into a 3% limit on the 12 @xmath0 m excess for 1500 k blackbody emission , restricting the k - band excess to less than 0.4% .
( the limits on cooler dust emission are lower still ) . two possible explanations for reconciling the observations of ciardi et al .
( 2001 ) with the mid - infrared observations are that the dust albedo is close to unity , implausible for hot dust , or that the dust particles are micron - sized or smaller and poor emitters at mid - infrared wavelengths .
our lack of detection of a silicate emission peak from such dust particles suggests that the second explanation is also untenable .
the authors are grateful to all people , and all peoples , that have made it possible for the summit of mauna kea to be a premier site of human scientific inquiry .
we thank the personnel of the william keck observatory , particularly randy campbell , for courteous and professional assistance in the observations .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
the simbad database , maintained by the centre de donnes astronomiques de strasbourg ( cds ) , and nasa s astrophysics data systems bibliographic services were used also extensively .
aumann , h. h. , beichmann , c. a. , gillett , f. c. , , dejong , t. , houck , j. , low , f. j. , neugebauer , g. , walker , r. , wesselius , p. r. , 1984 , , 278 , 23 .
backman , d. e. , paresce , f. , 1993 , in levy , e. h. , lunine , j. i. matthews , m. s. ( eds . ) protostars and planets iii , univ .
arizona press , tucson .
chyba , c. f. , thomas , p. j. , brookshaw , l. , sagan , c. , 1990 , science , 249 , 366 .
ciardi , d. r. , van belle , g. t. , akeson , r. l. , thompson , r. r. , lada , e. a. , howell , s. b. , 2001 , , 559 , 1147 .
fajardo - acosta , s. b. , telesco , c. m. , knnacke , r. f. , 1993 , , 417 , l33 .
fegley , b. , jr . ,
prinn , r. g. , hartman , h. , watkins , g. h. , 1986 , nature , 319 , 305 .
fernandez , j. a. , ip , w .- h .
, 1983 , icarus , 54 , 377 .
gaidos , e. j. , 1999 , , 510 , l131 .
gaidos , e. j. , 1998 , , 110 , 1259 .
gaidos , e. j. , henry , g. w. , henry , s. m. , 2000 , , 120 , 1006 .
gezari , d. y. , pitts , p. s. , schmitz , m. , 1993 , nasa rp-124 , nasa , washington dc .
habing , h. j. , dominik , c. , jourdain de muizon , m. , laurejis , r. j. , kessler , m. f. , leech , k. , metcalfe , l. , salama , a. , siebenmorgen , r. , trams , n. , bouchet , p. , 2001
, , 365 , 545 .
hanner , m. s. , brooke , t. y. , tokunaga , a. t. , 1998 , , 502 , 871 .
hanner , m. s. , tokunaga , a. t. , golisch , w. f. , griep , d. m. , kaminski , c. d. , 1987 , , 187 , 653 .
hayward , t. l. , hanner , m. s. , sekanina , z. , 2000 , , 538 , 428 .
jger , c. , molster , f. j. , dorschner , j. , henning , th . , mutschke , h. , waters , l. b. f. m. , 1998 , , 339 , 904 .
jones , b. , puetter , r. , 1993 , proc .
spie , 1946 , 610 .
knacke , r. f. , fajardo - acosta , s. b. , telesco , c. m. , hackwell , j. a. , lynch , d. k. , russell , r. w. , 1993 , , 418 , 440 .
koerner , d. w. , sargent , a. i. , ostroff , n. a. , 2001 , , 560 , l181 .
kuchner , m. j. , brown , m. e. , koresko , c. d. , 1998 , , 110 , 1336 .
kurucz , r. l. , 1992 , in barbuy , b. , renzini , a. , the stellar populations of galaxies ( proc .
149 ) , kluwer , dordrecht , p. 225 .
laureijs , r. j. , jourdain de muizon , m. , leech , k. , siebenmorgen , r. , dominik , c. , habing , h j. , trams , n. kessler , m. f. , 2002 , , 387 , 293 .
leinert , ch . ,
brahm , p. , acosta - pulido , j. , lenke , d. , siebenmorgen , r. , 2002 , , 393 , 1073 .
levison , h. f. , dones , l. , chapman , c. r. , stern , s. a. , duncan , m. j. , zhnle , k. 2001 , icarus , 151 , 286 .
melosh , h.j . ,
vickery , a.m. , 1989 , nature , 338 , 487 .
montes , d. , lopez - santiago , j. , galvez , m. c. , fernandez - fueroa , m. j. , de castro , e. , cornide , m. , 2001 , , 328 , 45 .
moshir , m. , et al . , 1990 ,
explanatory supplement to the _ iras _ faint source survey , version 2 , jpl d-100015 jpl , pasadena .
neugebauer , g. , soifer , b. t. , beichman , c. a. , aumann , h. h. , chester , t. j. , gautier , t. n. , lonsdale , c. j. , gillett , f. c. , hauser , m. g. , houck , j. r. , 2001 , science , 224 , 14 .
perryman , m. a. c. , and 18 co - authors , 1997 , , 323 , l49 .
roe , h. g. , 2002 , ph.d .
thesis , university of california , berkeley .
sitko , m. l. , grady , c. a. , lynch , d. k. , russel , r. w. , hanner , m. s. , 1999 , , 310 , 408 .
skinner , c. j. , barlow , m. j. , justtanont , k. , 1992 , , 255 , 31 .
skrutskie , m. , 2001 , baas , 198 , 33.01 .
strom , r. g. , 1987 , icarus , 70 , 517 .
waelkens , c. , and 20 co - authors , 1996 , , 315 , l245 .
wilner , d. j. , holman , m. j. , kuchner , m. j. , ho , p. t. p. , 2002
, , 569 , l115 .
llrlll 166 & k0 & 0.53 & 17/09/02 & 1.11 - 1.24 & 960 + 17925 & k1 & 0.69 & 17/09/02 & 1.31 & 480 + 20630 ( @xmath38 ceti ) & g5 & 1.40 & 17/09/02 & 1.04 - 1.06 & 960 + 60178 ( @xmath11 gem)@xmath39 & a2 & 7.22 & 19/09/02 & 1.08 - 1.10 & 360 + 77407 & g0 & 0.20 & 19/03/03 & 1.11 - 1.15 & 480 + 82443 & k0 & 0.27 & 18/03/03 & 1.04 - 1.25 & 720 + & & & 19/03/03 & 1.01 - 1.02 & 600 + 97334 & g0 & 0.25 & 18/03/03 & 1.04 - 1.12 & 792 + 114710 ( @xmath10 com)@xmath39 & g0 & & 19/03/03 & 1.01 - 1.08 & 720 + & & & 20/03/03 & 1.04 - 1.21 & 720 + 130948 & g1 & 0.46 & 18/03/03 & 1.01 - 1.04 & 720 + & & & 20/03/03 & 1.00 - 1.05 & 720 + 172167 ( @xmath11 lyra ) & a0 & 28.53 & 17/09/02 & 1.06 & 720 + & & & 18/03/02 & 1.21 - 1.35 & 360 + & & & 19/03/03 & 1.21 - 1.29 & 600 + 187642 ( @xmath11 aql)@xmath39 & a7 & 22.66 & 17/09/02 & 1.03 - 1.04 & 720 + & & & 18/09/02 & 1.28 - 1.35 & 216 + & & & 19/09/02 & 1.28 - 1.42 & 600 + 190771@xmath39 & g5 & 0.40 & 17/09/02 & 1.09 - 1.14 & 720 + llllll hd 166 & hd 20630 & [ 0.036 ] & [ 0.015 ] & & [ 0.18 ] + hd 17925 & hd 20630 & 0.070(0.10 ) & 0.030(0.04 ) & 1.1 & [ + hd 82443 & hd 114710 & [ 0.058 ] & [ 0.025 ] & & [ 0.27 ] + hd 97334 & hd 114710 & [ 0.053 ] & [ 0.023 ] & & [ 0.23 ] + hd 130948 & hd 114710 & & & & [ 0.18 ] + vega & altair & [ 0.011 ] & [ 0.005 ] & & 0.12@xmath40 + vega & castor & [ 0.012 ] & [ 0.005 ] & & + | we obtained low resolution ( r = 100 ) mid - infrared ( 8 - 13 @xmath0 m wavelengths ) spectra of 8 nearby young main sequence stars with the keck 1 telescope and long - wavelength spectrometer ( lws ) to search for 10 @xmath0 m silicate ( si - o stretch ) emission from circumstellar dust .
no stars exhibited readily apparent emission : spectra were then analyzed by least - squares fitting of a template based on a spectrum of comet hale - bopp . using this technique
, we were able to constrain the level of silicate emission to a threshold ten times below what was previously possible from space .
we found one star , hd 17925 , with a spectrum statistically different from its calibrator and consistent with a silicate emission peak of 7% of the photosphere at a wavelength of 10 @xmath0 m .
excess emission at 60 @xmath0 m from this star has already been reported . |
let @xmath17 be a submanifold of codimension @xmath18 contained in a hypersurface @xmath11 .
for a frame @xmath19 , @xmath4 tangent to @xmath11 and transversal to @xmath8 , @xmath20 transversal to @xmath11 , write @xmath21 where @xmath22 denotes the canonical connection of the affine @xmath2-space , @xmath23 are tangent to @xmath8 and @xmath24 is also tangent to @xmath8 .
we shall assume that the metric @xmath25 is non - degenerate , which is independent of the choice of the frame @xmath19 .
under this non - degenerate hypothesis , there exists a unique , up to multiplication by a scalar function , vector field @xmath4 tangent to @xmath11 and transversal to @xmath8 such that @xmath26 is tangent to @xmath11 , for any @xmath27 tangent to @xmath8 . any such vector field
is called a _
darboux vector field along @xmath8 _
( @xcite ) . for a fixed darboux vector field @xmath4
, one can define an unique affine invariant metric @xmath6 on @xmath8 such that @xmath28=\delta_{ij},\ ] ] for any @xmath6-orthonormal frame @xmath29 , where @xmath30 $ ] denote the canonical volume form in the @xmath2-space ( see @xcite ) . for @xmath20 transversal to @xmath11 , write @xmath31 where @xmath32 is tangent to @xmath8 .
it is proved in @xcite that there exists a vector field @xmath20 satisfying @xmath33 and @xmath34=1,\ ] ] for any @xmath6-orthonormal frame @xmath29 . moreover , any other @xmath35 with the same properties can be written as @xmath36 , for some @xmath37 , and so the plane @xmath38 , @xmath39 , generated by @xmath19 is unique .
the family @xmath40 is called the _
affine normal plane bundle_. a vector field @xmath41 is called _ parallel _ if @xmath42 is tangent to @xmath8 , for any @xmath43 .
some immersions admit a parallel darboux vector field and , when this is the case , we have some reasons to choose it : the transon planes coincide with the affine normal planes and the cubic form @xmath44 is apolar ( @xcite ) . in this paper
we give one more reason for the choice , namely , we prove that the @xmath6-laplacian of the position vector @xmath45 belongs to the affine normal plane if and only if @xmath4 is parallel .
we are particularly interested in the case that the immersion admits a parallel darboux vector field @xmath4 and also a parallel vector field @xmath46 linearly independent of @xmath4 .
we verify that these conditions are equivalent to the flatness of a certain connection @xmath47 , called normal connection .
when these conditions hold , we say that the immersion is _ normally flat_. when @xmath4 is parallel and umbilic , i.e. , @xmath48 is a multiple of the identity , then @xmath5 is a _
visual contour_. this means that there exists a point @xmath49 in the intersection of all tangent spaces of @xmath11 along @xmath8 .
if @xmath46 , linearly independent from @xmath4 , is also parallel and umbilic , then there exists another fixed point @xmath50 that belongs to the intersection of the normal planes @xmath38 , @xmath39 .
thus if the immersion is umbilic and normally flat , there exist fixed points @xmath13 and @xmath50 in the intersection of the tangent spaces and normal planes , respectively .
conversely , if such points @xmath13 and @xmath50 exist , the immersion is necessarily umbilic and normally flat . for @xmath39 ,
let @xmath51 be a @xmath6-orthonormal tangent frame for @xmath8 .
the affine distance function @xmath52 is given by @xmath53.\ ] ] then the singular set of @xmath54 is the affine normal plane @xmath7 ( @xcite ) and the bifurcation set @xmath55 of @xmath54 is called the _ affine focal set _ of the immersion @xmath5 .
we verify that for umbilic and normally flat immersions , the set @xmath55 reduces to a single line .
we also give conditions under which this set is locally a hypersurface of @xmath56 .
we consider in more detail curves @xmath45 contained in surfaces @xmath57 . in this case , the affine focal set @xmath55 is a developable surface @xmath55 in @xmath58 and we describe the stable singularities of @xmath55 , giving conditions under which they are equivalent to a cuspidal edge or a swallowtail .
we verify that there exist parallel vector fields @xmath4 and @xmath20 generating the affine normal plane bundle and so the immersion is normally flat . for curves ,
the umbilic condition must be replaced by the visual contour condition together with the existence of a fixed point @xmath50 in the intersection of the affine normal planes . we characterize the curves satisfying these conditions and apply the result for spatial curves
not necessarily contained in surfaces .
we also discuss two six vertex theorems , one affinely invariant for `` umbilic '' spatial curves and the other projectively invariant for general planar curves . for immersions
@xmath5 such that @xmath8 is contained in a hyperplane @xmath9 , we can choose @xmath4 parallel and @xmath20 the blaschke vector field of @xmath10 , which is also parallel . in this context
we verify that @xmath5 is an umbilic immersion if and only if @xmath10 is an affine sphere and the envelope of tangent spaces of @xmath5 is a cone . for immersions @xmath5 , where @xmath11 is a hyperquadric , we can choose a parallel darboux vector field @xmath4 @xmath59-unitary and orthogonal to @xmath8 , where @xmath59 is the blaschke metric of @xmath11 .
then the vector field @xmath60 , where @xmath50 is the center of @xmath11 , is umbilic , parallel and contained in the affine normal plane . in this context , we show that the immersion @xmath5 is umbilic if and only if @xmath8 is contained in a hyperplane .
the main result of this paper is a geometric characterization of umbilic and normally flat immersions .
consider a non - degenerate immersion @xmath61 and let @xmath62 denote its blaschke co - normal map .
we verify that the centro - affine immersion @xmath63 is umbilic and normally flat , for any choice of the origin @xmath64 .
the converse is also true : any umbilic and normally flat immersion is locally obtained by this construction . as a corollary
we show that a umbilic and normally flat immersion @xmath8 is contained in a hyperplane if and only if @xmath12 is a proper affine sphere .
the paper is organized as follows : in section 2 , we describe the darboux vector fields and other basic facts of the affine geometry of codimension @xmath18 submanifolds contained in a hypersurface . in section 3
we consider the case of parallel darboux vector field and prove that the laplacian of the position vector is in the affine normal plane . in section 4
we discuss conditions under which an immersion is normally flat and umbilic . in section 5
we describe the basic properties of the affine focal set , giving conditions for its regularity . in section 6
we consider curves contained in surfaces of @xmath58 , where the affine focal set coincides with the envelope of normal planes . in this case
, we give a complete classification of the stable singularities that appear in the affine focal set . in section 7
we study the particular cases of submanifolds contained in a hyperplane , when the immersion is umbilic if and only if the envelope of tangent spaces is a cone over an affine sphere , and of submanifolds contained in hyperquadrics , where the immersion is umbilic if and only if the submanifolds is contained in a hyperplane . in section 8
we prove the main result of the paper : given a hypersurface @xmath12 and a point @xmath13 in the @xmath14-space , the immersion @xmath15 , where @xmath16 is the co - normal of @xmath12 , is umbilic and normally flat , and , conversely , any umbilic and normally flat immersion is of this type .
denote by @xmath22 the canonical affine connection and by @xmath65 $ ] the canonical volume form of @xmath56 .
let @xmath66 be an immersion , where @xmath67 is a @xmath68-dimensional manifold and denote @xmath69 .
consider a transversal plane bundle @xmath38 , @xmath39 , and let @xmath70 be a frame for @xmath7 .
for @xmath71 , write @xmath72 then @xmath73 is a torsion free affine connection and @xmath74 , @xmath75 , are symmetric bilinear forms on @xmath67 . for @xmath76 , @xmath75 , write @xmath77 where @xmath78
are @xmath79-forms on @xmath67 and the linear maps @xmath80 are called _ shape operators _ of @xmath81 . in general , for @xmath41 , write @xmath82 and the linear map @xmath83 is called the shape operator of @xmath84 .
most of the time we shall consider that @xmath85 and @xmath86 is the inclusion map . the normal connection @xmath47 is defined by @xmath87 the curvature tensor of the normal connection , called _ normal curvature tensor _ , @xmath88 ,
is defined by @xmath89}\zeta.\ ] ] since @xmath90 has vanishing curvature , we obtain @xmath91 we have also ricci equations , @xmath75 , @xmath92 , @xmath93 @xmath94 @xmath95 @xmath96 the cubic form @xmath97 , @xmath75 , is defined by @xmath98 for @xmath99 .
it follows from codazzi equations that @xmath97 is symmetric in @xmath100 . for more details ,
see @xcite and @xcite .
from now on , we consider a codimension @xmath18 submanifold @xmath8 contained in a hypersurface @xmath11 of @xmath56 .
take vector fields @xmath101 tangent to @xmath11 and transversal to @xmath8 and @xmath102 transversal to @xmath11 .
we shall assume that @xmath25 given by equation is non - degenerate , which is independent of the choice of @xmath4 and @xmath20 .
under this non - degeneracy hypothesis , there exists a unique @xmath4 , up to multiplication by a scalar function , such that @xmath26 is tangent to @xmath11 , for any @xmath43 .
we call any such @xmath4 a _ darboux vector field _ of the immersion @xmath5 ( see @xcite )
. equation can be re - written as @xmath103 now equation implies that @xmath104 and so , by equation , @xmath105 is @xmath25-self adjoint .
consider an immersion @xmath106 and denote by @xmath4 a fixed darboux vector field tangent to @xmath11 along @xmath8 .
consider a tangent frame @xmath107 and let @xmath108.\ ] ] then @xmath109 is non - degenerate and we can define the metric .
@xmath110 where @xmath111 .
the metric @xmath6 is called the _ affine metric _ of the immersion @xmath5 ( @xcite,@xcite ) .
consider a tangent @xmath6-orthonormal frame @xmath29 and let @xmath20 be a transversal vector field satisfying @xmath112=1.\ ] ] then we obtain from equation that @xmath113 .
it turns out that there exists a transversal vector field @xmath20 such that @xmath114 is tangent to @xmath11 , for any @xmath27 tangent to @xmath8 .
moreover , the plane bundle generated by @xmath19 is unique , and it is called the _
affine normal plane bundle _ ( @xcite,@xcite )
. we shall denote it by @xmath38 , @xmath39 . from equation , @xmath33 , which , by equation , implies that @xmath115 is @xmath25-self adjoint .
we conclude that , in this setting , both @xmath105 and @xmath115 are @xmath6-self adjoint .
an immersion @xmath5 is _ affine semiumbilic _ with respect to @xmath7 if @xmath83 is a multiple of the identity , for some @xmath41 .
an immersion @xmath5 is affine semiumbilic with respect to @xmath7 at a point @xmath116 if @xmath117 is a multiple of the identity , for some @xmath41 .
if an immersion @xmath5 is affine semiumbilic with respect to @xmath7 at @xmath116 , then the shape operators @xmath117 , @xmath118 , commute . for @xmath119 ,
the converse also holds .
fix @xmath120 . if @xmath116 is affine semi - umbilic , we can write @xmath121 , for some real numbers @xmath122 .
this implies that @xmath123 . for the converse , write for @xmath119 @xmath124 where @xmath125 is a @xmath6-orthonormal frame of @xmath126 .
since both @xmath127 and @xmath128 are @xmath6-self adjoint , @xmath129 and @xmath130 .
the semiumbilic condition is then equivalent to @xmath131=0.\ ] ] on the other hand , the commutativity of @xmath127 and @xmath128 is also equivalent to this condition .
for immersions @xmath5 that admit a parallel darboux vector field @xmath4 , this choice of @xmath4 is ubiquitous .
in fact , it is proved in @xcite that the condition @xmath4 parallel is equivalent to the coincidence of the affine normal plane with the transon plane , and also equivalent to the apolarity of the second cubic form . in this section , we show that @xmath4 parallel is also equivalent to the condition that the laplacian of the position vector @xmath45 belongs to the affine normal plane . it is not always true that an immersion @xmath5 admits a parallel darboux vector field @xmath4 .
in fact , we have the following proposition : the immersion @xmath5 admits a local parallel darboux vector field if and only @xmath132 .
when it exists , the parallel darboux vector field is unique up to a multiplicative constant . if the darboux vector field @xmath4 is parallel , @xmath133 , and so equations and imply that @xmath132 .
conversely , assuming @xmath132 , the same equations imply that @xmath134 is locally exact . then @xmath135 is parallel , where @xmath136 .
finally , if @xmath4 is parallel , @xmath137 is parallel if and only if @xmath138 is constant . for more details
see @xcite . since @xmath139 , equation says that @xmath140 for @xmath99 .
equivalently , denoting @xmath141 , we have @xmath142 the cubic form @xmath143 is _ apolar _ with respect to @xmath25 if @xmath144 for any @xmath145 .
this condition is equivalent to @xmath146 . in @xcite ,
the following proposition is proved : the cubic form @xmath143 is apolar with respect to @xmath25 if and only if @xmath4 is parallel .
denote by @xmath54 the laplacian operator with respect to the metric @xmath6 . recall that @xmath147 is the trace with respect to @xmath6 of the hessian operator defined by @xmath148 where @xmath149 denotes the levi - civita connection of @xmath6 ( see @xcite , p.64 ) .
the laplacian of @xmath45 belongs to the affine normal plane if and only if @xmath4 is parallel . in this case @xmath150 where @xmath151
write @xmath152 where @xmath153 .
for @xmath4 parallel , the cubic form @xmath44 is apolar with respect to @xmath25 , i.e. , @xmath154 .
thus @xmath155 thus proving the proposition .
conversely , if @xmath156 belongs to the affine normal plane , @xmath154 , which implies that @xmath4 is parallel .
a submanifold @xmath5 is a _ visual contour _ if there exists @xmath49 such that the tangent space to @xmath11 at each point of @xmath8 passes through @xmath13 .
this class of submanifolds is the object of study of the _ centro - affine differential geometry _ and is important in computer graphics ( @xcite ) . for a visual contour
, one can choose the darboux vector field @xmath157 which is both parallel and also umbilic , i.e. , @xmath48 is a multiple of the identity , for any @xmath39 . to prove the converse , we need the following lemma : [ lem : constantcurvature ] assume @xmath158 . if the darboux vector field @xmath4 is both umbilic and parallel , then @xmath159 , for some constant @xmath160 . for @xmath39 ,
write @xmath161 . from equation @xmath162}\xi=0,\ ] ] we obtain @xmath163 taking @xmath23 linearly independent , we conclude that @xmath164 , which implies that @xmath160 is constant .
[ cor : visualcontour ] assume @xmath158 .
if there exists a darboux vector field @xmath4 which is both umbilic and parallel , then @xmath5 is a visual contour .
take @xmath4 umbilic and parallel . by the lemma [ lem : constantcurvature ] , @xmath165 , for some constant @xmath166 .
writing @xmath167 , we get @xmath168 and so @xmath13 is constant , which implies that @xmath5 is a visual contour .
an immersion @xmath5 is _ normally flat _ with respect to the affine normal plane bundle @xmath7 if @xmath169 .
an immersion @xmath5 is _ ( totally )
umbilic _ if @xmath83 is a multiple of the identity , for any @xmath41 , for any @xmath39 . in this section
we study the conditions under which the immersions are normally flat and umbilic .
fix a parallel darboux vector field @xmath4 .
then there exists a parallel vector field @xmath46 linearly independent with @xmath4 if and only if @xmath47 is flat .
when the parallel vector field @xmath20 exists , any other parallel vector field in @xmath7 is necessarily of the form @xmath170 , for some constant @xmath138 .
if @xmath46 is parallel , @xmath171 , and so equations and imply that @xmath172 . thus @xmath47 is flat .
conversely , if @xmath172 , the same equations imply that @xmath173 is locally exact .
then @xmath174 is parallel , where @xmath175 .
finally , if @xmath20 is parallel , @xmath170 is parallel if and only if @xmath138 is constant .
an immersion @xmath5 is normally flat if and only if it admits a parallel darboux vector field @xmath4 and a parallel vector field @xmath46 , linearly independent from @xmath4 .
the vector field @xmath4 is unique up to a multiplicative constant , while @xmath20 is unique up to the addition of @xmath137 , @xmath138 constant .
[ prop : unf ] assume that @xmath158 .
if @xmath5 is umbilic and normally flat , then it is a visual contour and there exists a point @xmath176 contained in the intersection of the affine normal planes @xmath38 , @xmath39 .
conversely , if @xmath5 is a visual contour and the intersection of the normal planes @xmath38 , @xmath39 , contains a point @xmath176 , then the immersion is umbilic and normally flat .
assuming @xmath5 is umbilic and normally flat , we can choose @xmath4 umbilic and parallel .
then corollary [ cor : visualcontour ] implies that @xmath5 is a visual contour , and so we can assume @xmath157 .
we can also choose @xmath20 umbilic and parallel .
so we may write @xmath177 , and , by lemma [ lem : constantcurvature ] , @xmath178 is constant .
writing @xmath179 , we have @xmath180 and @xmath181 , for any @xmath43 , which implies that @xmath50 is independent of @xmath39 .
conversely , if @xmath5 is a visual contour and @xmath182 , we may assume @xmath157 and define @xmath183 .
then @xmath20 is umbilic and parallel , which implies that @xmath5 is umbilic and normally flat .
in this section we define the affine distance and show that its singular set coincides with the affine normal plane .
we give also conditions for the regularity of the affine focal set , which is the bifurcation set of the affine distance .
finally , by comparing with the envelope of tangent spaces , we show that all simple singularities are realizable .
the affine distance @xmath184 is defined by @xmath185.\ ] ] the singular set of @xmath54 is @xmath186 . differentiating equation with respect
to @xmath187 we obtain @xmath188=0.\ ] ] differentiating @xmath189=1 $ ] with respect to @xmath187 we obtain @xmath190 we conclude that @xmath191.\ ] ] similarly , differentiating equation with respect to @xmath192 we obtain @xmath193.\ ] ] we conclude that the singular set of @xmath54 is defined by @xmath194 , @xmath195 , and the lemma is proved .
take a @xmath6-orthonormal basis @xmath29 formed by eigenvectors of @xmath115 , i.e. , @xmath196 .
write also @xmath197 [ prop : degreen ] write @xmath198 .
the bifurcation set @xmath199 of @xmath54 at @xmath116 is given by @xmath200 , where @xmath201 is a polynomial of degree at most @xmath68 in @xmath202 .
differentiating equations we obtain @xmath203 taking @xmath204 , the lemma is proved .
for an umbilic and normally flat immersion , the affine focal set reduces to a line . for umbilic and normally flat immersions , @xmath205 and @xmath206 . if the shape operators commute at @xmath116 , the affine focal set at this point consists of @xmath68 lines .
in particular this holds if @xmath116 is affine semiumbilic .
if the shape operators commute at @xmath116 , then we can find an orthonormal frame @xmath29 made of @xmath84-principal directions , for all @xmath41 . in this basis , @xmath207 , for @xmath208 , and we write @xmath209 . from we
conclude that the bifurcation set is given by @xmath210 which is equivalent to @xmath211 , for some @xmath212 .
assume that @xmath213 is a simple eigenvalue of @xmath128 associated with the principal direction @xmath214 .
we shall give conditions under which the affine focal set is smooth at @xmath215 .
[ lem : regular1 ] consider @xmath216 and @xmath217 in proposition [ prop : degreen ] . at this point , the vector @xmath218 is tangent to the curve @xmath200 .
differentiating @xmath200 and taking @xmath216 , @xmath217 we obtain @xmath219 and the lemma is proved .
[ lem : regular2 ] write @xmath220 assuming that @xmath20 is parallel , @xmath221 and for @xmath222 , @xmath223 denote by @xmath224 the eigenspace generated by the eigenvector @xmath214 of @xmath128 , and by @xmath225 the @xmath226-space @xmath6-orthogonal to @xmath224 .
assume that @xmath20 is parallel .
if @xmath213 is a simple eigenvalue of @xmath128 and @xmath227 , the affine focal set is smooth at @xmath215 and the tangent space of @xmath228 at this point is @xmath229 . by lemma [ lem : regular1 ] , @xmath84
is contained in the @xmath14-dimensional subspace @xmath229 . by lemma [ lem : regular2 ] , @xmath230 , @xmath231 ,
are also contained in this subspace .
thus we must only prove that these vectors are linearly independent .
but this follows from the hypothesis @xmath227 and @xmath232 , if @xmath222 .
the envelope of tangent spaces is the set @xmath233 or equivalently , @xmath234 if @xmath235 , for some non - zero eigenvalue @xmath166 of @xmath105 , then @xmath236 is regular ( for details , see @xcite ) .
if @xmath237 , for some non - zero eigenvalue @xmath166 of @xmath105 , then @xmath238 is a point of intersection of @xmath236 with @xmath228 . in @xcite
, one can find examples of such points where @xmath239 are versal deformations of the simple singularities , namely , @xmath240 , @xmath241 , @xmath242 , @xmath243 , @xmath244 , @xmath245 and @xmath246 ( for examples in case @xmath119 , see section [ sec : hyperplanar ] and @xcite , ch.8 ) .
these examples are also examples of singularities that appear in @xmath228 .
we conclude that any simple singularity appears as a singular point of the affine focal set @xmath228 for some immersion @xmath5 .
let @xmath160 and @xmath247 be a plane curves parametrized by arc length affine and we consider the surface @xmath8 parameterized by @xmath248 denote @xmath249 and let @xmath250 .
we shall assume that @xmath4 is tangent to a hypersurface @xmath251 . since @xmath252 is tangent to @xmath8
, @xmath4 is the parallel darboux vector fields of the immersion @xmath5 .
let @xmath253 , @xmath254 .
then @xmath255 and @xmath256 are parallel and belongs to the affine normal plane bundle .
moreover @xmath257=1.\ ] ] the affine distance function @xmath258 on product of curves is given by : @xmath259\ ] ] the bifurcation set is given by @xmath260 @xmath261\\ f_{v}&=-[x_1 , x-\phi(u_1,u_2 ) , \xi_1,\xi_2]\end{aligned}\ ] ] we have @xmath262 if , and only if , there are @xmath263 such that @xmath264 .
now for @xmath265 such that @xmath262 @xmath266 where @xmath267 denote the affine curvature of the plane curves @xmath160 and @xmath247 . therefore @xmath268 , if and only if , @xmath269 and @xmath270 or @xmath271 .
thus , at each point , the affine focal set is a pair of lines concurring at @xmath272 .
globally , we can write the affine focal set as @xmath273 , where @xmath274 denotes the affine planar evolute of the planar curve .
in this section , we shall consider a curve @xmath275 , where @xmath276 is an interval and @xmath11 is a surface of @xmath58 .
let @xmath275 be a curve .
the non - degeneracy hypothesis says that , at each point , the osculating plane of @xmath45 does not coincide with the tangent plane of @xmath11 . under this hypothesis
, there exists a reparameterization of @xmath45 such that @xmath277 is tangent to @xmath11 . from now on
, we shall assume that this property holds .
choose a vector field @xmath278 in the darboux direction satisfying @xmath279=1,\ ] ] where @xmath280 .
the vector field @xmath4 is parallel , @xmath281 is the affine metric and the affine normal plane bundle is spanned by @xmath282 .
differentiating equation we obtain that @xmath283 , and thus @xmath4 is parallel .
moreover , equation implies that @xmath284 , i.e. , @xmath281 . finally equation together with @xmath285 tangent to @xmath11 implies that @xmath286 belongs to the affine normal plane .
we can write @xmath287 take a real function @xmath288 such that @xmath289 . note that @xmath288 may not be globally defined . consider the vector field @xmath290 ( see figure [ fig : parallelvf ] ) .
then @xmath291=1 $ ] and @xmath292 defining @xmath293 , we obtain the following equations : @xmath294 note that @xmath4 and @xmath20 are both parallel and thus the immersion is normally flat with respect to the affine normal plane bundle @xmath295 . and @xmath20 .
] a curve @xmath296 is a visual contour if and only if @xmath166 is constant .
if @xmath166 is constant , then @xmath297 is constant and belong to the tangent plane to @xmath11 at @xmath116 , thus implying that @xmath45 is a visual contour .
conversely , if @xmath13 belongs to the tangent plane to @xmath11 at each @xmath116 , then @xmath298 is a parallel darboux vector field .
thus @xmath299 , for some constant @xmath138 , and hence @xmath166 is constant .
[ prop : sigmamuconstant ] a curve @xmath296 is a visual contour and there exists a constant point @xmath182 , for each @xmath39 , if and only if @xmath166 and @xmath178 are constant .
if @xmath178 is constant , then @xmath179 is constant and belongs to @xmath38 , for each @xmath116 .
conversely , if there exists a point @xmath182 , for each @xmath116 , then @xmath300 is parallel and belongs to @xmath38
. thus @xmath301 , for some constant @xmath138 , which implies that @xmath302 is constant . comparing the above proposition with proposition [ prop : unf ]
we see that the umbilic and normally flat condition for @xmath158 corresponds to @xmath166 and @xmath178 constant for curves .
the affine distance is given by @xmath303.\ ] ] thus @xmath304,\ ] ] and so @xmath305 is the equation of the affine normal plane .
we conclude that the affine focal set @xmath306 coincides with the envelope of @xmath307 , @xmath308 .
denote @xmath309 and @xmath310 and let @xmath311 be the line connecting the points @xmath312 and @xmath313 ( see figure [ fig : linel ] ) .
the envelope @xmath55 of the family of planes @xmath314 is the ruled surface formed by the lines @xmath311 , @xmath308 .
the equation of the plane @xmath307 is @xmath315 , where @xmath316.\ ] ] differentiating this equation with respect to @xmath317 we obtain @xmath318-\sigma [ x-\phi,\eta , t]-1.\ ] ] writing @xmath194 and substituting in this equation we get @xmath319 if and only if @xmath320 thus proving the proposition .
whose union is the affine focal set @xmath228 . ]
the ruled surface @xmath55 is developable . since @xmath321
we conclude that @xmath322=0,\ ] ] thus proving that @xmath55 is developable .
let @xmath323 be the affine focal set of @xmath5 .
each point of @xmath228 at @xmath39 belongs to the line @xmath324 then 1 .
@xmath323 is smooth if @xmath325 .
@xmath323 is locally diffeomorphic to a cuspidal edge if @xmath326 and @xmath327 .
@xmath323 is locally diffeomorphic to a swallowtail if @xmath328 and @xmath329 .
differentiating equation we obtain @xmath330-\sigma'\left[x-\phi,\eta , t\right]+(\lambda\sigma-\mu)f\ ] ] thus @xmath331 if and only if @xmath326 . differentiating equation and discarding the last parcel we obtain @xmath332-\sigma''\left[x-\phi,\eta , t\right]+(\lambda\sigma'-\mu')f\ ] ] thus @xmath333 if and only if @xmath328 . differentiating equation and substituting the values of @xmath202
we obtain @xmath334 the result follows now from @xcite , ch.6 . in this section
we describe those immersions @xmath335 for which @xmath166 and @xmath178 are constant .
for these curves , the affine focal set @xmath323 reduces to a single line .
we may assume , w.l.o.g .
, that @xmath336 and @xmath337 .
thus equation can be written as @xmath338 the condition @xmath339 can be written as @xmath340 and thus this equation reduces to @xmath341 or equivalently , @xmath342 for some constant vector @xmath50 . assuming @xmath343 and writing @xmath344 , for some planar curve @xmath345
, we obtain the discoupled equations @xmath346 we say that a planar curve @xmath347 is parameterized by affine arc - length if @xmath348=1 $ ] . in this case @xmath349 for a certain scalar function called the _ affine curvature _ of @xmath350 . the function @xmath351\ ] ] is called the _ affine distance _ , or _ support function _ , of @xmath350 with respect to the origin @xmath13 ( @xcite,@xcite ) .
next theorem will be generalized to higher dimensions in section [ sec : umbilic ] : [ thm : umbiliccurves ] given a planar curve @xmath350 parameterized by affine arc - length with affine curvature @xmath352 , let @xmath353 and @xmath354 denote the support function of @xmath350 with respect to an arbitrary origin @xmath13
. then @xmath355 satisfies equation and , conversely , any solution of is obtained in this way .
assume that @xmath353 and @xmath354 is the support function of @xmath356 with respect to @xmath13 .
then @xmath357 moreover @xmath358 $ ] and so @xmath359 = 1-\rho(u)z(u),\ ] ] thus proving that @xmath355 satisfies equation . since the general solution of the linear system of ode s is @xmath18-dimensional , we conclude that @xmath355 is a general solution of this system .
a flattening point of @xmath45 is a point @xmath360 such that @xmath361 .
consider a closed curve @xmath45 with constant @xmath166 and @xmath178 .
then @xmath45 has at least six flattening points . if @xmath45 is closed , then @xmath350 is a closed convex planar curve . by the affine six vertex theorem ( @xcite,@xcite ) ,
there exists at least six points of @xmath350 where @xmath362 .
this implies that @xmath361 at these points .
theorem [ thm : umbiliccurves ] is also interesting in the context of spatial curves , not contained in a surface .
assume @xmath363 is a smooth spatial curve parameterized by affine arc - length , i.e. , @xmath364=1 $ ]
. then we can write @xmath365 for some scalar functions @xmath352 and @xmath366 . the plane passing through @xmath367 and generated by @xmath368 and @xmath369 is called _ affine rectifying plane _ , while the envelope @xmath370 of these planes is called the _ intrinsic affine binormal developable _ of @xmath371 .
it is proved in @xcite that @xmath370 is a cylindrical surface if and only if @xmath372 .
writing @xmath373 in equation , we obtain equation .
theorem [ thm : umbiliccurves ] says that @xmath178 is constant if and only if @xmath374 , where @xmath353 , for some curve @xmath350 and @xmath375 $ ] .
thus we conclude that @xmath376 , where @xmath377 dv.\ ] ] the function @xmath378 is the area of the planar region bounded by the angle @xmath379 and the arc of @xmath350 , @xmath380 .
consider a spatial curve @xmath363 parameterized by affine arc - length , and let @xmath352 and @xmath366 be given by equation .
then the intrinsic affine binormal developable @xmath370 is cylindrical if and only if we can write @xmath376 , where @xmath350 is a planar curve and @xmath381 is given by equation , for some origin @xmath13 and some initial point @xmath382 .
we have that @xmath370 is cylindrical if and only if @xmath372 . by equation , this condition is equivalent to @xmath178 constant
. then theorem [ thm : umbiliccurves ] says that @xmath178 constant is equivalent to @xmath383 , where @xmath353 for some planar curve @xmath350 and @xmath354 the support function of @xmath350 with respect to some origin @xmath13 .
we conclude that @xmath370 cylindrical is equivalent to @xmath376 , where @xmath381 is given by equation .
the quantity @xmath384 is projectively invariant .
in fact , @xmath385 is the _ projective length _ of the curve @xmath45 and @xmath386 if and only if @xmath45 is contained in a quadratic cone ( for details , see @xcite ) .
the curve @xmath45 is projectively equivalent to a planar curve and in this case @xmath387 .
thus @xmath388 and so @xmath352 is the affine curvature of @xmath45 .
by the six vertex theorem for planar closed convex curves ( @xcite ) , @xmath45 admits at least six points where @xmath389 , or equivalently , @xmath386 .
this means that a closed convex curve @xmath45 admits at least six points with higher order contact with a quadratic cone .
in this section we consider two classes of normally flat submanifolds @xmath5 : ( 1 ) @xmath8 contained in a hyperplane @xmath9 and ( 2 ) @xmath11 hyperquadric . for @xmath8 contained in a hyperplane @xmath9 , take @xmath4 tangent to @xmath11 such that @xmath390 , where @xmath391 denotes inner product and @xmath392 is the euclidean normal of @xmath9
. then @xmath4 is the parallel darboux vector field of the immersion @xmath5 .
let @xmath20 and @xmath59 denote the blaschke normal vector field and metric of @xmath10 , respectively .
then @xmath20 is parallel and , for a @xmath59-orthonormal frame of @xmath126 we have @xmath393=1.\ ] ] since @xmath394 , @xmath29 is @xmath6-orthonormal and @xmath395=1.\ ] ] thus @xmath396 .
we conclude that @xmath397 is exactly the affine focal set of the immersion @xmath10 ( @xcite , ch.8 ) .
let @xmath5 be an immersion such that @xmath8 is contained in a hyperplane @xmath9
. then @xmath5 is umbilic if and only if @xmath10 is an affine sphere and the envelope of tangent spaces of @xmath5 is a cone .
the vector field @xmath20 is umbilic if and only if @xmath10 is an affine sphere .
the vector field @xmath4 is umbilic if and only the envelope of tangent spaces is a cone . in this section
we consider the case @xmath11 hyperquadric . by an affine transformation of @xmath56
, we may assume that @xmath11 is given by @xmath398 where @xmath399 .
denote by @xmath400 the non - degenerate , possibly indefinite , metric @xmath401 then the tangent space of @xmath11 is @xmath402 , where the orthogonality is taken with respect to @xmath400 .
moreover , one can verify that @xmath45 is the affine blaschke normal and @xmath400 restricted to the tangent space of @xmath11 is the affine blaschke metric @xmath59 of @xmath11 .
consider now @xmath5 .
the non - degeneracy hypothesis implies that @xmath403 , for any @xmath404 .
this implies that we can choose @xmath4 orthogonal to @xmath8 with @xmath405 we have that : 1 .
@xmath4 is a parallel vector field in the darboux direction .
the metric @xmath6 is the restriction of @xmath59 to @xmath8 . to prove 1 , write @xmath406 where @xmath407 is tangent to @xmath11 . since @xmath408
, we conclude that @xmath26 is tangent to @xmath11 and so @xmath4 is a darboux vector field of @xmath5 .
differentiating equation we obtain @xmath409 , for any @xmath27 tangent to @xmath8 , which implies that @xmath26 is tangent to @xmath8 and so @xmath4 is parallel . to prove 2 , let @xmath410 be a @xmath59-orthonormal frame of @xmath126 .
then @xmath411 is a @xmath59-orthonormal base of @xmath412 and so @xmath413=\delta_{ij}.\ ] ] this last equation implies that @xmath29 is a @xmath6-orthonormal frame of @xmath126 and thus @xmath414 .
the vector field @xmath20 belongs to the affine normal plane , is parallel and umbilic .
it is clear that @xmath20 is both parallel and umbilic , and we must verify that @xmath46 . consider a @xmath6-orthonormal frame @xmath29 . since the metric @xmath6 is the restriction to @xmath8 of the blaschke metric @xmath59 of @xmath11 , this frame is also @xmath59-orthonormal .
thus @xmath415 which implies that @xmath416=1.\ ] ] this last equation implies that @xmath46 .
fix @xmath417 and denote by @xmath418 the darboux vector field at @xmath116 .
the tangent space @xmath419 is orthogonal to both @xmath116 and @xmath418 . in particular , @xmath419 is contained in @xmath420 , for some @xmath421 , if and only if the vectors @xmath116 , @xmath418 and @xmath422 are coplanar
. let @xmath8 be a submanifold contained in a hyperquadric @xmath11 .
then the immersion @xmath5 is umbilic if and only if @xmath8 is contained in a hyperplane .
assume first that @xmath5 is umbilic and let @xmath9 be the hyperplane orthogonal to the line @xmath55 .
assuming that @xmath55 is parallel to @xmath422 , the above remark implies that @xmath419 is contained in @xmath420 , for some @xmath421 .
this implies that @xmath8 is contained in a hyperplane .
conversely , assume that @xmath423 , where @xmath9 is the hyperplane given by @xmath420
. then we can write @xmath424 , for some scalars @xmath425 and @xmath288 . from @xmath426 , @xmath427 , @xmath428
we obtain @xmath429 thus proving that @xmath288 is independent of @xmath39 .
we conclude that such an immersion is umbilic .
in this section , we give a geometric characterization of umbilic and normally flat immersions for @xmath158 and immersions with @xmath166 and @xmath178 constant , for @xmath430 . by propositions [ prop : unf ] and [ prop :
sigmamuconstant ] , these conditions are equivalent to @xmath45 being a visual contour with a constant point @xmath50 in the affine normal planes .
in order to keep the statements shorter , we shall refer to these immersions as umbilic and normally flat even in case @xmath430 .
consider a non - degenerate immersion @xmath61 and denote by @xmath431 its blaschke co - normal map .
fix an origin @xmath64 and define the immersion @xmath432 by @xmath433 where @xmath434 and @xmath435 is the _ affine distance _ , or _ support function _
, of @xmath12 with respect to @xmath13 ( @xcite ) .
take @xmath337 so that @xmath45 becomes a visual contour .
our main theorem is the following : [ thm : umbilic ] any immersion given by equation is umbilic and normally flat .
conversely , any umbilic and normally flat immersion @xmath296 is given by equation , for some immersion @xmath436 and origin @xmath64 . by proposition [ prop : unf ] , we have to show that @xmath437 , @xmath438 , belongs to the affine normal plane , for any @xmath308 .
take a tangent frame @xmath410 orthonormal with respect to the blaschke metric @xmath59 of the immersion @xmath12 .
since @xmath439 we have that @xmath440 the first parcel in the second member is tangent to @xmath11 : in fact , writing @xmath441 we obtain @xmath442 which is tangent to @xmath11 .
we conclude that @xmath443 , and so @xmath29 is @xmath25-orthonormal for the frame @xmath444 . to conclude that @xmath50 is in the affine normal plane we must verify that @xmath445\ ] ] is constant . but @xmath446=\left [ \nu_{*}x_1, ... \nu_{*}x_n , \nu \right]=-1,\ ] ] thus proving the direct part of theorem [ thm : umbilic ] . from the proof of theorem [ thm : umbilic ]
we obtain that the blaschke metric @xmath59 of @xmath12 coincides with the metric @xmath113 of the immersion @xmath45 .
thus the laplacian operators of @xmath12 and @xmath45 are the same .
if follows from @xcite , propositions 6.2 and 6.3 , that @xmath451 thus @xmath452 comparing with equation , the corollary is proved .
assume that @xmath45 is an umbilic immersion and normally flat and write @xmath453 .
define @xmath12 by the conditions @xmath454 for some origin @xmath64 .
these equations imply that @xmath455 , for any @xmath27 .
we conclude that @xmath456 , for some @xmath37 , where @xmath16 denotes the blaschke co - normal of the immersion @xmath12 . since by proposition [ prop : unf ] @xmath50 belongs to the affine normal plane of @xmath45 , we can find a local frame @xmath410 , @xmath25-orthonormal , such that @xmath457=-1.\ ] ] this equation implies @xmath458=-1,\ ] ] and then @xmath459=-\lambda^{n+1}.\ ] ] differentiating @xmath460 we obtain @xmath461 the same argument as above says that the first vector of the second member is tangent to @xmath11 , and so @xmath462 .
we conclude that @xmath463 now consider the @xmath59-orthonormal basis @xmath464 , where @xmath465 .
since @xmath466=-1,\ ] ] we conclude that @xmath467=-\lambda^{\frac{n}{2}}.\ ] ] from equations and we conclude that @xmath468 . the affine distance @xmath354 is constant if and only if @xmath12 is a proper affine sphere and @xmath13 its center ( @xcite , prop.5.10 ) .
thus , if @xmath12 is a proper affine sphere and @xmath13 is arbritary , @xmath45 is contained in a hyperplane .
conversely , if @xmath45 is contained in a hyperplane , @xmath354 is constant for some choice of @xmath13 .
thus @xmath12 is an affine sphere . | in this paper we study the affine focal set , which is the bifurcation set of the affine distance to submanifolds @xmath0 contained in hypersurfaces @xmath1 of the @xmath2-space .
we give condition under which this affine focal set is a regular hypersurface and , for curves in @xmath3-space , we describe its stable singularities . for a given darboux vector field @xmath4 of the immersion @xmath5
, one can define the affine metric @xmath6 and the affine normal plane bundle @xmath7 .
we prove that the @xmath6-laplacian of the position vector belongs to @xmath7 if and only if @xmath4 is parallel . for umbilic and normally flat immersions
, the affine focal set reduces to a single line .
submanifolds contained in hyperplanes or hyperquadrics are always normally flat
. for @xmath8 contained in a hyperplane @xmath9 , we show that @xmath5 is umbilic if and only if @xmath10 is an affine sphere and the envelope of tangent spaces is a cone . for @xmath11 hyperquadric , we prove that @xmath5 is umbilic if and only if @xmath8 is contained in a hyperplane
. the main result of the paper is a general description of the umbilic and normally flat immersions : given a hypersurface @xmath12 and a point @xmath13 in the @xmath14-space , the immersion @xmath15 , where @xmath16 is the co - normal of @xmath12 , is umbilic and normally flat , and conversely , any umbilic and normally flat immersion is of this type . |
in the last few years , a lot of effort has been invested to understand the physical properties of mnas - based thin films and multilayers .
bulk manganese arsenide is ferromagnetic at room temperature with a nias hexagonal structure ( @xmath2-phase ) .
@xcite at @xmath3313 k however , mnas displays a first - order phase transition that drives the system to a paramagnetic state . at this temperature ,
a structural transition occurs from the low - temperature @xmath2-phase to an orthorhombically - distorted ( mnp ) @xmath1-phase . at @xmath4398
k , a second structural transition takes place and the system regains a nias hexagonal structure .
mnas grows epitaxially on various standard semiconductor substrates such as gaas(100 ) , @xcite gaas(111)b , @xcite gaas(110 ) , @xcite inp(100 ) , @xcite si(100 ) @xcite and si(111 ) .
@xcite the excellent quality of mnas / gaas heterostructures makes this compound particularly interesting for hybrid metal / semiconductor spintronics .
@xcite the epitaxy introduces strain and defects in mnas thin films that significatively modify their structural and magnetic properties with repect to the bulk .
when mnas is grown on gaas(100 ) or gaas(111)b , epitaxial strain induces an @xmath5-phases coexistence in a wide range of temperature .
@xcite in addition , the @xmath6 of the thin films can be enhanced when they are grown on gaas(111)b substrates .
@xcite the orientation of the gaas substrate determines the growth behavior and epilayer orientation . when grown on gaas(111)b substrate , mnas lies with the hexagonal c axis aligned with the growth direction [ fig . [ figure1](b ) ] ; @xcite while for the growth on gaas(100 ) substrate , two orientations of the epilayer have been described .
@xcite by correctly adjusting the growth conditions , it becomes possible to obtain one single phase @xcite with the mnas @xmath7-axis parallel to the gaas [ 110 ] that lies inside the surface plane [ fig .
[ figure1](b ) ] .
due to the important misfit between the thin film and the substrate ( 7@xmath8 ) , the epilayer relaxes its strain from the very beginning of the growth .
above a certain thickness , the epilayer is expected to be fully relaxed at the growth temperature . however , during the after - growth cooling cycle , a residual strain is introduced due to the large dilatation coefficients difference between mnas and gaas .
@xcite in fact , as the @xmath9 phase transition proceeds mainly by a 1@xmath8 lattice parameter contraction in the hexagonal plane , strain will accommodate differently for mnas / gaas(100 ) or mnas / gaas(111)b .
another particularity of mnas epilayers grown on gaas(100 ) substrates is that , within the phase coexistence region , scanning tunneling microscopy ( stm ) images show the presence of stripes along the [ 0001 ] direction with two different heights , arranged alternated within the @xmath10 $ ] direction .
@xcite magnetic force magnetometry ( mfm ) studies indicate that the higher stripes are ferromagnetic ( @xmath2-phase ) while the shallower are paramagnetic ( @xmath1-phase ) .
@xcite x - ray diffraction ( xrd ) patterns also agree with the @xmath2-@xmath1 phase coexistence in the 280- 314 k temperature range .
no stripes are however observed for mnas epilayers grown on gaas(111)b .
instead , the surface morphology shows triangular and sticks features depending on the growth parameters like substrate temperature , as overpressure , etc . @xcite
even if no stripes are visible , the phase coexistence has been unambiguously observed between 296 and 328 k by xrd for this orientation .
@xcite in the present work , we perform a detailed study of the magnetism of mnas epilayers , grown on both gaas(100 ) and gaas(111)b substrate orientations .
the anisotropy of the films is analyzed in terms of magneto - crystalline and substrate - induced anisotropy terms .
the temperature dependence of the coercivity of the films is measured and explained in terms of thermally activated domain - wall movements and the appearance of new pinning centers .
mnas epilayers are grown by molecular beam epitaxy ( mbe ) with a substrate temperature of about @xmath11c on previously prepared gaas buffer layers .
after removal of the native oxide , a very thin buffer layer is grown on gaas(111)b substrates @xcite while a standard 100 nm thick buffer layer is grown on gaas(100 ) .
each mnas growth begins under as rich conditions on the gaas buffer layer .
different epilayer thicknesses , @xmath12 , ranging from 30 nm to 200 nm , are grown on both ( 100 ) and ( 111 ) substrate orientations .
the films are named m100-@xmath13 [ nm ] and m111-@xmath13 [ nm ] , respectively .
finally , the samples are capped _ in situ _ with an amorphous znse layer .
magnetization vs. temperature and magnetic field curves are measured in a vibrating sample magnetometer ( vsm ) and a superconducting quantum interference device magnetometer ( squid ) .
the temperature is varied from 5 k to 300 k and the magnetic fields swept up to 5 t. the magnetization loops are measured with the magnetic field rotating in the plane of the film ( ip ) @xmath14 and in the out - of - plane geometry ( oop ) , keeping the azimuthal angle @xmath15 fixed [ fig . [ figure1](a ) ] .
the ferromagnetic phase with hexagonal structure displays a strong anisotropy with a hard - axis parallel to the @xmath7-axis , that favors the orientation of the magnetization in a plane perpendicular to this direction .
@xcite as mentioned previously , the hexagonal @xmath7-axis is inside the plane of the films for m100 samples and perpendicular to it for m111 samples [ fig . [ figure1](b ) ] , which strongly affects the magnetism of the samples .
the magnetization loops reflect the structural stacking of mnas in the films .
[ figure2 ] shows the typical magnetization loops for m100 and m111 samples .
these have been measured in three different configurations of the applied field : in - the - plane of the films and parallel to the @xmath16-axis , ip - pa ( @xmath17 , @xmath18 ) , in - plane and perpendicular to the @xmath16-axis , ip - pe ( @xmath19 , @xmath20 ) and out - of - the - plane of the films with @xmath21 .
the @xmath16-axis is parallel to the [ 11@xmath220 ] axis for m100 samples , and is not clearly determined for m111 samples .
[ figure2](a ) evidences the existence of a strong uniaxial anisotropy in the plane of m100 films , with the easy - axis oriented along the mnas @xmath10 $ ] direction .
the saturation field of the in - plane curve , measured with the magnetic field perpendicular to the easy - axis direction , is much larger than the oop one , indicating that the anisotropy term overcomes the demagnetization effect in these films . in m111 films [ fig . [
figure2](b ) ] and due to the structure stacking onto the substrate , the hard axis strengthens the demagnetization effect favoring an almost isotropic in - plane magnetization .
the free energy density proposed is : @xmath23 @xmath24 for m100 and m111 samples , respectively .
the first term in each equation accounts for the zeeman energy and the second term includes a uniaxial anisotropy oriented in the plane of the film or out of the plane , depending on the samples substrate orientation .
the last term represents the magnetostatic energy . to describe the experimental results we have included an extra anisotropy term in the m100 energy expression , @xmath25 , that favors an in - plane magnetization .
we associate this contribution to substrate - induced strains onto the film structure . in ref .
das and coworkers already reported that the ferromagnetic phase is orthorhombically distorted at the coexistence region of @xmath2-mnas and @xmath1-mnas .
the existence of such anisotropy in m111 films could not be deduced from our measurements .
this contribution may be hidden by the demagnetizing and the uniaxial anisotropy terms .
the study of thinner films is needed to define the role of this contribution in the magnetization of mnas films .
corrections of higher order to the in - plane anisotropy of m111 samples are needed to describe the experimental results .
these terms are highly samples dependent , as discussed below .
the uniaxial anisotropy constant , @xmath26 was calculated from the area enclosed by the loops measured along the easy and the hard directions . the demagnetization term is calculated taking ms=870 emu/@xmath27 , as deduced from the magnetization vs. temperature measurements .
the calculated anisotropy constants are @xmath28 j / m@xmath29 and @xmath30 j / m@xmath29 for m100 samples and @xmath31 j / m@xmath29 for m111 ones .
the uniaxial anisotropy , @xmath26 , is of magneto - crystalline origin and agrees well with the values measured by j. lindner _
@xcite @xmath26 does not depend notably on the thickness of the samples or the substrate orientation and is very close to the bulk value .
@xcite in fig .
[ figure3 ] , the hysteretic zone of the loops measured with the magnetic field applied along the in - plane easy - axis is shown .
the shape of the loops of m111 and m100 samples are notably different . while in m111 samples [ fig .
[ figure3](b ) ] the magnetization reversal is smooth , the magnetization in m100 ones [ fig .
[ figure3](a ) ] shows a sharp jump at the coercive field where a partial inversion of the magnetization occurs , while higher magnetic fields are necessary to completely saturate the magnetization of the sample .
the shapes of the curves are associated with domains movements and are affected both by defects and other inhomogeneities and by the anisotropy of the samples .
@xcite the coercivity increases as the thickness of the films decreases , indicating that the magnetization reversal is affected by substrate - induced strains and topographic defects . in m100 samples , the coercivity of the loops measured along the easy - axis , @xmath32 follows the condition : @xmath33 indicating that the magnetization reversal is mainly driven by domain wall movement . in m111 samples , the in - plane anisotropy terms are much smaller and condition ( 3 ) is no longer valid .
the direction perpendicular to the film surface is a very strong hard - axis due to the superposition of the magneto - crystalline anisotropy and the demagnetizing field resulting in large domain walls .
for this reason we suggest that the walls in this system are nel - like , i.e. , the magnetization within the walls rotates in the film plane .
a simple calculation of domain wall widths , @xmath34 , in m111 films based on the effective anisotropy of the system results in @xmath35 values . according to the nel criteria , in this case the wall mode using an in - plane rotation has a lower energy than the classical bloch mode.@xcite
the in - plane angular dependence of the magnetization loops has been carefully measured by vsm .
we observe that the magnetization loops of m100 samples change remarkably as the magnetic field is rotated in the plane of the films while those of m111 samples vary very little .
the magnetic remanence vs. the azimuthal angle , @xmath36 curves underlines the symmetry of the magnetic anisotropy of the films .
m100 samples show a strong uniaxial anisotropy with a hard - axis oriented along [ 0001 ] direction .
m111 samples , on the other hand , present a much weaker anisotropy contribution with different symmetries , depending on the growth conditions .
uniaxial , three and four - fold anisotropy symmetries have been observed in different samples .
typical @xmath37(@xmath36 ) curves for m100 and m111 samples are shown in fig . [ figure4 ] .
the coercivity of m100 films shows two high peaks at both sides of the hard - axis angle .
the angular dependence of @xmath37 for @xmath38 can be fitted by : @xcite @xmath39 suggesting that the magnetization reversal is driven by 180@xmath40 wall movements [ inset fig .
[ figure4](a ) ] . in a recent article , f. schippan and coworkers @xcite report mfm images that show in similar samples the existence of 180@xmath40 walls , oriented along the @xmath10 $ ] direction .
as the uniaxial in - plane anisotropy is much larger than the demagnetizing term , the walls are bloch - like , i.e. the magnetization in the walls rotates out - of - the - plane of the films .
even if mfm measurements have been performed at room temperature where the @xmath2 and the @xmath1 phases coexist , these results are a strong indication of the domain configuration in these samples .
the @xmath36 dependence of the @xmath37 is much weaker for m111 samples and has a four - fold anisotropy in the case of the m111 - 66 film shown in fig . [
figure4](b ) .
the symmetry of the in - plane anisotropy in these samples can be three - fold and even two - fold and is strongly dependent on the sample growth parameters : substrate temperature during sample growth , surface stoichiometry , etc .
the origin of this term may be associated to surface - induced strains and/or surface reconstruction symmetries .
a study of the anisotropy of m111 ultrathin films is under progress to clarify this point . the oop angular dependence of m100 and m111 coercive fields is shown in fig .
[ figure5 ] .
an important increase of the coercive field with angle is observed , as the field approaches the film normal direction .
the coercivity abruptly decreases to zero at the hard - axis of magnetization .
m100 and m111 samples display a similar behavior , following a @xmath41 function .
the remanence monotonically decreases to zero as @xmath42 is varied from @xmath43/2 to 0 , keeping @xmath36 fixed and parallel to the in - plane easy axis direction . as condition ( 3 )
is fulfilled in this geometry for both group of samples , the reversal process can be thought as performed with only the component of the external field parallel to the easy - axis magnetization direction .
the temperature dependence of the coercive field , extracted from loops measured with the applied field oriented along the in - plane easy - axis was examined ( fig . [ figure6 ] ) . at low temperatures , t@xmath44200 k ,
the coercive field decreases smoothly with increasing temperature , as expected for an ordinary ferromagnet .
this behavior is understood in terms of thermal activation of domain walls .
@xcite at @xmath45200 k , this tendency is reversed and an increase of the @xmath37 is observed , up to @xmath46300 k. this anomalous behavior is associated to the appearance of new pinning centers .
these new pinning sites would progressively appear as the temperature increases at nucleation points of the @xmath1-phase or strained regions of the @xmath2-phase .
previous papers report the @xmath2-@xmath1 phase coexistence in both type of samples but at higher temperatures .
@xcite the magnetization vs. temperature curves do not show any particular feature in the t@xmath47 - t@xmath48 range ( see inset of fig .
[ figure6 ] ) .
ferromagnetic resonance measurements indicate that the magneto - crystalline anisotropy decreases with temperature as the system approaches the curie temperature , t@xmath49 , @xcite without showing noticeable signatures around 200 k. therefore , the magneto - crystalline anisotropy of mnas can not be the origin of the observed effect .
the m vs. t curve saturates at low temperatures to the bulk moment @xcite but an important increase of the curie temperature , t@xmath49 , from 314 k to 330 k ( m100 ) and 350 k ( m111 ) , is observed .
this result suggests that the strains induced by the gaas stabilize the @xmath2-hexagonal phase up to this high temperature .
in fact , x - ray @xcite and neutron @xcite diffraction patterns show a @xmath2-@xmath1 coexistence in the 314 k- t@xmath49 temperature range for m111 samples . thus , in this temperature range
, the magnetization would be given by the addition of an unsaturated ferromagnetic component that corresponds to the @xmath2-phase and a paramagnetic contribution , arising from the @xmath1 phase .
finally , above room temperature the thermal energy overcomes the effect of the new pinning centers and induces a strong reduction of the coercive field as the temperature approaches t@xmath49 .
data shown in fig . [ figure6 ] evidence that the reentrant effect is independent of the substrate orientation . the observed behavior in the coercivity
could then be associated to a former stage of the @xmath2-@xmath1 phase coexistence .
a. ney _ et al . _
@xcite first observed this anomalous behavior in m100 films and associated it to the existence of an array of @xmath2-mnas stripes .
the authors claim that the temperature dependence of the coercivity allows them to distinguish between homogeneous and stripes domains of @xmath2-mnas .
however , stm images show that m111 samples do not form ferromagnetic stripes at room temperature .
@xcite our finding of a similar behavior in both types of samples suggests that the origin of this effect does not reside in the appearance of the @xmath2-phase in the form of stripes but in the phase coexistence phenomena , beyond the epitaxy between mnas and gaas .
the number of pinning centers is temperature dependent due to the progressive nucleation of @xmath1-phase regions and/or the strain evolution in the @xmath2 phase towards the phase transformation .
[ figure7 ] shows the temperature dependence of the coercivity for m100 samples of different thicknesses .
the increase of the @xmath37 with temperature , in the t@xmath47 - t@xmath48 zone is more pronounced for thinner films .
however , the functional form of the temperature variation of the coercivity does not depend on thickness .
the results indicate , in summary , that even though the coercivity in this case is essentially given by extrinsic properties , the reentrant behavior seems to be an intrinsic characteristic of the magnetism of mnas thin films .
we have shown that the magnetization reversal in mnas films depends on the substrate orientation and is strongly affected by the magneto - crystalline anisotropy of the compound . on one hand , the angular dependence of the coercive field of films grown on gaas(100 ) shows that the in - plane magnetization reversal occurs by 180@xmath50 domain wall movements .
moreover , the shape of the hysteresis loops indicates that the magnetization alignment with the magnetic field is driven in two - steps .
this is explained by the existence of a distribution of potential barriers for domain movement , that have a very sharp threshold at the coercive field where more than 70% of the magnetization turns into the direction of the applied field .
the magnetization alignment is strongly affected by the in - plane hard uniaxial anisotropy of the system , that induces the formation of bloch walls along the films .
on the other hand the uniaxial anisotropy of samples grown on gaas(111)b is oriented perpendicular to the film surface , leading to the creation of nel domain walls and a smooth reversal of the magnetization in a low - field range .
summarizing , we have demonstrated that the coercivity of the films presents a general anomalous behavior as a function of temperature , independent of the substrate orientation and films thicknesses .
this effect is associated to the appearance of new pinning centers as a consequence of the progressive growth of a @xmath1-phase or distortion of the @xmath2-phase as the temperature is increased above 200 k.
the authors thank partial financial support from fundacin antorchas , foncyt 03 - 13297 , universidad
nacional de cuyo , conicet and the ecos - sud . ) the magnetic field applied in the plane of the films and perpendicular to the @xmath16-axis , ( @xmath52 ) in - plane and parallel to the @xmath16-axis and ( @xmath53 ) out - of- the plane of the films .
the measurements were performed at 5 k in the squid magnetometer . ] | the magnetic properties of mnas epilayers have been investigated for two different substrate orientations : gaas(100 ) and gaas(111 ) .
we have analyzed the magnetization reversal under magnetic field at low temperatures , determining the anisotropy of the films .
the results , based on the shape of the magnetization loops , suggest a domain movement mechanism for both types of samples .
the temperature dependence of the coercivity of the films has been also examined , displaying a generic anomalous reentrant behavior at t@xmath0200 k. this feature is independent of the substrate orientation and films thickness and may be associated to the appearance of new pinning centers due to the nucleation of the @xmath1-phase at high temperatures . |
understanding the phase diagram of quantum chromodynamics ( qcd ) , as a function of temperature @xmath11 and baryon chemical potential @xmath12 is of great interest in the phenomenology of strongly interacting dense matter @xcite .
there are many excellent reviews on the subject and some recent ones can be found in @xcite .
the only first principles approach to the subject is based on the lattice formulation of qcd in which one computes quantities using the monte carlo method .
unfortunately , due to a variety of algorithmic difficulties this has been difficult to accomplish . at intermediate and large chemical potentials and small temperatures
the numerical methods suffer from a severe sign problem .
thus , the most reliable lattice calculations can only be found at small @xmath0 where reasonable algorithms are available @xcite .
however , even these calculations become difficult especially for large lattices and realistic quark masses .
thus , it is fair to say that quantitatively many features of the @xmath13 phase diagram of qcd still remains unclear .
a recent review of the status of lattice calculations at finite temperature and density can be found in @xcite . given the difficulties of studying the phase diagram of qcd , it is interesting to consider qcd - like theories which do not suffer from sign problems at @xmath7 @xcite .
the sign problem is avoided due to special properties of the fermion action which makes the fermion determinant real and positive . as a price
, baryons turn out to be bosons . in spite of this difference
qcd - like theories provide interesting toy models for qcd . in certain cases
they are also interesting in their own right .
a famous example is two - color qcd and has been extensively studied over the years both theoretically @xcite and numerically @xcite .
two - color lattice qcd with staggered fermions ( 2clqcd ) is especially interesting due to an enhanced @xmath14 symmetry at zero quark mass and baryon chemical potential .
as we will see later , the long distance physics of this theory in the @xmath13 plane is described by an @xmath3 ( @xmath4 when @xmath2 and @xmath5 when @xmath6 ) field theory .
such field theories arise naturally in many condensed matter systems @xcite including the studies of spin and charge ordering in cuprate superconductors @xcite and superfluid transitions in @xmath15he @xcite .
more concretely , the normal - to - planar superfluid phase transition in @xmath16 is governed by an @xmath17 field theory @xcite , which is similar to the one that arises in 2clqcd at @xmath2 .
although some progress has been made in uncovering important qualitative features of the phase diagram of 2clqcd , many quantitative questions remain : ( 1 ) what is the order of the finite temperature chiral transition at zero and non - zero chemical potentials ? ( 2 ) can the low energy physics at small @xmath18 and @xmath0 be understood quantitatively by an appropriate chiral perturbation theory @xcite ?
( 3 ) what is the order of the phase transition that occurs when the lattice gets saturated with baryons at @xmath10 ?
the reason for the lack of quantitative progress can be traced to the fact that the difficulty of calculations are similar to those in qcd at zero chemical potential .
hence , all previous studies have been limited to small lattice sizes and relatively large quark masses .
further , diquark condensation occurs in 2clqcd and is difficult to study in the conventional approach .
one usually has to add a diquark source term and then extrapolate it to zero . fortunately , new monte carlo algorithms have emerged for lattice gauge theories in the limit of infinite gauge coupling @xcite where many of the conventional algorithmic problems disappear .
although this strong coupling limit has the worst lattice artifacts , the qualitative physics is expected to be preserved . since most of the current work is being done at couplings which can be considered rather strong , the phase diagram at these couplings
may not be altered significantly as compared to the infinite coupling theory . on the other hand , thanks to the new algorithms , one can study the chiral limit on large lattices with ease at infinite coupling .
thus , studying strong coupling 2clqcd should be a useful step in our general understanding of the subject .
the strong coupling limit attracted the attention of physicists in the eighties when a variety of qualitative results were obtained using mean field theory and numerical work @xcite .
interestingly , even today many qualitative questions continue to addressed in this limit @xcite .
the strong coupling limit of 2clqcd was originally considered in @xcite and has been recently reviewed in @xcite . however , many interesting questions , including the ones raised above , have remained unanswered even in this simplified limit . in this article
we extend the directed path algorithm invented in @xcite to study strong coupling 2clqcd in the chiral limit and attempt to answer many questions including those raised above .
our article is organized as follows . in section [ model ]
we discuss the model and the expected physics in detail . in section [ dimer ]
we explain the new algorithm which is followed by a section in which we discuss the observables and how we measure them .
section [ results ] contains our results which is followed by a section where we present a summary of our work along with conclusions .
a preliminary version of this work appeared in a recent conference proceedings @xcite .
the action of 2clqcd we study is given by @xmath19 .
\label{scact}\ ] ] the grassmann valued quark fields @xmath20 and @xmath21 , associated to the @xmath1 dimensional lattice site @xmath22 with coordinates @xmath23 , represent row and column vectors with @xmath24 color components . the components will be denoted as @xmath25 and @xmath26 .
the gauge fields @xmath27 are elements of @xmath28 group and live on the links between @xmath22 and @xmath29 where @xmath30 .
the factor @xmath31 for @xmath32 and @xmath33 . at weak couplings @xmath34 acts as the temporal lattice spacing ( assuming spatial lattice spacing is @xmath35 ) .
however there is no reason to expect this interpretation to hold at strong couplings .
thus , we think of @xmath34 as merely an asymmetry factor between spatial and temporal directions .
it allows us to study the effects of temperature on asymmetric lattices and was already used for this purpose in @xcite . in the dimer - baryon loop representation which we will construct later , the parameter @xmath36 is more natural . by choosing an @xmath37 lattice ( periodic in all directions ) we can study thermodynamics in the @xmath38 limit at a fixed @xmath39 by defining @xmath40 as the parameter that represents the temperature
zero temperature studies involve the limit @xmath41 with fixed @xmath18 .
the parameter @xmath0 represents the baryon chemical potential .
the absence of the gauge action shows that we are in the strong gauge coupling limit .
a detailed discussion of the symmetries of 2clqcd can be found in @xcite . for completeness
we review them briefly here .
we first rewrite eq .
( [ scact ] ) as @xmath42 \nonumber \\ & & - \sum_{x\ : { \rm even}}\frac{\eta_{t}(x)}{a_t}\bigg[\overline{x_{e}}(x ) % \left(\begin{array}{cc } % e^{a_t \mu } & 0 \cr % 0 & e^{-a_t \mu}\end{array}\right ) \mathrm{e}^{a_t\mu \sigma_3 } u_{t}(x)x_{o}(x+\hat{t } ) -\overline{x_{e}}(x ) % \left(\begin{array}{cc } % e^{-a_t\mu } & 0\cr % 0 & e^{a_t\mu}\end{array}\right ) \mathrm{e}^{-a_t\mu \sigma_3 } u_{t}^{\dagger}(x-\hat{t})x_{o}(x-\hat{t})\bigg ] \nonumber \\ \label{su2act}\end{aligned}\ ] ] where @xmath43 and @xmath44 are given by @xmath45 in our notation @xmath46 are pauli matrices that mix @xmath47 and @xmath48 present in @xmath49 and @xmath50 while @xmath51 are pauli matrices that act on the color space .
thus , @xmath52 since @xmath53 is an element of @xmath28 . clearly , when @xmath2 our model has a @xmath14 global symmetry : @xmath54 this symmetry is reduced to @xmath55 in the presence of a chemical potential : @xmath56 here @xmath57 is the baryon number symmetry @xmath58 and @xmath59 is the chiral symmetry of staggered fermions @xmath60 , @xmath61 where @xmath62 .
when @xmath2 one expects the chiral condensate , which is not invariant under the @xmath14 symmetry , to get a non - zero vacuum expectation value .
note that @xmath63x_o & x \in \mbox{odd } \cr \frac{i}{2}\bar{x}_e [ \sigma_i\otimes \tau_2]\sigma_2\bar{x}_e^t & x \in \mbox{even } \end{array}\right.\ \ i=1,2,3 \label{opara}\ ] ] transform as components of a three vector under the @xmath28 subgroup of the @xmath14 symmetry group .
it is easy to check that @xmath64 , \hat{\phi}_2 = [ -\chi_1\chi_2+\bar{\chi_1}\bar{\chi}_2 ] , \hat{\phi}_3 = \bar{\chi}\chi = [ \bar{\chi_1}\chi_1 + \bar{\chi}_2\chi_2],\ ] ] thus , the chiral condensate is the third component of a complex three vector .
in addition all the components carry the same non - zero @xmath59 chiral charge .
the above discussion makes it clear that the chiral condensate being non - zero is just a matter of choice .
more generally , the ground state of the theory is such that @xmath65 , where @xmath66 is some constant unit three vector . with this choice
, the ground state still remains invariant under a @xmath67 subgroup given by @xmath68 which implies that the @xmath14 symmetry is broken to a @xmath67 subgroup ( note that the @xmath67 subgroup must be a part of the @xmath28 subgroup of @xmath14 ) .
when one says that the chiral condensate is non - zero , one implicitly chooses @xmath66 along the third direction .
this then implies @xmath69 .
however when we study the effects of the chemical potential , it is natural to pick the ground state such that @xmath70 and @xmath71 which implies that the diquark condensate , @xmath72 while the chiral condensate vanishes .
note that even though @xmath73 the theory still breaks the @xmath59 symmetry since the diquark condensates carry a chiral charge .
when @xmath7 the @xmath14 global symmetry is explicitly broken to @xmath55 . at small @xmath0 both the @xmath67 symmetries
are expected to break spontaneously since the diquark condensate continues to be non - zero .
as @xmath0 increases the density of baryons increases and at a critical value @xmath74 the lattice becomes saturated with baryons which means that for @xmath75 the diquark condensate vanishes .
if this phase transition is second order , the low energy physics close to @xmath74 will be governed by a non - relativistic field theory .
renormalization group arguments indicate that this field theory is governed by mean field theory in @xmath76 ( @xmath77 represents spatial dimensions ) @xcite . at high temperatures
all the symmetries are expected to be restored .
this implies one must have a finite temperature phase transition for @xmath78 .
the order of this transition both at @xmath2 and @xmath6 in not known .
since the order parameter at @xmath2 is a complex 3-vector , its fluctuations are governed by the landau - ginzburg ( lg ) hamiltonian of the @xmath79-component complex field @xmath80 .
the @xmath14 symmetry is manifest as a @xmath17 symmetry .
theories with @xmath81-component complex scalar fields with @xmath3 symmetry are interesting in condensed matter physics in describing a variety of materials @xcite , and are described by the action ( or classical hamiltonian ) , @xmath82 when @xmath83 , depending on the sign of @xmath84 , two classes of ground states are allowed ( note @xmath85 is necessary for stability ) . when @xmath86 the ground state has a _ spiral _ or
_ helical _
order , while when @xmath87 the ground state has _ collinear _ or _ sinusoidal _ order .
since we found that @xmath88 , we discover that close to the finite temperature phase transition the long distance physics of 2clqcd with massless staggered fermions at @xmath2 is described by the above complex field theory with @xmath4 and collinear order ( @xmath89 ) .
this field theory is of interest in the study of the normal - to - planar superfluid transition in @xmath15he @xcite .
the question of whether the @xmath4 theory with collinear order can be second order still remains unresolved . while the @xmath90-expansion predicts a fluctuation driven first order transition @xcite recent results claim that a second order fixed point indeed exists @xcite .
as we will see later , in our work we find a weak first order transition .
it is easy to argue that in the presence of the baryon chemical potential the finite temperature phase transition must be governed by a landau - ginsburg hamiltonian similar to the one above but with @xmath5 .
note that the symmetry in the microscopic theory is reduced to @xmath91 which is manifest in the lg theory as an @xmath9 symmetry .
further , near a finite temperature phase transition the presence of a chemical potential does not usually break charge conjugation symmetries in the low energy effective theory .
for @xmath5 it is possible to define new fields @xmath92 such that the lg hamiltonian becomes @xmath93 + 2 ( u+2v ) |\varphi_1|^2|\varphi_2|^2 \bigg\ } \label{lg2}\end{aligned}\ ] ] it is then obvious that when @xmath89 there is always a decoupled @xmath8 fixed point at @xmath94 @xcite . using the knowledge of the critical exponents in the @xmath8 model
it can be established that this decoupled fixed point is stable @xcite .
however , the flow to this fixed point is rather slow so that corrections to the @xmath8 scaling can be substantial until one is very close to the critical point @xcite . at a non - zero value of @xmath0
we indeed find a second order transition .
a naive analysis indicates that the critical exponents are different from the @xmath8 exponents , however when the expected corrections to scaling are included the @xmath8 exponents can be used to fit our data .
one of the computational advantages of the strong coupling limit is that in this limit it is possible to rewrite the partition function , @xmath95[d\overline\chi d\chi ] \exp(-s),\ ] ] as a sum over configuration containing gauge invariant objects , namely monomers , dimers , and baryonloops @xcite .
monomers are absent in the massless theory . in the case of 2clqcd
a lattice configuration @xmath96 of dimers and baryonloops is constructed as follows : ( 1 ) every link of the lattice connecting the site @xmath22 with the neighboring site @xmath29 contains either a dimer @xmath97 or a directed baryon - bond @xmath98 .
when @xmath99 , it means that the link does not contain a dimer , while @xmath100 implies that the link contains a single ( double ) dimer .
similarly @xmath101 means the link does not contain a baryon - bond , while @xmath102 means the baryon - bond is directed from @xmath22 to @xmath29 and @xmath103 means it is directed from @xmath29 to @xmath22 .
we will also allow @xmath104 to be negative .
thus , if @xmath105 was positive , @xmath106 and @xmath107 will represent dimers and baryon - bonds connecting @xmath22 with @xmath108 .
( 2 ) if a site is connected to baryon - bonds then it must have exactly one incoming baryon - bond and one outgoing baryon - bond .
further it can not be connected to dimers .
thus baryon - bonds always form self - avoiding closed baryonloops .
( 3 ) every lattice site @xmath22 that does not contain a baryon - bond must satisfy the constraint @xmath109 where the direction @xmath105 in the sum takes negative values also .
this implies that sites connected by single dimers also form a loop which we call a dimer - loop .
an example of a configuration @xmath96 is given in figure [ fig0 ] .
an example of a dimer - baryonloop configuration.,scaledwidth=50.0% ] given the set @xmath110 of dimer - baryonloop configurations described above , eq .
( [ scact ] ) can be rewritten as @xcite , @xmath111 since the partition function is written as a sum over positive definite terms , a monte - carlo algorithm can in principle be designed to the study this system .
however , the algorithm needs to preserve many constrains . a method to do this was developed in @xcite in the context of quantum spin models and later extended to dimer models in @xcite .
we will now discuss an extension of these ideas to update the configurations @xmath96 .
in particular we consider three types of updates : a dimer - baryon loop flip update , a dimer update and a baryon update .
below , we discuss each of these updates in detail
. remember that we assume @xmath112 .
the dimer update and baryon updates have been described such that this redundant information also gets updated automatically .
this update is based on the observation that every baryon loop s orientation can be flipped without violating any constraints .
further a baryonloop can be converted into a dimer - loop and vice - versa .
thus , every loop can be in one of three states : a dimer - loop or a baryonloop with two different orientations .
let @xmath113 be the subset of lattice sites that are connected to either a dimer - loop or a baryonloop .
@xmath114 will be the number of these sites .
we pick a site from @xmath113 at random and change the state of the loop @xmath115 connected to that site to one of the three allowed states .
the change can be accomplished using a heat - bath ( or similar ) update if we assign the following weights to the three states : a dimer - loop carries a weight @xmath35 while the baryonloop ( in either orientation ) carries the weight @xmath116 $ ] , where @xmath117 note that @xmath118 changes sign if its orientation is flipped .
let @xmath119 be the set of sites connected by @xmath120 and @xmath121 be the number of such sites .
the dimer update changes the configuration on a subset of @xmath119 .
the update is as follows : 1 . a lattice site @xmath122 is selected randomly 2 .
if the site @xmath22 lies on a dimer - loop , then there will be two different directions @xmath104 such that @xmath123 .
one of these two directions is picked at random .
else there will be one direction @xmath104 such that @xmath124 . in that case
this direction is picked .
3 . if the update just started and @xmath22 is the first site , a virtual monomer is created at @xmath22 . if not one is added to @xmath125 where @xmath126 is the direction from which @xmath22 was reached .
one is subtracted from @xmath127 and the update moves to the neighboring site @xmath128 .
we will call @xmath22 as an `` active site '' and @xmath129 as a `` passive site '' in the notation of @xcite . see next subsection for more details .
with all the neighboring sites @xmath130 that belong to @xmath119 a non - zero weight @xmath131 is associated .
if @xmath132 is a temporal direction then @xmath133 , otherwise @xmath134 . if the neighboring site does not belong to @xmath119 then the weight is zero . for future reference
we define the total dimer weight on the site @xmath129 as @xmath135 .
now based on the weights @xmath131 a heat - bath ( or similar ) procedure is used to pick a new direction @xmath132 .
one is subtracted from @xmath136 and one is added to @xmath137 .
the update then moves to the neighboring site @xmath138 .
if the site @xmath22 is not the site which was picked in step 1 , the update moves to step 2 .
otherwise the site @xmath22 will contain one virtual monomer and one direction @xmath104 such that @xmath139 . with probability half the direction @xmath104
is picked and the update moves to step 3 and with the remaining probability half the virtual monomer on the site is removed and the update ends .
the third update is just a minor modification of the dimer update .
let @xmath140 be the set of sites connected to @xmath141 or containing a baryonloop and @xmath142 the number of such sites .
the baryon update changes the configuration on a subset of @xmath140 .
the update is as follows :
a lattice site @xmath22 in @xmath140 is selected randomly .
2 . if the site @xmath22 lies on a baryonloop , then there will be one direction @xmath104 such that @xmath143 .
this direction is picked . on the other hand
if @xmath22 is not on a baryonloop then there will be one direction @xmath104 such that @xmath124 . in that case
this direction is picked .
3 . if the update just started and @xmath22 is the first site , a virtual `` diquark '' is created at @xmath22 .
otherwise , one is subtracted from @xmath144 , where @xmath126 is the direction from which @xmath22 was reached . if @xmath145 after the subtraction then @xmath146 is set to 2 .
one is added to @xmath147 and if @xmath141 then it is set to zero .
the update moves to the neighboring site @xmath128 .
we will call @xmath22 as an `` active site '' and @xmath129 as a `` passive site '' in the notation of @xcite . see next subsection for more details .
4 . with all the neighboring sites
@xmath130 that belong to @xmath140 a non - zero weight @xmath131 is associated .
if @xmath132 is the positive temporal direction then @xmath148 , if it is along the negative temporal direction then @xmath149 , otherwise @xmath134 . if the neighboring site does not belong to @xmath140 then its weight is zero . for future reference
we define the total baryon weight on the site @xmath129 as @xmath150 .
now based on the weights @xmath131 an over - relaxation procedure is used to pick a new direction @xmath132 .
one is subtracted from @xmath147 and if @xmath124 then @xmath127 is set to zero .
one is added to @xmath144 and if @xmath151 after the addition then @xmath125 is set to two .
the update then moves to the site @xmath138 .
if the site @xmath22 is not the site which was picked in step 1 , the update moves to step 2 .
otherwise the site @xmath22 will contain one virtual diquark and it would have been reached from the direction @xmath132 .
one is subtracted from @xmath144 and if @xmath145 after the subtraction then @xmath146 is set to 2 .
the virtual diquark on the site is removed and the update ends . in the definition of the dimer and baryon updates we have defined active and passive sites .
the passive sites play an important role during the measurement of observables .
hence we clarify these two class of sites further .
both the dimer and baryon updates are directed loop updates .
they start on a site @xmath22 , which is called an active site .
then they go through a series of sites which are referred to as passive and active alternately .
thus the second site is a passive site , the third is an active site and so on . if the first site is such that @xmath152 then all passive sites @xmath129 , visited during the update , have @xmath153 and vice - versa .
the weights @xmath154 and @xmath155 for passive sites encountered during the updates will play an special role in the measurement of correlation functions as discussed below .
each of the three updates satisfy detailed balance .
the proof of detailed balance for the the dimer - baryon loop flip update is straight forward and conventional . on the other hand the proofs for the dimer and baryon updates
need some understanding of directed loop algorithms .
once this is clear , the proof is essentially straight forward .
we refer the reader to @xcite and do not prove detailed balance of these two algorithms in this article .
the combination of the three updates makes the algorithm ergodic . to see this
we note that there is always possible to flip all the baryonloops into dimer - loops .
once this is done one can use the proof given in @xcite to show that dimer updates are ergodic in the space of configurations that purely consist of dimers .
a variety of observables can be measured with our algorithm .
we will focus on the following : * the chiral two point function is given by @xmath156 and the chiral susceptibility @xmath157 is @xmath158 where @xmath159 .
both these observables can be measured easily during the dimer update .
it is possible to show that @xcite @xmath160 where @xmath22 is the first site of the dimer update and the sum is over all the passive sites @xmath129 encountered during the update . @xmath121 and
@xmath154 were defined in the previous section . * the diquark two point function is given by @xmath161 and the diquark susceptibility @xmath162 is given by @xmath163 these can be computed during the baryon update by using @xmath164 where @xmath22 is the first site picked during the update and the sum is over all the passive sites @xmath129 encountered during the baryon update . @xmath142 and @xmath155 were defined in the previous section .
* baryon density @xmath165 is defined as @xmath166 and in the dimer - baryon loop flip update it can be measured by the formula @xmath167 where @xmath115 is the dimer - baryon loop picked , @xmath168 is the fraction of the lattice sites that contain dimer - baryon loops , @xmath169 is the size of the loop and @xmath118(defined earlier ) is the temporal winding of the baryon loop . *
the helicity modulus associated with the @xmath67 chiral symmetry
@xmath170 @xmath171 where @xmath172 and @xmath173 . * the helicity modulus associated with the @xmath67 baryon number symmetry @xmath174
@xmath175 where @xmath176\ ] ] both @xmath177 and @xmath178 are observables that can be calculated easily for each configuration @xmath96 and averaged . note also that our definitions of @xmath177 and @xmath178 are natural at finite temperatures .
in a finite volume there is no spontaneous symmetry breaking .
however , the effects of symmetry breaking can still be studied by examining the large volume limit of @xmath179 and @xmath180 ; if the symmetry is broken in the infinite volume limit , we expect these susceptibilities to grow with the volume of the system . from symmetry considerations at @xmath2 it
is possible to show that @xmath181 which implies @xmath182 .
symmetry breaking can also be observed through the helicity modulus @xmath177 and @xmath178 ; both must reach a non - zero constant if the symmetry breaking pattern is as expected .
all these can be understood quantitatively using the low energy effective action @xmath183 where @xmath184 and @xmath185 are unit three and two vector fields respectively .
finite size scaling formula for various quantities can be obtained following the discussion in @xcite .
we note that this approach to low energy physics is equivalent to others found in the literature @xcite . at a fixed value of @xmath39
the parameter @xmath18 can be increased to induce a phase transition between a low temperature phase with spontaneous symmetry breaking and a high temperature symmetric phase . in order to study this phase transition
we have performed extensive calculations at a fixed @xmath186 for different spatial lattice sizes @xmath187 varying from @xmath188 to @xmath189 and for many different values of @xmath18 .
the low energy effective theory introduced in eq .
( [ eft ] ) with @xmath190 can then be used to predict the signatures of the broken phase .
we have studied two such signatures : ( 1 ) @xmath177 and @xmath178 go to non - zero constants at large @xmath187 .
extending the calculations of @xcite one can show @xcite @xmath191 ( 2 ) the finite size scaling of the chiral susceptibility is given by @xcite @xmath192 where @xmath193 is the shape coefficient for cubic boxes @xcite and @xmath194 .
figure [ fig1 ] gives our results at @xmath195 which is a value of @xmath18 in the broken phase .
the graph shows that the above expectations are satisfied well .
the inset shows the data for @xmath177 and @xmath178 as a function of @xmath187 .
the lines shown are fits to eq .
( [ fss2 ] ) which yields @xmath196 and @xmath197 .
the main figure shows the plot of @xmath179 versus @xmath187 for @xmath195 . the solid line is a fit to eq .
( [ fss1 ] ) with @xmath198 and @xmath199 fixed to the values quoted above .
the fit yields @xmath200 and @xmath201 with a @xmath202.,scaledwidth=80.0% ] figure [ fig2 ] shows the dependence of @xmath179 as a function of @xmath187 for different temperatures . using the data for @xmath203 we find that @xmath179 increases as @xmath204 for @xmath205 , but saturates for @xmath206 as @xmath187 becomes large .
thus , we think @xmath207 is between these two temperatures . for a second order transition , close to @xmath207 ,
one expects @xmath208 , \label{chictc } \\
y_c l & = & [ f_0 + f_1 ( t / t_c-1 ) l^{1/\nu}+ .... ] , \label{yctc}\end{aligned}\ ] ] which was recently observed in other strong coupling theories @xcite .
our data does not fit well to this form .
clearly , the data for @xmath179 shows more structure than can be captured by the above relations . we have also verified that @xmath209 does not seem to scale as a constant for large @xmath187 anywhere in the range @xmath210 .
hence we think that the transition is not second order .
interestingly , we are able to fit the non - monotonic behavior in @xmath179 at @xmath211 and @xmath212 using the relation @xmath213 as long as we use data for @xmath214 .
this form is natural in the presence of two phases ( one broken and one symmetric ) whose free energy densities differ by @xmath215 .
this leads us to conclude that the phase transition is indeed first order .
this figure shows the plot of @xmath179 versus @xmath187 for different values of @xmath18 across the phase transition .
the solid lines for @xmath216 are fits to eq .
( [ fss1 ] ) while those for @xmath211 and @xmath217 are fits to eq .
( [ fss3 ] ) .
we find @xmath218 , @xmath219 , @xmath220 and @xmath221 for the two temperatures .
all the fits have @xmath222 less than @xmath35.,scaledwidth=80.0% ] the existence of a first order transition implies that the correlation lengths at the critical point do not diverge . in that case
how big are these lengths at the transition ? in the high temperature phase one can compute screening masses @xmath223 , @xmath224 from the exponential decay of @xmath225 and @xmath226 respectively , for large spatial separations between @xmath227 and @xmath228 . at @xmath2
we expect @xmath229 . in the broken phase @xmath198 and
@xmath199 have dimensions of mass and are the relevant physical scales in the problem . in figure [ fig3 ]
we show the behavior of @xmath224 , @xmath198 and @xmath199 which have been obtained after extrapolations to infinite volumes . as can be seen ,
the correlation lengths close to the transition are extremely large , about @xmath230 lattice units , indicating that the transition is a weak order transition .
if the transition was second order , we would have expected @xmath231 , @xmath232 and @xmath233 .
interestingly , these relations describe the data reasonably well , but with different @xmath207 and @xmath234 in the two phases . in the low temperature phase a combined fit of @xmath198 and @xmath199 gives @xmath235 , @xmath236 , @xmath237 and @xmath238 with a @xmath239 . on the other hand
the fit of @xmath224 gives @xmath240 , @xmath241 and @xmath242 with a @xmath243 .
a combined fit of all the data on both sides of the transition with a single @xmath207 and @xmath234 does not fit well confirming our claim that the transition is first order .
this figure shows the plot of @xmath224 , @xmath198 and @xmath199 as a function of @xmath18 .
the solid lines are fits to the data as discussed in the text.,scaledwidth=80.0% ] having established a first order transition at zero chemical potential we next focus on the finite temperature transition at @xmath244 with @xmath186 .
the chemical potential reduces the symmetry of the theory to @xmath55 . at low temperatures both
the @xmath67 symmetries are broken leading to two goldstone bosons .
the two correlators @xmath245 and @xmath226 are no longer related : @xmath225 decays exponentially while @xmath226 remains non - zero for large @xmath246 .
this means @xmath179 saturates for large @xmath187 while @xmath180 grows with the volume and shows the presence of a diquark condensation and baryon superfluidity . @xmath177 and @xmath178 again go to non - zero constants at large @xmath187 .
in order to determine the finite size scaling formula , we use the same effective theory as given in eq .
( [ eft ] ) except that now both @xmath184 and @xmath185 are unit two vectors .
the large @xmath187 limits of @xmath177 and @xmath178 are now given by @xmath247 the effective field theory also predicts that @xmath180 is given by @xmath248 where @xmath249 .
interestingly , we find that the values of @xmath178 and @xmath177 come together as we increase the chemical potential , although for small temperatures and small chemical potentials we can still distinguish between them . on the other hand close to the finite temperature phase transition they become indistinguishable .
we note that the action in eq .
( [ lg2 ] ) predicts @xmath250 close to the phase transition .
the operator which splits them is irrelevant and goes to zero at the transition .
however , this explanation does not explain why @xmath177 and @xmath178 come close to each other as a function of @xmath0 . this behavior should be examined using effective field theory and may emerge naturally , but we have not attempted it so far .
figure [ fig6 ] gives our results at @xmath251 , a value of @xmath18 in the broken phase .
the inset shows the behavior of @xmath178 ( @xmath177 looks identical within errors ) .
fitting the data , we find @xmath252 with a @xmath222 of around @xmath253 . fixing @xmath198 and @xmath199 to these values , our data for @xmath180 fits well to eq .
( [ fss4 ] ) as long as we use @xmath254 .
we get @xmath255 and @xmath256 with @xmath257 .
the fit is shown as a solid line in the figure .
this figure shows @xmath180 as a function of @xmath187 at @xmath244 at three different temperatures close to the critical temperature .
the inset shows that @xmath178 goes to a constant as a function of @xmath187 at @xmath251 .
the solid line is a plot of eq .
( [ fss4 ] ) with @xmath258 and @xmath256 . clearly , @xmath180 grows as @xmath204 at @xmath251 but begins to saturate at @xmath259 indicating that there is a transition between these two temperatures.,scaledwidth=80.0% ] figure [ fig6 ] also shows @xmath180 at @xmath260 and @xmath261 as a function of @xmath187 . unlike the @xmath2 case , @xmath180 behaves monotonically suggesting that the transition could be second order . in order to check this
we can verify if eq .
( [ yctc ] ) is satisfied close to @xmath207
. a fit of our data to this relation gives @xmath262 , @xmath263 @xmath264 and @xmath265 with a @xmath266 . in figure [ fig7 ] ( plot on the left )
, we show our data and the fit . in the inset of the figure we plot @xmath180 at @xmath267 .
since this value of @xmath18 is close to @xmath207 we expect @xmath268 should describe the data reasonably well .
indeed a fit shows @xmath269 , @xmath270 with a @xmath222 of @xmath271 and is shown as a solid line in the inset .
the values of @xmath198 , @xmath272 and @xmath224 obtained from the infinite volume extrapolations must also satisfy @xmath273 figure [ fig7 ] ( plot on the right ) shows a combined fit of these three quantities as a function of @xmath18 using the value of @xmath207 obtained above .
we find @xmath274 , @xmath275 , @xmath276 , @xmath277 and @xmath278 with a @xmath279 .
note that these critical exponents do satisfy the hyper - scaling relation @xmath280 .
thus , our data strongly supports the existence of a second order transition .
unfortunately , the above results are in contradiction with the expectation from section [ ftpt ] , where it was argued that the critical behavior at @xmath6 must be governed by three dimensional @xmath8 universality class .
this implies that we should have obtained @xmath281 , @xmath282 and @xmath283 @xcite and not the exponents we found above . as discussed in section [ ftpt ] ,
the problem is that in a model with @xmath91 symmetry and collinear order , the corrections to the @xmath8 scaling , due to the irrelevant direction in the @xmath284 plane ( see section [ ftpt ] ) , are rather large .
taking into account the leading corrections to scaling one expects @xmath285 close to @xmath207 where @xmath286 @xcite .
the smallness of @xmath287 makes the corrections large for the lattice sizes we have explored .
in fact we find that our data also fits well to this corrected form if we use the @xmath8 critical exponents and the known value of @xmath287 in the fits .
a combined four parameter fit of our data close to @xmath207 now yields @xmath288 , @xmath289 , @xmath290 , @xmath291 with a @xmath292 .
this fit is shown in figure [ fig8 ] ( the plot on the left ) . at @xmath293
, @xmath180 fits well to the form @xmath294 when @xmath295 is fixed to @xmath296 .
the fit gives @xmath297 with a @xmath222 of @xmath298 ( solid line in the inset of figure [ fig8 ] ) .
when the scaling corrections are included in @xmath198 , @xmath272 and @xmath224 one gets [ fdtcxy ] @xmath299 fixing @xmath300 and using the @xmath8 critical exponents and @xmath287 as above , a combined fit again works very well and is shown in the plot on the right in figure [ fig8 ] .
we get @xmath301 , @xmath302 , @xmath303 , @xmath304 , @xmath305 and @xmath306 with a @xmath266 .
results at @xmath307 and @xmath308 .
note that @xmath180 grows with the volume while @xmath179 saturates .
both @xmath177 and @xmath178 grow linearly with @xmath187 .
the lines are drawn to guide the eye.,scaledwidth=80.0% ] the figure shows our results for @xmath309 , @xmath310 and @xmath311 as a function of @xmath0 . for @xmath312
the dotted line is the mean field result for @xmath309 given in eq .
( [ mfdelta ] ) and the solid line contains the one loop corrections . for @xmath313
the dashed line is the plot of @xmath314 and the solid line is the plot of @xmath315 .
we have used @xmath316 here .
the inset shows the baryon density @xmath317 as a function of @xmath0.,scaledwidth=80.0% ] next we turn to the physics at zero temperature . for this purpose
we compute quantities with @xmath318 at @xmath307 for various values of @xmath0 and @xmath187 .
we now expect @xmath319 where @xmath320
. note that since we use the same finite size scaling form in four and three dimensions , @xmath309 is normalized differently here as compared to the finite temperature case .
the chiral susceptibility @xmath179 , must again saturate with @xmath187 at any non - zero value of @xmath0 .
finally , due to our definitions ( see eqs.([yc ] ) , ( [ yb ] ) ) both the helicity modulus @xmath177 and @xmath178 , grow linearly with @xmath187 for large @xmath187 .
these expectations emerge nicely in our calculations as can be seen in figure [ fig4 ] . as the chemical potential increases the average number of baryons in the ground state increases .
at some critical chemical potential @xmath74 , the ground state has a baryon on every lattice site .
since the baryons behave as hard - core bosons , at @xmath74 there must be a phase transition to a phase where superfluidity is no longer present .
we now focus on this phase transition .
renormalization group arguments show that this phase transition must be governed by mean field theory @xcite .
a mean field analysis was performed recently in @xcite and the critical chemical potential was found to be @xmath321 .
the diquark condensate was shown to be @xcite @xmath322 in order to check these results we have computed the diquark condensate @xmath309 by fitting our @xmath180 data to the relation @xmath323 $ ] .
figure [ fig5 ] shows our data along with the mean field result @xcite and the result with one - loop corrections @xcite .
clearly , the one - loop corrections are necessary before connection with mean field theory can be established .
the fact that this phase transition is driven due to the saturation of the lattice with baryons can be seen in the inset of figure [ fig5 ] , where the baryon density is plotted as a function of @xmath0 .
for @xmath324 it costs an energy @xmath311 to remove a baryon .
this energy gap grows linearly as @xmath325 .
thus for @xmath313 one obtains a phase containing non - relativistic particles whose dispersion relation for small momenta looks like @xmath326 which leads to a spatial correlation length @xmath327 . since we expect @xmath328 to scale as @xmath329 close to @xmath74 one expects the kinetic mass @xmath330 to be a constant .
figure [ fig5 ] shows the plot of @xmath331 and @xmath310 as a function of @xmath0 .
we find @xmath332 with a @xmath333 and @xmath334 with a @xmath335 .
again @xmath74 is in excellent agreement with the mean field result which provides strong evidence that the phase transition is second order and belongs to the mean field universality class . from the behavior of @xmath311 and @xmath328 we find @xmath336 .
in this work we constructed an efficient cluster algorithm and studied the phase structure of two color lattice qcd with massless staggered fermions in the strong coupling limit .
we found that the finite temperature phase transition at zero chemical potential is weakly first order , while the same transition at an intermediate value of the chemical potential was second order .
this second order transition was found to be in the universality class of the three dimensional @xmath8 model as expected from theoretical arguments .
however , in order to show this we needed to include the large corrections to scaling expected in the theory .
the quantum phase transition at zero temperature between a baryon superfluid phase and a normal phase was also found to be second order in the mean field universality class .
the physics in the normal phase was that of interacting non - relativistic particles .
based on these observations , we can attempt to draw the full diagram of two color qcd in the strong coupling limit .
our proposal is shown in figure [ pdiag ] .
the phase diagram of two - color lattice qcd with staggered fermions at strong couplings.,scaledwidth=50.0% ] it is important to understand how this phase diagram will change as we go to weaker couplings and towards the continuum limit . clearly , when the four flavor nature of staggered fermions becomes important , it can begin to change significantly .
however , for couplings that are currently explored , universality arguments suggest that the phase diagram will qualitatively remain the same although quantitatively the values of the critical points will mostlikely get affected .
it is also interesting to ask why the transition at @xmath2 is so weak . as discussed in section [ model ] ,
renormalization group studies based on the @xmath90-expansion indicate that the transition will generically be a fluctuation driven first order transition @xcite .
such a transition could be weak as we observe .
on the other hand , recent work based on high order perturbation theory combined with resummation techniques do not rule out the possibility of a second order transition @xcite .
this means we are in the vicinity of a tricritical point and by changing some parameter in the theory we could change the transition to second order .
when this occurs , it is likely that the weakly first order line in the above phase diagram will disappear .
such a scenario can in principle be investigated by introducing more tunable parameters within our model and by studying their effects on the order of the transition .
it is amusing to note that the above phase diagram is different from the standard phase diagram in qcd where the baryon chemical potential induces a first order transition instead of weakening it .
this non - standard scenario was discussed in @xcite as a possibility in qcd and has also been seen in the potts model simulations @xcite .
finally , we note that the physics close to the quantum critical point is also interesting from the point of view of non - relativistic field theory .
since the model has a @xmath91 symmetry , the low energy effective field theory here is richer than in a theory with a simple @xmath67 particle number symmetry .
we would like to thank e. vicari for his time for explaining to us the existence of the decoupled @xmath8 fixed point and how the corrections to scaling may be important in our work .
we also thank ph .
de forcrand , s. hands , c. strouthos , t. mehen , r. springer , d. toublan and u .-
j.wiese for many helpful comments .
this work was supported in part by the department of energy grant de - fg02 - 03er41241 .
the computations were performed on the champ , a computer cluster funded in part by the doe .
below we list the exact expressions for various observables on a @xmath338 lattice .
we have tested the algorithm against these exact results .
tables [ tabe1 ] and [ tabe2 ] show the comparison of exact results against those obtained using the algorithm .
k. rajagopal and f. wilczek , arxiv : hep - ph/0011333 . t. schafer , arxiv : hep - ph/0304281 .
d. h. rischke , prog .
. phys . * 52 * , 197 ( 2004 ) [ arxiv : nucl - th/0305030 ] .
m. alford , prog .
suppl . * 153 * , 1 ( 2004 ) [ arxiv : nucl - th/0312007 ] .
m. a. stephanov , prog .
suppl . *
153 * , 139 ( 2004 ) [ arxiv : hep - ph/0402115 ] .
p. de forcrand and o. philipsen , prog .
suppl . * 153 * , 127 ( 2004 ) .
m. p. lombardo , prog .
theor . phys .
* 153 * , 26 ( 2004 ) [ arxiv : hep - lat/0401021 ] . c. r. allton , s. ejiri , s. j. hands , o. kaczmarek , f. karsch , e. laermann and c. schmidt , nucl .
suppl . * 141 * , 186 ( 2005 ) [ arxiv : hep - lat/0504011 ] .
o. philipsen , arxiv : hep - lat/0510077 .
j. b. kogut , m. a. stephanov , d. toublan , j. j. m. verbaarschot and a. zhitnitsky , nucl .
b * 582 * , 477 ( 2000 ) [ arxiv : hep - ph/0001171 ] .
j. b. kogut , m. a. stephanov and d. toublan , phys .
b * 464 * , 183 ( 1999 ) [ arxiv : hep - ph/9906346 ] .
j. wirstam , phys .
d * 62 * , 045012 ( 2000 ) [ arxiv : hep - ph/9912446 ] . c. ratti and w. weise , phys .
d * 70 * , 054013 ( 2004 ) [ arxiv : hep - ph/0406159 ] .
j. t. lenaghan , f. sannino and k. splittorff , phys .
d * 65 * , 054002 ( 2002 ) [ arxiv : hep - ph/0107099 ] .
j. b. kogut , j. polonyi , h. w. wyld and d. k. sinclair , nucl .
b * 265 * , 293 ( 1986 ) .
j. b. kogut , nucl .
b * 290 * , 1 ( 1987 ) .
o. kaczmarek , f. karsch and e. laermann , nucl .
suppl . * 73 * , 441 ( 1999 ) [ arxiv : hep - lat/9809059 ] .
s. hands , j. b. kogut , m. p. lombardo and s. e. morrison , nucl .
b * 558 * , 327 ( 1999 ) [ arxiv : hep - lat/9902034 ] . s. hands , i. montvay , s. morrison , m. oevers , l. scorzato and j. skullerud , eur .
j. c * 17 * , 285 ( 2000 ) [ arxiv : hep - lat/0006018 ] .
j. b. kogut , d. k. sinclair , s. j. hands and s. e. morrison , phys .
d * 64 * , 094505 ( 2001 ) [ arxiv : hep - lat/0105026 ] .
j. b. kogut , d. toublan and d. k. sinclair , phys .
d * 68 * , 054507 ( 2003 ) [ arxiv : hep - lat/0305003 ] .
j. i. skullerud , s. ejiri , s. hands and l. scorzato , prog .
theor . phys .
* 153 * , 60 ( 2004 ) [ arxiv : hep - lat/0312002 ] .
h. kawamura , phys .
rev . b*38 * , 4916 ( 1988 ) ; erratum b*42 * , 2610 ( 1990 ) .
d. h. adams and s. chandrasekharan , nucl .
b * 662 * , 220 ( 2003 ) [ arxiv : hep - lat/0303003 ] .
j. m. blairon , r. brout , f. englert and j. greensite , nucl .
b * 180 * , 439 ( 1981 ) .
n. kawamoto and j. smit , nucl . phys .
b * 192 * , 100 ( 1981 ) .
h. kluberg - stern , a. morel and b. petersson , nucl .
b * 215 * , 527 ( 1983 ) .
o. martin and b. siu , phys .
b * 131 * , 419 ( 1983 ) .
rossi and u. wolff , nucl .
b * 248 * , 105 ( 1984 ) .
u. wolff , phys .
b * 153 * , 92 ( 1985 ) .
e. dagotto , a. moreo and u. wolff , phys .
lett . *
57 * , 1292 ( 1986 ) .
e. dagotto , f. karsch and a. moreo , phys .
b * 169 * , 421 ( 1986 ) .
e. dagotto , a. moreo and u. wolff , phys . lett .
b * 186 * , 395 ( 1987 ) .
f. karsch and k. h. mutter , nucl .
b * 313 * ( 1989 ) 541 .
j. u. klatke and k. h. mutter , nucl .
b * 342 * , 764 ( 1990 ) .
g. boyd , j. fingberg , f. karsch , l. karkkainen and b. petersson , nucl .
b * 376 * , 199 ( 1992 ) .
b. bringoltz and b. svetitsky , phys .
d * 68 * , 034501 ( 2003 ) [ arxiv : hep - lat/0211018 ] .
b. bringoltz and b. svetitsky , phys .
d * 69 * , 014502 ( 2004 ) [ arxiv : hep - lat/0310032 ] . b. bringoltz , arxiv : hep - lat/0511058 . y. nishida , k. fukushima and t. hatsuda , phys . rept . * 398 * , 281 ( 2004 ) [ arxiv : hep - ph/0306066 ] .
s. chandrasekharan and f. j. jiang , pos * lat2005 * , 198 ( 2005 ) [ arxiv : hep - lat/0509117 ] .
fisher , p.b .
weichman , g. grinstein and d.s .
fisher , phys .
b*40 * , 546 ( 1989 ) .
p. hasenfratz and h. leutwyler , nucl .
b * 343 * , 241 ( 1990 ) . in the conference proceedings @xcite , eq .
( [ fss2 ] ) and eq .
( [ fss1 ] ) were incorrect leading to the wrong normalization of @xmath199 .
the correct normalization improves some of the fits .
s. chandrasekharan and c. g. strouthos , phys .
lett . * 94 * , 061601 ( 2005 ) [ arxiv : hep - lat/0410036 ] .
m. campostrini , m. hasenbusch , a. pelissetto , p. rossi and e. vicari , phys .
b63 , 214503 ( 2001 ) .
there is a factor of two mismatch in the formula quoted here and the one in @xcite .
the origin of this mismatch is the normalization of our kinetic term in eq .
( [ scact ] ) as compared to the one in @xcite . | we study thermodynamics of strongly coupled lattice qcd with two colors of massless staggered fermions as a function of the baryon chemical potential @xmath0 in @xmath1 dimensions using a new cluster algorithm .
we find evidence that the model undergoes a weak first order phase transition at @xmath2 which becomes second order at a finite @xmath0 .
symmetry considerations suggest that the universality class of these phase transitions should be governed by a @xmath3 field theory with collinear order , with @xmath4 at @xmath2 and @xmath5 at @xmath6 . the universality class of the second order phase transition at @xmath7 appears to be governed by the decoupled @xmath8 fixed point present in the @xmath9 field theory .
finally we show that the quantum ( @xmath10 ) phase transition as a function of @xmath0 is a second order mean field transition . |
for over a decade now , astronomers have attempted to estimate the ages of high - redshift galaxies using broad - band optical - infrared photometry ( _ e.g. _ lilly 1988 ; dunlop et al .
1989 , chambers & charlot 1990 ) .
unfortunately , however , the derived ages have been rendered virtually meaningless by disagreements between modellers over post main - sequence evolution ( _ e.g. _ charlot , worthey & bressan 1996 ) , and by the extreme susceptibility of such relatively crude broad - band data to dust reddening , emission - line contamination etc .
in contrast , it has long been anticipated that relatively robust age constraints for high - redshift galaxies could be derived given rest - frame near - ultraviolet spectra of sufficient quality .
this is because , for the potential ages of interest at @xmath0 ( i.e. ages @xmath5 5 gyr ) , the ultraviolet light of a stellar population is expected to be dominated by stars close to the turn - off point of the ` well - understood ' main sequence ( ms ) ( _ e.g. _ magris & bruzual 1993 ) . with the advent of deep optical spectroscopy on 10-m class telescopes , it has now proved possible to put this technique into practice . in particular , dunlop et al .
( 1996 ) were able to use a deep keck spectrum of the @xmath6 radio galaxy lbds 53w091 to first confirm that its near - ultraviolet spectrum was indeed dominated by starlight , and then to extract an age constraint of @xmath7gyr based primarily on comparison with a main - sequence only model of an evolving stellar population .
spinrad et al .
( 1997 ) explored further the reliability of this age estimate , and confirmed that the best agreement between ages derived using alternative evolutionary synthesis models was obtained if fitting was confined to the detailed shape of the near - ultraviolet spectral energy distribution .
not surprisingly , given its implications for cosmology ( for @xmath8 , the age of an einstein - de sitter universe at @xmath9 is only 2.3 gyr ) , this result has been the subject of subsequent close scrutiny , and claims that 53w091 is in fact less than 2 gyr old have been put forward by , for example , bruzual & magris ( 1997 ) . however , dunlop ( 1999 ) has argued that such young ages are only deduced using some models if the near - infrared photometry is also included in the fitting process , once again placing undesirable emphasis on the reliability of the modelling of post main - sequence evolution ( a point previously also explored by spinrad et al .
moreover , dunlop ( 1999 ) has shown that , certainly for the slightly redder @xmath10 galaxy 53w069 , if fitting is confined to the keck spectroscopic data ( dey et al .
2000 ) , the models of bruzual & charlot ( 1993 ) , worthey ( 1994 ) , and jimenez et al .
( 2000a ) _ all _ lead to the conclusion that its stellar population is @xmath11 gyr old ( assuming solar metallicity ) .
most recently , however , the reliability of even this near - ultraviolet spectroscopic age - dating has been called into question by yi et al . ( 2000 ) .
yi et al .
( 2000 ) , claim to have derived a much younger age for 53w091 , but also claim that this age is not due to differences in post - ms evolution , but rather to the fact that the spectrum produced by the main sequence in their models reddens much more rapidly than in the models of jimenez et al .
( 2000a ) . in support of their revised age estimates , yi et al .
cite the fact that their models can reproduce the spectrum of the sun at an age of 5 gyr , whereas the solar spectrum is not reproduced by the jimenez et al .
models until 8 - 10 gyr .
it is unclear to us why a stellar population should be expected to mimic the spectrum of the sun at its current age ( @xmath12 gyr ) ; even if the light from the stellar population is dominated by stars near the main - sequence turnoff the sun is not expected to leave the main sequence until an age of @xmath2 gyr ( jorgensen 1991 ) .
nevertheless this claim has motivated us to explicitly check the calibration of the main - sequence evolution in alternative evolutionary synthesis models of galaxy evolution .
this is the main subject of the present paper .
what we have done is to take the 3 alternative and independent models of galaxy evolution developed by yi et al .
( 2000 ) , jimenez et al .
( 2000a ) and worthey ( 1994 ) , and to check how rapidly they evolve to mimic the solar spectrum with and ( more importantly ) without inclusion of their post - ms components .
the models are summarized in section 2 , and the results of comparison with the solar near - ultraviolet spectrum are presented in section 3 .
we then proceed , in section 4 , to use these models ( again with and without post - ms components ) to check explicitly the extent to which the age estimates of 53w091 and 53w069 really are affected by different approaches to modelling post - ms evolution .
the main remaining uncertainty is the impact of having to assume a value for the metallicity of the stellar population , and in section 5 we explore whether , given near - ultraviolet data of sufficient quality , it may be possible to break the well - known age - metallicity degeneracy .
finally , our conclusions are summarized in section 6 .
this study was motivated by the apparent disagreement reported by yi et al .
( 2000 ) between their own models and those of jimenez et al .
( 2000a ) .
we are able to perform our own comparison of these models because sukyoung yi has kindly supplied us with his model seds up to an age of 5 gyr ( yi , private communication ) .
we also wanted to compare the predictions of a third independent set of evolutionary synthesis models , and have chosen the models of worthey ( 1994 ) for this purpose .
the reason for this choice was that we already knew , from our previous modelling of 53w091 ( dunlop et al .
1996 , spinrad et al . 1997 ) that the models of worthey ( 1994 ) also appear to yield younger ages than those of jimenez et al .
( 2000a ) , but for reasons which we suspected were primarily due to a different treatment of post - ms evolution ( see charlot et al .
1996 ) .
since the primary objective in this paper is to check the calibration of ms evolution using the sun , we have confined our attention only to models which assume solar metallicity .
however , we _ are _ interested in removing any potential confusion introduced by different treatments of post - ms evolution , because it is hard to be sure that the uv spectrum of an instantaneous starburst really is completely dominated by ms stars in all alternative models beyond an age of 2 - 3 gyr ( and beyond 5 gyr it is not expected to be ) .
we have therefore constructed a ms - only version of the models of jimenez et al .
( 2000a ) , and have also been supplied with a ms - only version of the models of worthey ( 1994 ) ( worthey , private communication ) . in both the jimenez and the worthey models
the isochrones have been cut off at the same point ( corresponding to point 55 in the vandenberg grid in the case of worthey s models ) . a pure ms - only version of the yi et al .
models was not available to us , but we do have access to a main - sequence + giant branch ( msgb ) version , which is stripped of horizontal branch and asymptotic giant branch contributions , and should be an excellent approximation to an ms - only model , at least in the near - ultraviolet for ages @xmath13 5 gyr , as yi et al . themselves claim .
since our main aim in this work was to calibrate the age - dating of distant stellar populations based on the rest - frame near - ultraviolet spectra , fitting was deliberately restricted to the spectral range @xmath15 .
the best fit was determined by binning the data to the same spectral resolution as the model in question , and then varying the age and normalization as free parameters until @xmath14 was minimized .
for the high - redshift galaxies 53w091 and 53w069 , the error on each binned spectral datapoint was derived from propagation of the original errors in the keck optical spectra ( see dunlop et al .
1996 ; dey et al .
2000 ) . in the case of the sun
, we are using the theoretical spectrum of kurucz as the best available representation of the true solar sed , and have simply assumed a constant flux - density error ( i.e. independent of wavelength ) , adjusted in size until reduced chi - squared ( @xmath16 ) equalled unity for the very best fitting model .
it is thus only possible to compare the relative ( rather than absolute ) ability of the different models to reproduce the solar ultraviolet spectrum as the age of each model is varied .
in figure 1a we show reduced @xmath17as a function of age for each of the full stellar population models when age is varied in an attempt to best reproduce the solar spectrum .
the models of yi et al .
predict the youngest age , indicating that the near - ultraviolet spectrum produced by these models best mimics that of the sun after an age of only 4 gyr .
the models of worthey yield a best - fit age of 5 gyr , while those of jimenez et al .
predict an age of 8 gyr . from this plot
it might appear that it is the models of jimenez et al .
that are most unusual , but it is important to note that i ) it is the models of jimenez et al .
which yield the best quality of fit to the solar spectrum , and ii ) it is to be expected that the full stellar population models will _ underestimate _ the main - sequence turn - off ( msto ) age of the sun , because of the inclusion of post - ms stars .
moreover , from this plot it is completely unclear how much of the ( substantial ) difference between the derived best - fit ages can be attributed to differing contributions of post - ms stars to the galaxy ultraviolet seds . for a meaningful comparison between galaxy synthesis models and the expected ms turn - off age of the sun , we really require to ensure that the ultraviolet spectra produced by the models are completely ms dominated .
therefore , in figure 1b we show reduced @xmath14 as a function of age in an analogous way to figure 1a , but this time using models , stripped as far as possible of post - ms contributions .
this allows not only a sensible comparison with the sun , but also makes it possible to assess the relative impact of post - ms contributions to the full model fits shown in figure 1a .
this plot makes it clear that it is the models of yi et al .
that are unusual , and apparently in error by a factor of two in terms of rate of ms evolution . stripped of hb and agb contributions the predictions of the yi et al .
models are little changed , indicating a best - fit age of 4 - 5 gyr .
in contrast , stripped of post - ms evolution the models of worthey and jimenez et al .
appear to be in good agreement not only with each other ( 9 gyr and 11 gyr respectively ) but also with the generally accepted ms turnoff age of the sun ( 10.5 gyr - jorgensen 1991 ) .
it is hard to assess the likely impact on the predictions of the yi et al .
models if the gb was also removed , but the lack of any dramatic change upon removal of hb and agb contributions does tend to support their own claim that their models are already highly ms dominated at ages @xmath5 5 gyr .
in contrast the contribution of post - ms stars in the models of worthey must be relatively strong , even at 5 gyr , because removal of hb + agb moves the derived age from 5 to 7 gyr , and subsequent removal of remaining post - ms stars completes the shift to 9 gyr as shown in figure 1b .
this therefore backs up the suggestion made in jimenez et al .
( 2000a ) that the main difference between the models of jimenez et al . and worthey lies in the strength of the agb and rgb , but that the ms evolution in both models is very similar , and yields sensible values for the turn - off age of the sun .
these results are tabulated in table 1 , and in figure 2 we show the best - fit model spectra superimposed on the solar spectrum .
it is worth noting ( from both figure 2 and figure 1 ) that , as one would hope , the ms - only models in every case yield a better _ quality _ of fit to the solar spectrum than do the full models which include some contribution from more evolved stars .
interestingly , the ms - only models of jimenez et al . yield both the best overall fit in terms of reduced @xmath14 , and a ms - turnoff age which most closely matches the accepted value of 10.5 gyr ( jorgensen 1991 ) . in summary , either the timescale of ms evolution in the models of yi et al .
is too short by a factor @xmath18 , or the gb contribution to the uv spectrum at an age of 4 gyr is unexpectedly large , in which case these models would again be extremely unusual and the comparison with the age of the sun as suggested by yi et al .
must inevitably be meaningless .
the age implications of the deep keck optical spectrum of the red mjy radio galaxy 53w091 have been previously discussed in some detail by dunlop et al .
( 1996 ) and spinrad et al .
however it is interesting and important to revisit the age - determination of this object for a number of reasons .
first , the models of jimenez et al . have been updated in the intervening years .
second , the yi et al .
models did not exist in 1996/1997 .
third , we have only recently obtained the ms - only versions of the worthey models .
we also felt it was important to re - analyze this spectrum given the claims made by bruzual & margris ( 1997 ) and yi et al .
( 2000 ) that the most recent models yield best - fit ages for the stellar population in 53w091 of less than 2 gyr .
therefore in figure 3 we show the results of fitting the same six models ( as fitted to the sun above ) , to the rest - frame near - ultraviolet sed of 53w091 .
the results , summarized in table 1 , are that both the jimenez et al .
( 2000a ) models ( full and ms only ) yield a best - fit age of 3 gyr ( with ages as young as 2 gyr formally excluded ) , both the yi et al . models ( full and msgb ) yield a best fit age of 2 gyr , while the best - fit age yielded by the worthey models changes from 2 gyr to 3 gyr if post - ms contributions are excluded .
given the fact that we have already shown that the ms clock in the yi et al .
models appears to be running up to a factor of 2 too fast , this means that the argument over whether the age of 53w091 can or can not be younger than 3 gyr comes down to a debate over the validity of the stronger agb and gb contributions in the worthey models compared with the more ms - dominated jimenez et al . models .
jimenez et al . ( 2000a ) argue that their models include a better treatment of mass - loss in evolved stars which results in less luminous rgb and agb contributions . in summary
, the available evidence still appears to favour a minimum age of 3 gyr for this galaxy , subject to remaining uncertaintities over the impact of possible non - solar metallicity ( see section 5 ) . as discussed by dunlop ( 1999 ) , the keck spectrum of the even redder mjy radio galaxy 53w069 ( @xmath10 ; dey et al .
2000 ) appears to offer the best example discovered to date of a highly - evolved coeval stellar population at a redshift as high as @xmath3 .
dunlop ( 1999 ) found , from a comparison of their near - ultraviolet seds , that 53w069 is significantly redder than 53w091 , and that the sed of the latter galaxy can be decomposed into that of 53w069 plus a low - level blue component which is approximately flat in @xmath19 .
it is therefore to be expected that model - fitting to the near - ultraviolet sed of 53w069 might yield even older age limits than those derived above for 53w091 . in figure 4
we show the results of fitting the same six models as before to the near - ultraviolet sed of 53w069 .
the results can again be found in tabulated form in table 1 . in summary , for this object 5 out of the 6 models yield a minimum age @xmath20 gyr , with only the yi et al .
full models allowing an age as young as 2 gyr .
since these models appear flawed based on the solar comparison discussed in section 3 , we conclude that it is extremely hard to escape the conclusion that 53w069 is at least @xmath21 gyr old . for this object the debate over whether the post - ms treatment of jimenez et al .
( 2000a ) is to be preferred over that of worthey ( 1994 ) translates into an uncertainty over whether the best - fit age is in fact @xmath11 or @xmath22 gyr ( see table 1 ) . for @xmath23 ,
the age of an einstein - de sitter universe at @xmath10 is only 2.5 gyr .
therefore , given the apparently flawed nature of the yi et al .
models , we conclude that the only way that the age of 53w069 at @xmath10 can be contained within an einstein - de sitter universe is if varying the assumed metallicity from solar can in fact reduce the best - fit age to @xmath24 gyr .
rlcc + + & & & + + + sun & j full & 8 & 1.67 + & j ms & 11 & 1.00 + & w full & 5 & 2.72 + & w ms & 9 & 1.57 + & y full & 4 & 2.06 + & y msgb & 4 & 1.67 + & j 3z & 8 & 1.37 + + + lbds 53w091 & j full & 3 & 1.25 + & j ms & 3 & 1.22 + & w full & 2 & 1.38 + & w ms & 3 & 1.22 + & y full & 2 & 1.30 + & y msgb & 2 & 1.25 + + + lbds 53w069 & j full & 5 & 1.63 + & j ms & 6 & 1.72 + & w full & 3 & 1.54 + & w ms & 4 & 1.78 + & y full & 2 & 1.63 + & y msgb & 3 & 1.69 + +
the potential severity of the age - metallicity degeneracy was highlighted in the work of worthey ( 1994 ) , and acknowledged by dunlop et al .
( 1996 ) in their original attempts to determine the age of 53w091 .
it is clear that assuming a metallicity of twice solar for the entire stellar population results in the lower limit to the derived age of both 53w091 and 53w069 falling below 2.5 gyr .
spinrad et al .
( 1997 ) presented arguments that , even assuming these galaxies possess a high - metallicity core , adoption of a universally high metallicity is inappropriate when analyzing the integrated light from the central @xmath2 kpc of an elliptical galaxy ( as sampled at @xmath0 by the keck spectroscopic slit ) .
moreover , both jimenez et al .
( 2000b ) and yi et al .
( 2000 ) have now attempted to produce more realistic mixed metallicity models and have independently found that such mixed - metallicity models in fact yield very similar ages to simple _ solar _ metallicity models ( despite the basic disagreement between model timescales detailed above ) .
this appears to be true , even when the average metallicity is twice solar , simply because it is the low metallicity component which dominates the light shortward of 3000 ( jimenez et al .
2000b ) .
ideally , however , we would like to be able to fit both metallicity mix and age to the data .
the extent to which this is possible obviously depends both on the quality of the spectroscopic data available , and on the presence of ( primarily ) metallicity - dependent and age - dependent features within the available spectral range .
this approach will be explored further in a separate paper , but here we present evidence that the age metallicity degeneracy can , at least in principle , be broken with data in the spectral range 2000 4000 , using the sun as a test case .
a mixed - metallicity model was constructed from the 0.2 , and 2.5 full models of jimenez et al . , with the relative contributions of the different metallicity models allowed to vary as free parameters .
the ability of this mixed - metallicity model to reproduce the metallicity and msto age of the sun is a test of the stellar population synthesis models ability to break the age - metallicity degeneracy given data in this near - ultraviolet spectral range .
the solar spectrum was rebinned in the same way as for the solar - metallicity only fitting process .
the mixed - metallicity model flux was built from normalised sed s , so that @xmath25 where @xmath26 is the mixed metallicity flux per unit wavelength in the bin centred on wavelength @xmath27 at @xmath28 gyr , @xmath29 is the flux per unit wavelength in the bin centred on wavelength @xmath27 of the model at @xmath28 gyr , and metallicity z , and @xmath30 is the fractional contribution to @xmath31 by @xmath32 . a @xmath17fit was used to determine the best - fit age , total normalization , and values of @xmath33 , @xmath34 and @xmath35 . in figure 5
we show the results of this mixed - metallicity fitting procedure .
comparison of the left - hand plot with the dotted line in figure 1a demonstrates the extent to which allowing metallicity to vary has weakened the constraint on age .
however , the best - fit age is still 8 gyr , as for the full solar - metallicity only models .
figure 5b shows the evolution of the fractional contributions to the flux of the different metallicity sed s . at the best - fit age of 8 gyr ,
the flux is clearly completely dominated by the solar metallicity sed , as it should be .
@xmath33 is 0.01 , @xmath34 is 0.99 and there is no contribution from the 2.5 model .
this gives a mean metallicity of 0.99 . in summary ,
given data of this ( obviously excellent ) quality , the mixed - metallicity model can return both the correct metallicity and the correct age to high accuracy ( where here the correct age does not mean the turnoff age of the sun , but rather the age of 8 gyr which was returned using the full model with the correct metallicity ) .
in this paper we have not attempted a model - maker s comparison of different evolutionary synthesis models , deliberately not entering into the debate over which model uses the most trustworthy components , such as isochrones , stellar atmospheres etc .
these issues will be addressed in a future paper ( jimenez et al . , 2000c ) .
instead , we have addressed the simple issue of whether the near - ultraviolet spectral energy distribution of the sun is reproduced by the main - sequence only components of three independent models at an age commensurate with current estimates of the ms - turnoff age of the sun .
this test was suggested by yi et al .
( 2000 ) , but in fact leads us to the conclusion that the evolutionary timescale of the main sequence in the yi et al .
models is anomalous when compared with the either the models of worthey ( 1994 ) , jimenez et al .
( 2000a ) or the current best estimate of the ms turn - off age of the sun .
we have also re - addressed the issue of the extent to which varying contributions from post - ms phases of stellar evolution can affect the age estimation of galaxies at @xmath3 from rest - frame ultraviolet spectra .
we find that ( assuming solar metallicity ) the minimum age of the dominant stellar population in the the @xmath36 radio galaxy 53w091 is 2 gyr if one invokes the relatively strong agb / rgb components included in the models of worthey ( 1994 ) , and 3 gyr is one adopts the weaker post - ms contributions included in the models of jimenez et al .
( 2000a ) . in the case of the even more passive @xmath10 radio galaxy 53w069 ,
these numbers become 3 gyr and 4 gyr respectively . finally ,
as stated above , these figures are derived on the assumption of approximately solar metallicity .
yi et al . ( 2000 ) and jimenez et al . (
2000b ) have in fact shown that the age metallicity degeneracy in a realistic mixed - metallicity population is probably not nearly as severe as previously feared .
nevertheless , ideally it is clearly desireable to determine both age and metallicity directly from the observations . using the spectrum of the sun
, we have shown that , given data of sufficient quality , it should in principle be possible to break the well - known age - metallicity degeneracy using data covering only the near - ultraviolet spectral range which is accessible when observing galaxies at @xmath0 with optical spectrographs .
the possibility of breaking the age metallicity degeneracy will be explored further in a subsequent paper .
* acknowledgements * + we would especially like to thank sukyoung yi for making his models available to us in exchange for the jimenez models , and also guy worthey for the willingness and speed with which he supplied us with specific versions of his model predictions , customised to meet our specific requirements .
louisa nolan acknowledges the support of a pparc studentship , and raul jimenez acknowledges the support of a pparc advanced fellowship .
* references * + bruzual , g. , charlot , s. , 1993 , apj , 405 , 538 + bruzual , g.a . ,
magris , g.c . , 1997 , astro - ph/9707154 + chambers , k.c . ,
charlot , s. , 1990 , apj , 348 , l1 + charlot , s. , worthey , g. , bressan , a. , 1996 , apj , 457 , 626 + dey , a. , et al .
2000 , apj , in preparation + dunlop , j.s . , 1999 , in : _ `
the most distant radio galaxies ' _ , knaw colloguium amsterdam , p.71 , eds . rottgering , h.j.a .
, best , p. , lehnert , m.d . , kluwer + dunlop , j.s . ,
guiderdoni , b. , rocca - volmerange , b. , peacock , j.a .
, longair , m.s . , 1989 , mnras , 240 , 257 + dunlop , j. , peacock , j. , spinrad , h. , dey , a. , jimenez , r. , stern , d. , windhorst , r. , 1996 , nature , 381 , 581 + jimenez r. , dunlop j.s .
, peacock j.a . , padoan p. , macdonald j. , j@xmath37rgensen u.g . , 2000a , mnras , in press + jimenez , r. , padoan , p. , juvela , m. , bowen , d.v . ,
dunlop , j.s .
, matteucci , f. , 2000b , apj , in press + jimenez r. et al .
, 2000c , in preparation
+ lilly , s.j . , 1988 , apj , 333 , 161 + jorgensen , u.g .
, 1991 , a&a , 246 , 118 + magris , g.c . ,
bruzual , g.b . , 1993 ,
apj , 417 , 102 + spinrad , h. , dey , a. , stern , d. , dunlop , j.s . , peacock , j. , jimenez , r. , windhorst , r. , 1997 , apj , 484 , 581 + worthey g. , 1994 , apjs , 95 , 107 + yi s. , brown t.m . , heap s. , hubeny i. , landsman w. , lanz t. , sweigart a. , 2000 , apj , in press ( astro - ph/9911067 ) . + | given sufficiently deep optical spectroscopy , the age estimation of high - redshift ( @xmath0 ) galaxies has been claimed to be a relatively robust process ( _ e.g. _ dunlop et al . 1996 ) due to the fact that , for ages @xmath1gyr , the near - ultraviolet light of a stellar population is expected to be dominated by ` well - understood ' main - sequence ( ms ) stars .
recently , however , the reliability of this process has been called into question by yi et al .
( 2000 ) , who claim to have developed models in which the spectrum produced by the main sequence reddens much more rapidly than in the models of jimenez et al .
( 2000a ) , leading to much younger age estimates for the reddest known high - redshift ellipticals . in support of their revised age estimates , yi et al .
cite the fact that their models can reproduce the spectrum of the sun at an age of 5 gyr , whereas the solar spectrum is not reproduced by the jimenez et al .
models until @xmath2 gyr . here
we confirm this discrepancy , but point out that this is in fact a _ strength _ of the jimenez et al .
models and indicative of some flaw in the models of yi et al .
( which , in effect , imply that the sun will turn into a red giant any minute now ) .
we have also explored the models of worthey ( 1994 ) ( which are known to differ greatly from those of jimenez et al . in the treatment of post - ms evolution ) and find that the main - sequence component of worthey s models also can not reproduce the solar spectrum until an age of 9 - 10 gyr .
we conclude that either the models of yi et al . are not as main - sequence dominated at 4 - 5 gyr as claimed , or that the stellar evolutionary timescale in these models is in error by a factor possibly as high as two .
our current best estimate of the age of the oldest galaxies at @xmath3 thus remains @xmath4 gyr , as we confirm with a new analysis of the existing data using the updated solar - metallicity models of both jimenez et al . and
worthey .
finally , by fitting a mixed metallicity model to the sun , we demonstrate that , given rest - frame ultraviolet data of sufficient quality , it should be possible to break the age - metallicity degeneracy when analyzing the spectra of high - redshift galaxies . |
the observation and modeling of the spectral line polarization produced by radiatively induced quantum coherences in atomic systems provides a novel diagnostic window to explore the magnetism of the sun and of other stars via the hanle effect ( e.g. , the recent monograph by landi deglinnocenti & landolfi 2004 ) .
for the moment , such scattering polarization signals have been observed without or with poor spatial and/or temporal resolution ( e.g. , stenflo & keller 1997 ; trujillo bueno et al . 2001 ; malherbe et al .
however , as shown in this work , there are important scientific reasons to try to observe them with a spatial resolution significantly better than 1 arcsec , so as to be able to see clearly the details of the solar granulation pattern . in the apparently quiet regions of the solar atmosphere
there are three main reasons for expecting spatial variability in the linear polarization signals produced by scattering processes .
first , the atmospheric inhomogeneities produced by the solar surface convection lead to horizontal fluctuations in the anisotropy of the radiation field , which must have their impact on the polarization of the emergent radiation ( trujillo bueno 2003 ; trujillo bueno et al .
second , the same atmospheric inhomogeneities should lead to local symmetry breaking effects , so that the radiation field loses the axial symmetry that is characteristic of a plane - parallel or spherically symmetric stellar atmosphere model . as a result , as pointed out by manso sainz & trujillo bueno ( 1999 ) through two - dimensional radiative transfer calculations in simplified solar model atmospheres , one can have scattering line polarization even at the center of the solar disk , without the need of an inclined magnetic field .
third , the scattering polarization signals are sensitive to magnetic fields via the hanle effect , and could therefore show fluctuations across the observed field of view if the statistical properties of the hidden field are not homogeneous , as is indeed indicated by a joint analysis of the hanle effect in atomic and molecular lines ( trujillo bueno et al .
2004 ; 2006 ) .
in this letter we present simulated polarimetric observations for the well - known case of the sr i 4607 line , which we have obtained by solving the radiative transfer problem of scattering line polarization in a realistic three - dimensional ( 3d ) hydrodynamical model of solar surface convection .
observations of the spatial variations of the scattering polarization in the sr i 4607 line have shown that these variations are very small and next to invisible with the used resolution ( e.g. , stenflo et al . 1997 ; trujillo bueno et al .
2001 ; malherbe et al .
2007 ) , in contrast to the large spatial variations in @xmath0 and @xmath1 seen in much stronger lines like ca i 4227 ( e.g. , fig .
5 in stenflo 2004 ) and ca ii k ( e.g. , fig . 4 in stenflo 2006 ) , which have been tentatively interpreted in terms of largely resolved magnetic fields in the solar chromosphere . while the chromosphere has prominent inhomogeneities on the supergranular scales , which are well resolved in the mentioned observations , the quiet photosphere varies on the scale of the solar granulation , which has so far been below the resolution that has been achieved in combination with high - precision spectro - polarimetry . in our previous 3d radiative transfer investigations ( shchukina & trujillo bueno 2003 ; trujillo bueno et al .
2004 ) only the influence of the mean intensity and anisotropy of the sr i 4607 line radiation was accounted for .
here we consider the full 3d scattering line polarization problem in the same realistic hydrodynamical model of the solar photosphere , including the hanle effect of a microturbulent magnetic field , but taking into account the above - mentioned symmetry breaking effects that we had previously neglected .
as we shall see , we confirm our conclusions concerning the presence of a substantial amount of hidden magnetic energy in the quiet regions of the solar photosphere .
in addition , we report on the new interesting results summarized in the abstract .
we have solved the 3d scattering polarization problem for the sr i 4607 line in a realistic model of the solar photosphere resulting from the hydrodynamical simulations of solar surface convection by asplund et al .
( 2000 ) .
first , we calculated the number density of sr i atoms at each grid - point of the 3d photospheric model and the overall population of the lower and upper levels of the 4607 line transition , whose total angular momentum values are @xmath2 and @xmath3 , respectively . to this end
, we solved the standard non - lte radiative transfer problem including all the allowed radiative and collisional transitions between the 15 bound levels of a realistic model of sr i and the ensuing ionizing transitions to the ground level of sr ii .
it is important to calculate the ionization balance of strontium without assuming lte and using a realistic atomic model in order to account properly for the uv overionization mechanism by means of which all the sr i levels become significantly underpopulated with respect to the lte case ( shchukina & trujillo bueno 2003 ) .
we point out that with the resulting self - consistently calculated populations , the computed spatialy - averaged emergent intensity profiles ( which take into account the doppler shifts of the convective flow velocities in the 3d model ) turn out to be in excellent agreement with the observations .
our second step consisted in calculating , at each grid - point of the 3d model , the self - consistent values of the six multipolar components of the atomic density - matrix that characterize the excitation state of the upper level of the sr i 4607 line transition , using the previously calculated lower level population values .
such density - matrix elements quantify the overall population of the upper level ( via the @xmath4 component ) , the level s population imbalances ( via the @xmath5 component ) and the quantum coherences between each pair of magnetic sublevels pertaining to the upper level ( via the real and imaginary parts of the @xmath6 and @xmath7 components ) .
to this end , we solved iteratively the set of equations that results from combining the radiative transfer equations for the stokes parameters @xmath8 , @xmath9 and @xmath10 and the following statistical equilibrium equations , whose derivation can be found in appendix a of trujillo bueno & manso sainz ( 1999 ) : @xmath11 and @xmath12 \left ( \begin{array}{l } { s^2_0 } \\ { \tilde{s}^2_1 } \\ { \hat{s}^2_1 } \\ { \tilde{s}^2_2 } \\ { \hat{s}^2_2 } \end{array } \right ) = ( 1-\epsilon ) \ , w^{(2)}_{j_uj_l } \ , { \cal { h}}^{(2 ) } \left ( \begin{array}{r } { j}^2_0 \\ \tilde{j}{}^{2}_{1 } \\ -\hat{j}{}^{2}_{1 } \\
\tilde{j}{}^{2}_{2 } \\
-\hat{j}{}^{2}_{2 } \end{array } \right)\,.\end{aligned}\ ] ] in these expressions , which are valid for the case of a resonance line without lower - level polarization , @xmath13 is the planck function , @xmath14 ( with @xmath15 the density - matrix elements of the upper level normalized to the overall population of the ground level ) and @xmath16 are the spherical components of the radiation field tensor ( e.g. , 5.11 in landi deglinnocenti & landolfi 2004 ) .
we point out that @xmath17 and @xmath18 ( with @xmath19 ) are complex quantities and that in eq .
( 2 ) @xmath20 and @xmath21 indicate the _ real _ parts , while @xmath22 and @xmath23 the _ imaginary _ parts .
the expressions we have used for @xmath24 , @xmath25 and @xmath18 ( with @xmath19 ) are those given by eqs .
( 4)-(7 ) in trujillo bueno ( 2001 ) , taking into account the doppler shifts caused by the convective velocities .
note also that @xmath26 is the hanle depolarization factor of a microturbulent magnetic field ( see eq .
a.16 in trujillo bueno & manso sainz 1999 ) , @xmath27 is the collisional destruction probability due to inelastic collisions with electrons ( e.g. , shchukina & trujillo bueno 2001 ) , @xmath28 is the collisional depolarizing rate due to elastic collisions with neutral hydrogen atoms measured in units of the einstein a@xmath29 coefficient ( e.g. , faurobert - scholl et al .
1995 ) , and @xmath30 is a coefficient whose value is unity for the sr i 4607 line under consideration .
the spherical components of the radiation field tensor are given by angular and frequency weighted averages of the stokes parameters .
thus , @xmath31 quantifies the mean intensity of the spectral line radiation , @xmath32 the anisotropy factor , and @xmath33 ( with @xmath19 ) the breaking of the axial symmetry of the radiation field through the complex azimuthal exponentials that appear inside the angular integrals ( i.e. , @xmath34 for @xmath35 and @xmath36 for @xmath37 , where @xmath38 is the azimuth of the ray ) .
we point out that while @xmath35 and @xmath37 are zero in a plane - parallel or spherically symmetric model atmosphere in the absence or in the presence of a microturbulent magnetic field , they turn out to fluctuate horizontally at each height in the above - mentioned 3d hydrodynamical model of the solar photosphere , with sizable positive and negative values mainly around the boundaries between the granular and intergranular regions .
the statistical equilibrium equations have a clear physical meaning .
in particular , eqs .
( 2 ) describe the transfer of the symmetry properties of the radiation field directly to the atomic system .
therefore , without the need of a magnetic field ( i.e. , for @xmath39 ) optical pumping processes can generate both _ population imbalances _
( if @xmath40 ) and _ coherences _ ( if @xmath41 and/or @xmath42 ) . in the presence of a microturbulent magnetic field
the population imbalances ( @xmath43 ) and the quantum coherences ( @xmath44 and @xmath45 ) are reduced because @xmath26 varies between 1 ( for @xmath46 g ) and 0.2 ( for @xmath47 , where @xmath48 , with @xmath49 the land factor of the upper level ) .
our numerical solution of this physical problem is based on the iterative methods developed by trujillo bueno & manso sainz ( 1999 ) , which require calculating the radiation field tensors @xmath50 at each iterative step . since the lower - level of the sr i 4607 line has @xmath2 , it can not be polarized .
therefore , the transfer equation for stokes @xmath51 ( with @xmath52 ) is simply given by @xmath53 , which we have solved by applying the 3d formal solver of fabiani bendicho & trujillo bueno ( 1999 ) .
the line contributions to @xmath54 and @xmath55 are the following ( e.g. , manso sainz & trujillo bueno 1999 ) @xmath56 @xmath57 @xmath58 and @xmath59 @xmath60 where the orientation of the ray is specified by @xmath61 ( with @xmath62 the polar angle ) and by the azimuthal angle @xmath38 . the exact expression for @xmath63 that we have used in our calculations is as complicated as the two previous ones ( see eq . 14 in manso sainz & trujillo bueno 1999 ) , but in a weakly anisotropic medium like the solar photosphere @xmath64 .
note also that at the line center of a significantly strong spectral line an approximate expression for estimating the emergent fractional linear polarization is @xmath65 ( for @xmath66 ) , with the corresponding source function values calculated at the atmospheric height where the line center optical depth is unity along the line of sight .
it is important to point out that the above - mentioned symmetry breaking effects imply non - zero values for @xmath67 and @xmath68 , which in turn imply non - zero values for @xmath69 and @xmath70 ( see eq .
2 ) . note also from eqs .
( 3 ) and ( 4 ) that their main observable effects would be non - zero stokes @xmath0 and @xmath1 signals at the solar disk center ( @xmath71 ) and non - zero stokes @xmath1 signals at any off - disk - center position .
let us show now how large such fractional polarization signals are expected to be at three on - disk positions ( @xmath72 , @xmath73 and @xmath71 ) , taking into account the diffraction limit effect of a 1-m telescope ( which we have accounted for through convolution with the airy function ) .
figure 1 shows the center - to - limb variation of the @xmath0 and @xmath1 line - center signals of the sr i 4607 line , which we have obtained by solving the scattering line polarization problem in the above - mentioned 3d hydrodynamical model without magnetic fields .
we point out that in each panel of fig . 1 there are some points of the field of view with signals outside the minimum and maximum values that we have chosen to optimize the visualization . as expected , at @xmath72 and @xmath73 ( see the left and middle top panels of fig .
1 ) the @xmath0 signals are almost everywhere positive , because far away from the solar disk center the term of eq .
( 3 ) proportional to @xmath74 makes the dominant contribution , which quantifies the anisotropy factor of the radiation field . in any case
, it is important to point out that the other terms of eq .
( 3 ) ( i.e. , those caused by the symmetry breaking effects quantified by the @xmath67 and @xmath68 tensors ) do influence the local values of @xmath0 . ] . in the absence of magnetic fields
@xmath0 at @xmath72 varies between about @xmath75 and @xmath76 , while the range of variation at @xmath73 lies between @xmath77 and @xmath78 .
the spatially averaged @xmath0 amplitude is about @xmath79 at @xmath72 and @xmath80 at @xmath73 , which is in agreement with the results of trujillo bueno et al .
the standard deviations ( @xmath81 ) of the @xmath0 fluctuations are approximately @xmath80 at @xmath72 and @xmath82 at @xmath73 . as shown in the corresponding bottom panels of fig .
1 the stokes @xmath1 signals are very significant , with a typical spatial scale of the fluctuation similar to that of @xmath0 , but with positive and negative values lying between about @xmath83 and @xmath75 at @xmath73 ( with a @xmath84 ) and between @xmath85 and @xmath86 at @xmath72 ( with a @xmath87 ) .
such @xmath1 signals are exclusively due to the symmetry breaking effects caused by the horizontal atmospheric inhomogeneities , which are quantified by the tensors @xmath88 and @xmath89 .
the spatially averaged @xmath1 amplitudes are not zero , although they are rather small ( e.g. , @xmath90 at @xmath73 ) .
the right panels of fig . 1 show that the @xmath0 and @xmath1 signals at the solar disk center ( @xmath71 ) are significant and that they have subgranular patterns . in this forward scattering geometry
both @xmath0 and @xmath1 have positive and negative values , which are exclusively due to the symmetry breaking effects ( see eqs . 3 and 4 for @xmath71 ) .
in the unmagnetized case of fig .
1 such values vary between @xmath91 and @xmath92 , approximately , but with most of the @xmath0 signals located between @xmath93 and @xmath94 ( with a @xmath95 ) .
as expected , the spatially averaged @xmath0 and @xmath1 values at the solar disk center are very small that is , of the order of @xmath96 .
figure 2 shows again results for the @xmath73 case , but assuming a particularly interesting magnetized model characterized by a horizontally fluctuating microturbulent field with @xmath97 g at all heights within the solid - line contours of the right panels of fig .
1 ( which delineate the granular upflowing regions ) and @xmath98 g at all heights outside such solid - line contours ( which correspond to the downflowing intergranular regions ) . this model implies saturation of the hanle effect for the sr i 4607 line in the intergranular regions of the photospheric plasma , as suggested by the hanle - effect investigation of trujillo bueno et al .
( 2004 ; 2006 ) .
the spatially averaged @xmath0 amplitude is now about @xmath82 ( instead of the @xmath80 value corresponding to the unmagnetized case of fig .
1 ) , while the amplitude of the spatially averaged @xmath1 profile is only @xmath99 ( instead of @xmath100 ) .
the most remarkable point is found when comparing the patterns of fig .
2 with those of the middle panels of fig .
now , in the magnetized case under consideration , the contrast of the @xmath0 and @xmath1 fluctuations across the field of view is significantly reduced , as a result of having assumed the largest possible hanle depolarization in the intergranular regions .
a similar conclusion is obtained for the other line - of - sights considered in fig . 1
. the use of a 1-m telescope equipped with adaptive optics should be sufficient to measure with a spatial resolution of 100 km and a temporal resolution shorter than 60 sec the predicted linear polarization signals with a polarimetric sensitivity better than @xmath101 .
a suitable instrument for detecting the reported small - scale fluctuations in @xmath0 and @xmath1 would be a fabry - perot interferometer combined with a modern polarimeter .
particular attention should be given to the spectral resolution of the instrument , since an instrumental profile with a too large full - width at half - maximum ( fwhm ) value would significantly reduce the amplitudes of the measured @xmath0 and @xmath1 fluctuations .
a fwhm@xmath102 m seems to be appropiate , given that for 25 m the minimum and maximum @xmath0 values at @xmath73 would be @xmath103 and @xmath104 , instead of the @xmath77 and @xmath78 values corresponding to the infinite spectral resolution case of fig .
1 . therefore
, a fabry - perot system based on two or three etalons would be advisable .
probably , the most interesting conclusion of this investigation is that the horizontal atmospheric inhomogeneities caused by the solar surface convection give rise to measurable linear polarization in the emergent radiation of the sr i 4607 line at the solar disk center .
such @xmath0 and @xmath1 disk - center signals have positive and negative values fluctuating horizontally with a sub - granular pattern .
they are exclusively due to the breaking of the axial symmetry of the radiation field that illuminates the atomic system , which induces quantum coherences between the magnetic sublevels of the upper level of the sr i 4607 line without the need of an inclined magnetic field .
actually , such symmetry breaking effects produce sizable @xmath1 signals at any on - disk position , with positive and negative values fluctuating across the field of view with an amplitude that increases towards the solar limb . on the other hand , with the exception of the previously mentioned forward scattering disk - center case ,
the @xmath0 signals far away from the solar disk center are instead dominated by the anisotropy of the radiation field and fluctuate around a non - zero mean value which is the larger the closer to the solar limb .
the local @xmath0 values are influenced by the symmetry breaking effects , but the spatially averaged @xmath0 amplitudes at each @xmath105 position turn out to be similar to those calculated by trujillo bueno et al .
( 2004 ) , which confirms their conclusion on the presence of a significant amount of hidden magnetic energy and ( unsigned ) magnetic flux in the quiet solar photosphere .
the reported patterns in the fractional linear polarization are of great diagnostic interest because they are sensitive to the thermal , dynamic and magnetic structure of the quiet solar atmosphere .
we urge our solar observer colleagues to carry out the proposed observations as soon as possible . while a 1-m telescope with adaptive optics should be sufficient for the strongly polarizing sr i 4607 line
, the observation of this type of linear polarization signals in most of the other lines of the second solar spectrum would require a larger aperture solar telescope and @xmath1 fluctuations are those of the solar granulation .
the design of the european solar telescope ( est ) and of the advanced technology solar telescope ( atst ) should be done thinking seriously in the scientific case of this letter . to detect and quantify the predicted fluctuations in @xmath0 and @xmath1 across a two - dimensional field of view
would be of fundamental importance to decipher the small - scale magnetic activity of our nearest star . | one of the greatest challenges in solar and stellar physics in coming years will be to observe the second solar spectrum with a spatial resolution significantly better than 1 arcsec .
this type of scattering polarization observations would probably allow us to discover hitherto unknown aspects of the sun s hidden magnetism . here
we report on some theoretical predictions for the photospheric line of sr i at 4607 , which we have obtained by solving the three - dimensional ( 3d ) radiative transfer problem of scattering line polarization in a realistic hydrodynamical model of the solar photosphere .
we have taken into account not only the anisotropy of the radiation field in the 3d medium and the hanle effect of a tangled magnetic field , but also the symmetry breaking effects caused by the horizontal atmospheric inhomogeneities produced by the solar surface convection .
interestingly , the @xmath0 and @xmath1 linear polarization signals of the emergent spectral line radiation have sizable values and fluctuations , even at the very center of the solar disk where we meet the forward scattering case . the ensuing small - scale patterns in @xmath0 and
@xmath1 turn out to be sensitive to the assumed magnetic field model , and are of great diagnostic value .
we argue that it should be possible to observe them with the help of a 1-m telescope equipped with adaptive optics and a suitable polarimeter . |
the method of registration of radio emission of ultrahigh - energy particles is based on askaryan effect .
the effect proposes that particle shower develops a negative charge excess , accompanying the passage of ultra - high energy particles through matter [ 1 ] .
according to this effect , excess of electrons in showers is caused by the annihilation of positrons with electrons of the earth s atmosphere .
the movement of the shower disc with negative charge excess at a speed greater than the phase velocity of light in that medium is the cause of cherenkov radiation emission at radio frequencies [ 2 ] . in the following years after this discovery ,
there have been many experimental studies of radio emission from air showers [ 3 , 4 ] . in paper
[ 7 ] pointed out the possibility of registration of extensive air showers ( eas ) with energies above 10@xmath0 ev , employing radio equipment placed on the surface of the earth and registering the radio emission by satellites on the earth orbit . in recent years ,
interest in the air shower radio emission , as an independent method to study the physics of the eas has grown significantly , and for registration of radio emission were built arrays of significant size [ 12 , 13 ] .
this method makes it possible not only to evaluate the energy , but also to reconstruct the longitudinal shower development , namely , the depth of maximum x@xmath2 [ 14 , 15 ] .
this is especially important for huge arrays where the uncertainty in the estimation of shower energy with different methods of detecting air showers reaches about ( 20 - 40)@xmath6 .
for example , auger and telescope array difference is 20@xmath6and the cause of differences is still remains unknown [ 16 ] .
thus , the radio emission , in conjunction with other methods of could be employed for intercalibration of huge arrays .
in the mid 80-es of the last century , the radio array with registration bandwidth 30 - 40 mhz was designed as an extension of main yakutsk particle array [ 5 ] .
air shower radio emission is registered by 20 receiving antennas , which are installed on 10 pillars as shown in fig .
[ petrovis - fig1 ] .
one pillar consists two independent half - wave dipoles with orientation e - w and n - s .
antennas installed at @xmath7 / 4 above ground , thus ensuring a maximum of the radiation pattern for the emission coming from the top . in the early 2009 at yakutsk array
the radio array was re - created [ 10 ] .
the array consisted 12 crossed at 90 @xmath8 receiving antennas , oriented in e - w and n - s , peripheral recording device and pc storage data .
the recording device was located directly underneath of antenna .
antennas are deployed near center of yakutsk array and consist 2 independent clusters that synchronized by gps .
the distances from the center are 50 - 100 m.
over 10 years of operation of the radio array , observation time is 50400 hours .
14700 events of air showers have been reported with the radio emission , whose axis is in a circle with a radius of 0.5 km and energy exceeding 10@xmath9 ev . during this time , showers with energy exceeding 10@xmath0 ev were registered . from this data set we found a correlation with air showers parameters [ 9 ] and demonstrated the possibility of using eas radio emission to study the physics and astrophysics of ultra - high energy cosmic rays [ 11 ] . fig .
[ petrovis - fig2 ] shows the cloud of points .
most of the showers has energies from 10@xmath0 to 4@xmath310@xmath0 ev .
noteworthy that some points located considerably higher than the total weight of the points .
we have not normalized the data for energy purposely in order to emphasize the special status of these points , namely , they are two showers with energy @xmath10 10@xmath11 ev and thus have almost an order of magnitude greater amplitude than the rest of the showers . as can be seen from fig .
2a , obtained dependence of the amplitude of a radio signal from a distance is close to the simple exponential mean of the type ( [ ip_eq1 ] ) : @xmath12 in fig .
2b points are normalized to average energy @xmath13e@xmath1@xmath14 = 3.7@xmath310@xmath0 ev and average zenith angle @xmath15 = 43.1@xmath8 .
it is seen that after the normalization the cloud of points indicates the rapid decay of radio signal and reflects the function of spatial distribution of radio signal depending on the distance from an antenna to air shower axis .
we derived formula to fit experimental data ( [ ip_eq2 ] ) , which in fig .
is shown as solid curve .
the fit explains well data on medium and large distances from shower axis .
@xmath16 } \right\ } \nonumber\end{aligned}\ ] ] @xmath17 here p@xmath18
parameter that characterizes longitudinal development of air shower x@xmath2 . using the interpolation method ( [ ip_eq3 ] )
[ 12 ] , and model calculations qgsjetii-04 for the proton and iron core evaluation of the atomic weight of primary cosmic ray particles that produced eas with such energy were made .
the value of @xmath19 = 1.5@xmath40,7 , which corresponds to radiation consisting of helium nuclei and average type cno nuclei [ 13 ]
. therefore registration of radio emission of the giant energy air showers can be independent method to learn physics as well as astrophysics of cosmic radiation .
long - term observation of radio emission at the yakutsk array proved the existence of radio emission at energies 10@xmath0 ev and it allowed obtaining the characteristics of radio emission at such energies : * attenuation function of air shower radio signal from distance at energies 3.7@xmath310@xmath0 ev and its gradient ; * proved the existence of radio emission at energies 10@xmath11 ev in most energetic air showers that were registered at yakutsk array [ 14 ] ; * significant signal in inclined showers and influence of magnetic field on ldf shape , which confirms once more [ 9 ] a role of geomagnetic mechanism of radio emission generation ; * estimation of mass composition of cosmic ray obtained in the frame of qgsjetii-04 model by radio data at energies above 10@xmath0 ev in a good agreement with our results presented in [ 12 , 13 ] and points on composition mostly he and cno nuclei .
this result does not contradict the result of the mass composition obtained at huge eas arrays [ 15 ] .
jetp , * 41 * , n 2(8 ) , 616 - 618 , 1961 .
jetp , * 48 * , n 3 , 988 - 990 , 1965 .
artamonov , v. b. atrashkevich , t. a. egorov et al .
publ.yakutsk : yp sb as ussr , 109 - 113 , 1987 .
a.d . filonenko
. physics - uspekhi , * 172 * , n 4 , 439 - 471 , 2002 .
physics of elementary particles and atomic nucleus , * 35 * , n 1 . 1 - 49 . 2004 .
p. schellart , a. nelles , s. buitink et al .
a&a , * 560 * , a98 , 2013 .
b. fuchs for the pierre auger coll .
a , * 692 * , ( 2012 ) , 93 .
apel et al .
physical review d , * 90 * , 062001 , 2014 .
doi:10.1103/physrevd.90.062001 s. knurenko , i. petrov .
j . phys .
: conf . ser. * 632 * , 012100 , 2015 .
doi:10.1088/1742 - 6596/632/1/012100 z.e.petrov et al .
vestnik of nefu , * 8 * , n 4 , 5 - 10 , 2011 .
, z. e. petrov , i.s .
petrov . proc . of xxi int .
symp . opt . of atm . and ocean .
phys.,d348-d351 , 2015 .
e.g. berezhko , s.p .
knurenko , l.t .
astroparticle physics , * 36 * , 31 - 36 , 2012 .
s. knurenko , i. petrov .
epj web of conference * 99 * , 04003 , 2015 .
doi : 10.1051/epjconf/20159904003 .
artamonov , t.e .
egorov , a.n .
efimov et al . proc .
21st icrc .
adelaide , * 9 * , 210 , 1990 .
e. barcikowski , j. bellido , j. belz , y. egorov , s. knurenko , v. de souza , y. tsunesada , m. unger .
epj web of conferences * 53 * , 01006 , 2013 .
doi : 10.1051/epjconf/20135301006 | in this paper , we present results obtained from the measurements of radio emission at frequency of 32 mhz with energy more than 10@xmath0 ev .
generalized formula that describe lateral distribution and depends on main characteristic of the air showers : energy e@xmath1 and depth of maximum x@xmath2 was derived .
the formula has a good agreement with data at average and large distances from shower axis . employing the ratio of radio emission amplitude at distances 175 m and 725 m we determined the depth of maximum x@xmath2 for air shower with energy 3.7@xmath310@xmath0 ev , which in our case is equal to x@xmath2 = 769@xmath434g@xmath3@xmath5 . |
accretion onto a black hole is observed to produce copious x ray emission . at high mass accretion rates ( approaching the eddington limit )
the spectra are dominated by a soft component at @xmath0 kev which is strongly ( very high state : vhs ) or weakly ( high state : hs ) comptonized by low temperature thermal ( or quasi
thermal ) electrons with @xmath1 kev ( e.g. et al . , 1999 ) .
there is also a rather steep power law tail ( @xmath2 ) which extends out beyond @xmath3 kev in the few objects with good high energy data ( e.g. grove et al . , 1998 ) .
at lower mass accretion rates , below @xmath4 per cent of eddington ( et al . , 1998 ; et al . , 1999 ) , there is a rather abrupt transition when the soft component drops in temperature and luminosity .
instead this ( low state : ls ) spectrum is dominated by thermal comptonization , with @xmath5 , rolling over at energies of @xmath6 kev ( see e.g. the reviews by tanaka & lewin 1995 ; van der klis 1995 ; nowak 1995 ) .
this spectral form seems to continue even down to very low luminosities ( @xmath7 per cent of eddington : the quiescent or off state e.g. kong et al .
these very different spectra can all be seen from a single object where the mass accretion rate changes substantially .
figure 1 ( left panel ) shows vhs , hs and ls spectra from the gbhc transient rxte j1550 - 564 . [ cols="^,^ " , ] -24pt the left panel of figure 3 shows the reflected spectrum from a slab of neutral , solar abundance material viewed at @xmath8 ( light grey / green ) compared to that from a disk extending down to 3 schwarzchild radii around a black hole .
the greens functions for the reflected continuum are taken from magdziarz & zdziarski ( 1995 ) , which include compton downscattering at all energies .
the iron line is calculated self consistently with the continuum reflection ( including compton downscattering on the line profile ) using the code of & czerny ( 1994 ) , and the whole spectrum ( line and continuum ) is smeared assuming that the flux irradiating the accretion disk @xmath9 ( see appendix a of et al . , 1999 ) .
the right panel of figure 3 shows the same for a highly ionised reflector .
the resulting spectrum now depends strongly on details of the assumed ionization structure , both as a function of height and of radius . here
we assume that the slab has a constant ionization parameter @xmath10 ( where @xmath11 is the illuminating luminosity , at a distance @xmath12 from the reflector of constant density @xmath13 ) .
the photo electric opacity is calculated by determining the ion populations ( balancing photo ionization with collisional recombination as as in done et al . , 1992 ) .
the electron temperature may then cause appreciable compton upscattering ( ross & fabian 1993 ; ross , fabian & young 1999 ) .
this is _ not _ included in the reflected continuum calculation ( magdziarz & zdziarski 1995 ) , although it is included to some extent in the line profile ( & czerny 1994 but see the discussion in ross et al . , 1999 ) .
as can be seen from these figures , the fraction of reflected flux in the 120 kev band is dependent on the ionization state of the material as well as the solid angle subtended by it .
the ionization state is determined mostly by the line and edge energies ( although the reflection continuum shape is also important ) . but these energies can be shifted by relativistic effects ( ross , fabian & brandt 1996 ) . thus all three parameters ( solid angle , ionization state and relativistic smearing ) should be fit simultaneously in order to determine the geometry of the system .
the quiescent spectra are too faint to do detailed spectral fitting or variability power spectra , but they do appear to be hard @xmath14 i.e. to be related to the ls spectra ( e.g. kong et al . , 2000 ) .
there is much more information on the ( much brighter ) ls spectra .
the ls spectra are relatively hard @xmath15 extending out to @xmath16 kev , which is an immediate problem for a ` disk
corona ' geometry where the magnetic flares are close enough together to form a quasi continuous corona above the disk .
in such a geometry , the hard x ray corona radiates about as much power down towards the disk as up towards the observer .
the downwards flux illuminates the disk , and some fraction ( given by the energy integrated reflection albedo ) is reflected but the rest is thermalised in the disk , forming an additional seed photon source ( haardt & maraschi 1993 ) . the reflection albedo can not approach unity even if the disk is completely ionised due to compton downscattering of the high energy spectrum . for @xmath17 the maximum albedo @xmath18 ,
so the _ minimum _ soft seed flux is approximately a quarter of the hard luminosity .
yet this ratio of soft to hard flux would result in a compton spectrum with @xmath19 ( haardt & maraschi 1993 ; poutanen 1998 ) .
thus the ` disk
corona ' geometry is ruled out .
if the magnetic corona exists , it must be patchy ( stern et al 1995 ; et al . , 1997 ) .
the solid angle subtended by a disk below a ( static ) patchy corona would be @xmath20 .
clearly this can be tested by looking at the amount of reflection , which is easily seen in the ls spectra ( e.g. done et al . , 1992 ) .
fitting simultaneously for the solid angle , ionization state and relativistic smearing gives that the reflector subtends a solid angle of @xmath21 , is predominantly neutral , and significantly broadened but _ not _ by as much as expected if the reflecting material extended down to the last stable orbit ( et al .
, 1997 ; 1998 ; 1999 ; done & 1999 ) .
this means that fitting without including the effects of relativistic smearing and/or ionization does not lead to large distortions of the derived parameters . fitting only for solid angle and ionization ( et al .
, 1997 ; zdziarski , & smith 1999 ) or for solid angle and smearing ( approximated by a gaussian : gilfanov et al . , 1999 ;
revnivtsev et al . , 2000 ) , show overwhelmingly that the solid angle is significantly less than @xmath22 , correlated with the spectral index in the sense that harder continuua show smaller reflected fractions ( zdziarski et al . , 1999 ;
gilfanov et al 1999 ; revnivtsev et al 2000 ) , _ and _ a smaller amount of smearing of the spectral features ( gilfanov et al 1999 ; revnivtsev et al 2000 ) .
these correlations extend to the time domain .
frequency resolved spectra show that the fastest variability is characterised by a flatter spectrum and lower reflected fraction ( revnivtsev et al . , 1999 ) . also
, the characteristic break frequency in the ls variability power spectra _ increases _ as the spectrum steepens and the reflected amplitude and smearing increase ( gilfanov et al 1999 ; revnivtsev et al 2000 ) . the static patchy corona model above a mainly neutral disk is ruled out by the observations of @xmath23 for hard spectra ( see beloborodov 2001 , these proceedings )
. for magentic coronae models to work then either the coronae are not static or the disk is ionized .
outflow from magnetic reconnection is seen in the sun , so it could also happen above the disk .
if the outflow velocity is transrelativistic then the hard x ray radiation from the reconnection is beamed away from the disk .
this reduces the disk illumination , and thus the reflected fraction and the reprocessed soft seed photons .
faster outflow velocities then give fewer reprocessed soft photons and smaller reflected fractions ( giving the @xmath24 correlation ) , and if these faster velocities are concentrated towards smaller radii then this would also suppress the relativistic smearing signature from the inner disk ( beloborodov 1999 ) .
however , the mechanism for the spectral transition ( ls hs ) is not entirely clear . given that it seems likely that the hs / vhs emission is powered by reconnection , then it requires that the spectral transition is associated with the magnetic loops mainly reconnecting inside the optically thick material , rather than buoyantly rising above the disk .
one potential trigger for this could be that the disk becomes radiation pressure dominated although agn disks are in general radiation pressure dominated yet can show hard spectra and low reflected fractions e.g. zdziarski et al . , ( 1999 ) ; chiang et al .
, ( 2000 ) ; done et al . , ( 2000 ) .
similarly the link to the variability power spectra is also not obvious : the particular radius picked out by the qpo / low frequency break can not be the radius at which the disk becomes radiation pressure dominated as this would _ increase _ as the luminosity increases from ls vhs while the power spectral features imply a _ decreasing _ radius .
this seems immediately ruled out by the observation that the reflected spectrum is mostly neutral in the ls .
however , one way to suppress both line and reflected continuum is from a radial dependance of ionization state . if the inner region of the disk were so highly ionised that even iron is completely stripped then it produces no atomic features .
its reflected continuum is unobservable in the 220 kev range and it appears instead to be part of the power law continuum .
the observed reflected spectrum could then come from material further out in the disk which is mostly neutral , and not highly relativistically smeared .
since the reflected fraction is the ratio of the observed reflection relative to the inferred incidence flux then @xmath25 is low both due to the inner disk solid angle being missed and because this unidentified reflected signature boosts the incident flux ( ross et al . , 1999 ) however , for a smooth ( power law ) distribution in ionization parameter with radius then between the extreme ionization inner disk and more neutral outer disk there must be an intermediate region of high ionization , where iron is mainly in h and he like ions .
these produce huge spectral features ( see figure 3 , right panel ) , which are unmistakable .
the ionised disk models can be split into radial zones of different ionization to test these ideas and in cyg x1 they can be shown to be incompatible with the data ( done & 1999 ) .
admittedly these models do not properly model the compton upscattering of the edge features ( and slightly underestimate the compton broadening of the line : ross et al . , 1999 )
associated with the high ionization material , but this is probably not a large distortion . similarly , the assumption of a constant ionization parameter slab can ( and should ) break down in the vertical direction .
illumination of the surface of the disk can result in an extreme ionization skin , which is completely reflective .
some of the incident flux is reflected in the skin , reducing the irradiating flux in the lower layers , so their ionization state is smaller .
deep enough in the disk then the material can be mostly neutral , giving the observed reflected signature which has @xmath26 as the illuminating flux for that layer is not the full incident flux and the apparent incident flux is boosted by the featureless reflected flux from the extremely ionised upper skin ( ross & fabian 1993 ; ross et al . , 1999 ) .
again though , there are problems with the intermediate ionization layers . if the ionization parameter varies smoothly as a ( power law ) function of height in the disk then between the extreme ionization and mostly neutral material there should be the highly ionised h and he like iron , with its unmistakable reflected spectral features .
however , comptonization is now much more of a problem as this spectrum has to be transmitted upwards through the extremely ionised , hot skin .
this additional compton broadening _ may _ be able to smear these features enough to disguise their presence .
such models have yet to be tested against the data but it seems likely that if the comptonization is strong enough to distort the h and he reflected spectrum then it should also distort the observed _ neutral _ reflection signature .
this also has to be transmitted through ( even more of ) the comptonizing layers , and yet it is easily detected . again in these models , as with the outflowing corona , there is no obvious trigger for the ls hs transition or particular radius to use for the power spectral variability .
a far more wideranging problem with the ionised reflection models considered to date is that they _
assume _ a given density structure as a function of height .
at best , the codes take this structure and then work out the temperature and ionization state of the material as a function of height .
but x ray illumination _
changes _ the density structure of the material .
material at the top of the disk is heated by the illumination , and so can expand , lowering its density .
deeper into the disk there is less x ray heating , so the material is cooler , and hence denser .
the self consistent density structure is especially important as there is a thermal ionization instability which affects x ray illuminated material in pressure balance ( field 1965 ; krolik , mckee & tarter 1981 ; kallman & white 1989 ; ko & kallman 1994 ; & czerny 1996 ; nayakshin , kazanas & kallman 2000 , hereafter nkk ) .
the disk is in hydrostic equilibrium in the vertical direction , so the gas pressure , @xmath27 must increase with depth into the disk . in these circumstances , it is clearer to talk about the ionization structure in terms of the pressure ionization parameter @xmath28 where @xmath29 is the radiation pressure of the illuminating radiation ( this relates to the density photo
ionization parameter @xmath30 ) .
the x ray heating clearly depends on height , with the material at the top of the layer being heated the most .
because the material is in pressure balance , the gas pressure is smallest at the top of the disk , so @xmath31 is at its maximum here .
this corresponds to the uppermost stable branch of the ionization equilibrium s
curve on an @xmath32 plot , with low density material which is highly or completely ionized .
atomic cooling is negligible so the temperature is a fraction of the local compton temperature ( see krolik et al . , 1981 and nayakshin 2000 ) ; compton heating is balanced by compton cooling and bremsstrahlung .
going deeper down into the disk , the x ray heating decreases with optical depth to electron scattering , @xmath33 , and the gas pressure increases . hence , both the temperature and ionization parameter @xmath31 decrease .
however , there is then a turning point on the s
curve , where the temperature and density are such that the rapidly increasing bremsstrahlung cooling causes the temperature to dramatically decrease .
this pulls the gas pressure down , but hydrostatic equilibrium requires that the gas pressure must monotonically increase as a function of depth into the disk .
the only way for the pressure to increase in a region with rapidly decreasing temperature is to rapidly increase the density , but this pushes up the cooling still further , and so decreases the temperature .
ionic and atomic species can now exist , so line transitions give a yet further increase in the cooling .
eventually this stabilises onto the bottom part of the s curve , where the x ray heating is balanced primarily by line cooling from low temperature , high density material .
this results in an almost _ discontinuous _ transition from a highly ionized , hot and relatively tenuous skin to mainly neutral , cool and relatively dense material .
reflection can then be suppressed as in the previous section , by the decrease in illuminating flux on the neutral layer because of scattering in the ionised skin , and because the scattered ( reflected ) spectrum from the ionised skin is interpreted as incident continuum .
the difference is that the ionization parameter can now make an _ abrupt _ rather than smooth transition from highly or extremely ionised to mostly neutral material ( nkk ) .
the resulting reflected spectrum is crucially determined by the optical depth where this instability occurs _ and _ the compton temperature of the skin ( which depends on the illuminating power law as well as the local thermal disk spectrum ) . for hard spectra , @xmath34 ,
then the compton temperature of the upper layer of the disk is high @xmath35 kev , its ionization is extreme so it produces no reflected spectral features .
if the instability occurs after an optical depth of @xmath36 then the apparent reflected fraction is @xmath37 from neutral material , as observed . for steeper spectra the compton temperature of the upper layers drops .
thus the instability point occurs rather higher in the disk , so the optical depth of the skin decreases , and the amount of reflected flux increases ( giving the @xmath24 correlation !
the right panel of figure 1 shows results from fitting simulated 220 kev ginga resolution spectra of these full vertical structure models with the fixed density reflection models .
the apparent @xmath24 is clearly seen for the ls @xmath14 spectra ( done & nayakshin 2000 ) .
these models concentrate on the vertical ionisation structure , but there should also be a radial dependance .
larger radii should be less ionised and the total disk spectrum is found by integrating over all these different ionisations . however , the meaning of less ionized is very different from that in the constant density models the skin remains completely ionized with a negligible amount of h and he like iron , but its thomson depth decreases with radius ( see nayakshin 2000 ) .
thus the inner disk can have a larger skin depth , so suppressing the most extreme relativistically smeared line components ( nayakshin 2000 ) .
perhaps this also gives a radius where the skin depth changes from @xmath38 to @xmath39 which could pick out a characteristic frequency for the power spectral variability , although the mechanism for the ls hs transition is again not clear ( see previous section ) .
all the observations can all be qualitatively explained in the truncated disk advective inner flow model , with the disk progressively penetrating further into the advective flow .
the geometry automatically gives a rather ` photon starved ' hard continuum . as the overlap between the two phases increases , the soft seed photon flux intercepted by the hot inner flow gets larger , and so leads to a steeper continuum .
but the increasing overlap also gives a larger solid angle subtended by the disk ( resulting in a @xmath24 correlation : poutanen , krolik & ryde 1997 ) .
the decreasing inner radius of the disk also gives the increasingly relativistically smeared spectral features , _ and _ higher frequency power spectral features .
the advective flow collapses at luminosities at a few per cent of eddington , giving a physical mechanism for the state transition which is observed at these fluxes .
the only problem is a quantitative one .
the inner disk radii deduced from the relativistic smearing are @xmath40 schwarzchild radii .
standard advective flow calculations assume that the transition radius between the ss disk and hot inner flow is large @xmath41 schwarzchild radii .
this is an assumption , and can be changed .
the transition radius can be physically determined by calculating the evaporation rate from the ionised disk skin ( & czerny 2000 ) , but irrespective of the mechanism , moving the disk inwards ( and especially increasing the overlap between the two flows ) increases the soft seed photon flux intercepted by the advective flow , leading to stronger compton cooling . bringing the disk in to @xmath40 schwarzchild radii
lowers the maximum luminosity at which the advective flow can exist by a factor of 23 ( esin 1997 ) , which then makes the ls
hs / vhs state transition at slightly lower luminosities than observed
. however , the ideas developed above about the ionization structure of the disk should also apply here . any hydrostatic disk illuminated by hard x
rays is subject to the thermal instability , and so should have an ionised skin .
the observed neutral reflected spectrum is transmitted through this skin , and is subject to compton broadening .
if some fraction of the width of the line is due to compton scattering rather than relativistic effects then this would result in an underestimate of the disk radius , and so extend the maximum luminosity of the advective flow back into the observed range .
much less work has been done on x ray reflection in the soft states .
the only spectra which have been analysed with both ionisation and relativistic smearing included in the reflected spectrum are nova muscae ( et al . , 1998 ) , rxte j1550 - 564 ( wilson & done 2000 ) and cyg x1 ( gilfanov et al .
, 1999 ; et al 1999 ) .
fitting with the fixed density / ionization structure models gives a solid angle of @xmath42 of strongly ionised material , and that the spectral features are strongly smeared .
fits which do not include ionisation of the reflector will _ overestimate _ the solid angle subtended by the reflecting material ( compare the left and right panels of figure 3 ) , so the results of gilfanov et al . , ( 1999 ) on the @xmath43 smearing correlations will probably change significantly for the steep spectrum objects in their sample when ionisation is included .
the true solid angle is again hard to infer due to ionization effects . even with the fixed density reflection models it depends on how the ionization varies as a function of radius ( et al . , 1998 ) .
obviously the ionization can also vary as a function of height , and the ionization instability again should operate . for steep spectral illumination this does not generally lead to suppression of the observed reflected flux as the compton temperature of the skin is not high enough to completely strip iron .
instead it is dominated by h and he like iron and so its reflected spectrum can be seen directly ( figure 1 , right panel ; done & nayakshin 2000 ) .
but yet another problem for the reflection models is that the disk should now be substantially collisionally ionised , as well as photo ionised .
even more uncertainty comes from the geometry .
if the thermal / non thermal electrons are co spatial then both should illuminate the disk .
but if only the non thermal electrons are associated with magnetic reconnection , while the thermal electrons are from a heated disk skin then most of the incident flux for reflection will be only from the non thermal power law , while the observed incident continuum below 20 kev can be dominated by the thermal compton component .
this would dilute the observed solid angle ( wilson & done 2000 ; et al . , 1999 ) , but the compton continuum produced from a non thermal electron distribution can be highly anisotropic , which could boost the observed solid angle ( ghisellini et al . , 1991 ) !
gbhc show different spectral states which roughly correspond to luminosity , with the ls
hs / vhs transition occurring at 24 per cent of eddington . in the ls ( @xmath44 ) , fitting the spectra with ( fixed density ) models of x ray reflection gives that the solid angle subtended by the disk is small @xmath45 , that the material is mainly neutral , and that the spectral features are broadened , but not by as much as expected from relativistic effects if the disk extended down to the last stable orbit around the black hole .
if the smearing is due to relativistic effects alone then it implies that the reflecting material only extends down to @xmath46 schwarzchild radii .
all these correlate , with steeper spectra having more reflection and more broadening ( and higher frequency features in the power spectra ) .
these observations rule out the continuous corona model ( from the overall hard spectral shape ) and a static patchy corona above a neutral disk ( from @xmath26 ) and a static patchy corona above an ionised disk ( as the reflector is mainly neutral ) and advective flows ( from the combination of small inner radius and high luminosity ) .
only the outflowing corona ( beloborodov 1999 ) can fit these ` observed ' parameters .
radial ionization structure is unlikely to change these conclusions , although vertical ionization structure can give some loopholes due to compton scattering of the observed reflected signature ( ross et al . , 1999 ) .
this _ may _ distort the reflected signature from a static patchy corona by enough to not see the high ionization zone , and compton broadening _ may _ allow the advective flow inner radius to be significantly larger .
however , the constant density models can be seriously misleading due to the ionization instability ( nkk ) .
this can cause a discontinuous transition between an extremely ionised skin overlying mostly neutral material , and can reproduce the observations ( and correlations ) with a static patchy magnetic corona .
but this instability should be present whatever the mechanism for illumination of the disk , so it can also co exist with either an advective flow or an outflowing corona .
this of course completely removes any diagnostic power from the fixed density reflection models ( done & nayakshin 2000 ) . for the hs and vhs the reflected spectrum
is plainly highly ionised and broadened . however , the solid angle is even less constrained than in the ls due to the complex continuum , which could represent a single thermal / non thermal electron population or could represent two entirely spatially distinct regions , with different solid angles for illumination of the disk material .
reflection _ may _ be able to live up to its promise of illuminating the geometry ( and so distinguishing between various ideas as to the underlying emission mechanisms ) but _ only _ with much more detailed modelling of the x ray irradiated disk .
much of this review includes ideas developed in collaborations with s. nayakshin , c. wilson and p. .
laurent p. , and l.g .
titarchuk , the converging inflow spectrum is an intrinsic signature for a black hole : monte carlo simulations of comptonization on free - falling electrons , _ apj _ , * 511 * , 289 - 297 , 1999 . | i review observations of the x ray spectra of galactic black hole candidates , and theoretical ideas as to how these can be produced .
x ray reflection should enable different source geometries to be distinguished , but the ionisation instability of x ray irradiated material in hydrostatic equilibrium gives rise to large ambiguities in interpretation .
it is not currently possible to determine whether the underlying emission mechanism in the low state is an advective flow or magnetic reconnection above the disk , but more detailed modelling of the ionisation instability _ may _ allow us to distinguish between these scenarios . |
under the assumption that the back - reaction to the spacetime geometry can be neglected , the propagation of scalar , electromagnetic and gravitational perturbations is described by linear wave equations that have to be solved on a given background spacetime . for this article
the background is chosen to be the schwarzschild spacetime .
the investigation of wave propagation in such a black - hole spacetime gives insight into typical phenomena like interference patterns and scattering , ( quasi- ) normal modes and black hole evaporation .
in a recent article @xcite we have shown how coordinate systems that are regular at the horizon can be used for constructing analytic solutions to the ( massless ) regge - wheeler equation . in particular , we have used eddington - finkelstein coordinates for constructing such solutions in terms of series that are convergent on the entire open interval @xmath10 , \infty [ \;$ ] .
we will use these results here and extend the treatment by i ) considering additionally a _ massive _ scalar field , which brings a new parameter into the game , and ii ) constructing solutions for the perturbation equations in the observer - associated painlev - gullstrand coordinates . considering a massive scalar field
establishes a connection to cosmology since it may serve as a dark - matter model @xcite .
we have chosen to present this treatment in eddington - finkelstein coordinates , because it is the best adapted coordinate system for this situation and allows to penetrate the horizon @xcite following the ingoing radiation .
giving the solutions to the general case of a massless bosonic field with spin @xmath2 on the other hand allows to investigate black hole physics in coordinates of a radially infalling observer and is therefore , at least in principle , connected to a real physical measurement situation .
the article is organized as follows : we start with some preparations in section [ sec : prep ] where , after introducing our unit conventions , we recall the basic features of two different coordinate systems for the schwarzschild spacetime , namely eddington - finkelstein and painlev - gullstrand coordinates . furthermore , the confluent heun equation that will be the central differential equation in this work is introduced and some of its properties are briefly discussed . in section [ sec : massive ] we consider a massive scalar field on a schwarzschild background spacetime and give the wave equation that describes the propagation of this field in eddington - finkelstein coordinates .
we reduce this wave equation to a purely radial differential equation and show how this radial equation can be transformed to a confluent heun equation .
thereupon , we give the generalized riemann scheme and the local solutions at the horizon in terms of the standard confluent heun function @xmath3 . in section [ sec : pg ] we consider a general massless bosonic spin @xmath2 field in painlev - gullstrand coordinates that are associated with observers which follow radial timelike geodesics .
starting from the wave equation in this coordinate system we arrive at a purely radial equation that belongs to the class of confluent heun equations as well .
we construct local solutions at the horizon in terms of the confluent heun function and asymptotic series at spatial infinity .
these local solutions have convergent representations only in a bounded domain of convergence , but we examine an analytic continuation procedure to overcome this problem .
for the present article we will use geometrical units such that the speed of light @xmath4 and newton s gravitational constant @xmath5 are set to unity . on top of that , it is convenient for the following analysis to measure all radial distances in units of the schwarzschild radius @xmath6 . with this convention the horizon of the schwarzschild black hole
is located at @xmath7 and the light - sphere , where circular lightlike orbits exist , is found at @xmath8 .
the schwarzschild metric , written in the usual schwarzschild coordinates , becomes then @xmath9 as an alternative to the usual schwarzschild coordinates @xmath10 we can introduce other coordinate systems to describe the same spacetime geometry .
two examples that we employ in this work are the eddington - finkelstein ( ef ) and the painlev - gullstrand ( pg ) coordinates which are both regular at the horizon . for details on regular coordinate systems for the schwarzschild spacetime
we refer the reader to ref .
and references therein .
the ingoing version of the ef coordinates is obtained using @xmath11 and the eddington - finkelstein coordinate @xmath12 is constant along all radially ingoing lightlike geodesics .
the painlev - gullstrand coordinates are in our units given by @xmath13 and are associated with infalling observers that follow radial timelike geodesics starting from spatial infinity @xcite .
both coordinate systems are obtained by a coordinate transformation starting with usual schwarzschild coordinates @xmath14 and @xmath15 , while the angular coordinates are not changed to preserve the symmetry of the situation .
the schwarzschild metric is not diagonal anymore in these coordinate systems but the apparent singularity at the horizon can be characterized as being only a coordinate singularity since the metric is invertible at @xmath7 in ef and pg coordinates . in the following
, we will omit the subscripts `` @xmath16 '' and `` @xmath17 '' at the new time coordinate @xmath12 since the two distinct cases are treated in separate sections . the singly confluent heun equation ( che ) , which belongs to heun s class of differential equations @xcite , can be given in the canonical form @xmath18{}{z } + \left ( a + \dfrac{b+1}{z } + \dfrac{c+1}{z-1 } \right ) \diff{}{z } + \left ( \dfrac{\mu}{z } + \dfrac{\nu}{z-1 } \right ) \right ] y(z ) = 0 \ , , \label{eq : che_mapleform}\end{aligned}\ ] ] where the two parameters @xmath19 and @xmath20 are defined in terms of all the other ones by @xmath21 the form is also called the maple - form of the che since it is implemented in the modern computer algebra system maple @xcite .
the che possesses regular singularities at @xmath22 and a single irregular singular point at @xmath23 @xcite . at the regular singularities local solutions
can be obtained in terms of frobenius series @xcite or be given using the standard confluent heun function heunc that is implemented in maple as well . for details about the definition of this confluent heun function the reader is referred to , e.g. , refs . .
at the irregular singularity asymptotic series , also called thom - type solutions , can be constructed @xcite .
the behavior of the local solutions depends on the characteristic exponents of the considered singularity . for details on the construction of these local solutions , that involves the so - called generalized riemann scheme ( grs ) as a very useful tool , the reader is referred to ref . and the excellent book on singular differential equations .
in this section we complement the recent work in ref . and give the solutions of the perturbation equation for a massive scalar field on schwarzschild background spacetime in ef coordinates .
the propagation of such a massive scalar field is described by the klein - gordon equation with a mass term @xmath24
\phi = 0 \ , , \label{eq : kleingordon}\end{aligned}\ ] ] where @xmath25 is the mass of the scalar boson .
the scalar klein - gordon field @xmath26 is a function of all the coordinates @xmath27 and can be expanded into scalar spherical harmonics @xmath28 as shown in many standard textbooks , see e.g. ref .
, @xmath29 where we introduce the tortoise coordinate @xmath30 with the ansatz we obtain the wave equation @xmath31{}{r _ * } - \diffp[2]{}{t } - v_{l}(r ) \right ] \psi_{l}(t , r _ * ) = 0 \ , , \quad v_{l}(r ) = \dfrac{(r-1)}{r } \left ( \dfrac{l(l+1)}{r^2 } + \dfrac{1}{r^3 } + m^2 \right ) \ , .
\label{eq : massive_waveeqn}\end{aligned}\ ] ] this equation is sometimes called the time - dependent regge - wheeler equation for a massive spin @xmath0 field . in their original work regge and
wheeler @xcite derived the analogous equation for the massless @xmath32 case . the lowest mode ( @xmath33 ) and the first excited mode @xmath34 of the potential @xmath35
are shown in fig .
( [ fig : potential ] ) for different values of the mass @xmath25 .
the maxima are located near the light - sphere radius @xmath8 . in the limit
@xmath36 the value of the potential is given by the squared mass .
this property distinguishes the present situation from the massless case which is obtained in the limit @xmath37 . a detailed numerical analysis of ( quasi- ) resonant states ( i.e. , the gray region shown in fig .
[ fig : potential ] ) in this potential can be found in ref .
, where schwarzschild coordinates were used in the discussion but ef coordinates were considered for the numerical evolution as well .
the potential as a function of the tortoise coordinate @xmath38 : the lowest mode @xmath33 ( left ) and the first excited mode @xmath39 ( right ) for @xmath40 ( dotted ) , @xmath41 ( dashed ) , @xmath42 ( dash - dotted ) and @xmath43 ( solid).,title="fig:",scaledwidth=49.0% ] the potential as a function of the tortoise coordinate @xmath38 : the lowest mode @xmath33 ( left ) and the first excited mode @xmath39 ( right ) for @xmath40 ( dotted ) , @xmath41 ( dashed ) , @xmath42 ( dash - dotted ) and @xmath43 ( solid).,title="fig:",scaledwidth=49.0% ] using the results of ref .
we can transform the wave equation into a wave equation in eddington - finkelstein coordinates that allows us to obtain analytical and regular solutions around the black hole horizon at @xmath7 .
there is no known exact analytic solution for the wave equation in the form due to the involved potential term .
the corresponding wave equation in ef coordinates is @xcite @xmath44{}{r } + 2 \dfrac{(r-1)}{r } \diffp{}{r v } + \dfrac{(r-1)}{r^3 } \diffp{}{r } -v_l(r ) \right ] \psi_l(v , r ) = 0 \ , .
\label{eq : massive_waveeqn_ef}\end{aligned}\ ] ] this equation can be further transformed into a radial equation using the ansatz @xmath45 that separates the ef time coordinate @xmath12 . the radial equation that the function @xmath46 has to fulfill
is then given by @xmath18{}{r}+ \left ( \dfrac{1 - 2i \omega r^2}{r(r-1 ) } \right ) \diff{}{r } - \left ( \dfrac{l(l+1)}{r(r-1 ) } + \dfrac{1}{r^2(r-1 ) } + \dfrac{m^2 r}{(r-1 ) } \right ) \right ] r_{\omega l}(r ) = 0 \ , .\end{aligned}\ ] ] we recognize regular singularities at the origin @xmath47 and at the horizon @xmath7 .
moreover , there is an irregular singularity at @xmath23 which can be seen by substituting @xmath48 and investigating the behavior at @xmath49 . with these properties the radial equation belongs to the class of singly confluent heun equations .
the transformation @xmath50 yields a che in the form of with the parameters @xmath51 hence , the grs for the radial equation becomes , using the general result of ref .
, @xmath52 0 & 0 & \left(1 - 2i\omega + \dfrac{2\omega^2 - m^2}{2 \sqrt{m^2-\omega^2 } } \right ) & \\[10pt ] 0 & 2i \omega & \left ( 1 - \dfrac{2\omega^2 - m^2}{2 \sqrt{m^2-\omega^2 } } \right ) & \\[10pt ] & & 0 & \\[2pt ] & & \left(-2\sqrt{m^2-\omega^2}\right ) & \end{pmatrix } \ , .
\label{eq : ef_grs}\end{aligned}\ ] ] we can give two independent solutions @xmath53 at the regular singularity at the horizon ( @xmath7 ) in terms of the standard confluent heun function [ eq : ef_frobius ] @xmath54 from this we recover indeed the previous solutions of the massless case @xcite in the limit @xmath55 . while the first solution @xmath56 is regular at the black hole horizon and corresponds to purely ingoing radiation at that radius , the second solution @xmath57 is not regular there since the phase performs infinitely many turns on the unit circle in the complex plane due to the term @xmath58 that has no well defined limit for @xmath59 .
when we want to imply causal boundary conditions , i.e. , no radiation should escape the black hole horizon , we only need to consider the first , regular , solution .
figure ( [ fig : frobenius1_ef ] ) shows the @xmath60 to @xmath61 modes of this regular solution @xmath56 .
the regular solution @xmath56 for a massive scalar field in ef coordinates for @xmath62 ( top ) , @xmath63 ( middle ) , @xmath64 ( bottom ) for the choice @xmath65 .
the modulus ( solid ) and the real part ( dashed ) are shown .
the horizon of the schwarzschild black hole is located at @xmath7.,title="fig:",scaledwidth=49.0% ] the regular solution @xmath56 for a massive scalar field in ef coordinates for @xmath62 ( top ) , @xmath63 ( middle ) , @xmath64 ( bottom ) for the choice @xmath65 .
the modulus ( solid ) and the real part ( dashed ) are shown .
the horizon of the schwarzschild black hole is located at @xmath7.,title="fig:",scaledwidth=49.0% ] + the regular solution @xmath56 for a massive scalar field in ef coordinates for @xmath62 ( top ) , @xmath63 ( middle ) , @xmath64 ( bottom ) for the choice @xmath65 .
the modulus ( solid ) and the real part ( dashed ) are shown .
the horizon of the schwarzschild black hole is located at @xmath7.,title="fig:",scaledwidth=49.0% ] the regular solution @xmath56 for a massive scalar field in ef coordinates for @xmath62 ( top ) , @xmath63 ( middle ) , @xmath64 ( bottom ) for the choice @xmath65 .
the modulus ( solid ) and the real part ( dashed ) are shown .
the horizon of the schwarzschild black hole is located at @xmath7.,title="fig:",scaledwidth=49.0% ] + the regular solution @xmath56 for a massive scalar field in ef coordinates for @xmath62 ( top ) , @xmath63 ( middle ) , @xmath64 ( bottom ) for the choice @xmath65 .
the modulus ( solid ) and the real part ( dashed ) are shown .
the horizon of the schwarzschild black hole is located at @xmath7.,title="fig:",scaledwidth=49.0% ] the regular solution @xmath56 for a massive scalar field in ef coordinates for @xmath62 ( top ) , @xmath63 ( middle ) , @xmath64 ( bottom ) for the choice @xmath65 .
the modulus ( solid ) and the real part ( dashed ) are shown .
the horizon of the schwarzschild black hole is located at @xmath7.,title="fig:",scaledwidth=49.0% ] the local solutions are , unfortunately , only useful in the rather small domain of convergence @xmath66 but can be analytically continued to the entire domain @xmath670,\infty[$ ] by the procedure that was demonstrated in ref .
for the massless case and a general bosonic spin @xmath2 in ef coordinates .
we will sketch how such a continuation procedure works for the case of massless fields in pg coordinates in the next section .
we start now with a brief reconsideration of the results derived in ref . that will be needed for this section . in that article
we have shown that the wave equation on schwarzschild background that describes the propagation of massless scalar ( @xmath0 ) , electromagnetic ( @xmath68 ) or gravitational ( @xmath32 ) perturbations written in pg coordinates is given by the partial differential equation @xmath69{}{r } - \dfrac{(r-1)}{r } \diffp[2]{}{v } + 2 \dfrac{(r-1)}{r^{3/2 } } \diffp{}{r v } \right . \\
- \dfrac{(r-1)}{2 r^{5/2 } } \diffp{}{v } + \dfrac{(r-1)}{r^3 } \diffp{}{r } - v_{s l}(r ) \right ] \psi(v , r ) = 0 \ , , \label{eq : waveeqn_pg}\end{gathered}\ ] ] where the spin - dependent potential is given by @xmath70 we can separate now the pg time coordinate @xmath12 using the separation ansatz @xmath71 thereby we introduce the frequency @xmath72 of the perturbation that has the dimension of an inverse length in our units ; so the combination @xmath73 is a dimensionless quantity .
the harmonic time dependence in the ansatz leads to a purely radial perturbation equation , that is @xmath74{}{r } + \left ( 1- 2i \omega r^{3/2 } \right ) \diff{}{r } \right .
+ \left ( r^2 \omega^2 + \dfrac{i\omega r^{1/2}}{2 } - l(l+1 ) - \dfrac{1-s^2}{r } \right ) \right ] r_{\omega s l}(r ) = 0 \ , .\label{eq : radialeqn_pg}\end{gathered}\ ] ] this radial equation is a second order , ordinary , linear and homogeneous differential equation .
we further recognize that equation possesses regular singularities at @xmath75 .
using the transformation @xmath76 we are able to find an irregular singularity at @xmath77 , which correspond to @xmath78 . with these singularities the radial equation could belong to the class of singly confluent heun equations @xcite .
indeed , a transformation that maps into the maple - form of the confluent heun equation is given by @xmath79 the function @xmath80 has to fulfill the differential equation with the parameters @xmath81 hence , the grs for this equation can be given by using the general result @xcite , which yields @xmath82 we can construct two independent solutions at the regular singularity at the horizon ( @xmath7 ) using the confluent heun function @xcite .
these two solutions are [ eq : pg_frobenius ] @xmath83 the first solution @xmath84 describes a wave that propagates into the black hole horizon and is totally regular when approaching this radius .
the second solution @xmath85 on the other hand describes outgoing radiation at @xmath7 and is not well defined when approaching the horizon since the phase of this solution performs infinitely many turns on the unit circle in the complex plane in the limit @xmath59 , as we have seen in section [ sec : massive ] before .
figure ( [ fig : pg_frobenius ] ) shows the real part and the modulus of both solutions and clearly demonstrates the corresponding behavior at the horizon .
the two local solutions around the horizon : the regular solution @xmath84 ( left ) and the second solution @xmath85 ( right ) for @xmath86 . the modulus ( solid ) and real part ( dashed ) are shown and the horizon is located at @xmath7.,title="fig:",scaledwidth=49.0% ] the two local solutions around the horizon : the regular solution @xmath84 ( left ) and the second solution @xmath85 ( right ) for @xmath86 . the modulus ( solid ) and real part ( dashed ) are shown and the horizon is located at @xmath7.,title="fig:",scaledwidth=49.0% ] using the entries of the third column of the grs we can construct the thom - type solutions ( asymptotic series ) at the irregular singularity at spatial infinity .
these solutions are @xcite [ eq : pg_asymptotic ] @xmath87 and the series expansion coefficients @xmath88 and @xmath89 can be calculated in terms of recurrence relations that are obtained by inserting the solutions above into the differential equation .
the first solution , together with the separated time dependence @xmath90 corresponds to radiation that is propagating into the singularity at @xmath78 ( outgoing ) , while the second solution describes waves that emerge from this singularity and propagate to smaller radii ( ingoing ) .
both local frobenius solutions at the horizon @xmath91 are convergent within the unit circle centered at @xmath7 in the complex plane as explained in ref . .
however , it is possible to perform an analytic continuation that combines the frobenius solutions with the asymptotic series and in this way to obtain solutions on the entire open interval @xmath670 , \infty[$ ] .
this procedure can be partly traced back to work of jaff @xcite and is demonstrated in detail for the radial perturbation equation for massless fields in ef coordinates in ref . using the general formalism introduced by slavyanov and lay @xcite in terms of the central two point connection problem ( ctcp ) .
we refer the reader to these works and restrict ourselves here to a brief sketch of how to apply this procedure to the present case of pg coordinates .
the five - step procedure that we apply is the following : we perform the simple transformation @xmath92 to shift the involved regular singularity from @xmath7 to @xmath93 and reorder the resulting differential equation in the form @xmath18{}{z } + \left ( \dfrac{a_2}{z+z _ * } + \dfrac{a_1}{z } + g_0 \right ) \diff{}{z } + \left ( \dfrac{c_2}{z+z _ * } + \dfrac{c_1}{z } \right ) \right ] y_{\omega s l}(z ) = 0\end{aligned}\ ] ] with the parameter @xmath94 to match the general formulas given in ref . .
note that the irregular singularity at @xmath78 remains at @xmath23 .
the next step is to perform a special transformation of the dependent variable @xmath95 .
the transformation we need is a so - called s - homotopic transformation and given by @xmath96 the three new exponents @xmath97 are connected to characteristic exponents at the two involved singularities .
similar to the treatment in ref .
we obtain the possible values * @xmath98 is the indicial exponent of the chosen frobenius solution at @xmath99 .
@xmath100 or @xmath101 according to the grs .
* @xmath102 is the characteristic thom exponent of order one and depends on the chosen thom solution at @xmath23 .
the possibilities are @xmath103 or @xmath104 . * for large @xmath105
we get @xmath106 .
thus , @xmath107 is the characteristic exponent of order zero of the chosen thom solution .
@xmath108 , @xmath109 or @xmath110 . as a summary of this step we sketch in figure ( [ fig : exponents ] ) the appropriate choice of the three exponents according to the behavior at the singularities that is determined by the frobenius solutions and the asymptotic series . a scheme showing the proper choice of the exponents @xmath111 and @xmath112 related to local solutions for the connection problem.,scaledwidth=80.0% ] we apply now a transformation of the independent variable .
the necessary transformation in this step is given by the mbius transformation @xmath113 that maps the irregular singularity from @xmath23 to @xmath114 at the boundary of the unit circle .
the other singularity involved is now located at @xmath115 and we have mapped the infinite interval @xmath116 $ ] to the finite interval of unit length @xmath117 $ ] .
now , we solve the resulting differential equation for @xmath118 with the series expansion ansatz @xmath119 this yields a three term recurrence relation for the series expansion coefficients @xmath120 in the form @xmath121 where the @xmath122 and @xmath123 depend on the parameters and , therefore , also on @xmath124 and @xmath125 .
the actual form of these quantities can be derived using the results given in ref . .
the series representation of the function @xmath118 is absolutely convergent on the interval @xmath126@xcite . to construct the total solution to the radial equation in pg
coordinates we have to recover the radius variable @xmath15 that we started with .
this is done by using the inverse mbius transformation @xmath127 now , we construct the total solution using again the transformation .
hence , we obtain @xmath128 where the exponents @xmath129 and @xmath102 must be chosen according to the physical boundary conditions with the help of figure ( [ fig : exponents ] ) and the @xmath130 follow from the mentioned three - term recurrence relation . the solution is now absolutely convergent in the interval @xmath670.5,\infty[$ ] @xcite and we have , thus , solutions on the the entire interval @xmath670,\infty[$ ] since for the region @xmath670,2[$ ] the frobenius solutions in terms of the standard confluent heun function can be used and the analytically continued solutions coincide with these local frobenius solutions in the overlapping region .
we have shown that the propagation equations for bosonic spin @xmath2 fields on schwarzschild background spacetime can be reduced to purely radial differential equations for eddington - finkelstein and painlev - gullstrand coordinates . in these coordinate systems
the radial equations admit ( local ) solutions in terms of convergent series .
we complemented our recent work @xcite , where massless fields were investigated in ef coordinates , by additionally analyzing a massive scalar field in ef coordinates and extending the treatment of the massless bosonic fields to the case of pg coordinates . in each of these cases it is possible to derive two local solutions around the regular singularity at the black hole horizon .
one of these solutions is regular at @xmath131 and connected to causal boundary conditions of purely ingoing radiation , and the other solution describes radiation emerging from the black hole .
while the regular solution can be used to investigate scattering of radiation or to calculate quasi - normal modes , the second solution describes hawking radiation and is , therefore , connected to black hole evaporation as shown in ref . .
we have further shown that using an analytic continuation procedure the bounded domain of convergence of the local solution ( i.e. , the unit circle around the horizon in the complex plane ) can be expanded to the entire open interval between the origin and spatial infinity .
it is a remarkable feature that the structure of the radial equations in ef and pg coordinates is the same .
both cases belong to the class of confluent heun equations and can be solved using frobenius series at the regular singularity at @xmath131 that can analytically be continued up to , but not in general including , spatial infinity .
the equations in pg coordinates are somewhat more complicated and we recommend , thus , to treat all these perturbations in ef coordinates .
if , however , the situation should be linked to the measurement of a radially infalling observer on a timelike geodesic , one should use the solution we provided in eqn . .
this solution can be used to investigate various phenomena concerning massless bosonic waves on schwarzschild background and can be adopted to a broad range of initial or boundary value conditions .
the most interesting feature of the regular coordinate systems is the possibility to penetrate the horizon and thereby to follow the radiation that falls into the black hole .
the authors would like to thank eugen radu , claus lmmerzahl , domenico giulini and norman grlebeck for valuable discussions during the preparation of the manuscript .
support from the deutsche forschungsgemeinschaft within the research training group 1620 models of gravity " is gratefully aknowledged .
the first author is financially supported by the german collaborative research center 1128 `` geo - q '' and acknowledges support from the aveiro university in portugal for accommodation during the vii black hole workshop . | in a previous paper we have considered the regge - wheeler equation for fields of spin @xmath0 , 1 or 2 on the schwarzschild spacetime in coordinates that are regular at the horizon . in particular ,
we have constructed in eddington - finkelstein coordinates exact solutions in terms of series that are regular at the horizon and converge on the entire open domain from the central singularity to infinity . here
we extend this earlier work in two different directions .
firstly , we consider in eddington
finkelstein coordinates a _ massive _ scalar field that can serve as a dark matter candidate .
secondly , we extend the treatment of the massless case to painlev gullstrand coordinates , which are associated with radially infalling observers . |
the notion of mesh patterns was introduced by brndn and claesson @xcite to provide explicit expansions for certain permutation statistics as , possibly infinite , linear combinations of ( classical ) permutation patterns .
this notion was further studied in @xcite .
kitaev and remmel @xcite initiated the systematic study of distribution of quadrant marked mesh patterns on permutations .
the study was extended to 132-avoiding permutations by kitaev , remmel and tiefenbruck in @xcite , and the present paper continues this line of research .
kitaev and remmel also studied the distribution of quadrant marked mesh patterns in up - down and down - up permutations @xcite .
let @xmath10 be a permutation written in one - line notation .
then we will consider the graph of @xmath4 , @xmath11 , to be the set of points @xmath12 for @xmath13 .
for example , the graph of the permutation @xmath14 is pictured in figure [ fig : basic ] . then if we draw a coordinate system centered at a point @xmath12 , we will be interested in the points that lie in the four quadrants i , ii , iii , and iv of that coordinate system as pictured in figure [ fig : basic ] . for any @xmath15 and any @xmath16 , the set of all permutations of length @xmath17
, we say that @xmath2 matches the quadrant marked mesh pattern @xmath18 in @xmath4 if , in @xmath11 relative to the coordinate system which has the point @xmath12 as its origin , there are at least @xmath5 points in quadrant i , at least @xmath6 points in quadrant ii , at least @xmath7 points in quadrant iii , and at least @xmath8 points in quadrant iv . for example , if @xmath14 , the point @xmath19 matches the marked mesh pattern @xmath20 since in @xmath11 relative to the coordinate system with the origin at @xmath21 , there are 3 points in quadrant i , 1 point in quadrant ii , 2 points in quadrant iii , and 2 points in quadrant iv . note that if a coordinate in @xmath18 is 0 , then there is no condition imposed on the points in the corresponding quadrant .
in addition , we shall consider patterns @xmath18 where @xmath22 . here when a coordinate of @xmath18 is the empty set , then for @xmath2 to match @xmath18 in @xmath16 , it must be the case that there are no points in @xmath11 relative to the coordinate system with the origin at @xmath12 in the corresponding quadrant .
for example , if @xmath14 , the point @xmath23 matches the marked mesh pattern @xmath24 since in @xmath11 relative to the coordinate system with the origin at @xmath25 , there are 6 points in @xmath11 in quadrant i , 2 points in @xmath11 in quadrant ii , no points in both quadrants iii and iv .
we let @xmath26 denote the number of @xmath27 such that @xmath2 matches @xmath18 in @xmath4 .
note how the ( two - dimensional ) notation of lfarsson @xcite for marked mesh patterns corresponds to our ( one - line ) notation for quadrant marked mesh patterns .
for example , @xmath28 \useasboundingbox ( 0.0,-0.1 ) rectangle ( 1 + 1.4,1 + 1.1 ) ; { \foreach \x/\y in { } \fill[pattern color = black!65 , pattern = north east lines ] ( \x,\y ) rectangle + ( 1,1 ) ; } { \draw ( 0.01,0.01 ) grid ( 1 + 0.99,1 + 0.99 ) ; } { \foreach \x/\y/\z/\w/\a in { 0/0/1/1/k } { \fill[color = white!100 , opacity=1 , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \draw[color = black , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \fill[black ] ( \x/2+\z/2,\y/2+\w/2 ) node { $ \scriptstyle\a$ } ; } } { \foreach \x/\y in { 1/1 } \filldraw ( \x,\y ) circle ( 6pt ) ; } \end{tikzpicture } } } \hspace{-0.25cm},\ { \mathrm{mmp}}(k,0,0,0)= { \raisebox{0.6ex } { \begin{tikzpicture}[scale=0.35 , baseline=(current bounding box.center ) , scale=2.3 ] \useasboundingbox ( 0.0,-0.1 ) rectangle ( 1 + 1.4,1 + 1.1 ) ; { \foreach \x/\y in { } \fill[pattern color = black!65 , pattern = north east lines ] ( \x,\y ) rectangle + ( 1,1 ) ; } { \draw ( 0.01,0.01 ) grid ( 1 + 0.99,1 + 0.99 ) ; } { \foreach \x/\y/\z/\w/\a in { 1/1/2/2/k } { \fill[color = white!100 , opacity=1 , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \draw[color = black , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \fill[black ] ( \x/2+\z/2,\y/2+\w/2 ) node { $ \scriptstyle\a$ } ; } } { \foreach \x/\y in { 1/1 } \filldraw ( \x,\y ) circle ( 6pt ) ; } \end{tikzpicture } } } \hspace{-0.25cm},\ ] ] @xmath29 \useasboundingbox ( 0.0,-0.1 ) rectangle ( 1 + 1.4,1 + 1.1 ) ; { \foreach \x/\y in { } \fill[pattern color = black!65 , pattern = north east lines ] ( \x,\y ) rectangle + ( 1,1 ) ; } { \draw ( 0.01,0.01 ) grid ( 1 + 0.99,1 + 0.99 ) ; } { \foreach \x/\y/\z/\w/\a in { 0/1/1/2/a } { \fill[color = white!100 , opacity=1 , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \draw[color = black , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \fill[black ] ( \x/2+\z/2,\y/2+\w/2 ) node { $ \scriptstyle\a$ } ; } } { \foreach \x/\y in { 1/1 } \filldraw ( \x,\y ) circle ( 6pt ) ; } \end{tikzpicture } } } \hspace{-2.07 cm } { \raisebox{0.6ex } { \begin{tikzpicture}[scale=0.35 , baseline=(current bounding box.center ) , scale=2.3 ] \useasboundingbox ( 0.0,-0.1 ) rectangle ( 1 + 1.4,1 + 1.1 ) ; { \foreach \x/\y in { } \fill[pattern color = black!65 , pattern = north east lines ] ( \x,\y ) rectangle + ( 1,1 ) ; } { \draw ( 0.01,0.01 ) grid ( 1 + 0.99,1 + 0.99 ) ; } { \foreach \x/\y/\z/\w/\a in { 0/0/1/1/b } { \fill[color = white!100 , opacity=1 , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \draw[color = black , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \fill[black ] ( \x/2+\z/2,\y/2+\w/2 ) node { $ \scriptstyle\a$ } ; } } { \foreach \x/\y in { 1/1 } \filldraw ( \x,\y ) circle ( 6pt ) ; } \end{tikzpicture } } } \hspace{-2.07 cm } { \raisebox{0.6ex } { \begin{tikzpicture}[scale=0.35 , baseline=(current bounding box.center ) , scale=2.3 ] \useasboundingbox ( 0.0,-0.1 ) rectangle ( 1 + 1.4,1 + 1.1 ) ; { \foreach \x/\y in { } \fill[pattern color = black!65 , pattern = north east lines ] ( \x,\y ) rectangle + ( 1,1 ) ; } { \draw ( 0.01,0.01 ) grid ( 1 + 0.99,1 + 0.99 ) ; } { \foreach \x/\y/\z/\w/\a in { 1/0/2/1/c } { \fill[color = white!100 , opacity=1 , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \draw[color = black , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \fill[black ] ( \x/2+\z/2,\y/2+\w/2 ) node { $ \scriptstyle\a$ } ; } } { \foreach \x/\y in { 1/1 } \filldraw ( \x,\y ) circle ( 6pt ) ; } \end{tikzpicture } } }
\ \mbox { and } \ \ \ { \mathrm{mmp}}(0,0,\emptyset , k)= { \raisebox{0.6ex } { \begin{tikzpicture}[scale=0.35 , baseline=(current bounding box.center ) , scale=2.3 ] \useasboundingbox ( 0.0,-0.1 ) rectangle ( 1 + 1.4,1 + 1.1 ) ; { \foreach \x/\y in { 0/0 } \fill[pattern color = black!65 , pattern = north east lines ] ( \x,\y ) rectangle + ( 1,1 ) ; } { \draw ( 0.01,0.01 ) grid ( 1 + 0.99,1 + 0.99 ) ; } { \foreach \x/\y/\z/\w/\a in { 1/0/2/1/k } { \fill[color = white!100 , opacity=1 , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \draw[color = black , rounded corners = 1.5pt ] ( \x+0.125,\y+0.125 ) rectangle ( \z-0.125,\w-0.125 ) ; \fill[black ] ( \x/2+\z/2,\y/2+\w/2 ) node { $ \scriptstyle\a$ } ; } } { \foreach \x/\y in { 1/1 } \filldraw ( \x,\y ) circle ( 6pt ) ; } \end{tikzpicture } } } \hspace{-0.25cm}.\ ] ] given a sequence @xmath30 of distinct integers , let @xmath31 be the permutation found by replacing the @xmath27-th largest integer that appears in @xmath4 by @xmath27 .
for example , if @xmath32 , then @xmath33 . given a permutation @xmath34 in the symmetric group @xmath35
, we say that the pattern @xmath36 _ occurs _ in @xmath16 provided there exists @xmath37 such that @xmath38 .
we say that a permutation @xmath4 _ avoids _ the pattern @xmath36 if @xmath36 does not occur in @xmath4 .
let @xmath39 denote the set of permutations in @xmath1 which avoid @xmath36 .
in the theory of permutation patterns , @xmath36 is called a _
classical pattern_. see @xcite for a comprehensive introduction to the study of patterns in permutations .
it has been a rather popular direction of research in the literature on permutation patterns to study permutations avoiding a 3-letter pattern subject to extra restrictions ( see ( * ? ? ? *
subsection 6.1.5 ) ) . in @xcite
, we started the study of the generating functions @xmath40 where for any @xmath41 , @xmath42 for any @xmath9 ,
we will write @xmath43 for the coefficient of @xmath44 in @xmath45 . there is one obvious symmetry in this case which is induced by the fact that if @xmath46 , then @xmath47 .
that is , the following lemma was proved in @xcite .
[ sym ] ( @xcite ) for any @xmath41 , @xmath48 in @xcite , we studied the generating functions @xmath49 , @xmath50 , and @xmath51 where @xmath52 can be either the empty set or a positive integer as well as the generating functions @xmath53 and @xmath54 .
we also showed that sequences of the form @xmath55 count a variety of combinatorial objects that appear in the _ on - line encyclopedia of integer sequences _ ( oeis ) @xcite .
thus , our results gave new combinatorial interpretations of certain classical sequences such as the fine numbers and the fibonacci numbers as well as provided certain sequences that appear in the oeis with a combinatorial interpretation where none had existed before .
another particular result of our studies in @xcite is enumeration of permutations avoiding simultaneously the patterns 132 and 1234 .
the main goal of this paper is to continue the study of @xmath56 and combinatorial interpretations of sequences of the form @xmath55 in the case where @xmath57 and exactly two of these parameters are non - zero .
the case when at least three of the parameters are non - zero will be studied in @xcite .
next we list several results from @xcite which we need in this paper .
[ thm : qk000 ] ( ( * ? ? ? * theorem 4 ) ) @xmath58 and , for @xmath59 , @xmath60 hence @xmath61 and , for @xmath62 , @xmath63 [ thm : q00k0 ] ( ( * ? ? ?
* theorem 8) ) for @xmath59 , @xmath64 and @xmath65 it follows from lemma [ sym ] that @xmath66 for all @xmath59 .
thus , our next theorem ( obtained in @xcite ) gives an expression for @xmath50 .
[ thm : q0k00 ] ( theorem 11 of @xcite ) @xmath67 for @xmath68 , @xmath69 and @xmath70 as it was pointed out in @xcite , _ avoidance _ of a marked mesh pattern without quadrants containing the empty set can always be expressed in terms of multi - avoidance of ( possibly many ) classical patterns .
thus , among our results we will re - derive several known facts in permutation patterns theory
. however , our main goals are more ambitious aimed at finding distributions in question .
throughout this paper , we shall classify the @xmath73-avoiding permutations @xmath0 by the position of @xmath17 in @xmath4 .
that is , let @xmath74 denote the set of @xmath46 such that @xmath75 .
clearly each @xmath76 has the structure pictured in figure [ fig : basic2 ] .
that is , in the graph of @xmath4 , the elements to the left of @xmath17 , @xmath77 , have the structure of a @xmath73-avoiding permutation , the elements to the right of @xmath17 , @xmath78 , have the structure of a @xmath73-avoiding permutation , and all the elements in @xmath77 lie above all the elements in @xmath78 .
it is well - known that the number of @xmath73-avoiding permutations in @xmath1 is the _ catalan number _ @xmath79 and the generating function for the @xmath80 s is given by @xmath81 if @xmath59 , it is easy to compute a recursion for @xmath71 for any fixed @xmath82 .
it is clear that @xmath17 can never match the pattern @xmath83 for @xmath59 in any @xmath46 . for @xmath84 ,
it is easy to see that as we sum over all the permutations @xmath4 in @xmath85 , our choices for the structure for @xmath77 will contribute a factor of @xmath86 to @xmath71 since none of the elements to the right of @xmath17 have any effect on whether an element in @xmath77 matches the pattern @xmath83 and the presence of @xmath17 ensures that an element in @xmath77 matches @xmath83 in @xmath4 if and only if it matches @xmath87 in @xmath77 . similarly , our choices for the structure for @xmath78 will contribute a factor of @xmath88 to @xmath71 since neither @xmath17 nor any of the elements to the left of @xmath17 have any effect on whether an element in @xmath78 matches the pattern @xmath83 .
thus , @xmath89 multiplying both sides of ( [ k0l0rec ] ) by @xmath90 and summing over all @xmath91 , we obtain that @xmath92 so that we have the following theorem .
[ thm : qk0l0 ] for all @xmath93 , @xmath94 note that by theorem [ thm : q00k0 ] , we have an explicit formula for @xmath95 for all @xmath82 so that we can then use the recursion ( [ k0l0gf ] ) to compute @xmath96 for all @xmath59 . note that @xmath98 since @xmath99 , it follows from the recursions ( [ x=0qk000 ] ) and ( [ x=0k0l0gf ] ) that for all @xmath62 , @xmath100 .
this is easy to see directly .
that is , it is clear that if in @xmath46 , @xmath101 matches @xmath102 , then there is an @xmath103 such that @xmath104 so that @xmath2 matches @xmath105 .
vice versa , suppose that in @xmath46 , @xmath101 matches @xmath105 where @xmath62 . because @xmath4 is @xmath73-avoiding this means the elements in the first quadrant relative to the coordinate system with @xmath106 as the origin must be increasing .
thus , there exist @xmath107 such that @xmath108 and , hence , @xmath109 matches @xmath102 .
thus , the number of @xmath46 where @xmath110 is equal to the number of @xmath46 where @xmath111 for @xmath62 . in @xcite , we computed the generating function @xmath112 for small @xmath52 . thus , we have that @xmath113 note that @xmath114 by theorem [ thm : q00k0 ] .
thus , by ( [ x=0k0l0gf ] ) , we can compute that @xmath115 we note that @xmath116 is the sequence of the pell numbers which is a000129 in the oeis .
this result should be compared with a known fact @xcite that the avoidance of @xmath117 , @xmath118 and @xmath119 simultaneously gives the pell numbers ( the avoidance of @xmath120 is equivalent to avoiding simultaneously @xmath121 and @xmath122 ) . find a combinatorial explanation of the fact that in @xmath1 , the number of ( 132,2134,1234)-avoiding permutations is the same as the number of ( 123,2143,3214)-avoiding permutations .
can any of the known bijections between @xmath73- and @xmath117-avoiding permutations ( see ( * ? ? ?
* chapter 4 ) ) be of help here ?
the sequence @xmath123 is sequence a052963 in the oeis which has the generating function @xmath124 .
that is , @xmath125 .
this sequence had no listed combinatorial interpretation so that we have now given a combinatorial interpretation to this sequence .
similarly , @xmath126 .
thus , by ( [ x=0k0l0gf ] ) , we can compute that @xmath127 in this case , the sequence @xmath128 is sequence a077938 in the oeis which has the generating function @xmath129 .
that is , @xmath130 .
this sequence had no listed combinatorial interpretation so that we have now given a combinatorial interpretation to this sequence .
we can also find the coefficient of the highest power of @xmath131 that occurs in @xmath71 for any @xmath72 .
that is , it is easy to see that the maximum possible number of matches of @xmath83 for a @xmath132 occurs when @xmath133 is a @xmath73-avoiding permutation in @xmath134 and @xmath135 is an increasing sequence
. thus , we have the following theorem . [ maxcoeff0k0l ] for any @xmath72 and @xmath136 ,
the highest power of @xmath131 that occurs in @xmath71 is @xmath137 which appears with a coefficient of @xmath138 .
given that we have computed the generating functions @xmath139 , we can then use ( [ k0l0gf ] ) to compute the following .
@xmath140 @xmath141 @xmath142 we can explain several of the coefficients that appear in the polynomials @xmath143 for various @xmath52 .
@xmath144 for @xmath91 .
this follows immediately from the fact that @xmath145 .
we can also give a simple inductive proof of this fact .
clearly @xmath146 .
assume that @xmath147 for @xmath148 .
then suppose that @xmath149 and @xmath75
. then it must be the case that the elements to the left of @xmath2 are decreasing so that @xmath150 .
but then the elements to the right of @xmath2 must form a 132-avoiding permutation of @xmath151 which has no occurrence of the pattern @xmath152 . thus ,
if @xmath153 , we only have one such @xmath4 and if @xmath154 , we have @xmath155 choices for @xmath156 by induction .
it follows that @xmath157 the sequence @xmath158 is the sequence a000337 in the oeis which has the formula @xmath159 , and the following theorem confirms this fact . for @xmath160 ,
@xmath161 to prove ( [ q1010x ] ) , we classify the @xmath132 such that @xmath162 according to whether the @xmath2 which matches @xmath152 occurs to the left or right of position of @xmath17 in @xmath4 . first
, suppose that @xmath163 and the @xmath164 matching @xmath152 in @xmath4 is such that @xmath165 .
it follows that @xmath166 is an element of @xmath167 such that @xmath168 .
we proved in @xcite that @xmath169 so that we have @xmath170 choices for @xmath171 .
it must be the case that @xmath172 so that we have @xmath155 choices for @xmath156 .
it follows that there are @xmath173 permutations @xmath46 where the unique element which matches @xmath152 occurs to the left of the position of @xmath17 in @xmath4 .
next suppose that @xmath132 , @xmath162 , @xmath163 and the @xmath164 matching @xmath152 is such that @xmath174 .
then the elements to the left of @xmath2 in @xmath4 must be decreasing and the elements to the right of @xmath2 in @xmath4 must be such that @xmath175 .
thus , we have @xmath176 choices for @xmath156 by induction .
it follows that there are @xmath177 permutations @xmath46 where the unique element which matches @xmath152 occurs to the right of the position of @xmath17 in @xmath4 .
thus , @xmath178 here the last equality can easily be proved by induction or be verified by mathematica .
we also can find explicit formulas for the second highest coefficient in @xmath179 for @xmath59 .
@xmath180 for all @xmath181 .
we proceed by induction on @xmath52 .
first we shall prove that @xmath182 for @xmath183 .
that is , suppose that @xmath132 and @xmath184 .
if @xmath185 , then @xmath186 must be strictly increasing .
similarly , if @xmath187 so that @xmath188 , then @xmath189 must be strictly increasing .
it can not be that @xmath190 where @xmath191 because in that case the most @xmath152-matches that we can have in @xmath4 occurs when @xmath192 is an increasing sequence and @xmath156 is an increasing sequence which would give us a total of @xmath193 matches of @xmath152 .
thus , the only other possibility is if @xmath194 in which case @xmath195 .
we proved in @xcite that @xmath196 .
thus , if @xmath194 we have that @xmath197 choices for @xmath189 .
it follows that @xmath198 for @xmath183 .
assume that @xmath62 we have established ( [ secondk010 ] ) for @xmath199 .
we know that the highest power of @xmath131 that occurs in @xmath143 is @xmath200 which occurs with a coefficient of 1 for @xmath201 . now
@xmath202 since the highest power of @xmath131 that occurs in @xmath203 is @xmath204 and the highest power of @xmath131 that occurs in @xmath205 is @xmath206 , @xmath207 unless @xmath208 .
thus , we have 3 cases . + + * case 1 .
* @xmath209 . in that case , @xmath210 * case 2 . * @xmath211 . in this case , we are considering permutations of the form @xmath212 .
then we must have @xmath213 so that there is only one choice for @xmath214 .
thus , in this case , @xmath215 * case 3 . * @xmath216 . in this case ,
@xmath217 for @xmath218 .
+ + thus , it follows that @xmath219 for @xmath181 .
similarly , we have computed the following .
@xmath220 @xmath221 in this case , the sequence @xmath222 is a052963 in the oeis which satisfies the recursion @xmath223 with @xmath224 , @xmath225 and @xmath226 , and has the generating function @xmath124 .
@xmath227 we have also computed the following .
@xmath228 @xmath229 @xmath230 we can also find a formula for the second highest coefficient in @xmath231 for @xmath232 . for all @xmath59 , @xmath232 and @xmath233 , @xmath234 first we establish the base case which is when @xmath235 and @xmath232 .
in this case , @xmath236 since the highest power of @xmath131 that can appear in @xmath237 is @xmath238 for @xmath239 and the highest power of @xmath131 that can appear in @xmath240 is @xmath241 for @xmath242 , it follows that the highest power of @xmath131 that appears in @xmath243 will be less than @xmath244 for @xmath245 .
thus , we have three cases to consider . + + * case 1 . * @xmath209 . in this case ,
@xmath246 and we know that @xmath247 + * case 2 . * @xmath211 . in this case ,
@xmath248 and it was proved in @xcite that @xmath249 + * case 3 . * @xmath216 . in this case ,
@xmath250 and it was proved in @xcite that @xmath251 thus , it follows that @xmath252 for example , for @xmath253 , we get that @xmath254 and , for @xmath255 , we get that @xmath256 which agrees with the series that we computed .
now assume that @xmath68 and we have proved the theorem for @xmath199 and all @xmath232 .
then @xmath257 since the highest power of @xmath131 that can appear in @xmath258 is @xmath259 for @xmath260 and the highest power of @xmath131 that can appear in @xmath231 is @xmath261 for @xmath262 , it follows that the highest power of @xmath131 that appears in @xmath263 will be less than @xmath264 for @xmath245 .
thus , we have three cases to consider .
+ + * case 1 . *
@xmath209 . in this case ,
@xmath265 and we know that @xmath266 + * case 2 . * @xmath211 . in this case ,
@xmath267 and we know that @xmath268 + * case 3 . * @xmath216 . in this case ,
@xmath269 and we know by induction that @xmath270 thus , it follows that @xmath271
by lemma [ sym ] , we know that @xmath273
. thus , we will only consider @xmath274 in this section .
suppose that @xmath275 .
it is clear that @xmath17 can never match the pattern @xmath276 for @xmath59 in any @xmath46 . for @xmath277 ,
it is easy to see that as we sum over all the permutations @xmath4 in @xmath85 , our choices for the structure for @xmath77 will contribute a factor of @xmath278 to @xmath274 .
that is , since all the elements @xmath77 have the elements in @xmath78 in their fourth quadrant and @xmath78 consists of at least @xmath279 elements so that the presence of @xmath17 ensures that an element in @xmath77 matches @xmath276 in @xmath4 if and only if it matches @xmath280 in @xmath77 .
similarly , our choices for the structure for @xmath78 will contribute a factor of @xmath281 to @xmath274 since neither @xmath17 nor any of the elements to the left of @xmath17 have any effect on whether an element in @xmath78 matches @xmath276 .
now suppose @xmath282 and @xmath283 . in this case , @xmath78 consists of @xmath284 elements .
in this situation , an element of @xmath77 matches @xmath276 in @xmath4 if and only if it matches @xmath285 in @xmath77 .
thus , our choices for @xmath77 contribute a factor of @xmath286 to @xmath274 .
similarly , our choices for the structure for @xmath78 will contribute a factor of @xmath281 to @xmath274 since neither @xmath17 nor any of the elements to the left of @xmath17 have any effect on whether an element in @xmath78 matches the pattern @xmath276 . note that since @xmath287 , we know that @xmath288 .
it follows that for @xmath275 , @xmath289 multiplying both sides of ( [ q - k00l ] ) by @xmath90 , summing for @xmath275 and observing that @xmath290 for @xmath291 , we see that for @xmath93 , @xmath292 thus , we have the following theorem .
[ thm : k00l ] for all @xmath93 , @xmath293 note that we can compute generating functions of the form @xmath49 by theorem [ thm : qk000 ] and generating functions of the form @xmath294 by theorem [ thm : q0k00 ] so that we can use ( [ qk00lgf- ] ) to compute @xmath295 for any @xmath296 . by theorem [ thm : k00l ]
, we have that @xmath298 we note that @xmath299 so that @xmath300 .
as described in the previous section , we have computed @xmath112 for small values of @xmath52 in @xcite . plugging those generating functions into ( [ qk00lgf ] )
, one can compute that @xmath301 it is easy to see that the maximum number of @xmath302-matches occurs when either @xmath4 ends with @xmath303 or @xmath304 .
it follows that for @xmath160 , the highest power of @xmath131 in @xmath305 is @xmath306 and its coefficient is @xmath307 .
more generally , it is easy to see that the maximum number of @xmath308-matches occurs when @xmath46 ends with a shuffle of @xmath309 with @xmath310 . thus , we have the following theorem .
[ highqk001 ] for @xmath311 , the highest power of @xmath131 in @xmath312 is @xmath313 and its coefficient is @xmath314 .
we can also compute @xmath315 it is easy to explain some of these coefficients .
that is , we have the following theorem .
* @xmath316 for all @xmath91 , * @xmath317 for all @xmath160 , and * @xmath318 for all @xmath160 . to see that @xmath319 for @xmath91 note that the only permutations @xmath46 that have no @xmath302-matches are the identity @xmath320 plus the permutations of the form @xmath321 for @xmath322 . for @xmath160
, we claim that @xmath323 this is easy to see by induction .
that is , there are three ways to have a @xmath46 with @xmath324 .
that is , @xmath4 can start with @xmath17 in which case we have @xmath325 ways to arrange @xmath186 or @xmath4 can start with @xmath326 in which case there can be no @xmath327 matches in @xmath328 which means that we have @xmath329 choices to arrange @xmath328 or @xmath4 can end with @xmath17 in which case @xmath189 must have exactly one @xmath330-match so that by our previous results in @xcite , we have @xmath331 ways to arrange @xmath332 .
thus , @xmath333 . for @xmath334 , we note that @xmath335 thus , @xmath336 it was proved in @xcite that @xmath337 for @xmath338 and , by theorem [ highqk001 ] , + @xmath339 for @xmath338 . thus , for @xmath160 , @xmath340 one can also compute that @xmath341 and @xmath342 here the sequence @xmath343 which starts out @xmath344 is the sequence a083329 from the oeis which counts the number of partitions @xmath345 of @xmath17 , which when written in _ increasing form _ , is such that the permutation @xmath346 avoids the permutations 213 and 312 . for the increasing form of a set partition @xmath345 , one write the parts in increasing order separated by backslashes where the parts are written so that minimal elements in the parts increase
. then @xmath346 is just the permutation that results by removing the backslashes .
for example , @xmath347 is written in increasing form and @xmath348 .
find a bijection between the @xmath46 such that @xmath349 and the set partitions @xmath345 of @xmath17 such that @xmath346 avoid 231 and 312 .
none of the sequences @xmath350 for @xmath351 appear in the oeis .
similarly , one can compute that @xmath352 then one can use this formula to compute that @xmath353 it is easy to see that permutations @xmath46 which have the maximum number of @xmath354-matches in @xmath4 are those permutations that end in either @xmath355 , @xmath355 , @xmath356 , @xmath357 or @xmath358 .
thus , the highest power of @xmath131 that occurs in @xmath359 is @xmath360 which has a coefficient of @xmath361 .
@xmath362 it is easy to see that permutations @xmath46 which have the maximum number of @xmath363-matches in @xmath4 are those permutations that end in either a shuffle of @xmath364 and @xmath326 or @xmath365 , @xmath366 , and @xmath367 .
thus , the highest power of @xmath131 that occurs in @xmath368 is @xmath369 which has a coefficient of @xmath370 .
@xmath371 it is easy to see that permutations @xmath46 which have the maximum number of @xmath363-matches in @xmath4 are those permutations that end in either a shuffle of @xmath364 and @xmath372 or @xmath373 , @xmath374 , @xmath375 , and @xmath376 .
thus , the highest power of @xmath131 that occurs in @xmath377 is @xmath378 which has a coefficient of @xmath379 .
none of the series @xmath380 for @xmath381 appear in the oeis .
by lemma [ sym ] , we know that @xmath383
. thus , we will only consider @xmath384 in this section .
suppose that @xmath385 .
it is clear that @xmath17 can never match the pattern @xmath386 for @xmath59 in any @xmath46 . for @xmath387 , it is easy to see that as we sum over all the permutations @xmath4 in @xmath85 , our choices for the structure for @xmath77 will contribute a factor of @xmath388 to @xmath384 since none of the elements to the right of @xmath77 have any effect on whether an element of @xmath77 matches @xmath386 .
the presence of @xmath17 and the elements of @xmath77 ensures that an element in @xmath78 matches @xmath386 in @xmath4 if and only if it matches @xmath389 in @xmath78 .
thus , our choices for @xmath78 contribute a factor of @xmath390 to @xmath384 .
now suppose @xmath391 and @xmath283 . in this case ,
@xmath77 consists of @xmath392 elements . in this situation ,
an element of @xmath78 matches @xmath386 in @xmath4 if and only if it matches @xmath393 in @xmath78 .
thus , our choices for @xmath78 contribute a factor of @xmath394 to @xmath384 .
as before , our choices for the structure for @xmath77 will contribute a factor of @xmath388 to @xmath384 but in such a situation @xmath395 .
it follows that for @xmath385 , @xmath396 multiplying both sides of ( [ recqk00l ] ) by @xmath90 , summing for @xmath275 and observing that @xmath397 for @xmath291 , we see that for @xmath93 , @xmath398 it follows that we have the following theorem .
[ thm : q0kl0 ] for all @xmath72 , @xmath399 since we can compute @xmath95 by theorem [ thm : q00k0 ] , we can use ( [ q0kl0gf ] ) to compute + @xmath400 for all @xmath93 .
it follows from theorem [ thm : q0kl0 ] and theorem [ thm : q00k0 ] that @xmath402 thus , one can compute that @xmath403 similarly , one can compute @xmath404 note that @xmath405 thus , it follows that @xmath406 one can use ( [ q0kl0gf ] ) and our previous computations for @xmath95 to compute + @xmath407 .
@xmath408 @xmath409 @xmath410 @xmath411 we can explain the highest and second highest coefficients that appear in @xmath412 for all @xmath82 .
that is , we have the following theorem . [ h2ndhq01l0 ] * for all @xmath82 and @xmath275 , the highest power of @xmath131 in @xmath412 is @xmath413 and its coefficient is @xmath138 .
* @xmath414 for all @xmath183 .
* for all @xmath415 , @xmath416 for all @xmath417
. for ( i ) , it is easy to see that the maximum number of @xmath418 matches occurs for a @xmath46 if @xmath4 starts with @xmath17 followed by any arrangement of @xmath419 followed by @xmath420 in increasing order .
thus , the highest power of @xmath131 in @xmath407 is @xmath413 and its coefficient is @xmath138 . for parts ( ii ) and ( iii )
, we use the fact that @xmath421 it was proved in @xcite that for @xmath422 , the highest power of @xmath131 that occurs in @xmath423 is @xmath424 and its coefficient is @xmath425 .
it follows that for @xmath426 , the highest power of @xmath131 that appears in @xmath427 is less than @xmath428 .
thus , we have three cases to consider . + + * case 1 .
* @xmath209 . in this case
@xmath429 so that we get a contribution of @xmath430 . + + * case 2 . *
@xmath431 . in this case
@xmath432 so that we get a contribution of @xmath433 .
+ + * case 3 . * @xmath216 . in this case
@xmath434 so that we get a contribution of @xmath435 .
+ thus , it follows that @xmath436 then parts ( ii ) and ( iii ) follow from the fact that it was proved in @xcite that * @xmath196 for @xmath338 and , for all @xmath62 , * @xmath437 for @xmath311 .
similarly , one can compute the following .
@xmath438 @xmath439 @xmath440 @xmath441 in this case , we can explicitly calculate the highest and second highest coefficients that appear in @xmath442 for sufficiently large @xmath17 .
that is , we have the following theorem . [ h2ndhq02l0 ] * for all @xmath82 and @xmath443 , the highest power of @xmath131 that appears in @xmath442 is @xmath444 which appears with a coefficient of @xmath445 .
* for all @xmath446 , @xmath447 .
* for all @xmath415 and @xmath448 , @xmath449 . for ( i )
, it is easy to see that the maximum number of @xmath418-matches occurs for a @xmath46 if @xmath4 starts with @xmath326 or @xmath450 followed by any arrangement of @xmath419 followed by @xmath451 in increasing order . thus , the highest power of @xmath131 @xmath452 is @xmath453 and its coefficient is @xmath445 . for parts ( ii ) and ( iii )
, we use the fact that @xmath454 it was proved in @xcite that for @xmath422 , the highest power of @xmath131 that occurs in @xmath455 is @xmath424 and its coefficient is @xmath425 .
moreover , it was proved in @xcite that @xmath456 and , for @xmath415 , @xmath457 it follows that for @xmath458 , the highest power of @xmath131 that appears in + @xmath459 is less than @xmath460 .
thus , we have four cases to consider when computing @xmath461 .
+ * case 1 .
* @xmath462 . in this case , by theorem [ h2ndhq01l0 ] , we have that , @xmath463 + * case 2 . * @xmath431 . in this case
@xmath464 and @xmath465 * case 3 . * @xmath466 . in this case
@xmath467 so that we get a contribution of @xmath468 for @xmath446 . + + * case 4 . * @xmath216 . in this case
@xmath469 so that we get a contribution of @xmath470 for @xmath446 .
+ + thus , it follows that @xmath471 similarly , we have four cases to consider when computing @xmath472 for @xmath415
. + + * case 1 .
* @xmath473 . in this case , by theorem [ h2ndhq01l0 ] , we have that @xmath474 + * case 2 . * @xmath431 . in this case @xmath475 and @xmath476 * case 3 . *
@xmath466 . in this case
@xmath477 so that we get a contribution of @xmath478 for @xmath448 .
+ + * case 4 . * @xmath216 . in this case
@xmath479 so that we get a contribution of @xmath480 for @xmath481 .
+ + thus , it follows that for @xmath481 , @xmath482 for example , when @xmath483 , we obtain that @xmath484 and , when @xmath485 , we obtain that @xmath486 which agrees with the series that we computed .
suppose that @xmath488 . it is clear that @xmath17 can never match the pattern @xmath489 for @xmath59 in any @xmath46 .
there are three cases that we have to consider when dealing with the contribution of the permutations of @xmath74 to @xmath490 . + + * case 1 .
* @xmath491 .
it is easy to see that as we sum over all the permutations @xmath4 in @xmath85 , our choices for the structure for @xmath77 will contribute a factor of @xmath492 to @xmath487 since no element in @xmath77 can match @xmath489 .
the presence of @xmath17 plus the elements in @xmath77 ensure that an element in @xmath78 matches @xmath489 in @xmath4 if and only if it matches @xmath493 in @xmath78 .
hence our choices for @xmath78 contribute a factor of @xmath494 to @xmath487 .
thus , in this case , the elements of @xmath85 contribute @xmath495 to @xmath487 . + + * case 2 .
* @xmath496 .
note that in this case , there are at least @xmath52 elements in @xmath497 and at least @xmath279 in @xmath78 .
the presence of the elements in @xmath78 ensure that an element in @xmath77 matches @xmath489 in @xmath4 if and only if it matches @xmath498 in @xmath77 .
hence our choices for @xmath77 contribute a factor of @xmath499 to @xmath487 .
the presence of @xmath17 plus the elements in @xmath77 ensures that an element in @xmath78 matches @xmath489 in @xmath4 if and only if it matches @xmath500 in @xmath78 .
thus , our choices for @xmath78 contribute a factor of @xmath501 to @xmath487 .
thus , in this case , the elements of @xmath85 contribute @xmath502 to @xmath487 .
+ + * case 3 .
* @xmath282 .
let @xmath503 so that @xmath287 .
it is easy to see that as we sum over all the permutations @xmath4 in @xmath85 , our choices for the structure for @xmath78 will contribute a factor of @xmath504 to @xmath487 since no element in @xmath78 can match @xmath489 .
the presence of the elements in @xmath78 ensures that an element in @xmath77 matches @xmath489 in @xmath4 if and only if it matches @xmath505 in @xmath77 .
hence our choices for @xmath77 contribute a factor of @xmath506 to @xmath487 .
thus , in this case , the elements of @xmath85 contribute @xmath507 to @xmath487 .
+ it follows that for @xmath508 , multiplying both sides of ( [ q0k0lrec ] ) by @xmath90 and summing , we see that + + @xmath510 + @xmath511 + @xmath512 + + thus @xmath513 note the first term of the last term on the right - hand side of ( [ q0k0lgf ] ) is @xmath514 so that we can bring the term @xmath515 to the other side and solve @xmath516 to obtain the following theorem .
it is easy to see that the maximum number of @xmath489-matches occurs for a @xmath46 if @xmath4 starts with some @xmath73-avoiding rearrangement of @xmath533 and ends with some @xmath73-avoiding rearrangement of @xmath534 . in the middle of such a permutation
, we can choose any @xmath73-avoiding permutation of @xmath535 .
it follows that the highest power of @xmath131 which occurs in @xmath490 is @xmath531 which occurs with a coefficient of @xmath532 .
we proved in @xcite that for all @xmath539 , @xmath540 .
in addition , we proved that for @xmath91 , the highest power of @xmath131 that occurs in @xmath541 is @xmath542 and @xmath543 and that for @xmath338 , @xmath544 .
it follows that for @xmath183 , @xmath545 we can also get explicit formulas for @xmath546 , @xmath547 , and @xmath548 based on the fact that we know that @xmath549 then one can use the above formulas to compute that @xmath550 it is then easy to compute @xmath551 for @xmath338 .
this is a known fact ( * ? ? ?
* table 6.1 ) since avoidance of the pattern @xmath552 is equivalent to avoiding the ( classical ) pattern 321 ( thus , here we deal with avoidance of 132 and 321 ) .
h. lfarsson , a unification of permutation patterns related to schubert varieties , a special issue of _ pure mathematics and applications _ ( _ pu.m.a .
_ ) , _ pure mathematics and applications _ ( _ pu.m.a .
_ ) , special issue on `` permutation patterns '' , to appear . | given a permutation @xmath0 in the symmetric group @xmath1 , we say that @xmath2 matches the marked mesh pattern @xmath3 in @xmath4 if there are at least @xmath5 points to the right of @xmath2 in @xmath4 which are greater than @xmath2 , at least @xmath6 points to the left of @xmath2 in @xmath4 which are greater than @xmath2 , at least @xmath7 points to the left of @xmath2 in @xmath4 which are smaller than @xmath2 , and at least @xmath8 points to the right of @xmath2 in @xmath4 which are smaller than @xmath2 .
this paper is continuation of the systematic study of the distribution of quadrant marked mesh patterns in 132-avoiding permutations started in @xcite where we mainly studied the distribution of the number of matches of @xmath3 in 132-avoiding permutations where exactly one of @xmath9 is greater than zero and the remaining elements are zero . in this paper
, we study the distribution of the number of matches of @xmath3 in 132-avoiding permutations where exactly two of @xmath9 are greater than zero and the remaining elements are zero .
we provide explicit recurrence relations to enumerate our objects which can be used to give closed forms for the generating functions associated with such distributions . in many cases ,
we provide combinatorial explanations of the coefficients that appear in our generating functions .
the case of quadrant marked mesh patterns @xmath3 where three or more of @xmath9 are constrained to be greater than 0 will be studied in @xcite .
+ * keywords : * permutation statistics , quadrant marked mesh pattern , distribution , pell numbers |
the coupling between carriers and phonons is known or believed to be important for many materials , including cuprates,@xcite manganites , @xcite , nickelates @xcite and bismuth perovskites.@xcite these materials display a variety of interesting phenomena , including , but not limited to , high - temperature superconductivity ( cuprates , babio@xmath2 ) , layered ferromagnetism ( manganites ) and a spin / charge density wave ( nickelates ) .
the carrier - phonon coupling leads to the formation of a polaron , a coherent quasi - particle ( qp ) consisting of the charge carrier and the cloud of phonons surrounding it and moving coherently with it .
polarons have been studied extensively especially in the holstein model , @xcite the simplest model where local phonons modify the on - site energy of the carrier , but also to generalizations with short - range and long - range couplings of similar origin , such as the breathing - mode ( bm ) model @xcite , the double - well potential model@xcite and the frhlich model @xcite .
the other possibility is that the coupling to phonons modulates the carrier s hopping integrals , a scenario described by the ssh model , @xcite which has seen an increased amount of interest in recent years.@xcite this is because the ssh model exhibits a sharp transition in the properties of its polaron , one signature being the change of the polaron ground state ( gs ) momentum from @xmath1 ( at weak coupling ) to a finite value that smoothly evolves toward @xmath3 ( at strong coupling ) .
such transitions were shown to be impossible for models where the phonons modulate the on - site energy @xcite.a study of a model which includes both types of carrier - phonon coupling was carried out by herrera _ et . al _ @xcite and found that in addition to the transition observed in the ssh model , a second transition of the gs momentum also takes place . whether other such transitions can occur and what are their characteristics , is currently an open question .
efforts to understand polaron physics have , so far , focused almost exclusively on single - band models .
it is therefore a natural question to ask whether the polaron properties of multi - band models are similar , or whether they are qualitatively different from those of single band models . in this article
, we answer this question based on a study of the single polaron properties of the two - band model depicted in fig . [
fig : model ] , with two different atoms per unit cell , one of which is light and thus supports lattice vibrations ( optical phonons ) .
the coupling to this phonon mode modulates the hopping integrals , in direct analogy with the one - band ssh model.(for the vanishing carrier concentration of interest to us , the ssh - model is a one - band model because the peierls dimerization only occurs at half - filling @xcite ) .
due to a phonon on site @xmath4 is also shown . ]
given this similarity , one may expect our two - band model to behave like the ssh model , and indeed we find a sharp transition at strong carrier - hole coupling , where the polaron gs momentum changes its value .
however , unlike in the one - band ssh model where the gs momentum changes continuously with coupling , in our model the gs momentum jumps discontinuously from @xmath5 to @xmath1 .
furthermore , this transition leads to an extreme flattening of the polaron band , unlike in the ssh model .
we conclude that this transition is qualitatively different from that of the one - band ssh model .
we argue that the mechanism for the transition is a competition between the bare carrier hopping which favors a gs momentum @xmath5 , and the phonon - modulated hopping which favors @xmath1 . while some of this physics is similar to that explaining the transition in the ssh model , we find strong indications that the two - band nature of our model plays a vital role .
this makes it unlikely that such a discontinuous transition can occur in a one - band model .
the physics described in this article is therefore , to our knowledge , fundamentally new and our findings hint at the possibility that there is still more , new polaron physics to be discovered in other multi - band models .
the physical inspiration for our model is the perovskite babio@xmath2 , which is known to have strong electron - phonon coupling and to exhibit superconductivity up to surprisingly high temperatures upon hole doping ( @xmath6 in the case of ba@xmath7k@xmath8bio@xmath2 @xcite and @xmath9 for bapb@xmath7bi@xmath10o@xmath2 ) @xcite , widely believed to be due to a phononic glue @xcite .
the relevant valence orbitals are the bi 6@xmath11 orbitals and the o 2@xmath12 orbitals .
previous work on polarons and bipolarons in this material was carried out by allen _
_ @xcite and is based on the rice - sneddon model @xcite which assumes that the bi atoms undergo charge disproportionation , bi@xmath13bi@xmath14bi@xmath15bi@xmath16 . this scenario has received wide - spread attention @xcite , and as a consequence many model hamiltonians only take into account the bi 6@xmath11 orbitals .
polaronic signatures in agreement with this work have also been found experimentally @xcite .
a different scenario is provided by a recent study of foyevtsova _ et .
_ @xcite and the experimental as well as theoretical work of menushenkov _ et al _ @xcite .
_ argue that the bio@xmath17 octahedra undergo a breathing distortion due to strong hybridization between the bi 6@xmath11 and o 2@xmath12 orbitals , with the holes being located primarily on the o. this picture is similar to that proposed recently for the nickelates @xcite and the inspiration for our `` toy model '' . in this picture
both the bi 6@xmath11 and o 2@xmath12 orbitals need to be taken into account . since o atoms are much lighter than bi atoms , only optical phonons on the o atoms are considered and are allowed to modulate the hopping integrals . when compared to other perovskites such as the cuprates , manganites and nickelates , babio@xmath2 is appealing because of its comparatively simple electronic structure and absence of magnetic properties . to simplify things even more , instead of considering a model describing such a material at or near half filling ,
as it is in reality , we investigate the single polaron physics in its almost fully compensated case , _
i.e. _ like in labio@xmath2 with one extra hole .
generalizing to one carrier ( polaron ) per unit cell will be left for future work . moreover , as indicated above , we restrict ourselves to study a 1d bio - like chain , instead of treating the full 3d system .
there are two practical reasons for these simplifications : ( i ) they make comparison to the ssh model , where polaron results are currently available only for the 1d model , possible ; ( ii ) they allow us to use exact diagonalization ( ed ) to find essentially exact results very efficiently,@xcite which in turn also allows us to probe a wide range of coupling strengths and phonon frequencies to understand the relevant physics . beside using ed to understand the polaron properties of this model
, we also develop and validate here two simple versions of the variational momentum average ( ma ) approximation,@xcite which capture the relevant polaronic physics qualitatively and even quantitatively .
ma approaches are very useful because their accuracy improves in higher dimensions while maintaining similar computational efficiency .
this is in contrast to ed and most other numerical methods that become very costly due to the significant increase of the hilbert space in higher dimensions .
this work can therefore be seen as a first step towards a study of the 3d systems .
note that apart from their usefulness in treating more complex problems , developing such approximations also leads to a better understanding of the nature of the polaron s cloud .
to summarize , the research presented in this article serves two main purposes : ( i ) to reveal surprising , new polaron physics whose nature appears to be tied to the two - band nature of our model , and ( ii ) to serve as a test - ground for approximations which will be useful in solving the 3d many - band problem , and thus help to improve our understanding of the fascinating compound babio@xmath2 , if the scenario of foyevtsova _ et
_ turns out to be valid , or of other many - band materials with hopping - modulated carrier - phonon coupling .
the remainder of this article is organized as follows : in sec .
[ sec : model ] we introduce the model and in sec .
[ sec : methods ] we introduce the ed algorithm , ma and a perturbation theory .
the results are presented in sec .
[ sec : results ] and sec .
[ sec : conclusions ] contains our conclusions .
we study the single polaron properties of the 1d , two - band model sketched in fig .
[ fig : model ] .
there are two atoms per unit cell , one hosting valence electrons in an @xmath11-orbital , and the other in a @xmath12-orbital .
the latter atom is assumed to be sufficiently light so that it is a good approximation to ignore the motion of the heavier ones . in other words , we assume that the lighter atoms oscillate inside the `` cage '' made of heavier atoms , giving rise to an optical phonon mode that modulates the @xmath11-@xmath12 hopping of the carriers .
we consider the limit of a very lightly doped insulator , _
i.e. _ all states are filled except for a single hole present on the chain . the inspiration to study such a model
was discussed in the previous section .
the kinetic energy of the hole is described by a nearest neighbor tight - binding hamiltonian : @xmath18 here @xmath19 creates a hole on the @xmath11 ( @xmath12 ) orbital of the atoms in the @xmath20 unit cell ( the spin is an irrelevant degree of freedom here , and we do not write it explicitly ) .
their fourier transforms are : @xmath21 and @xmath22 , where @xmath23 is the location of the unit cell @xmath24 , and the number of unit cells @xmath25 .
the momentum @xmath26 is restricted to the first brillouin zone ( bz ) , @xmath27 . for an undistorted chain ( no phonons ) , and keeping in mind that @xmath28 and @xmath29 are hole operators , for the choice of lobe orientation shown in fig .
[ fig : model ] we have @xmath30 , leading to the kinetic energy : @xmath31 the @xmath32 hopping parameter is given by the overlap of the @xmath11 and @xmath12-orbitals when the atoms are in their equilibrium positions .
changing its sign corresponds to changing the sign convention for the @xmath12-orbitals , and therefore does not have any physical effect . as a consequence ,
the results we present below for a hole - doped chain remain identical for an electron - doped chain as well . when phonons are excited , the hopping amplitudes change from this equilibrium value , resulting in the hole - phonon coupling term discussed below . the hole s on - site energy depends on whether it sits on an @xmath11- or a @xmath12-orbital and leads to a charge transfer term : @xmath33 the difference in on - site energies can have either sign , favoring the @xmath12 ( if @xmath34 ) or @xmath11 ( if @xmath35 ) orbitals . the on - site energy for the @xmath11 orbitals
is set to zero .
the phonons are assumed to be described by a dispersionless einstein mode with energy @xmath36 ( we set @xmath37 ) : @xmath38 where @xmath39 creates a phonon on the @xmath12-orbital of the lighter atom of the @xmath20 unit cell .
as discussed , the heavier atoms are taken to be frozen in their equilibrium positions .
this model allows for two types of hole - lattice coupling .
one comes from the modulation of on - site energies , because when the distance between neighbor atoms changes , so do the corresponding coulomb interactions .
we are not aware of a comprehensive study of this type of coupling for this two - band model , although the asymptotic cases with @xmath40 ( the breathing - mode model @xcite ) and @xmath41 ( the double - well model@xcite ) have been studied and have revealed interesting polaronic behavior , although still rather conventional . instead , we focus here on the hole - lattice coupling arising from the fact that changes in the distance between adjacent atoms also modulate the hopping integrals .
as we show below , this modulation of the kinetic energy leads to qualitatively new physics of a kind that , so far as we know , has not been revealed before .
thus , the hole - phonon coupling that we study arises from the linear expansion of the hopping amplitudes @xmath42 and @xmath43 as a function of the small oxygen displacement @xmath44 .
if we choose coordinates such that @xmath45 for a displacement toward the left , to linear order this expansion gives : @xmath46 = t[1+\tilde{\alpha } ( b_i + b_i^\dagger ) ] \\ & t_{i , i-1 } \approx -t[1-\alpha x_{i-1 } ] = -t[1-\tilde{\alpha } ( b_{i-1 } + b_{i-1}^\dagger)]\end{aligned}\ ] ] within this approximation @xmath47 , where after absorbing all the constants into the coupling @xmath48 , we find : @xmath49\end{aligned}\ ] ] we choose @xmath50 . note that because the sign of @xmath48 is controlled by the choice of the coordinate system , a change @xmath51 is equivalent with choosing @xmath45 for displacements toward the right .
consequently the polaron properties only depend on the magnitude of @xmath48 , not on its sign . and @xmath52 for the lower band of energy @xmath53 .
in both panels @xmath54 , @xmath55 and @xmath56 . ] the hamiltonian studied here is , therefore , @xmath57 before moving on , we briefly comment on its main limitations . keeping only linear terms in the expansion of the hopping integrals
is a valid approximation when @xmath48 is sufficiently small , so that the polaron cloud creates rather small local distortions . for large values of @xmath48 ,
the effects of higher order coupling terms @xmath58 with @xmath59 , need to be considered because now the local distortions become large so the displacements @xmath60 may no longer be assumed to be small .
this is true for all models with electron - phonon coupling . for the holstein and double - well potential models , the effect of such non - linear terms
was shown to be significant in the strong coupling limit ( as defined by the linear term).@xcite another approximation is to consider only nearest neighbor ( nn ) hopping . coming back to the inspiration for this model , for
a bio - like chain one could argue that the bi 6@xmath11 orbitals are quite broad and therefore next nearest neighbor ( nnn ) hopping between them may play a role .
furthermore , in 3d models @xmath61 hopping between nn o 2@xmath12 orbitals needs to be included . in 1d , however , this type of hopping is much less relevant because o atoms are separated by at least on bi atom . finally , as already mentioned , hole - phonon coupling ( linear or to higher order ) resulting from the modulation of the on - site energies
could also be included in this model .
nevertheless , here we study polaron properties for the linear model discussed above , for a wide range of @xmath48 values , to explore its physics and to provide a baseline from which the effect of such additional terms can be gauged .
we also note that all such further extensions can be studied with the methods we use in this work .
the free - hole dispersion is obtained by diagonalizing @xmath62 .
this results in two bands with eigenenergies @xmath63 , shown in fig . [
fig : h0](a ) . the lower band , which will be referred to as @xmath64 , has its minimum at @xmath5 where the hybridization between @xmath11 and @xmath12-orbitals is maximal .
in contrast , the hybridization vanishes at @xmath1 . the free - hole weights @xmath65 are defined as @xmath66 and measure the overlap between the low - energy eigenstate @xmath67 and the free hole in a pure @xmath68 or @xmath12-state , respectively , and
are given by : @xmath69 note that @xmath70 confirms that for @xmath1 , the low - energy free hole sits only on the @xmath12-orbitals .
we are interested in the evolution of this low - energy band as the coupling to phonons is turned on , and in the nature of the resulting quasiparticle the polaron .
a lot of insight can be gained from studying the anti - adiabatic limit @xmath71 , where at sufficiently small @xmath48 the phonon cloud of the polaron is very small , _
i.e. _ with at most one phonon present . the energy correction must be of order @xmath72 because @xmath73 changes the phonon number and thus has vanishing average value in the free hole ground state .
consider the effects of @xmath73 .
for an @xmath11-orbital hole , it allows it to hop onto the adjacent @xmath12-orbital while also emitting a phonon at this @xmath12-orbital .
the phonon needs to be reabsorbed , which can occur in two ways : ( i ) by hopping to the next @xmath11-orbital , or ( ii ) by hopping back to the original @xmath11-orbital .
the first process results in an effective @xmath11-@xmath11 hopping of amplitude @xmath74 , whereas the second changes the on - site @xmath11 energy by @xmath75 . for a hole starting from a @xmath12-orbital , emission of a phonon
requires the hole to hop to an adjacent @xmath11-orbital . from there the phonon
can only be reabsorbed if the hole returns to the original @xmath12-orbital .
this process gives an additional on - site @xmath12 energy of @xmath76 .
thus , the effective hamiltonian is : @xmath77 & t(1-{\mathrm{e}^{-i k a } } ) \\
t(1-{\mathrm{e}^{i k a } } ) & -2 t \lambda - \delta \\
\end{array } \right ) \label{eq : heff_matrix}\end{aligned}\ ] ] where we introduced the dimensionless , effective coupling @xmath78 .
the lowest eigenenergy , _
i.e. _ the polaron dispersion , is given by : @xmath79 -\frac{\delta}{2 } \nonumber \\ & -\sqrt{\left [ t \lambda \cos(ka)-\frac{\delta}{2 } \right]^2 + 4t^2 \sin^2(\frac{ka}{2 } ) } \label{eq : eeff}\end{aligned}\ ] ] this gives a good approximation for the polaron band in the limit @xmath80 , as discussed below .
the polaron eigenstates can also be obtained by ed .
our implementation is a direct extension of the method proposed in ref .
, which has already been successfully applied to a variety of polaronic models : the holstein model in various dimensions,@xcite a generalized holstein model with longer range interactions,@xcite the @xmath81-@xmath82 model with hole - phonon coupling , @xcite and the breathing mode model,@xcite to name a few .
we briefly review it here .
the hilbert space is spanned by the following translationally invariant basis states : @xmath83 here @xmath84 is an index identifying the orbital , such that @xmath85 and @xmath86 .
@xmath87 defines specific phonon cloud configurations , and @xmath26 is the total momentum .
following bona _
_ , we construct the hilbert space by acting @xmath88 times with the full hamiltonian @xmath89 on the free carrier states @xmath90 and @xmath91 and all the states which are created in this process .
this quickly generates a large enough hilbert space to accurately calculate the ground state energy of the polaron with the lanczos technique.@xcite convergence is reached when an increase in the value of @xmath88 no longer produces a change in the eigenenergy .
the number of states contained in this hilbert subspace for different values of @xmath88 is listed in table [ tab : nstates ] .
.number of states in the hilbert space [ cols="^,^",options="header " , ] ma is an accurate variational method for calculating propagators of single polaron hamiltonians , from which polaron properties such as its energy and quasiparticle weight can be obtained .
its simplest version was introduced for the holstein model,@xcite and then it was shown that it can be systematically improved by increasing the size of the variational space , _ i.e. _ which phonon configurations are included.@xcite apart from the holstein model , ma has also been shown to be very accurate for many other lattice polaron models including the bm model,@xcite the ssh model,@xcite and the double - well model.@xcite here we propose two versions of ma : ma@xmath92 for a variational space containing states with a one - site phonon cloud , and ma@xmath93 which also includes states with one additional phonon on a site adjacent to this one - site phonon cloud .
ma@xmath93 is a restricted version of ma@xmath94 which allows the additional phonon to be at any distance from the cloud.@xcite it can also be viewed as a simplified version of the two - site cloud version described in ref . for the breathing mode model .
the latter can be implemented easily for this model as well but is more cumbersome to generalize to higher dimensions .
as we argue below , these simpler versions already suffice for our purposes .
the green s functions ( gf ) of interest are : @xmath95 where again @xmath84 identifies the orbitals .
@xmath96^{-1}$ ] is the resolvent of @xmath89 and @xmath97 is a small positive imaginary number indicating that we are computing retarded gfs .
a complete derivation of the ma solution can be found in appendix a.
we first present the ed results and discuss their meaning and implications .
we then use them to gauge the accuracy of the ma@xmath92 and ma@xmath93 results . in fig .
[ fig : egs_vs_g ] we plot the highest and lowest ed values of the polaron energy , @xmath98 and @xmath99 , vs. the effective coupling @xmath100 for two values of the phonon frequency .
results are qualitatively similar for all other tested values of @xmath36 . for small @xmath100 , the polaron gs energy ( at @xmath0 ) is close to the free hole energy , @xmath101 . at @xmath1 ,
the polaron band lies just below the polaron+phonon continuum , so @xmath102 . with increasing @xmath100 ,
the polaron band moves to lower energies and its bandwidth narrows considerably , as the polaron becomes heavier .
all this is standard polaronic physics .
the surprise is that at sufficiently large @xmath100 , @xmath98 and @xmath99 cross , indicating that the gs momentum changes its value . for @xmath103 , in fig .
[ fig : egs_vs_g](a ) , the scale of the graph makes it difficult to see the actual crossing , but its occurrence is verified below , in fig . [
fig : dispersion_close_to_gc ] .
it is not a priori obvious that the new gs momentum is necessarily at @xmath1 , but we will show below that this is the case .
consequently , we define the critical coupling @xmath104 ( or @xmath105 ) as the value for which @xmath106 . and
@xmath99 vs. the effective coupling @xmath100 for @xmath103 ( left panel ) , and @xmath107 ( right panel ) , at @xmath108 .
convergence was reached for @xmath109 , respectively . ] and @xmath110 for @xmath103 ( left panel ) and @xmath107 ( right panel ) .
note that @xmath111 , because the @xmath11 and @xmath12-orbitals do not hybridize at @xmath1 .
the vertical , dashed line marks @xmath104 .
convergence was reached for @xmath109 , respectively . ] changing @xmath112 results in a shift of the polaron band .
the polaron energy must be smaller than that of the free hole , and the latter is shifted downwards ( upwards ) for @xmath34 ( @xmath35 ) .
a similar shift is therefore expected at least for small @xmath100 , and it is seen to appear for all @xmath100 .
the effect of @xmath112 on @xmath104 is discussed below . before discussing the nature of the transition in detail
, we quickly analyze the qp weight , @xmath113 , defined as the overlap between the polaron eigenstate and the @xmath11 and @xmath12 free - hole states , respectively .
they are shown in fig .
[ fig : zgs_vs_g ] . note that @xmath114 for any finite @xmath100 , because the @xmath11 and @xmath12 orbitals do not hybridize at @xmath1 ( see eq .
( [ eq : heff_matrix ] ) ) and is therefore not shown .
this is in stark contrast to the free hole case , where at @xmath1 all the free - hole weight is on the @xmath12-orbital .
consequently there is a discontinuous change in the @xmath1 qp weight when the hole - phonon coupling is turned on .
for @xmath0 the qp weight falls off rapidly as @xmath100 increases , whereas for @xmath1 its projection on the @xmath11-orbitals first increases and then falls off more slowly .
as pointed out above , this initial increase in qp weight occurs because at @xmath1 and for sufficiently small @xmath100 , the polaron band lies just below the polaron+phonon continuum .
the vertical dashed lines mark @xmath104 , the value of @xmath100 at which the @xmath0 and @xmath1 polaron energies cross .
note that there is no sudden change in the qp weights at @xmath104 .
this is a strong hint that the crossover is not due to a change in the nature of the phonon cloud .
( left panel ) and @xmath115 ( right panel ) for @xmath103 .
note that @xmath111 , because the @xmath11 and @xmath12-orbitals do not hybridize at @xmath1 .
the vertical , dashed line marks @xmath104 .
convergence was reached for @xmath116 the dependence of the qp weight on @xmath112 is shown in fig .
[ fig : zgs_vs_g_diff_delta ] for @xmath103 . as expected
, @xmath34 increases the amount of @xmath12 character and decreases the @xmath11 character ; @xmath117 has the opposite effect . for larger values of the coupling @xmath100 ,
the change in qp weights due to @xmath112 becomes negligibly small . at these values of @xmath100
the phonon cloud is already quite sizable , as indicated by the small values of the qp weights .
together , fig .
[ fig : egs_vs_g ] and fig .
[ fig : zgs_vs_g_diff_delta ] show that for strong coupling , @xmath112 shifts the energy of the polaron but does not change the nature of its phonon cloud .
this , in turn , suggests that @xmath118 strongly favors a specific kind of phonon cloud . in other words
, there appears to be only one dominant mechanism that allows the hole to lower its energy via the emission and absorption of phonons . and values of @xmath119 .
the curves have been shifted by @xmath98 for better comparison and the curve at @xmath104 is highlighted in red .
the green , dashed lines in the rightmost panel are the perturbative results from eq .
( [ eq : eeff ] ) .
all curves are for @xmath120 and the cutoff is @xmath121 , respectively . ]
the ( shifted ) polaron dispersion @xmath122 vs @xmath26 , close to the critical coupling @xmath105 , is shown in fig . [
fig : dispersion_close_to_gc ] for @xmath123 and @xmath124 .
note that close to @xmath105 , the polaron bandwidth is extremely narrow .
these results show that the gs momentum does indeed change discontinuously from @xmath125 to @xmath126 , justifying our definition of @xmath105 . for @xmath127 , the prediction of eq .
( [ eq : eeff ] ) ( dashed lines ) agrees quite well with ed and clearly reproduces the transition .
as we show now , a closer examination of this perturbative result indeed unveils a lot of the physics governing this transition .
when @xmath128 , the polaron dispersion given by eq .
( [ eq : eeff ] ) is not much different from the free hole dispersion @xmath129 , with the ground state at @xmath0 . for @xmath130 ,
on the other hand , the polaron dispersion can be approximated as : @xmath131 - \frac{\delta}{2},\end{aligned}\ ] ] with the ground state at @xmath1 and a flat band for @xmath132 .
consequently there is a critical value , @xmath104 , where the ground state momentum changes .
indeed it can be verified that for @xmath133 the @xmath26-dependence in the square root of eq .
( [ eq : eeff ] ) exactly cancels the @xmath134 term and the band becomes completely flat , @xmath135 .
for @xmath136 the ground state is at @xmath0 , and for @xmath137 it is at @xmath1 .
it is clear that this change of the ground state momentum is due to a competition between the antisymmetric , bare hopping @xmath138 and the symmetric , phonon - modulated hopping , @xmath118 . to our knowledge
this is the first report of a change in the ground state momentum caused by a symmetric phonon - modulated hopping ; such changes have been seen before only for models with antisymmetric , phonon - modulated hopping.@xcite however , while for the latter case the ground - state momentum changes _ smoothly _ for @xmath139 , here we observe a
_ discontinuous jump _ from the edge to the centre of the bz .
note , furthermore , that the hybridization between the two bands which leads to the square root in eq .
( [ eq : eeff ] ) plays an important role for the nature of the transition and appears to be responsible for the flatness of the band close to @xmath104 .
although perturbation theory explains many features of the transition , there are some differences between the large @xmath36 case and the cases where @xmath36 is comparable to @xmath81 .
( [ eq : eeff ] ) predicts a completely flat band at @xmath104 .
while this is seen for @xmath127 , it is no seen for smaller @xmath36 .
instead , here higher order processes leading to phonon - mediated , longer range effective hopping , cause the bandwidth to remain finite at all values of @xmath100 .
we can attribute these processes to longer range effective hopping because they lead to a maximum at @xmath140 , halving the bz .
their contribution is very small , as indicated by the small bandwidth close to @xmath104 .
this leads us to conclude that the phonon - assisted hopping process described in section [ sec : perturbation - theory ] is indeed primarily responsible for most of the mobility of the polaron . to make these arguments more compelling and valid for phonon clouds with more than one phonon
, we now analyze the nature of the polaron cloud in more detail . for @xmath1 ,
the polaron has qp weight on @xmath141 but not @xmath142 .
acting with @xmath118 on @xmath143 ( and ignoring normalization factors ) gives @xmath144 , which we therefore expect to be a part of the polaron wavefunction .
acting on this again with @xmath118 gives @xmath145 | 0\rangle$ ] .
this pattern continues and we find that for @xmath1 the polaron eigenfunction contains states of the type @xmath146 , _
i.e. _ with an odd number of phonons and the hole on the same @xmath12-orbital , and @xmath147 | 0\rangle$ ] , _ i.e. _ a symmetric state with an even number of phonons and the hole on the adjacent @xmath11-orbitals . while we only considered states with a one - site phonon cloud adjacent to the hole , this structure of the phonon cloud
is indeed verified by ed , which finds very small weight for other configurations . for @xmath0 ,
consequently we need to start the construction of the eigenstate from @xmath149 .
acting with @xmath118 gives @xmath150 .
this pattern generalizes and mixes states of the form @xmath151 , _
i.e. _ with an even number of phonons and the hole on the same @xmath12-orbital , and @xmath152 |0\rangle$ ] , _ i.e. _ an antisymmetric state with an odd number of phonons and the hole on the adjacent @xmath11-orbitals .
this also is verified by ed .
thus , the polaron cloud structure is very different at @xmath1 and @xmath125 , and this has important consequences .
consider the configuration latexmath:[$\sum_j { \mathrm{e}^{ikr_j } } s_{j}^\dagger b_{j}^{\dagger,2n } times with @xmath118 ; among other states , this links to @xmath154 .
acting another @xmath155 times with @xmath118 gives ( among many other states ) @xmath156 , _
i.e _ the original state translated by one site .
clearly , such processes contribute to the mobility of the polaron .
similarly , states like @xmath157 can be moved by first acting @xmath158 times with @xmath118 linking to @xmath159 , and then to @xmath160 after applying @xmath118 another @xmath158 times . for these processes to work ,
it is crucial that the number of phonons is even ( odd ) if the carrier is on an @xmath161 orbital .
this is the case for @xmath1 , but for @xmath0 we found that exactly the opposite is the case .
this explains why the phonon - modulated effective hopping , which dominates at large couplings , favours @xmath1 and this eventually becomes the ground - state .
note , furthermore , that this discussion also illustrates why an ma version restricted to a one - site phonon cloud is expected to capture well the phenomenology of this model . defined in eq .
( [ eq : nph ] ) for @xmath103 ( left panel ) and @xmath107 ( right panel ) at different values of @xmath100 .
other parameters are @xmath0 and @xmath120 . convergence was reached for @xmath162 and @xmath163 , respectively . ] to analyze the spatial extent of the polaron cloud we calculate the phonon distribution @xmath164 where @xmath165 is the polaron - state with momentum @xmath0 and @xmath166 is the distance between the carrier at site @xmath167 and the phonon at site @xmath168 .
this means that @xmath169 when the carrier is on a @xmath12-orbital , and @xmath170 when it is on an @xmath11-orbital .
the results are shown in fig .
[ fig : nph ] for @xmath103 and @xmath107 .
we see that the polaron cloud is located in the immediate vicinity of the carrier , _
i.e. _ we are dealing with a very small polaron .
if the carrier is on an @xmath11-orbital the majority of phonons are hosted by the two oxygens next to it , while for a carrier on a @xmath12-orbital the three closest oxygen sites contribute . as @xmath100
is increased past the critical coupling ( @xmath171 for @xmath103 and @xmath172 for @xmath107 ) , there is no noticeable change in the nature of the cloud .
we also find that for @xmath1 and sufficiently large @xmath100 the phonon cloud is qualitatively similar ( not shown ) to that for @xmath0 . for small @xmath100
this is not true as here the polaron state at @xmath1 is not well - separated from the polaron + phonon continuum . reducing @xmath36 drastically changes the overall number of phonons , but not the spatial extent of the cloud .
normalized by the free hole mass @xmath173 for @xmath107 ( left panel ) and @xmath127 ( right panel ) and @xmath120 .
the dashed line marks @xmath104 .
the values of @xmath88 are 14 and 10 , respectively . ] apart from the small bandwidth , the heavy nature of the polaron also becomes apparent from the value of its inverse effective mass , @xmath174 , shown in fig .
[ fig : meff ] for @xmath175 at @xmath107 and @xmath127 .
note that we normalize this with the free hole mass .
as @xmath176 from below , @xmath177 decreases indicating that the polaron becomes heavier .
this is expected because as the discussion above shows , phonon - modulated hopping can not move the polaron cloud at this momentum , and the bare hole hopping is renormalized to smaller values due to the presence of the phonon cloud . at @xmath178
the ground state changes momentum and consequently we now need to follow @xmath179 which increases because here the phonon - modulated hopping is active and increases the mobility of the polaron . for large values of @xmath36 the increase of @xmath179 for @xmath180
is substantial resulting in a much lighter polaron than at @xmath181 . for smaller values of @xmath36 , on the other hand ,
@xmath179 levels off quite rapidly and the polaron remains heavy .
this is to be expected because for smaller @xmath36 , the phonon cloud is quite large at @xmath104 , suppressing the polaron s mobility . for @xmath71 , on the other hand
, the average phonon number is small and an increase in @xmath100 results in an increase of the effective hopping between @xmath11-orbitals .
this effective hopping can become larger than the direct hopping , resulting in a lighter polaron . a major difference between the two panels of fig .
[ fig : meff ] is that for @xmath107 the inverse effective mass at @xmath1 has a minimum at finite @xmath100 .
again , this is due to the presence of the polaron+phonon continuum , which for sufficiently small @xmath100 forces the polaron band to flatten out near @xmath1 .
indeed , the location of this minimum agrees with the location of the maximum in @xmath182 in fig .
[ fig : zgs_vs_g ] .
let us now discuss the role of the charge transfer energy @xmath112 . in fig .
[ fig : lambda_c ] we show the critical coupling @xmath104 for different values of @xmath112 and @xmath36 .
the dashed lines show the perturbative result , eq .
( [ eq : lambda_c ] ) , which for sufficiently large @xmath36 is in good agreement with the ed results . as a function of @xmath36 and @xmath112 .
different values of @xmath88 were used to ensure convergence .
dashed lines are the perturbative result from eq .
( [ eq : lambda_c ] ) . ] the critical coupling increases with increasing @xmath112 .
this is not surprising since a large value of @xmath112 favors the @xmath12 character , whereas the phonon - modulated hopping favors the @xmath11 character .
consequently , a negative value of @xmath112 facilitates the transition . however , this effect is significant only for relatively large values of @xmath36 .
as @xmath36 decreases the variation of @xmath104 with @xmath112 becomes very small , suggesting that here @xmath112 has little influence on the nature of the polaron cloud .
this agrees with the conclusions drawn from fig .
[ fig : zgs_vs_g_diff_delta ] . at small values of @xmath36
the value of @xmath104 is also much larger than that predicted by perturbation theory .
however , @xmath183 and in fact the transition from @xmath1 to @xmath0 is actually achieved at smaller values of @xmath48 for small @xmath36 ( see inset of fig . [
fig : ma_gc ] ) . for these values of @xmath36
the phonon cloud is quite large and therefore it is to be expected that the bare hopping , favoring @xmath0 , is renormalized substantially and the phonon - modulated hopping , favoring @xmath1 , wins already at smaller values of @xmath48 . and ma@xmath93 .
( a ) @xmath184 for @xmath185 at different values of @xmath100 .
( b ) the absolute value of the difference in % .
( c ) the dispersion @xmath184 at @xmath186 .
( d ) @xmath184 at @xmath187 . in all cases @xmath103 and @xmath120 . ]
this concludes our analysis of the ed results .
we now briefly compare the ma predictions with the ed results , in order to validate our choice of the variational space .
this is useful because ma can be much more easily and efficiently extended to higher dimensions than ed calculations .
moreover , prior work@xcite has shown that any version of ma becomes more accurate in higher dimensions , where the bare propagators decay faster with distance outside the free - hole continuum .
note also that ma accuracy improves with increasing phonon frequency @xmath36;@xcite this is why we present results for @xmath103 where the phonon cloud is very large at strong couplings , posing a challenging test for this ( and any other ) approximation . in panel ( a ) of fig .
[ fig : ma_1 ] we compare the ma results for the polaron energies @xmath188 at different values of @xmath100 , to the ed results .
as expected for a variational approach , both ma@xmath92 and ma@xmath93 values are always larger than the ed ones .
panel ( b ) of fig .
[ fig : ma_1 ] shows the relative difference between ma and ed . at small values of @xmath100 both ma@xmath92 and ma@xmath93
are very accurate , but at larger values of @xmath100 , ma@xmath93 is clearly superior , showing that its additional configurations acquire finite weight . obviously , adding more cloud configurations will further increase accuracy , but it is clear that this rather small set already suffices to capture quantitatively quite accurately the polaron properties .
a comparison for the dispersion @xmath184 is shown in panel ( c ) of fig .
[ fig : ma_1 ] for the intermediate coupling @xmath186 and in panel ( d ) for the critical coupling @xmath189 .
panel ( c ) shows that both ma@xmath92 and ma@xmath93 give the best results for @xmath0 .
this is probably due to the fact that at @xmath186 the gs is still at @xmath0 , and a variational method like ma is expected to perform best for the gs .
note that panel ( b ) also shows that for @xmath190 the relative error is smaller for @xmath0 than for @xmath1 .
the dispersion shown in panel ( d ) is very narrow ( on this scale ) .
ma@xmath92 and ma@xmath93 both reproduce the small bandwidth quite well , but shifted to higher energies by 2% and 6% , respectively . however
, neither ma@xmath92 nor ma@xmath93 predict the exact value of @xmath189 .
instead , ma@xmath92 predicts a smaller value of 1.1 , whereas ma@xmath93 predicts a larger value of 1.45 .
this trend is true for all values of @xmath36 , as shown in fig .
[ fig : ma_gc ] . here
we also see that ma@xmath93 gives very good predictions for @xmath191 .
the agreement becomes gradually worse as @xmath192 , where because of the low cost of phonons , the spatial extent of the phonon cloud increases beyond two sites .
it also needs to be pointed out that for small values of @xmath36 small differences in the hole - phonon coupling @xmath48 are amplified in the effective coupling @xmath193 . for small @xmath36
we therefore show the values of @xmath105 in the inset of fig .
[ fig : ma_gc ] . obtained with ed for @xmath120 , and the predictions of ma@xmath92 and ma@xmath93 . ] and ma@xmath93 .
( a ) the qp weight @xmath194 for @xmath125 for different values of @xmath100 . note that @xmath111 for all @xmath100 .
( b ) @xmath194 for @xmath195 . in all cases @xmath103 and @xmath120 ] a comparison of the qp weights obtained with ma and ed is shown in fig .
[ fig : ma_2 ] .
panel ( a ) shows @xmath197 ( as pointed out above , @xmath198 for all @xmath100 ) .
panel ( b ) shows @xmath199 for @xmath1 and @xmath0 .
as expected for a variational approximation , the ma weights are larger than the ed values everywhere .
for small @xmath100 both ma@xmath92 and ma@xmath93 perform very well .
the largest quantitative disagreement appears for intermediate values of @xmath100 . here
, @xmath200 falls off quite rapidly , and this is captured only qualitatively by ma .
note that ma@xmath93 performs much better than ma@xmath92 in this @xmath100-range , indicating that a further improvement of ma by adding phonons on adjacent sites is possible .
the results improve substantially for @xmath139 . at these values ,
the qp weight does not change as rapidly anymore .
we find that here ma@xmath92 and ma@xmath93 are on equal footing , with ma@xmath92 actually outperforming ma@xmath93 at @xmath201 .
this could be related to the fact that ma@xmath93 predicts a value that is too large for @xmath104 , whereas ma@xmath92 predicts a value that is smaller than the exact result .
we have investigated the effect of phonon - modulated hopping in a two - band model describing a chain with alternating atoms with @xmath11 and @xmath12 valence orbitals , respectively . as discussed in the introduction ,
part of the motivation for this work is to study a toy 1d model inspired by the perovskite labio@xmath2 ( or nearly fully compensated babio@xmath2 ) in the scenario where holes are primarily located on the o atoms , to understand the properties of the resulting polaron and to develop a good approximation that can be easily generalized for similar 3d system .
the key finding is that at sufficiently large hole - phonon coupling , the polaron undergoes a sharp transition where its gs momentum jumps from @xmath125 to @xmath126 .
sharp polaron transitions have been observed previously in the ssh and related models with phonon - modulated hopping , @xcite however there are qualitative differences between our results and those of these studies . in these other models
the momentum changes _ smoothly _ as the effective coupling is increased .
also , the polaron mass diverges at the transition but remains finite ( and can be surprisingly light ) in the limit of infinite effective coupling . in our model , on the other hand , there is a _ discontinuous _ jump of the momentum from @xmath0 to @xmath1 . to the best of our knowledge ,
such a discontinuous transition has not been discussed before .
these results suggest that there may be more new physics to be discovered in models of carrier - phonon coupling different from that of the much studied one - band models of the holstein and fr " ohlich type .
we argued that this transition is due to a competition between the bare hopping @xmath138 and the phonon - modulated hopping @xmath118 : the former favors a @xmath0 ground state , while the latter increases polaron mobility at @xmath1 but not at @xmath0 .
this analysis suggests that a transition can only occur when @xmath138 and @xmath118 have different symmetries .
this is supported by the findings of zhang _ et .
@xcite who report a transition in an ssh - like model for symmetric bare hopping and antisymmetric phonon - modulated hopping , but not for symmetric phonon - modulated hopping .
we have used arguments based on perturbation theory , but also more general symmetry arguments , to support the view that the discontinuous nature of this transition is closely tied to the two - band nature of the model . in perturbation
theory the hybridization between the @xmath11 and @xmath12 bands leads to the square root dependence in eq .
( [ eq : eeff ] ) , which is the main ingredient needed for this type of discontinuous jump of the gs momentum .
we believe that it is highly unlikely that a transition with such a discontinuous jump can be found in a one - band model .
this suggests that the polaron properties of multi - band models are qualitatively different from those of one - band models , and that more work needs to be done to exhaustively study all possible types of polaron transitions . whether such transitions also occurs in 2d and 3d models where nn hopping between oxygens
is also included is not a priori clear ; we believe this to be the case and we are currently investigating such models .
when extended to one - site phonon cloud configurations with an arbitrary number of phonons , the arguments mentioned above allowed us to elucidate the mechanism behind the transition .
they also suggested two types of ma variational approximations , where the phonon cloud is restricted to be one - site in extent ( ma@xmath92 ) and one - site cloud with at most one additional phonon on an adjacent site ( ma@xmath93 ) .
we find that ma@xmath92 reproduces the gs energy with an accuracy of 6% or better while ma@xmath93 yields even better results with an accuracy of at least 2% .
these values were found for @xmath103 where the phonon cloud is already very large , and therefore show that ma performs well even under difficult conditions . of course , the ma accuracy is significantly less for the qp weights at intermediate coupling @xmath100 .
this is not surprising , because the part of the eigenstates ignored by the variational calculation does contribute to the wavefunctions normalization ( and therefore serves to lower the qp weight ) even if it may not have much effect on the eigenenergy .
it is important to note , however , that both ma approaches reproduce the qualitative behavior correctly , and that adding more variational states will further improve quantitative agreement .
ma is also successful in predicting an accurate value for the critical coupling @xmath104 for not too small phonon frequencies . at small @xmath36 , @xmath104
is underestimated by ma@xmath92 and overestimated by ma@xmath93 , so at least they provide bounds for its true value . the major advantage of ma , when compared to ed , is its numerical cost . for ma@xmath92 one merely needs to multiply @xmath202-matrices , whereas for ma@xmath93 the largest matrix size is @xmath203 .
in fact one can also store all of the @xmath204 and @xmath205 in one large sparse matrix , which further increases performance .
the essential aspect is that in both cases the number of states for a given number @xmath206 of phonons is independent of @xmath206 .
this is why generalizing these ma approaches to 3d will be significantly more efficient than for more general ma schemes , like that of ref . .
we have shown that in 1d these schemes already work quite well , and since ma improves its accuracy in higher dimension , these generalizations should suffice to accurately and efficiently obtain polaron properties in the multi - band , 3d model .
this work is now in progress .
we thank george sawatzky for suggesting this problem , and c.p.j .
adolphs for sharing his experience with ed .
this work was supported by nserc , qmi and the ubc 4yf ( m.m.m . ) .
the equations of motion ( eom ) for @xmath207 can be constructed by repeatedly using dyson s identity @xmath208 , where @xmath209 and @xmath210 is the resolvent for @xmath211 .
choosing @xmath212 and doing this once yields : @xmath213 where we defined the generalized gfs : @xmath214 these describe the projection onto states with a one - site phonon cloud , and are therefore the only gfs included in ma@xmath92 .
note that we suppressed the @xmath215 index in the definition of the generalized gfs since the eom do not couple to gfs with different @xmath215 .
for ma@xmath93 , we also include the contribution from the generalized gfs : @xmath216 with one additional phonon either to the left or to the right of the phonon cloud .
note that for @xmath217 , @xmath218 .
we now apply dyson s identity again , to obtain the eom for the generalized gfs . suppressing the @xmath219 argument for simplicity , we find : @xmath220 @xmath221 @xmath222 where we introduced the free hole real - space gfs : @xmath223 for simplicity its @xmath224 argument
was again suppressed in the equations above . in 1d , these free hole real - space gfs can be obtained analytically , see appendix b. the right hand side of these eom only links to generalized gfs with specific values of @xmath225 .
let us first consider the case of ma@xmath92 .
here we only consider generalized gfs of the type @xmath226 . by inspection of eq .
( [ eq : eom_f ] ) we find that the eom only link to @xmath227 and @xmath228 .
furthermore since @xmath229 and @xmath230 and @xmath231 ( see appendix b ) the eom simplify to : @xmath232 we define the column vector @xmath233 and recast the eom in the form @xmath234 where @xmath235 and @xmath236 are @xmath237-matrices which for @xmath238 can be read off directly from the equations above . to get @xmath239 we use @xmath240 and @xmath241 , where @xmath215 is the same as in eq .
( [ eq : eom_g ] ) .
the eom can now be solved with the ansatz @xmath242 , @xcite which is justified because for large @xmath206 , @xmath243 goes to zero and therefore @xmath244 must go to zero as well .
plugging this ansatz back into eq .
( [ eq : recurrence ] ) yields @xmath245^{-1}\tilde{\alpha}_n$ ] which can be calculated recursively starting with @xmath246 , where @xmath247 is chosen sufficiently large so that its further increase has no effect on the results .
once we have calculated @xmath248 linking @xmath249 to @xmath250 , we can rewrite eq .
( [ eq : eom_g ] ) in matrix form : @xmath251 note that this requires identifying the first row of @xmath252 with @xmath253 and the second row with @xmath254 .
@xmath255 is the @xmath202-matrix with @xmath256 .
the self - energy @xmath257 is defined by the @xmath258-matrix equation @xmath259^{-1}$ ] . in the ma@xmath92 approximation
it is therefore given by : @xmath260 we now show how to use the ma@xmath92 eom to rigorously derive the perturbation result of sec .
[ sec : perturbation - theory ] . for a cutoff of @xmath261 , we find that @xmath262 and consequently @xmath263 for @xmath264 we have @xmath265 and @xmath266 ( see appendix b ) and we recover the result of eq .
( [ eq : heff_matrix ] ) .
the ma@xmath93 case is treated in exactly the same manner , but is slightly more tedious . for @xmath267 phonons
we include the following 17 gfs : @xmath268 with @xmath269 ; @xmath270 with @xmath271 ; @xmath272 with @xmath273 ; @xmath274 with @xmath275 ; @xmath276 with @xmath277 ; and @xmath278 with @xmath279 . for @xmath280 ,
some of these gfs are identical so that their number is reduced to 12 .
similarly , for @xmath217 one only needs to keep 7 gfs .
the generalized gfs are again arranged in a vector @xmath281 and the eom recast as a recurrence equation : @xmath282 the matrices @xmath204 and @xmath205 can be read off from eqs .
( [ eq : eom_f ] ) , ( [ eq : eom_f(1,n ) ] ) and ( [ eq : eom_f(n,1 ) ] ) .
furthermore , eq . ( [ eq : def_f ] ) indicates that @xmath283 which we use to read off @xmath284 . again
the eom are solved with the ansatz @xmath285 @xcite , yielding @xmath286^{-1}\alpha_n$ ] .
we then rewrite eq .
( [ eq : eom_g ] ) in matrix form : @xmath287 here @xmath288 is a reduced version of @xmath289 .
it is a @xmath290-matrix which only contains the rows of @xmath289 linking @xmath291 and @xmath292 to @xmath293 .
this is necessary since the other 4 generalized gfs contained in @xmath294 do not appear in eq .
( [ eq : eom_g ] ) .
the matrix @xmath88 is a @xmath295-matrix whose only non - zero elements are @xmath296 and @xmath297 .
the self - energy @xmath257 in the ma@xmath93 approximation is , then : @xmath298
the real - space gfs of @xmath299 are defined as @xmath300 .
this can be rewritten as @xmath301 , _
i.e. _ @xmath302 measures the probability amplitude that a hole injected at site 0 will be removed at site @xmath225 .
the eom are obtained using dyson s identity . for @xmath303 , and suppressing the @xmath219-dependence , we find : @xmath304 where we defined the shorthand @xmath305 . for @xmath306 , the general form of the eom
is @xmath307 eliminating @xmath308 , the eom for @xmath309 , with @xmath310 , can be recast as @xmath311 = -t^2 g_{n-1}^{p , p } - t^2 g_{n+1}^{p , p}\end{aligned}\ ] ] similarly we find @xmath312 in the time - domain the small imaginary part @xmath313 corresponds to a finite lifetime of the hole .
therefore the probability that a hole travels from site 0 to site @xmath314 falls off exponentially as @xmath315 . and we use the ansatz @xmath316 , for @xmath317 . for @xmath318 we need to use @xmath319 . in both cases , plugging the ansatz back into the eom gives : @xmath320 we need to choose the solution which satisfies @xmath321 .
using the ansatz in eq .
( [ eq : eom_g0pp ] ) we obtain @xmath322 from this all the other @xmath309 are obtained as @xmath323 .
we can apply the same procedure to find analytical expressions for the @xmath308 .
however , we need to be careful since both @xmath324 and @xmath325 link to @xmath326 and therefore need to be treated separately . after some algebra
we find @xmath327 @xmath328 similarly one finds : @xmath329 @xmath330 note that for @xmath331 , eq . ( [ eq : z ] ) implies that @xmath332 . using this in the expressions for @xmath333 and @xmath334
we find that they scale as @xmath335 .
the diagonal real - space gfs , @xmath336 and @xmath337 on the other hand scale as @xmath338 .
all real - space gfs with larger values of @xmath339 go to 0 since @xmath340 .
54ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevlett.92.146403 [ * * , ( ) ] \doibase 10.1023/a:1007792713484 [ * * , ( ) ] link:\doibase 10.1038/nmat3378 [ * * , ( ) ] http://stacks.iop.org/0953-8984/20/i=4/a=043201 [ * * , ( ) ] link:\doibase 10.1080/00018730210140805 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.112.106404 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.144101 [ * * , ( ) ] link:\doibase 10.1103/physrevb.50.7222 [ * * , ( ) ] link:\doibase 10.1103/physrevb.91.121114 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.256402 [ * * , ( ) ] link:\doibase 10.1103/physrevb.60.1633 [ * * , ( ) ] link:\doibase 10.1103/physrevb.65.174306 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.104304 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.115114 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.165113 [ * * , ( ) ] link:\doibase 10.1103/physrevb.90.125131 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.036402 [ * * , ( ) ] link:\doibase 10.1103/physrevb.76.165109 [ * * , ( ) ] link:\doibase 10.1103/physrevb.76.174305 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.035453 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.235120 [ * * , ( ) ] link:\doibase 10.1103/physrevb.89.035122 [ * * , ( ) ] link:\doibase 10.1103/physrevb.90.085149 [ * * , ( ) ] link:\doibase 10.1080/14786445008521794 [ * * , ( ) ] link:\doibase 10.1080/00018735400101213 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.42.1698 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.266605 [ * * , ( ) ] link:\doibase 10.1063/1.4733986 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.223002 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.63.63 [ * * , ( ) ] link:\doibase 10.1038/332814a0 [ * * , ( ) ] link:\doibase 10.1016/0038 - 1098(75)90327 - 0
[ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.017001 [ * * , ( ) ] link:\doibase 10.1103/physrevx.3.021011 [ * * , ( ) ] link:\doibase 10.1103/physrevb.88.224509 [ * * , ( ) ] link:\doibase 10.1103/physrevb.65.115112 [ * * , ( ) ] link:\doibase 10.1103/physrevb.56.3105 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.47.689 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0038-1098(76)90632-3 [ * * , ( ) ] link:\doibase 10.1107/s0567740879002417 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.61.2713 [ * * , ( ) ] link:\doibase 10.1103/physrevb.76.174103 [ * * , ( ) ] link:\doibase 10.1016/j.physc.2015.02.012 [ * * , ( ) ] , link:\doibase 10.1103/physrevlett.95.176403 [ * * , ( ) ] @noop ( ) , http://stacks.iop.org/0953-8984/12/i=16/a=303 [ * * , ( ) ] link:\doibase 10.1134/1.1410607 [ * * , ( ) ] link:\doibase 10.1103/physrevb.61.11263 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.156402 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.126404 [ * * , ( ) ] http://stacks.iop.org/0295-5075/102/i=4/a=47003 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.054519 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.66.763 [ * * , ( ) ] link:\doibase 10.1103/physrevb.74.245104 [ * * , ( ) ] | we present exact diagonalization and momentum average approximation ( ma ) results for the single polaron properties of a one - dimensional two - band model with phonon - modulated hopping . at strong electron - phonon coupling
, we find a novel type of sharp transition , where the polaron ground state momentum jumps discontinuously from @xmath0 to @xmath1 .
the nature and origin of this transition is investigated and compared to that of the su - schrieffer - heeger ( ssh ) model , where a sharp but smooth transition was previously reported .
we argue that such discontinuous transitions are a consequence of the multi - band nature of the model , and are unlikely to be observed in one - band models .
we also show that ma describes qualitatively and even quantitatively accurately this polaron and its transition .
given its computationally efficient generalization to higher dimensions , ma thus promises to allow for accurate studies of electron - phonon coupling in multi - band models in higher dimensions . |
metals become superconducting , when their electrons form singlet cooper pairs via an attractive interaction . on the other hand , electrons in metals
can also undergo another type of singlet formation , namely to form a kondo screening cloud , when magnetic impurities are present . the fascinating interplay of cooper pair and kondo singlet formation @xcite
can be microscopically observed when magnetic atoms or molecules are adsorbed on superconducting surfaces .
the ground state of such a combined system has been predicted to be either a kondo screened singlet state ( @xmath0 ) , if the kondo scale @xmath1 is much larger than the superconducting pairing energy @xmath2 , or an unscreened multiplet state ( @xmath3 ) for @xmath4 .
characteristic features include subgap states with bound state energies @xmath5 - often called shiba states .
for the singlet ground state with screened impurity spin , the bound states are @xmath3 excitations at @xmath6 , where the kondo singlet is broken , and for the multiplet ground state the bound states with @xmath7 are singlet excitations including kondo screening .
when the energy scale for kondo singlet formation becomes smaller and comparable with the cooper pairing energy , @xmath8 , the bound state energies at @xmath5 go to @xmath9 , and the @xmath3 state becomes the ground state . at this point
a quantum phase transition ( qpt ) occurs .
the cooper pair breaking effect is expected to behave like @xmath10 and thus is most effective here leading to a strong suppression of the superconducting @xmath11 for larger impurity concentrations.@xcite for a long time the accurate resolution of the subgap bound states and their dependence on @xmath12 and @xmath13 have remained elusive .
recently , the bound states have been analyzed in tunable mesoscopic superconductor - quantum dot - normal lead structures.@xcite scanning tunneling microscopy ( stm ) has been used to detect the local influence of single magnetic atoms on a superconducting substrate.@xcite to study the interplay , an experiment with variable magnetic interaction strengths is desirable .
manganese - phthalocyanine ( mnpc ) molecules on a pb(111 ) substrate form a moire - like superstructure .
tunneling spectroscopy on different mn sites reveals two kondo screening channels and pairs of bound states of varying energy @xmath5 .
the large number of different adsorption sites leads to a variety of magnetic interactions and kondo scales .
the smaller kondo scale lies in the interesting regime @xmath8 .
shiba states crossing the fermi level and the predicted qpt could be observed.@xcite a number of important question remained however unresolved . what is the role of the second kondo screening channel and does it give rise to a shift of the critical point of the qpt ? which microscopic parameter drives the behavior of the system across the qpt ?
does the asymmetry in the stm intensity of the bound states reveal particular physical properties of the system ?
our theoretical approach to address those questions is a combination of scaling arguments and numerical renormalization group ( nrg ) calculations .
@xcite the former are used to connect the complex experimental situation to an effective anderson impurity model with one relevant channel and its model parameters .
nrg calculations for this model demonstrate that the experimental behavior can be quantitatively understood by _ only _ varying the hybridization between mnpc and the substrate .
the results are reliable as the nrg is known to capture the kondo effect accurately in contrast to many other methods , used to describe impurities in superconductors , which contain mean field aspects like classical spins.@xcite the accuracy of the theoretical modeling is tested by the direct comparison of the experimental and theoretical results for the point of the qpt and the positions and weights of the bound states .
= 100mv ) , c ) tunneling spectra taken on different molecules on top of the mn center ( feedback opened at i=450pa , @xmath14=8.6mv ) . also shown are the fits used to derive the deconvoluted spectra below in ( d ) .
e ) examples of nrg spectra with broadened subgap peaks for similar @xmath15 . ]
as detailed in ref . , the mnpc molecules ( see fig . [ fig:1](a ) ) have been deposited on an atomically clean pb(111 ) substrate at room temperature under ultra - high vacuum conditions .
stm at 4.5k resolves highly ordered islands ( fig .
[ fig:1](b ) ) .
tunneling spectroscopy has been used to resolve the superconducting gap structure and its subgap states , as well as kondo resonances on different molecules .
the superconducting state of the pb(111 ) substrate and pb tip shows as pronounced differential conductance peaks at @xmath16 with @xmath17mev .
three differential conductance spectra taken on different mnpc molecules are shown in fig .
[ fig:1](c ) .
two larger peaks as well as the two smaller peaks are located at symmetric bias voltages within the gap of the superconductor - superconductor tunneling barrier .
the larger peaks are an expression of the shiba states and indicate the magnetic interaction with the superconducting substrate .
the smaller peaks are a result of thermal excitations at the measurement temperature of 4.5k across the gap . in order to remove the effect of the superconducting tip and finite temperature on the tunneling spectra
, we developed a deconvolution method @xcite .
this procedure consists of extracting the superconducting density of states of the tip from spectra on the bare surface and using the result for fitting the differential conductance spectra of the mnpc spectra assuming a set of shiba states ( for details see ref . , and the appendix ) .
the result is representative for the quasi - particle density of states ( dos ) of the mnpc molecule on the superconducting pb surface ( fig .
[ fig:1](d ) ) . from these plots
we can deduce the energy of the shiba states and their intensity .
@xmath18 ( @xmath19 ) is the energy for the bound state with larger ( smaller ) weight @xmath20 ( @xmath21 ) , where @xmath22 .
we observe a gradual increase in the asymmetry of the weights when @xmath18 shifts from negative values to positive ones .
this is shown in fig .
[ fig : wbgowbs ] as ratio @xmath23 .
vs bound state energy @xmath24 .
comparison of experimental results with theory .
in all the range of bound state energy we used the model parameters @xmath25 , and @xmath26 and the overlap @xmath27 was varied . ]
we additionally identify a broad peak at the mnpc center around the fermi energy ( see fig .
[ fig : ebvstk](b ) ) , which can be fitted by two fano lineshapes , representing two different kondo screening processes with @xmath28.@xcite @xmath29 with @xmath30mev scales with @xmath31mev ( for details the appendix ) .
the occurrence of two kondo screening channels can be related to the spin state of the mnpc molecule . for the isolated mnpc complex ,
density functional theory ( dft ) calculations have found that the high spin configuration of mn ( @xmath32 ) is reduced to @xmath33 and the unpaired electrons occupy the @xmath34 , the @xmath35 , and the @xmath36 orbital ( @xmath37-axis along an arm of mnpc ) and are aligned due to hund s rule coupling.@xcite when adsorbed on the pb surface ( in @xmath38 direction ) , the @xmath39 orbital hybridizes strongly with the pb states and is therefore assumed to be quenched.@xcite hence , the observation of two screening channels suggests a spin state of @xmath40 .
since @xmath41 , kondo screening dominates for this channel.@xcite therefore , we only correlate @xmath42 with the appearance of the bound states inside the gap .
[ fig : ebvstk ] shows the dependence of @xmath15 on @xmath43 . at large @xmath44 ,
the kondo screening is efficient and the many - body ground state is a singlet . here
tunneling can occur via the doublet state which involves breaking of the kondo singlet and rearrangement of cooper pairs . with decreasing @xmath44
this requires less energy and we find @xmath45 .
the level crossing occurs at @xmath46 .
this point can be regarded as the critical point of the qpt and is a universal feature .
further decrease of @xmath44 leads to the unscreened doublet ground state .
= 8.6mv ) .
b ) tunneling spectrum on the same molecule as in ( a ) in a larger energy scale evidencing a broad background , which can be fitted by two kondo resonances ( feedback opened at i=470pa , @xmath14=130mv ) .
c ) bound state energies @xmath47 vs kondo temperature @xmath48 .
comparison of experimental results with theory for the same choice of model parameters as in fig .
[ fig : wbgowbs ] . ]
for the theoretical description we use an anderson impurity model ( aim ) hamiltonian of the form @xmath49 the superconducting medium reads , @xmath50 where @xmath51 creates a band electron with momentum @xmath52 , spin @xmath53 and band index @xmath54 , where @xmath55 , and there are @xmath56 available channels .
@xmath57 is the corresponding electronic dispersion and @xmath2 the gap parameter chosen real .
the band electrons hybridize with the impurity states via @xmath58 where @xmath59 creates a d - level impurity electron with spin @xmath53 and index @xmath60 . in the present situation ,
the number of conduction channels which hybridize is equal to the d - orbital states , i.e. @xmath61 .
we will assume different matrix elements @xmath62 due to different overlapping integrals .
these matrix elements determine the energy scale for the hybridization of d - states with the substrate through @xmath63 , where @xmath64 is the dos of the conduction band at the fermi level @xmath65 . despite the experimental observation of two kondo screening channels
, we will now show that the behavior of the shiba states can be well described by a single channel model ( @xmath66 ) , suggesting a low energy decoupling of the kondo channels . for the single channel case the `` d - orbital '' term @xmath67 simply reads ,
@xmath68 with the d - level position @xmath69 relative to @xmath70 and the on - site coulomb interaction with strength @xmath71 , where @xmath72 . for this model
we perform nrg calculations @xcite to calculate the lowest energy excitations and their spectral weights , which characterize the subgap bound states .
examples for the low energy spectra can be seen in fig .
[ fig:1 ] ( e ) .
we now explain how to choose the model parameters for the different mnpc molecules on the pb(111 ) surface .
we expect that the main difference for the mnpc molecules in the different adsorption sites is the magnitude of the hybridization @xmath27 , which in turn leads to different @xmath73 .
the energy level alignment of the d - states @xmath69 and coulomb energy @xmath71 are , on the contrary , expected to change little with the site.@xcite we therefore _ only _ vary @xmath27 to explain the data .
the superconducting gap @xmath17mev sets the energy scale .
the relation of values for @xmath69 , @xmath27 and @xmath71 is constrained to give suitable values for the kondo temperature @xmath74 .
their actual value can be fixed by matching the strong experimental weight asymmetry @xmath23 of the shiba states for the maximal @xmath75 on the doublet side in fig .
[ fig : wbgowbs ] .
this yields @xmath76 , @xmath26 , and @xmath77 .
this corresponds to an asymmetric aim , with @xmath78 .
a variation of @xmath79 from @xmath80 then reproduces accurately the weight asymmetry in fig .
[ fig : wbgowbs ] on decreasing @xmath18 and also the variation of @xmath81 with @xmath82 in fig .
[ fig : ebvstk ] .
nrg calculations for an asymmetric one - channel aim can thus account for the experimentally observed universal features such as @xmath83 and non - universal ones like @xmath23 .
most important for the understanding of the physical processes is the correct description of the qpt .
for the one channel non - degenerate aim and the kondo model , nrg studies @xcite have estimated that the phase transition occurs when @xmath84 . a deviation from this
would indicate that a different number of channels contribute to the kondo screening of the same electron spin state.@xcite in particular , this may illustrate the role of the second kondo channel observed in the experiment . at first sight the experimental result for the qpt , @xmath85 ( fig .
[ fig : ebvstk ] ) , seems to suggest a more complicated situation than a one - channel model .
however , we find that the origin of this discrepancy is the use of different definitions of @xmath73 , which can vary by a prefactor . in the theoretical works @xcite the definition @xcite @xmath86^{1/2}{\mathrm e}^{\frac{\pi{\varepsilon}_{\rm d}({\varepsilon}_{\rm d}+u)}{2\gamma u } } \label{tk4}\ ] ] was used . for the experimental values of @xmath73 , we employ the widely used definition,@xcite based on the width of the kondo resonance @xmath87 ( half width at half maximum ) in the limit @xmath88 .
we adopt the same definition in our nrg calculations ( @xmath89 ) and find that @xmath87 and the definition in eq .
( [ tk4 ] ) can differ by a factor 4 ( see also ref . ) . taking this into account our result for the qpt in fig .
[ fig : ebvstk ] is in excellent agreement with the theoretical prediction for a one - channel model .
we can thus conclude that the second kondo screening channel does not shift the transition point .
we now discuss the emergence of the low energy effective one - channel model and its parameters using scaling arguments .
first notice that the magnitudes of @xmath69 , @xmath71 and @xmath27 do not correspond to usual atomic values @xmath90 , but rather to values of the order @xmath91mev .
, @xmath71 @xmath90 with varying @xmath27 do not reproduce the behavior of @xmath23 very well . ]
this is related to the fact that the aim under consideration is an effective model valid for low energies . using insights from the dft calculations for mnpc,@xcite and the observation of two kondo channels , we start with a model of the form of equation ( [ eq : ham ] ) with the two d - levels coupled to two bands ( @xmath56@xmath92@xmath93@xmath92@xmath94 ) with the hybridization terms @xmath95@xmath96@xmath97 .
channel 1 thus describes the physical processes related to the smaller kondo temperature .
we assign it to the @xmath98 orbital , since due the spatial orientation the overlap with the substrate is much larger for the @xmath99 orbital .
the impurity term can be written in terms of the level positions @xmath100 relative to @xmath65 , intra orbital coulomb energy @xmath101 , inter orbital coulomb interaction @xmath102 and a hund s rule interaction @xmath103 , @xmath104 where @xmath105 with the pauli matrix @xmath106 .
the complete set of bare parameters of this model can not be extracted from existing dft calculations,@xcite or experimental observations .
therefore , the following are qualitative arguments on general grounds . and @xmath107 in the two regimes @xmath108 and @xmath109 .
scaling stops at @xmath110 , and the effective one band model is used to compute the low energy properties . ]
we now derive scaling equations,@xcite which connect the original model and its parameters at a high energy scale @xmath111 of the order of the electronic bandwidth to the low energy effective ones at @xmath112 , where the second channel decouples due to its complete kondo screening .
we focus on the quantities in channel 1 , where the shiba states occur .
we only include processes with matrix elements @xmath113 explicitly , where the level occupations of @xmath114 changes .
there are also contributions @xmath115 , whose inclusion leads to quantitative changes in the equations but does not alter the conclusions ( for details see the appendix ) .
band structure calculations @xcite for pb suggest a situation , where the band @xmath116 is asymmetric with respect to @xmath117 , @xmath118 .
then , we have two scaling regimes ( i , ii , see fig .
[ fig : scaling ] ) : i , where the scaling process has no counterpart in the occupied states , and ii , where the residual band structure is symmetric around @xmath117 .
the scaling equations for the energies @xmath114 and @xmath71 in regime i , i.e. , in @xmath119 read,@xcite @xmath120 where @xmath121 . different from the usual approaches we also use a scaling equation for @xmath95 , which is derived from @xmath122 , assuming @xmath123 as a scaling invariant .
such a scaling procedure can be continued as long as the levels @xmath124 , @xmath125 lie within @xmath126 and do not interfere with the kondo scale .
we find @xmath127 and @xmath128 , such that @xmath129 is shifting away from @xmath130 when scaling to lower energy and @xmath71 decreases . the model becomes more asymmetric in this regime .
when @xmath131 , we reach the scaling regime ii , with the corresponding scaling equations @xmath132 for starting values , @xmath133 , @xmath134 , one has @xmath135 . in this case
@xmath114 is shifting towards @xmath130 . due to the term in the denominator
the effect becomes strong when @xmath136 [ see fig .
[ fig : scaling ] ( right ) ] . for the same effective starting values we have @xmath128 .
hence , @xmath137 and ( @xmath138 decrease further under the scaling . this scaling can be continued until we reach @xmath139 .
we have @xmath140ev , and for usual estimates for @xmath103 we expect @xmath141 . at this scale
the spins lock into the high spin @xmath40 configuration due to the dominating hund s coupling . as shown in ref .
, this leads to a reduction of the magnetic coupling @xmath142 and the kondo scale . in our approach
it can be included in @xmath95 , such that the scaling is slowed down.@xcite the scaling can then be continued until @xmath143 .
then one part of the impurity spin becomes kondo screened and decouples . at this scale
we have an effective single band model with @xmath144 .
its parameters are substantially reduced from the bare values and generically asymmetric , which clarifies our choice in the earlier nrg calculations .
since nrg calculations for multi - channel models are very challenging , our approach combined with scaling equations could be useful in other situations for complex molecules on surfaces . in conclusion , we provide a unified experimental and theoretical perspective of the microscopic interplay of kondo screening and superconducting pairing for mnpc on lead , as manifested in the subgap bound states .
we identify the change of hybridization as the relevant quantity and show that in spite of the complex spin state of mnpc and two kondo screening channels , an effective description based on the one - channel aim captures both universal aspects like @xmath145 and non - universal ones like the asymmetry of the weights very well . in the future
it would be interesting to envisage situations , where different impurity spins can interact , such that different kinds of quantum phase transitions can occur .
* acknowledgment * we wish to thank p. coleman , m. haverkort , a.c .
hewson , a. subedi for helpful discussions and g. schulze for the deconvolution program of tunneling spectra .
jb acknowledges financial support from the dfg through ba 4371/1 - 1 .
we also thank the focus area nanoscale of freie universitt berlin and the deutsche forschungsgemeinschaft through sfb 658 for financial support .
in conventional stm experiments , to first order approximation the @xmath146 spectra resemble the density of states of the sample .
this relies on a constant density of states of the stm tip . in the presented experiment ,
the tip is coated with the superconducting material of the substrate , thus exhibiting a bardeen - cooper - schrieffer ( bcs ) like density of states . to interpret the measured @xmath146 spectra as the density of states of the sample , the influence of the tip dos has to be extracted .
we do this by a deconvolution procedure , which has been described and tested in the supplementary material of ref . .
the stm current @xmath147 as a function of voltage @xmath14 has the following form , @xmath148,\ ] ] where @xmath149 is the local density of states of the sample at position @xmath150 , @xmath151 the one of the tip and @xmath152 is the fermi function .
the deconvolution method consists of a two step process based on fitting the experimental spectra with a simulated density of states . in the first step ,
the experimental spectrum of the clean surface @xmath153 is measured and fitted .
the density of states of sample and tip are equal , @xmath154 , and follow a bcs density of states broadened by a lorentzian function . to reproduce the @xmath146 spectra
, we calculate the tunneling current at each bias voltage and numerically derive the @xmath146 function .
this function is then fitted to the experimental data to extract the temperature , superconducting gap and broadening . since the same tip and experimental conditions
are used to measure the @xmath155 spectra on the mnpc molecules , these parameters remain fixed for modeling the tip density of states @xmath151 , which is necessary in the next step .
the shiba states are simulated by two lorentz peaks inside a symmetric gap ( modeled by two broadened step functions ) around the fermi level . with this model density of states , the tunneling current
is again calculated at each bias voltage and the @xmath146 signal derived numerically .
the fitting procedure then gives a result for the local @xmath156 from which the position , width , and amplitude of the shiba states can be extracted .
to show the validity and precision of the procedure , we present in fig .
1c ) a direct comparison of the measured and fitted spectra for a number of examples .
for a quantitative comparison between experiment and theory it is important that the same definition of the kondo temperature @xmath73 is used . in the literature one can find different definitions for the kondo temperature @xmath73.@xcite
theoretically , it is convenient to define it via the magnetic susceptibility of the impurity in the limit @xmath88 , @xmath157 with a suitable prefactor @xmath37 .
this quantity is however , not experimentally measured in the present context .
the quantity which is most easily accessible experimentally is the width of the kondo resonance @xmath87 ( half width at half maximum ) . to make the comparison of experiment and theory consistent
we have used this definition also in the theory , i.e. @xmath158 . the experimental data is well understood in terms of an overlap of two fano functions for the two kondo channels with two different kondo temperatures @xmath123 , @xmath159 . for details ref . can be consulted . in table
[ table : tk1tk2 ] we display a number of values found on different molecules .
.exemplary values for @xmath123 and @xmath159 on different molecules with energy of the larger shiba state @xmath160 .
[ cols="^,^,^",options="header " , ]
here we present the complete set of scaling equations for the two - channel model .
we will argue that the conclusions in the main text are not changed qualitatively due to additional terms proportional to @xmath161 .
the starting point is the effective two orbital model in equation ( [ eq : ham ] ) together with ( [ eq : twochannel ] ) . in the following we focus on the complete scaling equations in regime ii .
the scaling equations in regime i only contain a part of the terms and show how the asymmetry is increased for an asymmetric band .
they can be derived and discussed in a similar fashion . for each orbital
there are 4 possible states , such that there are 16 atomic states which can be described by the occupation numbers @xmath162 , @xmath163 for orbital 1 and 2 , respectively . due to the hund s rule term ( @xmath164 ) , @xmath165 the atomic eigenbasis for the singly occupied situation , @xmath166 is given by singlet and triplet states , @xmath167 , and hence we use those quantum numbers , where @xmath168 is the total spin with @xmath38-component @xmath169 .
one has @xmath170 i.e. , the triplet state state has the lowest energy , and the energy difference is @xmath171 for our definition of @xmath172 , which does not include a factor @xmath173 . for the derivation of the scaling equations it is useful to define hubbard operators @xmath174 for the atomic states.@xcite the atomic terms can be given in diagonal energy representation with @xmath175 and @xmath176 .
the hybridization term constitutes of a lengthy expression in terms of those hubbard operators taking into account all possible processes of different occupation .
the single particle occupation parameters of the hamiltonian can be expressed as @xmath177 . by projecting out high energetic particle and hole states in the bath @xcite we find the equations , @xmath178 \label{eq : scalinged1compl}\\ & + & \frac{\gamma_2}{\pi } \big[\frac{3/2}{\lambda+{\varepsilon}_{d,2}+u_{12}-j_{\rm h}}+\frac{1/2}{\lambda+{\varepsilon}_{d,2}+u_{12}+3 j_{\rm h}}-\frac{2}{\lambda+{\varepsilon}_{d,2 } } \big ] , \nonumber\end{aligned}\ ] ] @xmath179 \label{eq : scalingu1compl } \\ & + & \frac{\gamma_2}{\pi}\big[\frac{2}{\lambda+{\varepsilon}_{d,2}+2u_{12}}-\frac{3}{\lambda+{\varepsilon}_{d,2}+u_{12}-j_{\rm h } } -\frac{1}{\lambda+{\varepsilon}_{d,2}+u_{12}+3j_{\rm h}}+\frac{2}{\lambda+{\varepsilon}_{d,2 } } \big ] .
\nonumber\end{aligned}\ ] ] similar equations are found for @xmath180 and @xmath181 , where one has to interchange the indices @xmath182 in @xmath183 , @xmath184 and @xmath185 on the right hand side . an equation for @xmath103 follows from @xmath186 , @xmath187.\end{aligned}\ ] ] we can represents @xmath102 as @xmath188 , and find @xmath189.\end{aligned}\ ] ] we can use the kondo scales in the two channels @xmath190 as scaling invariants . there is an approximate form @xmath191 for the situation with hund s rule coupling,@xcite where @xmath192 is the kondo scale for @xmath193 . from this
we can also derive scaling equations for @xmath194 .
if one neglects @xmath103 in the denominator as being smaller than the other scales then one finds @xmath195 , such that the hund s coupling varies little.@xcite the eqs .
( [ eq : scalii1 ] ) and ( [ eq : scalii2 ] ) in the main text for regime ii correspond to ( [ eq : scalinged1compl ] ) , ( [ eq : scalingu1compl ] ) without the terms proportional to @xmath97 . in order to have kondo physics in both channels ,
we need the atomic ground state to be one where both orbitals are singly occupied , which implies @xmath196 and @xmath197 .
we also expect a deviation from particle - hole symmetry @xmath198 . in the main text
we have argued that from the term proportional to @xmath95 one finds @xmath135 .
it is easy to see that for @xmath199 , @xmath200 and @xmath201 the term proportional to @xmath97 is also negative , such that @xmath135 is reinforced .
similarly , the additional term @xmath202 in eq .
( [ eq : scalingu1compl ] ) has the same sign as the first term and therefore @xmath203 is maintained .
hence , the qualitative picture remains unchanged , when these additional terms are taken into account .
there will however be quantitative changes in the scaling .
the quantities in channel 2 , @xmath180 , @xmath181 possess a similar scaling flow , but are not of major interest here as the corresponding kondo temperature @xmath159 largely exceeds the superconducting gap . a more complete discussion of the different regimes of the scaling equations and the different resulting behavior is beyond the scope of this work , but can be subject of a separate study . | magnetic molecules adsorbed on a superconductor give rise to a local competition of cooper pair and kondo singlet formation inducing subgap bound states . for manganese - phthalocyanine molecules on a pb(111 ) substrate , scanning tunneling spectroscopy resolves pairs of subgap bound states and two kondo screening channels .
we show in a combined approach of scaling and numerical renormalization group calculations that the intriguing relation between kondo screening and superconducting pairing is solely determined by the hybridization strength with the substrate .
we demonstrate that an effective one - channel anderson impurity model with a sizable particle - hole asymmetry captures universal and non - universal observations in the system quantitatively .
the model parameters and disentanglement of the two screening channels are elucidated by scaling arguments . |
at the large hadron collider ( lhc ) top quark pairs ( @xmath0 ) will be produced copiously .
the next - to - leading order cross section at lhc energies has been estimated to be 833 pb @xcite . due to this large cross section , at fairly low integrated luminosities ranging from approximately 10 @xmath1 to 100 @xmath1 of lhc data , a top quark signal could be established and the first top quark pair production cross section measurements could be performed .
more details are provided in section 2 .
in addition , top quark events can be used for the calibration of residual jet energy scale corrections , as it is presented in section 3 , or b - jet identification efficiencies @xcite . for all samples used in the different studies presented here
@xmath2 collisions were generated at @xmath3 tev . in order to mimic the realistic conditions during the early data taking ,
effects due to tracker misalignment and calorimeter miscalibration were taken into account in the reconstruction .
the effect of multiple proton - proton interactions was not included .
for the different analysis reviewed in this section the event selection strategy is based on a sequential cut procedure based on robust variables . among them ,
lepton and jet transverse momentum or energy , lepton isolation variables , electron identification related variables , or missing transverse energy .
all relevant backgrounds , except the single top production have been taken into account .
different sources of systematic errors have been considered , including those associated to lepton energy scale , trigger efficiencies , jet reconstruction and energy scale , missing transverse energy , parton distribution functions and background cross sections .
the observability and cross section measurements of the @xmath0 production are reviewed in the following subsections for different possible final states of the @xmath0 decay .
the semileptonic muon channel , the dilepton channel and the @xmath4 and @xmath5 channels are considered in sections 2.2 , 2.3 and 2.4 respectively .
more details are provided in @xcite , @xcite , @xcite and @xcite .
a simple and robust event selection was developed , requesting exactly one tightly isolated muon with @xmath6 gev and at least four jets with @xmath7 gev for the leading jet and @xmath8 gev for the other jets . for 10 @xmath1 of integrated luminosity ,
128 signal events are expected , corresponding to a selection efficiency of 10.3@xmath9 , together with 25 other @xmath0 final state events , as well as 45 ( 7 ) w+jets ( z+jets ) events .
the number of qcd background events is determined as 11 with a large uncertainty .
figure [ semifig ] shows the jet multiplicity distribution for events passing the final selection except the requirement of @xmath10 and the invariant mass of the three jets with the highest vectorially summed @xmath11 for the final selection . a clear signal stands out in the first 10 @xmath1 of cms data in events selected with two high momentum leptons , missing transverse energy , and at least two jets with expected signal - to - noise ratio of 7 to 1 in all events combined and 25 to 1 in electron - muon final state alone .
the most important background comes from z+jets events . for an integrated luminosity of 100 @xmath1 the detector
is assumed to be calibrated and aligned with an amount of data corresponding to 10 @xmath1 .
an additional cut on a robust b - jet identification discriminator based on a track counting algorithm leads to an almost background free event selection where about 160 signal events are expected . for this scenario
the cross - section can be extracted by means of a robust event counting method and the associated statistical uncertainty is found to be @xmath12 .
figure [ dilfig ] shows the expected number of dilepton events in the first 10 @xmath1 passing the final selection except the requirement of @xmath13 and the missing transverse energy for the final selected events in @xmath14 channel for a total integrated luminosity of 100 @xmath1 . for this study taus
are defined as tracks approximately collinear with a jet @xmath15 and isolated of other tracks in the cone @xmath16 . using an event selection strategy similar to the ones described in 2.2 or 2.3 , approximately 86 events are found in a simulated event sample corresponding to the first 100 @xmath1 of integrated luminosity .
a signal - to - background ratio @xmath17 is found for 1-prong tau decays , where the most important background comes from the w+jets events .
a least - square kinematic fit using lagrange multipliers can be applied to enforce the mass constraint in the reconstructed @xmath18 events .
both the w boson mass and the top quark mass in the hadronic top quark decay , @xmath19 , are constrained to agree with their measured values .
residual corrections are estimated on the energy scale of the jets arising from both the light quarks in the w boson decay and the heavier bottom quark in the top quark decay . utilizing the first integrated luminosity of 100 @xmath1 of cms data , an uncertainty smaller than 1@xmath9 can be obtained on the jet energy scale for both light and heavy jets .
more details can be found in @xcite
observability of top quark signal for different final states of @xmath0 production has been presented .
it has been shown that the top quark signature can be established with the first 10 @xmath1 in both the semileptonic muon channel , where about 128 signal events are expected with a signal - to - noise ratio of about 1.5 , and in the dilepton channel , where about 80 signal events are expected in the 3 possible final states with a combined signal - to - noise ratio of 7 to 1 .
in addition the semileptonic muon channel final state can be used to obtain the jet energy scale corrections for both light and heavy jets with an uncertainty smaller than 1@xmath9 for an integrated luminosity of 100 @xmath1 . | the top quark was discovered at the tevatron in 1995 . for the last decade
the study of its properties has been a major theme in the worldwide experimental high energy physics program .
the advent of the lhc opens up a new era in top quark physics ; because of the large @xmath0 cross - section and the high luminosity , the lhc can be thought of as a top factory . here
we present the prospects and plans for ttbar physics at cms at an early stage of the experiment , covering from the initial establishment of the top signal , to the first measurements that become possible for an integrated luminosity of 100 @xmath1 , considering a realistic detector performance . |
ultraluminous x - ray sources ( ulxs ) are bright , point - like , off - nuclear extragalactic x - ray sources . since their discovery over two decades ago ( fabbiano et al .
1989 ) , the true nature of these sources has been a mystery .
their isotropic x - ray luminosities range from a few @xmath4 @xmath5 ergs s@xmath6 to as high as @xmath7 ergs s@xmath6 .
luminosities at the high end of this range are found in a handful of sources which have been dubbed hyperluminous x - ray sources ( hlxs , e.g. , m82 x-1 : matsumoto et al .
2001 , kaaret et al . 2001
; hlx eso 243 - 49 x-1 : miniutti et al .
2006 , farrell et al .
the apparent x - ray luminosities of these sources exceed the eddington limit for any known stellar - mass black hole , hence the term ultraluminous " . based on their variability on short timescales (
some ulxs are known to vary on timescales of the order of a few minutes ) and high luminosities , some ulxs are almost certainly powered by accretion onto black holes ( excluding the population of x - ray bright supernovae ; immler & lewin 2003 ) .
the controversy , however , is with their mass range .
black holes , in principle , can exist in three mass ranges .
stellar - mass black holes ( stbh : mass @xmath8 3 - 50 @xmath9 ) are the remnants of stellar evolution of massive normal stars ( stars excluding the pop iii category ) . on the opposite end of the mass scale
are the super - massive black holes ( smbhs : mass @xmath10 @xmath9 ) - which are currently theorized to form either via accretion onto `` seed '' black holes in the early universe ( @xmath8 few myrs after the big bang ) with subsequent mergers ( volonteri et al .
2003 ; davies et al . 2011 ) or via direct collapse of dense gas ( begelman et al . 2006 ) .
their exact formation scenario is still somewhat uncertain . finally , one has the intermediate - mass black holes ( imbhs : mass range 100 - few @xmath4 1000 @xmath9 ) , whose formation mechanisms are the subject of considerable debate .
there are a number of processes whereby an imbh can form .
for example , they can be formed at the centers of globular clusters via accretion onto a massive black hole ( mass @xmath8 50 @xmath9 ) over a time period of @xmath11 years ( miller et al .
2002 ) , by accretion onto a primordial black hole ( e.g. , mack et al .
2007 ) , or by mergers of seed black holes .
the off - centered nature of ulxs rules them out as candidates for smbhs .
given the ages of their host galaxies , if they were smbhs , they should have sunk to the centers of their galaxies due to dynamical friction ( miller & colbert 2004 ) . clearly , this is not the case .
this leaves the possibility of their being either stbhs or imbhs .
theories supporting the idea of a stbh as the central engine often employ either beamed x - ray emission or some sort of super - eddington emission mechanism to explain the observed high luminosities .
current ideas include geometric beaming ( king et al .
2001 ) , relativistic beaming similar to the phenomenon of blazars ( krding et al .
2002 ) , and super - eddington emission as a result of a photon - bubble instability in radiation - dominated accretion disks ( begelman et al .
more recently , based on high - quality @xmath12-@xmath13 spectra of 12 ulxs , gladstone et al .
( 2009 ) have suggested that ulxs might be accreting stbh systems in a new _ ultraluminous accretion state _
( an addition to the current classification of black hole accretion states : thermal ( high / soft ) , hard ( low / hard ) & steep power law ( very high ) as in mcclintock et al . 2006 ) , where accretion is again super - eddington .
alternatively , it has been suggested that these sources might be powered by gravitational energy release from accretion onto an imbh ( colbert & mushotzky 1999 ) .
this presents a relatively simpler solution , where the energetics proceed similarly to that of the stbh systems ( at sub - eddington accretion rates ) with the corresponding physical properties scaling appropriately with the mass of the compact object .
it is also possible that ulxs constitute an inhomogeneous population of both stbhs and imbhs .
it is now known that certain characteristic timescales in accreting black hole systems scale with the mass of the black hole .
for example , mchardy et al . (
2006 ) have established that in soft - state stbhs and agn , the break timescale of the power density spectrum ( pds ) , the mass of the black hole and the luminosity are strongly correlated .
their work was further extended to the hard - state stbhs by krding et al .
these studies strongly demonstrate that certain physical timescales of accreting compact objects scale directly with the mass of the source ( after accounting for differences in accretion rates , i.e. , luminosity ) .
therefore , timing studies can play a key role in extracting valuable information about the mass of the compact source , which is crucial in the case of ulxs in order to better understand the nature of the emission processes that result in such high apparent luminosities .
of particular interest here are the low frequency quasi - periodic oscillations ( lfqpos ) . within the context of stbhs , the lfqpos ( those in the range @xmath8 0.1 - 15 hz ) are broadly classified into three categories based on their properties and the overall nature of their pds ( casella et al . 2005 ) .
the pds with type - a and b qpos is characterized by weak red noise ( noise at the low frequency end of the pds ) with the type - a qpos occurring with relatively low coherence ( quality factor , q : centroid frequency / fwhm , @xmath14 3 ) compared to the type - b ( q @xmath15 6 ) . finally , one has the type - c qpos which are most relevant to the present work .
the pds accompanying these qpos in stbh systems can be described by a flat - topped , band - limited noise breaking to a power law with the qpos evident on the power - law portion of the spectrum , close to the break .
further , the type - c qpos are fairly coherent with the quality factor , q from 5 - 15 and amplitudes ( % rms ) ranging from 2 - 20 . in stbhs
, they are known to occur in the frequency range from @xmath16 0.1 - 15 hz . in light of the work unifying the stellar - mass and the super - massive black holes ,
if some ulxs were to host imbhs then the qualitative behavior of their pds should be comparable to stbhs with the characteristic variability times scaling according to the mass of the putative imbh .
this idea has been explored by a number of authors to search for the `` qpo analogs '' in ulxs ; and qpos have now been detected in a handful of them .
the qpos detected in m82 x-1 ( strohmayer & mushotzky 2003 ) , ngc 5408 x-1 ( strohmayer et al .
2007 ) and ngc 6946 x-1 ( rao et al . 2010 ) resemble the type - c qpos , while those detected in the m82 source x42.3 + 59 more closely resemble the type - a or -b qpos ( feng et al .
the crucial difference here is that the qpo centroid frequencies of the ulx sources appear to be scaled down by a factor of a few@xmath4(10 - 100 ) ( @xmath8 few mhz ) compared to the lfqpos in stbhs . under the assumption that the ulx mhz qpos are analogs of the lfqpos in stbhs
, it is reasonable to assume that their characteristic timescales / frequencies ( e.g. , qpo centroid frequencies , pds break frequencies ) scale with the mass of the accreting source and vice versa .
however , the type - c qpos occur with a wide range of centroid frequencies ( @xmath17 hz ) in stbh systems . therefore , timing information alone is not sufficient to accurately estimate ulx masses in this way , but combining timing and spectral information has proven to be a valuable tool .
for example , in stbh systems , the power - law photon index and disk flux are correlated with the qpo centroid frequency .
the general trend is an increase in power - law photon index and disk flux with the qpo centroid frequency ( e.g. , sobczak et al .
2000 ; vignarca et al .
2003 ; shaposhnikov & titarchuk 2009 ) , with evidence for saturation ( constancy of the power - law photon index and disk flux with a further increase in qpo centroid frequency ) beyond a certain frequency .
using a reference stbh system with a measured qpo centroid frequency - photon - index relation , one can then scale the qpo centroid frequencies detected in ulx systems at a given power - law spectral index to get an estimate of the mass .
for example , with archival @xmath12-@xmath13 data from m82 x-1 , dewangan et al . ( 2006 ) extracted the energy and power spectra of the source .
the pds was strikingly reminiscent of a stbh with type - c qpos , i.e. , flat - top noise breaking to a power - law with qpos on the power - law portion of the spectrum .
however , the respective timescales were scaled down by a factor of @xmath8 10 .
the qpo in this case was centered around a frequency of @xmath8 114 mhz and the energy spectrum had a power law photon index of @xmath8 2.0 .
using the qpo frequency - photon index correlations from two stbh reference sources , grs 1915 + 105 and xte j1550 - 564 , they estimated the mass of the ulx in m82 x-1 by scaling its qpo centroid frequency ( @xmath8 114 mhz ) at the given photon index ( @xmath8 2.0 ) .
they estimated the mass of the black hole to be in the range 25 - 520@xmath9 .
similar scaling arguments were used by rao et al .
( 2010 ) to estimate the mass of the black hole in the ulx ngc 6946 x-1 to be in the range ( 1 - 4 ) @xmath4 1000 @xmath9 . based on both the pds and the energy spectrum of ngc 5408 x-1 , strohmayer et al .
( 2009 ) ( s09 ) argued that the source behavior was consistent with the steep power - law state ( spl ) often seen in stbhs .
they compared the available data from ngc 5408 x-1 to five different stbh reference sources and estimated the mass of the black hole to be a few @xmath4 1000 @xmath18 .
feng et al .
( 2010 ) detected 3 - 4 mhz qpos from the ulx x42.3 + 59 in m82 and identified them as either type a / b analogs of stbhs .
they estimated the mass of the black hole to be in the range 12,000 - 43,000 @xmath9 by scaling the qpo frequency to that of the type a / b qpos in stbhs .
it is important to note that the above scaling arguments have several caveats .
first , the mass estimates were established under the assumption that the mhz qpos detected from these ulx systems are analogous to a particular type of lfqpo detected in stbh systems , i.e. , a , b or c. these identifications were based on the qualitative nature of the power spectrum ( pds ) alone in some cases and both the pds and the energy spectrum in the case of ngc 5408 x-1 .
second , in the case of the ulx mhz qpos , the observed range of qpo centroid frequencies and photon indices has been limited . to gain a more secure identification
one would like to see the qpo frequencies and photon spectral indices _ correlate _ in a similar fashion as for the stbhs .
lcccc + obsid & t@xmath19 & net countrate & effective + & ( ks ) & ( counts s@xmath6 ) & exposure(ks ) + + + 0302900101 ( 2006 ) & 132.25 & @xmath20 & 99.94 + + 0500750101 ( 2008 ) & 115.69 & @xmath21 & 48.55 + + 0653380201 ( 2010a ) & 128.91 & @xmath22 & 80.68 + + 0653380301 ( 2010b ) & 130.88 & @xmath23 & 110.0 + + 0653380401 ( 2011a ) & 121.02 & @xmath24 & 107.75 + + 0653380501 ( 2011b ) & 126.37 & @xmath25 & 98.66
+ + + + + in this paper we further explore these issues using extensive new observations of ngc 5408 x-1 with _ xmm - newton_. we describe the properties of the mhz qpos over a wider range of centroid frequencies than previous data allowed .
more specifically , we study the nature of the mhz qpos from ngc 5408 x-1 through a systematic search for timing - spectral correlations similar to those seen in the stbhs .
the article is arranged as follows . in
2 we describe the data used for the present study . in
3 we show the results from our timing analysis , while in 4 we give details of the energy spectral analysis . in 5 we describe our search for timing - spectral correlations similar to those seen in accreting stbh systems .
the search is conducted using two different spectral models , a phenomenological model of a @xmath26-@xmath27 @xmath28 + @xmath29-@xmath30 and then with a model describing comptonization by bulk motion ( titarchuk et al .
1997 ) that has been used previously to derive black hole masses from qpo scaling arguments ( shaposhnikov & titarchuk 2009 ) .
finally , in 6 we discuss the implications of our results on the mass of the black hole in ngc 5408 x-1 .
_ xmm - newton _ has now observed ngc 5408 x-1 on multiple occasions .
we use data from six of the most recent observations ( @xmath8 100 ks each ) spread over a time span of five years ( 2006 - 2011 ) .
the details of the observations are outlined in table 1 .
results from the first two observations ( in 2006 and 2008 , respectively ) were summarized in strohmayer et al .
( 2007 ) ( s07 ) & strohmayer et al .
( 2009 ) ( s09 ) . using the 2006 data ,
s07 reported the first detection of quasi - periodic variability from this source .
s09 s analysis of the 2008 observation again showed evidence for the presence of quasi - periodic oscillations .
however , the most prominent qpo during the 2008 observation was at a lower frequency ( qpo at @xmath8 10 mhz compared to 20 mhz in 2006 ) .
further , s09 noted that the disk contribution to the total flux and the power - law index of the energy spectrum decreased slightly compared to its state in 2006 .
this is analogous to a trend often seen in stbh systems where the disk flux and the power law index of the energy spectrum positively correlate with the centroid frequency of the most dominant qpo ( e.g. , viagnarca et al .
2003 , sobczak et al .
these findings were used to propose additional observations with _ xmm - newton _
, with the goal of detecting qpos over a range of frequencies and hence to further explore for correlations between timing and energy spectral properties .
a large program was approved for cycle 9 ( pi : strohmayer ) , and four observations ( @xmath8 100 ks each ) were made under this program .
two of the observations were carried out in 2010 ( 2010a & 2010b ) and the rest in 2011 ( 2011a & 2011b ) . here
we present results from these new observations as well as a reanalysis of the earlier pointings , so as to facilitate a consistent comparison of all the available data .
[ fig : figure1a ] for the present work , we use the data acquired by the european photon imaging camera ( epic ) , i.e , both the _ pn _ and _ mos _ to get a higher signal - to - noise ratio in the power and energy spectra .
we used the standard sas version 11.0.0 to reduce the images and filtered event - lists from all the epic data . the standard filter with _ ( pattern@xmath31=4 ) _ , to include only single and double pixel events ,
was applied to the event - lists and events only in the energy range 0.3 - 12.0 kev were considered for further analysis .
background flaring was prominent for brief periods during certain observations .
the power and energy spectra were carefully extracted taking into account both the background flaring and the instrumental dead time effects ( specific details in the next sections ) . in all the observations the source was easily detectable and
we did not face any source confusion problems .
we extracted source events from a region of radius 32 centered around the source .
this particular value was chosen to roughly include 90% of the light from the source ( estimated from the fractional encircled energy of the epic instruments ) .
a background region , free of other sources , was extracted in a nearby region .
the size of the background region was chosen to be consistent with the source region .
further , the size of the source and the background region was chosen to be consistent for all the six observations ( i.e. , 32 ) .
we present the specific details of the timing and spectral analysis in the following sections .
[ fig : figure1b ]
we first produced the source and the background light curves to assess the quality of the data .
shown in the left panels of figure 1a and 1b are the background count rates ( _ pn+mos1+mos2 _ ) for each of the six observations .
overlaid are the time intervals during which a given instrument ( _ pn , mos1/2 _ ) was continuously active for more than 5 ks , i.e. , all the instrumental good time intervals ( gtis ) longer than 5 ks . for a given instrument
, a horizontal line ( offset to an arbitrary value for each instrument ) indicates the active time , while a vertical line marks the beginning or the end of a continuous data interval .
such an insight is important as we are using both the _ pn _ and the _ mos _ data to achieve higher count rates compared to , say , the _ pn _ alone .
in other words , this ensures that the combined _ pn _ and _ mos _ event - lists contain only events corresponding to the times during which all the three instruments were active .
furthermore , inspection of the light curves reveals background flaring at either the beginning or the end , or both , of all the observations .
therefore , the final combined _ ( pn+mos1+mos2 ) _ light curves were extracted by taking into account the gtis of all the instruments and excluding the periods of background flaring .
the filtered light curves ( 100s bins ) of the source and the non - flaring background in the energy range 0.3 - 12.0 kev are shown in the right panels of figures 1a and 1b .
the combined mean count rates ( 0.3 - 12.0 kev ) during each observation are listed in table 1 .
notice that the gaps in the data are either due to exclusion of background flaring times , dead time intervals or the absence of gtis longer than 5 ks .
lccccccc + dataset & 2006 & 2008 & 2010a & 2010b & 2011a &
2011b + + + exposure(ks ) & 27.6@xmath43 & 15.5@xmath42 & 23.2@xmath43 & 21.0@xmath44 & 67.4@xmath41 & 55.0@xmath41 + + energy range(kev ) & 0.30 - 2.0 & 1.0 - 10.0 & 0.85 - 10.0 & 1.0 - 4.0 & 1.10 - 8.0 & 1.20 - 2.0 + + countrate(cts s@xmath6 ) & @xmath32 & @xmath33 & @xmath34 & @xmath35 & @xmath36 & @xmath37 + + c & @xmath38 & @xmath39 & @xmath40 & @xmath41 & @xmath40 & @xmath42 + + a & @xmath43 & @xmath44 & @xmath45 & @xmath46 & @xmath47 & @xmath48 + + @xmath49 & @xmath50 & @xmath51 & @xmath52 & @xmath53 & @xmath54 & @xmath55 + + @xmath56(mhz ) & @xmath57 & @xmath58 & @xmath59 & @xmath60 & @xmath61 & @xmath62 + + @xmath63 & @xmath64 & @xmath65 & @xmath66 & @xmath67 & @xmath68 & @xmath69 + + n@xmath70 & @xmath71 & @xmath72 & @xmath73 & @xmath74 & @xmath75 & @xmath76 + + @xmath77(mhz ) & @xmath78 & @xmath79 & @xmath80 & @xmath81 & @xmath82 & @xmath83 + + @xmath84(mhz ) & @xmath85 & @xmath86 & @xmath87 & @xmath88 & @xmath89 & @xmath90 + + n@xmath91 & - & @xmath92 & - & - & - & - + + @xmath93(mhz ) & - & @xmath94 & - & - & - & - + +
@xmath84(mhz ) & - & @xmath95 & - & - & - & - + + + @xmath96/dof & 309.83/208 & 753.75/764 & 178.14/137 & 128.05/97 & 327.11/329 & 268.86/267 + ( continuum ) & ( 618.71/211 ) & ( 868.30/770 ) & ( 257.33/140 ) & ( 207.54/100 ) & ( 669.58/332 ) & ( 543.03/270 ) + + + + + + + it is evident from the light curves ( right panels of figure 1a and 1b ) that the source varies significantly during all the observations . to quantify the variability ,
we construct a power density spectrum from each of the light curves . to achieve higher signal - to - noise in the pds , we break the gtis into shorter segments ( s / n @xmath97 @xmath98number of individual spectra , e.g. , van der klis 1989 ) and derive an average power spectrum . for a given observation , the size and the number of such segments
are shown in the first row of table 2 .
figures 2a and 2b show the pds and the best - fitting model _ ( thick solid line ) _ for each of the six observations in two different energy bands . shown in the left panel
is a pds in an energy band in which a qpo is detected with high statistical significance and in the right panel is a pds derived from photons in the energy band of 1.0 - 10.0 kev ( the reason for constructing power spectra in two different energy bands is discussed in the next paragraph ) .
all the power spectra shown here use the so - called leahy normalization , with the poisson noise level being 2 ( leahy et al .
. clearly , in each spectrum the power rises below @xmath8 0.1 hz with evidence for a qpo in the range of 10 - 40 mhz and essentially poisson noise at higher frequencies .
[ fig : figure1a ] to quantify this behavior , we fit a bending power law to the continuum and a lorentzian to model the qpo ( belloni et al .
the model fits well with the reduced @xmath99 @xmath16 1 ( last row of tables 2 & 3 ) in three cases ( 2008 , 2011a , 2011b ) and gives acceptable fits with the reduced @xmath99 in the range 1.3 - 1.5 ( last row of tables 2 & 3 ) in the other three cases ( 2006 , 2010a , 2010b ) .
a careful analysis of the residuals in the latter three cases indicates that multiple weak features contribute significantly to increase the overall @xmath99 .
for example in the case of the pds derived from the 2010a data ( bottom two panels of figure 2a ) , the weak qpo - like feature at @xmath8 55 mhz and the excess at @xmath8 0.2 hz contribute about 30 to the total @xmath99 .
however , their individual statistical significance is rather low . nevertheless , for the purposes of analyzing the qpo properties and studying the overall qualitative nature of the power spectra , the fits are adequate .
the best - fitting model parameters ( derived from a fit in the frequency range 1.0 mhz - 0.5 hz ) for each observation are shown in tables 2 & 3 for the two different energy bands ( highlighted in the tables ) .
also shown are the @xmath99/degrees of freedom ( dof ) values for each of the fits along with the @xmath99/dof corresponding to the continuum model ( in braces ) .
the change in @xmath99 serves as an indicator of the statistical significance of the qpos . for a given observation
, we choose two different energy bands : the first energy band ( second row of table 2 ) is the bandpass in which the qpo is detected with a very high confidence and the second bandpass of 1.0 - 10.0 kev was chosen to consistently compare the properties of the qpo across all the observations ( overall variability is energy dependent ) .
we confirm the overall qualitative nature of the pds , that is , flat - topped noise breaking to a power law with a qpo on the power law portion ( near the break ) .
this is consistent with stbhs showing type - c qpos but with the characteristic frequencies scaled down by a factor of a few@xmath410 .
this is in agreement with the results reported by s07 and s09 , except with regard to possible close qpo pairs , as here all our pdss ( figure 2a & 2b ) are averaged to the extent of smearing out weak features ( to clearly identify the strongest qpo ) .
the effect of averaging is evident from the best - fit parameters of the qpos ( tables 2 & 3 ) .
more specifically , since we have smeared out the possible close pair qpos ( as reported by s07 & s09 using the 2006 and the 2008 datasets , respectively ) the average quality factors of the qpos reported here are relatively lower than those typical of the type - c qpos from stbhs .
moreover , variation of the qpo centroid frequency over the timescale of the observation can further decrease its coherence .
furthermore , we analyzed the pdss of the backgrounds from each of the six observations and note that they are all consistent with a constant poisson noise .
[ fig : figure1a ] one of the main goals of the present work is to better characterize the mhz qpos seen from ngc 5408 x-1 within the context of the known classes of lfqpo seen in stbhs ( casella et al . 2005 ) .
an important diagnostic for this purpose is the study of the variation of the rms amplitudes of these qpos with their centroid frequency . in the case of lfqpos from stbh systems
this behavior is fairly well - established ( e.g. , revnivtsev et al . 2000 ; sobczak et al .
2000 ; vignarca et al .
2003 ; mcclintock et al .
the typical behavior of the stbh lfqpos , in the energy range 2.0 - 20.0 kev ( the nominal rxte bandpass ) , can be described as follows : as the source evolves along the low / hard state towards the spl state , the qpo frequency increases and the rms amplitude increases . before reaching the spl state
, the source traverses an `` intermediate '' state in which the qpo frequency continues to increase , but the rms amplitude decreases . upon reaching the spl state
the correlation tends to break down , showing more scatter in the rms amplitude ( see figure 11 in mcclintock et al .
2009 ) .
lccccccc + dataset & 2006 & 2008 & 2010a & 2010b & 2011a & 2011b + + + exposure ( ks ) & 27.6@xmath43 & 15.5@xmath42 & 23.2@xmath43 & 21.0@xmath44 & 67.4@xmath41 & 55.0@xmath41 + + energy range ( kev ) & 1.0 - 10.0 & 1.0 - 10.0 & 1.0 - 10.0 & 1.0 - 10.0 & 1.0 - 10.0 & 1.0 - 10.0 + + countrate(cts s@xmath6 ) & @xmath100 & @xmath33 & @xmath101 & @xmath102 & @xmath103 & @xmath103 + + c & @xmath41 & @xmath39 & @xmath104 & @xmath105 & @xmath106 & @xmath107 + + a & @xmath108 & @xmath44 & @xmath109 & @xmath110 & @xmath111 & @xmath112 + + @xmath49 & @xmath113 & @xmath51 & @xmath114 & @xmath115 & @xmath116 & @xmath117 + + @xmath56(mhz ) & @xmath118 & @xmath58 & @xmath119 & @xmath120 & @xmath121 & @xmath122 + + @xmath63 & @xmath123 & @xmath65 & @xmath124 & @xmath125 & @xmath126 & @xmath127 + + n@xmath70 & @xmath128 & @xmath72 & @xmath129 & @xmath130 & @xmath131 & @xmath132 + + @xmath77(mhz ) & @xmath133 & @xmath79 & @xmath134 & @xmath135 & @xmath136 & @xmath137 + + @xmath84(mhz ) & @xmath138 & @xmath86 & @xmath139 & @xmath140 & @xmath141 & @xmath142 + + rms@xmath70 & @xmath143 & @xmath144 & @xmath145 & @xmath146 & @xmath147 & @xmath148 + + n@xmath91 & - & @xmath92 & - & - & @xmath149 & - + + @xmath93(mhz ) & - & @xmath94 & - & - & @xmath150 & - + + @xmath84(mhz ) & - & @xmath95 & - & - & - & - + + rms@xmath91 & - & @xmath151 & - & - & @xmath152 & - + + + @xmath96/dof & 268.34/208 & 753.75/764 & 200.60/137 & 135.27/97 & 350.16/326 & 291.45/266 + ( continuum ) & ( 755.28/211 ) & ( 868.30/770 ) & ( 242.10/140 ) & ( 190.64/100 ) & ( 851.45/332 ) & ( 677.00/269 ) + + + + + + + the x - ray variability of ngc 5408 x-1 is known to depend on energy ( middleton et al .
the overall rms variability increases with an increase in the energy of the photons at least up to 2.5 kev ( given the poor signal - to - noise at higher energies , it is not clear whether the variability strength levels off or decreases ) .
this is similar to the energy dependence of lfqpo ( @xmath16 few hz ) detected from the galactic micro - quasar grs 1915 + 105 ( rodriguez et al .
2004 ) , where the qpo amplitude increases with energy , before rolling over .
we used the qpo fit parameters from all the observations to explore the dependence of the qpo rms amplitude with centroid frequency . to be consistent across all the observations , we chose an optimum bandpass of 1.0 - 10.0 kev to ensure good statistics and significant detection of the variability ( the variability of the source increases at higher energies while the countrate decreases ) .
in addition , to explore whether our results might depend on energy , we derived the qpo properties in a different energy band ( though with some overlap of the optimum band defined above ) of 0.5 - 5.0 kev .
the rms amplitude versus qpo frequency results for the 1.0 - 10.0 kev and 0.5 - 5.0 kev are shown in the left and the right panels of figure 3 , respectively ( compare with figure 11 of mcclintock et al .
given that the qpo amplitude is higher at higher energies ( middleton et al .
2011 ) , not all power spectra derived from the lower energy band ( 0.5 - 5.0 kev ) yielded statistically significant qpo features . this is why the right panel of figure 3 has one less data point .
clearly , the variation is similar in both the energy bands .
we further compare with the results from stbhs in 5 .
another prime driver for the long observations of ngc 5408 x-1 was to search for high frequency qpo ( hfqpo ) analogs of stbhs .
these qpos are observed in the range from about 50 - 450 hz in stbhs ( mcclintock & remillard 2006 ; strohmayer 2001a ) .
simple mass scaling arguments would suggest that in imbhs they would be expected in the frequency range of 0.1 - 1 hz , with a low rms amplitude of @xmath8 2% .
we do not detect any obvious qpo - like feature at these frequencies in any of the individual observations .
further , given that hfqpos appear to be reasonably stable in frequency in stbhs ( e.g. , strohmayer 2001b ) , we averaged the power spectra derived from the individual observations to improve the signal - to - noise , however , this did not lead to a detection .
we estimate the upper limit on the rms amplitude of a qpo - like feature in the frequency range of 0.1 - 1.0 hz to be @xmath8 4% .
we fit the x - ray spectra of ngc 5408 x-1 using the xspec ( arnaud 1996 ) spectral fitting package and the epic response files were generated using the @xmath153 and @xmath154 tools which are part of the _ xmm - newton _ science analysis system ( sas ) software . since , a primary goal of the present work is to search for timing - spectral correlations similar to those seen in accreting stbhs , we began our analysis by characterizing the energy spectra using the same phenomenological models often used to describe the x - ray spectra of accreting stbhs . in terms of xspec models
, we used _
diskpn + power - law_. we also explored the _ bmc _ model , comptonization by bulk motion ( titarchuk et al . 1997 ) , because it has been used to derive mass estimates from qpo frequency scaling arguments , and reference spectral - timing correlations have been derived for a significant sample of stbhs using the parameters derived from this model ( see shaposhnikov & titarchuk 2009 ) .
we start by describing some of the specifics of our data extraction and reduction methods .
lccccccc + & & & _
tbabs@xmath155(diskpn+apec+cutoffpl ) _ & & & + + + dataset & n@xmath156 & t@xmath157 & kt@xmath158 & @xmath159 & e@xmath160 & @xmath96/dof + + + 2006 & 12.03@xmath161 & 0.153@xmath162 & 1.03@xmath163 & 1.88@xmath164 & 3.80@xmath165 & 375.48/264 + + 2008 & 12.93@xmath166 & 0.146@xmath167 & 0.97@xmath168 & 1.92@xmath169 & 4.29@xmath170 & 214.02/209 + + 2010a & 11.96@xmath171 & 0.160@xmath172 & 1.04@xmath173 & 1.84@xmath174 & 4.18@xmath175 & 357.17/276 + + 2010b & 10.53@xmath176 & 0.171@xmath177 & 1.02@xmath168 & 1.92@xmath178 & 4.92@xmath179 & 385.27/287 + + 2011a & 12.30@xmath180 & 0.160@xmath172 & 0.98@xmath181 & 1.92@xmath182 & 4.28@xmath183 & 389.94/266 + + 2011b & 11.54@xmath184 & 0.167@xmath185 & 1.05@xmath186 & 1.98@xmath187 & 5.56@xmath188 & 307.22/255 + + + since we are interested in exploring the correlations between the timing and the spectral behavior , we elected to extract energy spectra from the same time intervals as those used to extract the pds ( see the first row of tables 1 & 2 ) .
such synchronous measurements will tend to minimize any offsets that could be induced by variations in the system properties ( both spectral and timing ) within a given observation .
in addition to the standard filters described in 2 , the _ ( flag==0 ) _ filter was imposed to get the highest quality spectra .
the combined ( _ pn+mos _ ) average count rate of the source varied in the range 1.19 - 1.46 cts s@xmath6 over all the observations . given such relatively low count rates , source pileup was not an issue in any of the six observations . since we are using both the _ pn _ and the _ mos _ data , we individually reduce the energy spectra from each of the detectors using the same good time intervals .
this gave us three energy spectra ( _ pn , mos1 , mos2 _ ) for each of the six observations .
a given model was fit simultaneously to all the three spectra to derive the tightest constraints on the best - fit parameters .
the _ pn _ and the _ mos _ spectra were binned to 1/3 of the fwhm of the pn and the mos spectral resolution , respectively .
we used the sas task _ specgroup _ for this purpose .
given the high number of total counts ( @xmath16 ( 6 - 12)@xmath410,000 ) , this gave us high - quality spectra in each case .
we began by fitting the spectra with a multi - colored disk ( _ diskpn _ in xspec , gierlinski et al .
1999 ) + power law ( _ powerlaw _ in xspec ) + an x - ray hot plasma ( _ apec _ in xspec ) as in s09 .
we used the tuebingen - boulder ism absorption model ( _ tbabs _ in xspec ) to account for interstellar absorption .
the hydrogen column density was set as a free parameter .
this model is very similar to that often used to model the spectra of stbhs .
indeed , it is common to use this or similar phenomenological models to determine the spectral state of a stbh .
however , here in addition to a thermal disk and a corona , there is an x - ray emitting plasma in the model .
this was realized by s07 after identifying systematic emission - like features in the residuals of the _ diskpn + power - law _ fit to the 2006 data .
we confirm their result that the @xmath96 improves significantly ( by @xmath8 40 - 100 over all the six observations ) with addition of the _ apec _ component to the standard _ multi - colored disk + power - law _ model .
such a feature is not exclusive to the ulx in ngc 5408 , as ngc 7424 ( soria et al .
2006c ) and holmberg ii x-1 ( dewangan et al .
2004 ) also show evidence for the presence of an x - ray emitting plasma .
in addition , the high - resolution vla radio observations at 4.9 ghz of the counterpart of the x - ray source in ngc 5408 x-1 revealed a radio nebula of @xmath16 40 pc extent ( lang et al .
based on the value of the power - law index of the radio spectrum , it was suggested that the radio emission is likely from an optically thin synchrotron emitting gas .
given this , it is possible that the putative x - ray hot plasma is coincident with the radio nebula .
the best - fitting parameters of the above model for each of the six observations are shown in the top panel of table 4 .
we note that the x - ray temperature of the plasma has remained constant across the suite of our _ xmm - newton _ observations .
[ fig : figure1a ] we also fit the spectra with a model based on componization by bulk motion ( laurent & titarchuk 1999 , _ bmc _ in xspec ) .
as mentioned above , this model has been used within the context of quantifying timing - spectral correlations in stbhs ( shaposhnikov & titarchuk 2009 ) .
the underlying physical basis of this model is similar to that of the multi - colored disk + power - law model , i.e. , the presence of a thermal disk and a hot electron corona .
however , in this case the power - law is produced by the comptonization of soft photons ( from the disk ) within a converging inflow onto the black hole . in this model ,
the normalization parameter derived from the fit mimics the disk flux from the _ diskpn _ component in the other model we have employed .
we also include the _ apec _ component in this case .
the best - fitting model parameters including the absorbing hydrogen column density using the _ tbabs _ in xspec for each of the six datasets are shown in the bottom panel of table 4 .
it can be noted straight away that , for a given observation , the plasma temperatures ( _ apec _ ) found from fitting the two spectral models are consistent with each other ( within the error bars ) .
the reduced @xmath96 values for the _ diskpn+powlaw+apec _ and the _ bmc+apec _ model are in the range of 1.0 - 1.6 and 1.0 - 1.3 ( @xmath8 210 - 290 degrees of freedom : last column of table 4 ) , respectively .
clearly , they both give acceptable values of reduced @xmath96 .
the latter model ( _ bmc+apec _ ) fits the data slightly better ( see table 4 ) than the simple _ diskpn+apec+powlaw _ model .
however , the improvement in @xmath96 is not statistically significant in all the cases and therefore it is not possible to rule out one model over the other based on the reduced @xmath96 alone . a sample energy spectrum ( using the 2010b data ) is shown in figure 4 . also shown in the figure are the ratios of the data to the folded model in the two cases ( _ diskpn+apec+powlaw _ and _ bmc+apec _ ) .
we use the spectral parameters derived here and the timing properties from the previous section to search for correlations between the two .
the results of which are presented in the next section .
it has been recently suggested ( gladstone et al .
2009 ) that the high - quality x - ray spectra of ulxs can be better characterized by complex comptonization models that predict a rollover at higher energies ( @xmath15 3 kev ) . a detailed study of the validity of such models in the case of ngc 5408 x-1 is beyond the scope of the present work
, however , for the sake of completeness we also fit all our energy spectra with a cut - off power law component rather than just the simple power law model .
in xspec we used the model , _
tbabs*(diskpn+apec+cutoffpl)_. ( the _ cutoffpl _ model describes a power law with a high energy exponential rollover . )
the best fitting parameters for this model are shown in table 5 .
we note that this model leads to a significant improvement in the @xmath96 compared to the _
tbabs*(diskpn+apec+pow ) _ model ( compare the last column of table 5 with the last column of table 4 ) .
we also searched for possible timing - spectral correlations within the context of the _ cutoff powerlaw _
model , viz . , qpo frequency vs the cutoff energy .
we do not detect any clear correlation between the qpo frequency and the rollover energy .
this is mainly due to poor statistics at the high end ( @xmath15 5 kev ) of the energy spectra which result in large uncertainties on the rollover energies .
the measurement of fluorescent lines from elements like fe , ni , cr , ca etc .
is another important probe of the vicinity of the x - ray emitting region . in principle , these reflection features can be produced by hard x - rays , presumably from the comptonizing corona , irradiating cold material , e.g. , in the disk . among all such emission lines with their rest - frame energies in the _ xmm - newton _ bandpass ( 0.2 - 10.0 kev ) , the iron feature is the most prominent ( e.g. , matt et al .
1997 ) . given the subtle changes in the energy spectra over the suite of observations ( see table 4 ) , it does not seem unreasonable to combine all the data while searching for weak emission features .
we searched for iron emission line features in the energy range of 6.0 - 7.0 kev , but did not detect any obvious features . using an effective exposure of @xmath16 540 ks ( all datasets ,
see table 1 ) , we were able to place tight constraints on the equivalent width of any feature in the energy range of 6.0 - 7.0 kev . assuming an unresolved narrow feature of width 10 ev , the upper limit ( 90% confidence ) on the equivalent width of an emission line at 6.4 kev is 5.4 ev
we then check for the possibility of broad emission lines in the energy range 6.0 - 7.0 kev .
the upper limits on the equivalent width assuming a broad emission line of width 0.3 kev and 1.0 kev are 11.6 ev and 11.0 ev , respectively
. the apparent weakness of iron emission from the disk can be due to a number of factors : an accretion disk that is completely ionized , scattering of the reflected iron - line photons in an optically thick corona ( matt et al .
1997 ) , especially when the corona is a thin layer above the accretion disk or a low iron abundance in the accretion disk ( matt et al .
further , a high inclination angle for the accretion disk combined with one or more of the above factors can decrease the intensity of the iron - line .
recently , gladstone et al .
( 2009 ) have analyzed energy spectra of a sample of ulxs ( including ngc 5408 x-1 ) and have suggested that these sources might be operating in a new accretion state , with rollover at higher energies ( @xmath189 2.0 kev ) as a characteristic signature of such a state .
thus , the nature of the spectrum might be such that there are not enough photons at higher energies to generate a detectable iron - line fluorescent feature ( cole miller , private communication ) .
+ + + + + + +
a primary goal of the new observations of ngc 5408 x-1 is to further test the preliminary classification of the mhz qpos detected from ngc 5408 x-1 as the analogs of the 0.1 - 15 hz low - frequency , type - c qpos ( lfqpos ) detected from accreting stbhs .
if a strong connection can be demonstrated , then scaling of the characteristic frequencies ( e.g. , qpo centroid frequency ) can be better justified to estimate the mass of the black hole within this source .
as mentioned earlier , scaling relations using the pds break frequency have been successfully tested using power spectra of both stellar and super - massive black holes of known mass ( mchardy et al .
2006 and krding et al .
based on the qualitative nature of the pds and the energy spectra derived using the 2006 and 2008 data , s09 suggested that the mhz qpos from ngc 5408 x-1 may indeed be lfqpo analogs .
more specifically , they proposed that its x - ray state was analogous to that of a stbh in the so - called steep power - law ( spl ) state exhibiting type - c qpos , but at the same time emitting a few@xmath410 higher x - ray flux .
having analyzed the data from all six observations ( 3 & 4 ) we confirm that the derived power- and energy - spectral parameters are qualitatively consistent with the source being in a spl state exhibiting type - c qpos ( based on the state descriptions in mcclintock & remillard 2006 ) .
a summary of the working definition of the spl state with type - c qpos and the pds & energy spectral parameters of ngc 5408 x-1 are given in table 6 .
a further step in understanding the nature of these qpos is to determine if they show the same evolutionary behavior in the qpo parameters as those exhibited by the stbh lfqpos .
one such characteristic behavior is the qpo rms amplitude versus centroid frequency relationship .
our results for this relation from ngc 5408 x-1 are shown in the left and right panels of figure 3 in two different energy bands ( 1.0 - 10.0 kev and 0.5 - 5.0 kev , respectively ) . comparing these plots with those from stbhs , as in figure 11 of mcclintock et al .
( 2009 ) , the results appear to be at least qualitatively consistent with the behavior of spl data from some stbhs ( see for example , the green triangles in figure 11 of mcclintock et al .
however , because a tight correlation appears to break down in the spl state it is hard to conclude definitively that the ngc 5408 x-1 behavior is exactly analogous .
[ fig : figure 5a ] another important signature of the type - c lfqpos in stbhs is their dependence on the energy spectral features , i.e. , the timing - spectral correlations . using the x - ray state classifications as in mcclintock et al .
( 2006 ) , lfqpos in stbhs are usually detected in the low / hard state , spl state and during the transition between these two states . within the context of the multi - colored disk + power - law model parameters
, it is known that the centroid frequency of these lfqpos is correlated with the disk flux and the photon index of the power - law component . in stbhs ,
the typical behavior is that the qpo centroid frequency is positively correlated with the disk flux and the photon index of the power law . at a certain higher qpo centroid frequency ( @xmath16 5 - 10 hz ) the relationship seems to flatten ( saturate ) and even appears to start reversing in some cases ( vignarca et al .
furthermore , the correlation is not exactly the same for all stbhs , i.e. , the value of the slope of the correlation and the photon index / disk flux at saturation are different for different sources and can be different for the same source in a different outburst ( vignarca et al . 2003 ; shaposhnikov & titarchuk 2009 ) . in figure 5
we show plots of several derived spectral parameters and fluxes ( using the _ diskpn _ + _ power - law _ fits ) as a function of the qpo centroid frequency .
these can be compared with similar plots derived from stbh systems ( see , for example , figures 8 & 9 in mcclintock et al . 2009 ; figure 10 of vignarca et al .
again , there appears to be qualitative consistency between the behavior exhibited by ngc 5408 x-1 and stbhs , but the full range of behavior is not yet seen , again because the range of spectral variations in ngc 5408 x-1 ( most notably in the power - law index ) is too modest .
additional measurements of qpo properties at lower values of the power - law photon index , for example , could provide a more definitive test .
[ fig : figure1a ] one can also investigate the spectral - temporal correlations in the context of the spectral parameters derived using the _ bmc _ model .
indeed , shaposhnikov & titarchuk ( 2009 ) explored such correlations in detail for a sample of stbhs using rxte data .
they also developed a set of fitting functions to quantify the observed correlations and used these to obtain mass estimates by scaling arguments .
figure 5b compares the results from ngc 5408 x-1 with those from stbhs using the _ bmc _ spectral fits .
we used the published data from shaposhnikov & titarchuk ( 2009 ) to produce representative plots for a stbh ( xte j1550 - 564 , top panels ) .
shown from left to right are the qpo centroid frequency versus the power - law photon index ( @xmath159 ) , the _ bmc _ normalization ( disk flux ) versus @xmath159 and the comptonized fraction , @xmath190 versus @xmath159 , respectively .
the data points are color coded to highlight the low / hard state ( blue ) and the spl state ( green ) .
we used the data from the 1998 outburst of xte j1550 - 564 which clearly demonstrates the typical behavior of stbhs ( top panels ) .
comparisons of these plots with those in , for example , shaposhnikov & titarchuk ( 2009 ) lead to similar conclusions as with the results in figure 5a .
there is qualitative consistency with the type - c identification , but the classification is not definitive . in order to make a more secure association of the qpos in ngc 5408 x-1 with type - c ( spl state ) qpos in the stbhs , we need , at a minimum , to obtain observations of the source over a greater range of power - law spectral indices .
it has been established that black hole masses scale with the break frequency of their pds ( mchardy et al .
2006 , krding et al . 2007 ) and this relation is known to hold over 6 orders of magnitude in mass , i.e. , from stbh to super - massive black holes . furthermore , at least in stbhs , the break frequency of the pds is known to strongly correlate with the centroid frequency of the lfqpos ( wijnands et al .
therefore , it is reasonable to assume that the mass of the black hole scales with the centroid frequency of the lfqpos . this can be especially useful in cases where black hole mass measurements are otherwise difficult , viz . , in ulxs . however , an important caveat is to ensure that the qpos used for comparison are similar in nature .
our new results confirm the earlier result from s09 that the qualitative nature of the pds and the energy spectrum of ngc 5408 x-1 are very similar to that of stbhs in the spl state ( see table 6 ) . at the same time , the characteristic timescales within the pds are lower by a factor of @xmath0 , while the x - ray luminosity is higher by a factor of a few@xmath410 , when compared to a typical stbh in such a state .
however , based on the scaling between the break frequency of the pds and the centroid frequency of the lfqpos ( centroid frequency of lfqpos @xmath16 12@xmath4break frequency of the pds ) , the association of the mhz qpos from ngc 5408 x-1 with the lfqpos from stbhs has been questioned ( middleton et al .
2011 ) a striking feature of the behavior of ngc 5408
x-1 is that the energy spectrum of the source has remained roughly constant over the current suite of observations , while the properties of the most prominent qpo ( e.g. , rms amplitude , centroid frequency ) have changed significantly .
these results appear consistent with the source being in a spectral state similar to the spl state in stbhs .
more specifically , in stbhs , certain spectral properties like the disk flux , the photon index of the power - law correlate positively with the qpo centroid frequency up to a certain frequency ( @xmath16 2 - 10 hz ) beyond which the relationship turns around or remains roughly constant
. detailed analysis of the timing - spectral correlations in stbh sources such as xte j1550 - 564 and h1743 - 322 has revealed a comprehensive picture ( mcclintock et al .
2009 ) of their behavior .
tracking these two sources as they evolved from the low / hard state into the spl state , it was realized that the spectral properties like the disk flux and the power - law photon index are tightly correlated with the qpo centroid frequency during the low / hard state and during a transition phase intermediate between the low / hard and the spl state ; and the relationships seem to saturate / breakdown ( spectral properties remain roughly constant with increase in the qpo frequency ) as the sources entered the spl state
. it may be that we are in fact seeing ngc 5408 x-1 in a similar state , that is , we may be seeing the `` saturated '' portion of the relation .
given that the overall accretion timescales change with the mass of the black hole ( timescales increase as the mass increases from stbhs to super - massive black holes , e.g. , mchardy et al .
2006 ) , it is possible that we are seeing a similar timing - spectral behavior , and that the apparent longevity of the source in the spl state might also be due to its having a higher mass
. there may be a useful analogy to agn in that their spectra are known to remain roughly constant on timescales of at least a few years ( markowitz et al . 2003 ) while transient stbh systems undergo strong energy spectral changes on timescales of the order of a few days ( mcclintock et al .
2006 and references therein )
. it could be that the persistence of ngc 5408 x-1 in the approximately same spectral state is related to it being of intermediate mass between stbhs and agn .
alternatively , perhaps the source is in an accretion state , like the so - called ultraluminous state of gladstone et al .
( 2009 ) , that is just not well sampled by the behavior of galactic stbhs .
we note that the indication for a rollover in the power - law at 4 - 5 kev ( see 4 and table 5 ) provides some support for this interpretation , in which case the spectral state of ngc 5408 x-1 may not be exactly analogous to the spl state of stbhs
. however , either conclusion will remain tentative until ngc 5408 x-1 is observed in a state with a lower value of the power - law photon index , i.e. , a state similar to the low / hard state in stbhs .
such an observation would clearly be important in order to make a more definitive identification of its qpos in the context of those observed in stbhs .
assuming the constancy of the energy spectral features , i.e. , disk flux , @xmath159 , with the increasing qpo centroid frequency ( see figure 5 ) is analogous to the saturation observed in the timing - spectral correlations in stbhs , one can estimate the minimum mass of the black hole in ngc 5408 x-1 .
if we are indeed seeing only the saturated portion of the complete timing - spectral correlation curve , then the minimum qpo frequency observed in ngc 5408 x-1 serves as an upper limit on the so - called transition frequency , i.e. , the frequency beyond which the timing - spectral correlations tend to saturate . in order to derive a mass estimate , this particular frequency can be scaled to a reference stbh of known mass with a measured transition frequency ( @xmath191 ) .
we can use the transition frequency as defined by shaposhnikov & titarchuk ( 2009 ) in their fits of qpo and spectral parameters .
we find a sample of three stbhs with mass and transition frequency measurements in shaposhnikov & titarchuk ( 2009 ) that are suitable for this purpose .
further , it is known that the same source can exhibit different tracks ( timing - spectral correlation curves ) , i.e. , different transition frequencies , in various outburst episodes . given that we are interested in a lower limit on the mass of the black hole in ngc 5408 x-1 , we consider the lowest transition frequency for a given source .
the minimum mass of the black hole in ngc 5408 x-1 , can then be calculated as @xmath191@xmath4mass@xmath192/(minimum qpo frequency ) .
the mass estimates using the three reference sources , gro 1655 - 40 , xte j1550 - 564 and gx 339 - 4 are reported in table 7 ( @xmath8 1000 m@xmath193 ) , and show substantial overlap with those reported by s09 .
lcccc + & & & minimum & + reference & mass & @xmath191 & mass@xmath194 & references + & ( m@xmath193 ) & ( hz ) & ( m@xmath193 ) & + + + gro j1655 - 40 & @xmath195 & @xmath196 & 1720 & 1 , 4 + + xte j1550 - 564 & @xmath197 & @xmath198 & 1490 & 2 , 4 + + gx 339 - 4 & @xmath189 6 & @xmath199 & 820 & 3 , 4 + + + + - ( 1 ) greene et al . 2001 ; ( 2 ) orosz et al . 2002 ( 3 ) muoz - darias et al .
2008 ( 4 ) shaposhnikov et al . 2009 .
the ulx in the irregular dwarf galaxy , ngc 5408 has been suggested to harbor an intermediate - mass black hole ( mass@xmath200 @xmath16 100 - a few@xmath41000 m@xmath193 ) ( s09 ) .
this mass estimate was strictly based on the assumption that the mhz qpos seen from this source are the type - c analogs of the low - frequency qpos ( 0.1 - 15 hz ) from stbhs .
here we have presented results from new observations of ngc 5408 x-1 in which we searched for timing - spectral correlations similar to those often exhibited by the type - c lfqpos from stbhs .
our analysis of multi - epoch _ xmm - newton _ data from ngc 5408 x-1 reveals that certain characteristic features of the power spectra , especially the qpo centroid frequency , changed significantly .
however , the energy spectrum has remained roughly constant .
these results can be interpreted in two ways .
this could be due to complete independence of the timing properties on the energy spectra , unlike stbhs , in which case mass scalings derived from the qpos are likely to be problematic ; or there is in fact a correlation , but we are seeing only the saturated part of the correlation behavior ( constancy of the @xmath159/disk flux with increasing qpo frequency ) .
such saturation is often seen in stbhs ( e.g. , vignarca et al .
assuming we are seeing this saturated portion of the correlation curve , we estimate the lower limit on the mass to be @xmath16 800 m@xmath193 .
at least one observation with the source in a low / hard - like state ( @xmath159 @xmath31 2.0 ) is necessary to resolve the issue of whether the timing and spectral properties are correlated as in stbhs or not . | we present results of new _ xmm - newton _ observations of the ultraluminous x - ray source ( ulx ) ngc 5408 x-1 , one of the few ulxs to show quasi - periodic oscillations ( qpo ) .
we detect qpos in each of four new ( @xmath0 ks ) pointings , expanding the range of frequencies observed from 10 - 40 mhz .
we compare our results with the timing and spectral correlations seen in stellar - mass black hole systems , and find that the qualitative nature of the timing and spectral behavior of ngc 5408 x-1 is similar to systems in the steep power - law state exhibiting type - c qpos .
however , in order for this analogy to quantitatively hold we must only be seeing the so - called _ saturated _ portion of the qpo frequency - photon index ( or disk flux ) relation . assuming this to be the case , we place a lower limit on the mass of ngc 5408 x-1 of @xmath1 .
alternatively , the qpo frequency is largely independent of the spectral parameters , in which case a close analogy with the type - c qpos in stellar systems is problematic .
measurement of the source s timing properties over a wider range of energy spectral index is needed to definitively resolve this ambiguity .
we searched all the available data for both a broad fe emission line as well as high frequency qpo analogs ( 0.1 - 1 hz ) , but detected neither .
we place upper limits on the equivalent width of any fe emission feature in the 6 - 7 kev band , and of the amplitude ( rms ) of a high frequency qpo analog of @xmath2 ev and @xmath3 , respectively . |
the interplay between gravitational attraction and non - gravitational forces produces systems with various shapes in the universe , including elongated structures , spheroids , discs , shells and rings .
all of them participate in variety of processes and act on different scales , ranging from cosmological dimensions down to planetary sizes .
some configurations , spheroidal in particular , hold by themselves , while other are maintained by an imposed control of a central body or an external field
. this paper deals with discs and tori typical configurations of rapidly circulating fluids that are governed by a central object .
its gravitational attraction supports global stability of the configuration , at least below a certain critical density of the disc , while compactness of the centre suggests prospective observational methods .
high density configurations suffer from instabilities and their mass contributes to general - relativity effects jointly with the central mass . nowadays
it should be self - evident that black - hole discs are of indisputable relevance to different research communities , ranging from pure relativists to astro - physicists and observational astronomers .
this diversification is reflected in varied approaches that people adopt to the problem .
a mathematically minded theorist would take more interest in non - trivial ( but still manageable ) analytical solutions to einstein s equations endowed with axial symmetry .
an observer will want to actually _ see _ gas flows swirling in an accretion disc near a black hole ( bh ) or a compact star .
astronomers often invoke accretion discs to explain spectral features observed in galactic nuclei and in close ( interacting ) binary stars , although detailed physics of accretion flows still remains uncertain .
therefore , it is tempting to search for common intersections in such a broad subject .
however , we will argue that serious distinctions can not be neglected which prevent one to develop an all - encompassing scheme that would describe inherently different configurations .
in particular , effects of general relativity ( gr ) have a chance to govern only the innermost regions of accretions discs and their role is quickly diminished with distance from the centre .
local ( non - gravitational ) physics plays crucial role and it often dominates over gravitation and governs observable properties of real systems ( perhaps the most profound example arises if one attempts to compare accretion discs around stars with those around super - massive bhs ) . in order to see where limitations of any unified scheme stem from , it is useful to discuss both simplified and general analytical models and some particular astronomical objects in more detail . in this paper , two - dimensional models are discussed .
needles to say , this is already enormous restriction . obtained solutions offer a convenient tool to elucidate physical mechanisms connected with non - spherical sources of strong gravitational field .
compact toroids represent a non - trivial model of such objects that require gr , although newtonian approach provides a sufficiently accurate description in many cases .
interaction with the central field is especially intriguing if the latter is generated by an ultra - compact body .
the paper starts with a summary of astrophysical motivations ( section [ sec : motivations ] ) .
our attention is focused on discs around super - massive black holes , because in this case the importance of self - gravity has been traditionally acknowledged , but we also mention compact binary stars , in which case it has been recognised relatively recently that self - gravity may also play a role , especially in the context of neutron star collisions , tidal disruption and subsequent emission of high - energy gamma - rays and , possibly , gravitational waves . in section [ sec : relativity ] , the problem of massive discs around black holes is discussed within gr . whereas gr is mainly relevant for the inner parts of self - gravitating discs ( within one hundred schwarzschild radii , @xmath0 ) , the field of actual discs is equally important for their structure at larger distances from the centre where the newtonian regime is adequate ( section [ sec : astrophysical ] ) .
evidence for self - gravity in accretion discs is still rather tentative and incomplete ; some possibilities and observational implications are explored in section [ sec : consequences ] .
there we also indicate the potential role of other ingredients , such as a compact nuclear star cluster in which the accretion flow may be embedded . however , the reader should be aware that important parts of the whole debate remain untouched in the present paper .
in particular , we do not discuss magneto - hydrodynamical processes , neither do we address non - axially symmetric modes of the disc flow .
one should bear in mind that these processes are probably crucial for understanding the mechanism of viscosity as well as launching of jets in astronomical objects .
disc structures are ubiquitous in such objects where the fluid circulates around and inflows onto a central compact body . the central mass , @xmath1 , can be considered as one of the model parameters and varied by many orders . for the black - hole mass in order to maintain clear distinction from the disc mass .
both contributions add to the total mass of the system .
similar situation arises with angular momentum , as discussed later in the text . ]
its value provides primary classification of cosmic objects .
physical characteristics of accretion discs scale roughly with @xmath1 . indeed
, accretion discs around super - massive black holes in active galactic nuclei and quasars share _ some _ properties with circumstellar discs in close binary systems ( e.g. , cataclysmic variable stars , x - ray binaries and micro - quasars ) . for example , both the discs around stellar - mass black holes and accretion discs around super - massive black holes may develop local and global instabilities which involve gravity . however , there are important distinctions between these two kinds of objects which prohibit any simplistic scaling .
it turns out that galactic nuclear discs tend to be cooler and less dense compared to circumstellar discs , which translates to different opacity of the material , its ionization structure and , hence , different spectral properties .
there has been also a very lively and still open discussion , whether the presence of a black hole in stellar - mass bh candidates ( or , rather , the absence of any hard surface ) can be revealed in the radiation properties of their discs .
environmental effects bring more distinctions between different objects with accretion discs .
unfortunately , at this point a bias is always inevitably introduced to our models by assumptions about turbulent transport of matter and momentum .
most solutions are model dependent . neglecting this dependence
can hardly be avoided when showing results of computations , nevertheless , we attempt to accentuate generic features when possible .
we refer to monographs by frank , king & raine @xcite and kato , fukue & mineshige @xcite for exposition of accretion theory , to krolik @xcite and peterson @xcite for discussion of accretion flows with the emphasis on active galactic nuclei , and to various review papers @xcite summarizing recent advances as well as the problems encountered by traditional scenarios .
accretion plays a crucial role in the processes of energy liberation and mass accumulation that take place in the core of astronomical objects .
disc - type accretion represents an important mode which is realized under suitable circumstances , defined by the global geometrical arrangements , local microphysics of the fluid medium and the initial conditions as well .
rotation is crucial and it puts discs , toroids and rings in the same category of centrifugally supported systems . by accretion discs one understands topologically toroidal fluid configurations orbiting around a central body and gradually drifting towards it , provided a mechanism to redistribute the angular momentum .
the centre governs or at least substantially contributes to the gravitational field in which the medium evolves .
the formation of the central body is presumed as being completed , although the accretion process gradually transfers mass and angular momentum onto the central body and thus contributes to its subsequent evolution .
it is worth noting that the genesis of our own solar system with a star in the centre was imagined , by pierre - simon laplace and immanuel kant , as emerging out of a turbulent , slowly spinning nebula with flattened , disc - like shape .
this was in the final years of the 18th century .
the potential associated with a gravitating ring was examined in detail by f w dyson @xcite .
later , considerations associated with star and planet formation were the main driving force behind the initial development of accretion disc theory .
modern evidence for such configurations was reported as early as in 1930s by baxandall @xcite in the context of binary stars . in the attempt to interpret spectrograms of the famous @xmath2 lyrae binary system and in order to explain the origin of secondary lines , visibly displaced in their wavelength , `` circulating currents in the outer envelope '' were tentatively invoked together with a gaseous envelope that surrounds the larger component . since then , interacting binary stars have been the prototypical systems for studying accretion discs .
the treatment of self - gravitating discs was pursued in ostriker s newtonian equilibria of uniformly rotating , polytropic , slender rings @xcite .
later , self - gravitating configurations with realistic equation of state and opacity were constructed and the basic formalism for self - gravitating bh discs was given in gr @xcite .
in stellar binary systems , the gas in the disc is provided by a donor star via roche - lobe overflow or stellar wind @xcite .
accretion process helps to feed the central body .
if conditions are suitable , the gas radiates and provides a way to observe and study the system .
historically , limiting cases were distinguished for methodological reasons and for the sake of simplification . when speaking of hydrodynamical non - self - gravitating accretion , the main assumptions concern stationarity , spatial symmetries that are imposed on the flow and on the gravitational field ( spherical flows with negligible angular momentum , axially symmetric accretion onto a moving body , and disc - like planar accretion ) , and the sound speed profile ( subsonic versus supersonic flows ) .
see , e.g. @xcite and references cited therein .
it turns out that behaviour of truly three - dimensional unsteady flows is rather different and more complex @xcite .
nearly spherical accretion occurs if the angular momentum content of accreted material is negligible ( this case appears appropriate mainly for accretion onto an isolated body ) . in another limit ,
one ignores radial transport of the material and considers an axially symmetric toroidal configuration that preserves perfect rotation about the symmetry axis .
both these cases were treated in many works , assuming the test - fluid approximation ( flow lines are determined by the central gravity and pressure gradients only ) .
toroidal topology of the fluid system introduces the main difference when compared with the physics of rotating stars ( for the latter , see a text - book exposition of the subject by tassoul @xcite ) . on large scales ,
the accretion mechanism was proposed as the energy source in bright galactic nuclei in late 1960s ( lynden - bell @xcite ) . since then
the idea has been continuously developed and brought to a text - book level .
self - gravity has global consequences on the disc shape , namely , it influences the location of its inner and outer edges , as well as the disc geometrical thickness .
the effect can be roughly parameterised by the ratio of the disc mass to the central bh mass , @xmath3 @xcite .
wilson @xcite speculated about the importance of self - gravity ( especially in cataclysmic binary systems ) , which would have direct observational consequences , but this possibility was later rejected on the basis of more detailed arguments ; it turns out that non - standard ( massive ) discs are ruled out in cataclysmic binaries @xcite .
self - gravity is normally unimportant in binary star systems also because their discs have limited space to occupy . under typical conditions ( i.e. among well - known cosmic systems ) , the disc mass plays a role in the process of gradual fragmentation of proto - planetary nebulae , formation of planetary bodies , and their subsequent migration .
( recent observations , in particular with the hubble space telescope , have demonstrated that discs are a by - product of star formation and also present during the birth of planets . )
also planetary rings are subject of self - gravitational instabilities but we can not tackle all these lively topics here ( see @xcite ) .
there has been a recent interest in high - density ( neutron ) tori , in which case the disc own gravity must be important .
this kind of massive discs could be formed in the course of binary star coalescence or as a product of collapse to a bh @xcite .
the question of stability and time evolution of these heavy , transient tori is essential .
the idea was advanced in the framework of scenarios of failed supernovae and strongly magnetised tori @xcite , and an interesting possibility of intermittent accretion was hypothesised in the context of massive neutron tori that may be formed from a tidally disrupted compact star @xcite .
as mentioned above , gravitational field of heavy discs influences the global structure of the system , which in turn determines possible figures of equilibrium .
coming back to accretion discs in galactic nuclei , vertical structure of these discs is affected by their self - gravity especially in the middle and the outer regions ( beyond a few hundreds @xmath4 , say ) where the vertical component of the disc gravitational field exceeds the corresponding @xmath5 component of the central field @xcite .
the main aspect of discs non - negligible gravitational effect on their own structure concerns the stability with respect to perturbations .
self - gravity was first examined in the context of accretion discs around super - massive black holes by paczyski @xcite , who proposed that local ( jeans - type ) self - gravitational instabilities are invoked if the density exceeds a threshold value .
turbulent fragmentation may contribute to viscosity and to heating of the disc medium .
even the formation of stars may be induced in this way @xcite .
collin & zahn @xcite explored the possibility and the impact of star - formation in galactic - centre discs .
recently , also sirko & goodman @xcite discussed consequences which the formation of massive stars in the disc may have on its heating and spectrum .
non - axisymmetric instabilities of accretion discs and tori have been widely investigated ( see e.g. @xcite and references cited therein ) .
their importance for modelling realistic accretion discs has been soon recognised , but we do not examine these types of instability here . if one considers strictly axisymmetric and stationary rotational motion , one finds idealized configurations , which provide useful insight into the problem of accretion discs ; for example , the question of the role of mean stationary solutions .
however , there is no doubt that these solutions must be rather different from real accretion flows ; investigations of the latter has culminated in recent three - dimensional numerical simulations @xcite .
it is noteworthy here that we are allowed to treat both accretions discs in binary stars ( with an accreting stellar - mass black hole ) and in galactic nuclei , because common physical mechanisms operate in both cases .
however , there are important distinctions which become particularly essential when realistic configurations are discussed .
we will point out limitations of simple scaling between the two limiting cases later in the text , but great part of this paper concerns very simplified situations in which rough scaling according to the system size and mass is possible . in the next section
we summarize techniques that have been employed to derive spacetimes of gravitating discs around black holes within gr framework and then we proceed to possible astrophysical applications of massive discs .
we start the discussion by briefly recalling a uncomplicated case of a light non - self - gravitating newtonian disc that can be formed at a certain evolutionary stage of a binary star when the donor component dumps material on the other one .
a text - book analysis of matter transfer employs the notion of roche potential in which the actual gravitational field of a binary is approximated by the field of two massive points orbiting around a common centre of mass @xcite ( see the left panel of figure [ fig : v0 ] ) .
an accretion disc forms when the donor fills its roche lobe and starts overflow near the lagrange l@xmath6 point .
the gas transferred in this way possesses high angular momentum and creates a disc - like structure before being captured by the accreting body .
the roche model gives a basic classification of binary stars into three main categories according to the sizes of individual components : detached , semi - detached and contact systems . as mentioned above , their discs
can usually be treated as non - gravitating .
in contrast to this transparent scheme , a sufficiently massive disc could change the potential field near the cusp . in the right panel of figure
[ fig : v0 ] , distortion of equipotential surfaces is depicted with the disc contribution to the gravitational field taken into account ( here , the disc mass is comparable to that of the secondary star ) .
it is relevant to remark that equilibria of relativistic tori also show the structure of critical lobes , which is similar to the roche geometry , even if the fluid is non - self - gravitating and the only source of gravity is the central black hole ( see section [ testfluid ] ) .
this structure facilitates accretion onto the central body .
equilibrium figures of critical ( lobe - filling ) tori represent a starting point for the discussion of accreting black holes in compact binary systems .
naturally , the roche lobe geometry is applicable for the mass overflow onto a black hole as much as it is in other systems of close binaries , in which the mass is exchanged between the components . however , topologically similar configurations of critical tori can arise also when accretion proceeds onto an isolated black hole . in this case
the torus can develop a cusp and the overflow for a purely relativistic reason .
the latter situation is particularly important for accretion onto super - massive bhs in galactic nuclei .
an issue of global stability of such configurations must be raised : if a small amount of material is transferred through the inner cusp onto the central body , then a new , modified distribution of mass and of angular momentum is established .
geometrical shape of the critical equilibrium is changed accordingly .
a question arises whether the new configuration is stable or whether , instead , it undergoes further accretion in a catastrophically increasing ( runaway ) rate .
abramowicz @xcite claimed the presence of this kind of instability in pseudo - newtonian massive tori , while wilson @xcite demonstrated that mass - less tori ( i.e. test fluid ) around a kerr bh are not run - away unstable .
answer to the problem of this instability depends on the radial distribution of angular momentum throughout the initial configuration ( before the mass exchange takes place ) and , in the relativistic case , also on bh angular momentum parameter .
it was demonstrated that increasing bh rotation and a positive slope of the angular momentum distribution both have a stabilizing effect @xcite .
increasing @xmath7 , on the other hand , acts against stability @xcite .
comparisons @xcite are not easy because different approaches have been employed in papers ( numerical techniques vs. analytical calculations ; proper gr treatments vs. approximations and pseudo - newtonian models ) . the black - hole
disc systems are difficult to treat exactly , even in the test - fluid approximation .
severe simplifications have thus been imposed on theoretical models , ranging from presumed symmetries and boundary conditions to various assumptions about microphysics of the medium . simplifying
assumptions are the more unavoidable when the disc gravity ( playing a role of the external source to the gravitational field ) can _ not _ be neglected and when one intends to proceed analytically .
further restrictions are dictated by applicability of the results to real cosmic objects .
there are at least three aspects in which the system of a rotating black hole with an external source attracts relativists .
first , due to the non - linearity of einstein s equations , the field of a multi - body system is a traditional challenge where , in most cases , one does not manage with a simple superposition . on the first post - newtonian level , the `` celestial mechanics '' of gravitationally
interacting bodies can be kept linear ( see damour @xcite ) , but in the strong - field regime the interaction may bring surprising features that have only been described in very few cases yet ( for a two - body problem , the current `` state of art '' is the third post - newtonian approximation @xcite ) . the second point concerns manifestation of frame dragging due to the rotation of the sources .
contrary to the newtonian treatment ( that does not discriminate directly between static and stationary situations ) , the relativistic field is determined not only by mass - energy configuration , but also by its motion within the bodies : the inertial space can loosely be imagined as a viscous fluid mixed by the sources . in today
s `` gravito - electromagnetic '' language : gravity has not only an electric component , generated by the mass , but also a magnetic one , generated by mass currents @xcite .
let us only refer to somewhat casual selection @xcite for further details at this point .
needless to say , the third point is the presence of a bh itself .
near such an extreme body , the deviations from newtonian theory become dominant ; in particular , the above mentioned implications of non - linearity and rotation reveal themselves prominently .
the class of stationary axisymmetric solutions of einstein equations is the appropriate framework for the attempts to include the gravitational effect of an `` external '' source in an exact analytical manner . at the same time , such spacetimes are of obvious astrophysical importance , as they describe the exterior of a body in equilibrium @xcite ( see @xcite for thorough expositions on stationary and axially symmetric spacetimes in general relativity ) . in the weyl
papapetrou coordinates ( @xmath8 ) of the cylindrical type , the stationary axisymmetric metric can be written as ; the signature of the metric tensor @xmath9 is ( @xmath10+++ ) .
greek indices run from 0 to 3 and latin indices ( @xmath11 , @xmath12 , ) run from 1 to 3 ; indices from the beginning of the latin alphabet ( @xmath13 , @xmath14 , ) will represent cyclic coordinates @xmath15 and @xmath16 .
partial differentiation is denoted by @xmath17 or by a subscript comma , covariant derivative is denoted by @xmath18 . ] @xmath19 where the unknown functions @xmath20 , @xmath21 , @xmath22 and @xmath23 only depend on @xmath24 and @xmath5 ; @xmath22 is interpreted as the angular velocity of inertial - frame dragging with respect to observers at rest at spatial infinity .
the metric coefficients @xmath25 have invariant meaning , they can be expressed in terms of killing fields @xmath26 and @xmath27 : @xmath28 @xmath29 @xmath30 the corresponding sub - determinant is @xmath31 , so @xmath32 the einstein equations read @xmath33 @xmath34 @xmath35 where @xmath36 and @xmath37 stand for the gradient and divergence in a flat three - dimensional space with coordinates ( @xmath24,@xmath16,@xmath5 ) ; thus @xmath38 and @xmath39 $ ] in the axially symmetric case .
once @xmath21 , @xmath22 and @xmath20 are known , @xmath23 can be integrated from equations @xmath40 @xmath41 \\
\label{lambda2 } + \textstyle{\frac{1}{2}}{r}^{3}b^{3}e^{-4\nu } \left(\omega_{,{r}}^{2}-\omega_{,z}^{2}\right)= 8\pi{r } b\left(t_{{r}{r}}-t_{zz}\right).}\ ] ] the unknown metric functions are subject to boundary conditions on the horizon ( if there is one ) , on the symmetry axis and at spatial infinity
. the metric must be regular on the horizon and on the axis , and in the case of an isolated source one also requires asymptotic flatness ( the conditions are discussed in @xcite , for example ) . outside of the sources , where @xmath42 , the simplest solution of equation ( [ eqb ] ) with a satisfactory asymptotic behaviour ( equation ( [ b , asymp ] ) below ) is @xmath43 .
the two remaining field equations ( [ eqomega])([eqnu ] ) reduce to @xmath44 @xmath45 and the relations ( [ lambda1])([lambda2 ] ) yield @xmath46 @xmath47 equations ( [ eqomega , vac])([eqnu , vac ] ) are often presented in the form of the ernst equation @xcite @xmath48 for a complex ( ernst ) potential @xmath49 whose imaginary part @xmath50 is given by @xmath51 @xmath23 is found by a line integration as above .
a black - hole horizon is a null hypersurface below which the spacetime is dynamical ( see e.g. @xcite for thorough accounts ) . in weyl - lewis - papapetrou coordinates ,
the horizon is located where @xmath31 vanishes . in the vacuum case
we can put @xmath43 , so the horizon lies on the axis then , @xmath52 .
the bh interior thus has to be studied in different coordinates , e.g. in spheroidal coordinates of the boyer
lindquist type ( @xmath53 ) , introduced by the transformation @xmath54 where @xmath55 and @xmath56 , @xmath1 being a scale parameter ( it represents the black - hole mass ) and @xmath57 determining where the horizon reaches up along the @xmath5 axis .
the horizon is now given by @xmath58 , i.e. @xmath59 ( hence , @xmath60 ; there exist more black - hole horizons in general , but we mean the outermost one here ) .
important parameters of the horizon are its surface area @xmath61 , surface gravity @xmath62 and the angular velocity @xmath63 relative to infinity , @xmath64 @xmath62 and @xmath63 are constant all over the horizon . in order to learn a `` true shape '' of the bh , not distorted by coordinates , the horizon must be represented properly as a two - dimensional surface in a three - dimensional euclidean space .
one can e.g. follow @xcite .
the two - dimensional metric is first rewritten as @xmath65,\ ] ] where @xmath66 .
an isometric embedding of the two - surface @xmath67 in @xmath68 with coordinates @xmath69
is then given by @xmath70 if the horizon s gaussian curvature @xmath71 is negative in some place , the global embedding is impossible . on the symmetry axis ( @xmath52 )
, the regularity condition requires @xmath72 , and therefore the metric adopts the form @xmath73 here .
simplification of the formulae can also be expected in the equatorial plane ( @xmath74 ) if the spacetime is reflectionally symmetric with respect to it .
this is usually the case for astrophysically motivated considerations . in the static case @xmath75 .
if @xmath76 ( which is fulfilled with zero pressure ) , the metric ( [ wlp ] ) acquires the weyl canonical form @xmath77 equations ( [ eqb])([lambda2 ] ) reduce themselves to poisson s equation @xmath78 for @xmath20 , and to relations @xmath79 for @xmath23 .
the linearity of laplace equation allows for simple superposition of vacuum weyl fields . from the astrophysical point of view , a given analytical solution is problematic if it contains features which should not be present in a realistic source , such as ( i ) physical singularities on or above the horizon , or ( ii ) bad asymptotic behaviour .
the existence of a singularity follows according to its type @xcite e.g. from the divergence of scalars constructed from the metric tensor or from its derivatives , or from the divergence of physical ( tetrad ) components of the riemann tensor .
the kretschmann invariant @xmath80 is usually checked first .
however , the spacetimes are known which do contain singularities , although all of their curvature invariants vanish .
it is thus difficult to verify the _
non-_existence of singularities ; the only proof of regularity at a given point is to find a coordinate system in which the metric is smooth enough locally . astrophysically relevant
is the region of outer communications outside the horizon , but a relativist is interested in the black hole interior as well , because even there it is possible to perform physical observations . in the weyl
papapetrou coordinates the metric is not defined below the horizon ( this would correspond to imaginary radius @xmath24 ) and must be extended there .
one asks about how is the bh interior ( and the singularity in particular ) influenced by the presence of external sources . the answer is unknown even for quite simple superpositions . far away from the horizon , one inquires for the asymptotic behaviour of the field . that of an isolated stationary source falls to zero in a specific manner
the spacetime is said to be asymptotically flat ( e.g. chapter 11 in r wald s book @xcite , or section 3 in j ehler s festschrift @xcite ) . in case of the metric ( [ wlp ] )
, it must hold , for @xmath81 , @xmath82 @xmath83 @xmath84 @xmath85 although a suitable combination of solenoidal motions can satisfy the assumption of stationarity and axial symmetry , it is natural to assume that the elements of the source follow the simplest type of motion that along spatially circular orbits ( @xmath86 , @xmath87 ) with steady azimuthal angular velocity @xmath88 .
such a motion follows the symmetries , so an observer on a circular orbit experiences time - independent field around . in order that the corresponding four - velocity @xmath89 points inside the light cone , @xmath90 must be positive ( we denoted @xmath91 the linear speed with respect to the local zero - angular - momentum observer ) .
the latter condition is fulfilled if @xmath92 falls within the range @xmath93 on the horizon ( where @xmath94 ) , the permitted range narrows down to a unique value @xmath63 . in the above - given equations , the four - velocity covariant components @xmath95 stand for the specific energy and the specific azimuthal angular momentum with respect to an observer at spatial infinity . in case of the circular orbit
, the four - acceleration @xmath96 can be written as @xmath97 it has at most two non - zero components , @xmath98 and @xmath99 ; in general , @xmath99 is zero in the equatorial plane and non - zero elsewhere , while @xmath98 vanishes if @xmath100 where the upper / lower sign corresponds to a co - rotating / counter - rotating orbit ( these terms may be somewhat misleading in spacetimes with large angular momentum ) . among equatorial geodesics ,
three particular cases are important : the photon , the marginally bound and the marginally stable circular orbits .
the photon orbits delimit the regions where the circular motion is time - like ; they are given by equalities @xmath101 .
the marginally bound orbits demarcate the regions where particles on circular orbits have lower energy than necessary for the existence at spatial infinity ; the limiting case is given by @xmath102 .
the marginally stable orbits also bound the regions where circular motion is stable with respect to small perturbations acting within the orbital plane .
rayleigh s criterion of linear stability gives the location of the marginally stable orbits where @xmath103_{,{r}}=0 $ ] .
these orbits play a key role in the theory of accretion discs : the matter is swept away from unstable sectors . in particular , the innermost marginally stable orbit should represent the disc inner rim .
however , perturbations _
perpendicular _ to the orbital plane can also be important .
for the above - mentioned exact solutions of einstein s equations , the frequencies of vertical oscillations were given explicitly @xcite .
a peculiar feature of rotating fields are the dragging effects . in accordance with mach s ideas of relativity of motion and inertia ( and his interpretation of newton s bucket experiment in particular ) , lense & thirring showed , in the early times of general relativity , that dragging is present in einstein s theory .
machian inspiration remains alive within contemporary relativity see , for example , the last - decade references @xcite or the tbingen conference @xcite and references therein .
the effect has obvious analogy in electromagnetism , where ( electric ) currents generate the magnetic component of the field .
an extreme implication of dragging is the occurrence of the ergosphere in the vicinity of ultra - compact rotating objects . in this region
@xmath104 , thus the stationary observer can not remain static ( i.e. at rest relative to infinity ) , albeit he still withstands the radial attraction .
the static - limit surface , given by @xmath105 and hence by @xmath106 , also represents the set of points from where signals emitted by static observers ( @xmath107 ) reach infinity with an infinite redshift .
several processes working in the ergosphere have been suggested , by means of which the rotational energy of the black hole could be extracted without violating the second law of bh dynamics @xcite . with the metric ( [ wlp ] )
, @xmath106 corresponds to @xmath108 which can only be satisfied by @xmath109 , so the static limit really lies outside the horizon ; in particular , @xmath52 implies @xmath110 , so the static limit touches the horizon at the axis . in order to complete the relativistic solution
, one must also describe the interior of sources where @xmath111 .
it is difficult to find a realistic interior solution , especially if it has to match smoothly the vacuum exterior and if we restrict ourselves to analytical approaches .
the problem can be substantially simplified by assuming that the source is infinitely thin : the solution is then vacuum - type everywhere , with the energy - momentum tensor @xmath112 found from the discontinuity of the normal field across the source as in electrodynamics ( the appropriate covariant method is known as the israel s formalism @xcite ) . the case of a thin equatorial layer ( a disc ) in a stationary axisymmetric spacetime has notably been studied by @xcite .
the disc lying on @xmath87 has zero radial pressure inside ( @xmath113 ) . at the same time , the vertical pressure vanishes ( @xmath114 ) if the disc mid - plane coincides with the equatorial plane ( @xmath115 ) of the reflectionally symmetric spacetime . in the weyl
papapetrou coordinates , the non - zero @xmath116 components read @xcite @xmath117 using ( [ wlpinverse ] ) and ( [ lambda , z , vac ] ) , this yields @xmath118,\ ] ] @xmath119 @xmath120,\ ] ] @xmath121.\ ] ] the total mass and angular momentum of the disc are fixed by the komar integrals that , respectively , lead to @xmath122 @xmath123 @xmath124 is the position of the disc inner rim .
all the @xmath5-derivatives are understood to be calculated in the limit @xmath125 .
if a bh is present in the disc s centre , its contribution to the total mass and angular momentum need to be included . for a given type of spacetime
, the aggregate values of the parameters can be written as sums , @xmath126 and @xmath127 ( cf .
@xcite , chapter 6.6.1 ) .
the black - hole parts can be calculated as komar integrals , or they can be inferred by subtracting @xmath128 , @xmath129 from the totals , derived from the metric asymptotes ( [ nu , asymp],[omega , asymp ] ) . @xmath130 and @xmath131 are related through smarr s formula , @xmath132 further , if the term @xmath133 is non - negative , then a stationary observer exists with respect to whom the energy - momentum tensor assumes a diagonal ( `` isotropic '' ) form @xmath134 where @xmath135 and @xmath136 are , respectively , observer s four - velocity and the unit basis vector in the @xmath16-direction .
angular velocity of this observer is @xmath137 @xmath138 and @xmath139 denote the corresponding specific energy and angular momentum at infinity . represents discs with non - zero heat flow .
then the energy - momentum tensor must be written in a more general form , @xmath140 , with @xmath141 and @xmath142 .
see @xcite . ]
the observer measures the surface density @xmath143 and tangential pressure @xmath144 if @xmath145 , the energy - momentum tensor can represent two equal streams of particles moving on ( accelerated ) circular orbits in opposite directions at the same speed @xmath146 .
this `` speed of sound '' was shown @xcite to be equal to the geometric mean of the local prograde and retrograde circular geodesic speeds @xmath147 as measured by the observer @xmath148 : @xmath149 where @xmath150 @xmath151 are the angular velocities of the prograde and retrograde circular geodesics ( [ omegapm ] ) .
although counter - rotating relativistic streams are hardly applicable to any realistic astrophysical situation , they can be very illuminating especially in the discussion of physical sources of the kerr metric and other spacetimes .
further physical requirements can be raised , namely , the weak energy condition ( @xmath152 , @xmath153 ) , the dominant energy condition ( @xmath152 , @xmath154 ) and the non - negativity of pressure ( @xmath155 ) .
their combination gives @xmath156 .
the requirements are very simple in the static case : if @xmath157 , then @xmath155 is equivalent to @xmath158 which is satisfied below the lagrangian point of zero field ( where @xmath159 and @xmath160 ) . in regions with tension ( @xmath161 ) the particles
would be attracted more by the outer parts of the disc than by the central bh , thus no circular geodesics exist ( then hoop stresses have to be employed to interpret the disc matter ) ; this typically happens in discs with much matter on large radii .
the other condition , @xmath162 , is equivalent to @xmath163 .
to summarize , the above constraints are in fact ensured by the obvious condition @xmath164 which can be written explicitly as @xmath165 .
relativistic spacetimes are being found in three ways : ( i ) by numerical solution of einstein s equations , ( ii ) by perturbation of previously known spacetimes , and ( iii ) by exact analytical solution of einstein s equations .
let us mention several results of the above approaches on stationary axisymmetric spacetimes that describe rotating black holes with additional matter .
teukolsky @xcite lists today s computational resources and the appropriate , hyperbolic formulation of the field equations as the main `` reasons why we are on the verge of important advances in the computer solution of einstein s equations '' .
whereas analytical prospects are restricted , the present - day computers can handle almost any situation .
however , covering a sufficiently representative time - span is still a problem .
furthermore , it is often impossible to recognise whether a given numerical solution represents a typical situation or just a marginal case . due to this lack of generality ,
it is difficult to discern , analyse and interpret different classes of solutions across boundless ranges of possibilities .
nevertheless , numerical solutions provide explicit _ examples _ of spacetimes and processes that might otherwise remain only conjectural @xcite . numerical spacetimes with a rotating bh surrounded by a stationary axisymmetric source were already constructed by lanza @xcite ( a hole with a thin finite equatorial disc ) and by nishida & eriguchi @xcite ( a hole with a thick toroid ) ; recently , a new ( `` multidomain spectral '' ) method has been utilized to spheroidal as well as toroidal fluid configurations @xcite , with the possibility of inclusion of a central object in future . to mention just one particular point , contrary to a common experience ( also gained by @xcite ) that the horizons inflate towards the external sources , @xcite ended with a prolate horizon ( stretched along the rotation axis ) in certain cases ( when the disc was strongly counter - rotating with respect to the hole ) .
it would be an interesting consequence of the interplay between dragging from the hole and from the disc if it were confirmed that this can indeed happen , even though under extreme circumstances only .
a great deal of literature and several formalisms have been devoted to perturbations of bh spacetimes .
will @xcite gave an explicit result for the perturbation of the schwarzschild metric by a ( rotating ) axisymmetric weakly gravitating thin equatorial ring .
this direct approach is however not practicable for a rotating hole and/or for an extended source with pressure ( not mentioning more complicated situations ) .
the rotating case ( specifically , the algebraically special vacuum case ) was discussed notably by teukolsky @xcite who succeeded in separating the decoupled equations for the first order perturbations of a kerr bh into the second - order ordinary differential equations . by solving the `` gravitational '' teukolsky equation ,
the perturbative deformation of the kerr horizon was calculated by @xcite .
the stationary axisymmetric green s function of the teukolsky equation was provided by @xcite .
the most comprehensive expositions of the first - order bh perturbation theory were given by chandrasekhar @xcite in the newman
penrose formalism . in a related approach to the first - order perturbations of rotating fields @xcite ,
the perturbation components are expressed in terms of a single ( debye ) potential that obeys a wave - like equation ( adjoint of the teukolsky master equation ) .
the problem of perturbation of the kerr horizon was treated this way by chrzanowski @xcite .
yet another approach , not restricted to algebraically special backgrounds , was advanced in @xcite .
note that all these perturbative methods deal with curvature components , as opposed to the earlier metric formalisms of regge & wheeler , zerilli and moncrief . in the charged case , the perturbation problem introduces coupling between gravitational and electromagnetic quantities .
conversion between gravitational and electromagnetic waves takes place . in the case of reissner
nordstrm bh , these interacting perturbations were studied by bik @xcite and , more recently , by torres del castillo @xcite .
a gauge invariant derivation of the basic equations was given by fernandes & lun in schwarzschild space - time @xcite , while @xcite discussed a rotating analogy and provided its link with the teukolsky equation .
the second - order perturbation theory has also been under development since the 1970s . for schwarzschild black holes ,
it is surveyed in gleiser @xcite , while the rotating case is tackled in campanelli & lousto @xcite .
it should be remarked that solutions describing stationary sources around black holes are not the primary aim of the perturbation strategy .
historically , attention has rather been devoted to the stability of black - hole solutions , to the relaxation of black holes into a stationary no - hair state , to non - stationary processes of astrophysical significance ( such as the behaviour of weakly gravitating particles or waves in bh backgrounds ) and to gravitational waves ( see thorne @xcite for a survey ) .
one can add that the approximate ( though non - perturbative ) method for superposing two ( rotating ) black holes was proposed by krivan & price @xcite to fix initial data for both analytic and numerical computations of bh collisions .
a perturbative approach is adequate in situations where the outside source has only a very small effect on the field of the main body . whenever this is not the case , the superpositions have to be described by solutions of the full , non - linear theory . at the end of bik s
survey @xcite , one is encouraged `` not to cease in embarking upon journeys for finding them , and perhaps even more importantly , for revealing new roles of solutions already known '' , because `` is there another so explicit way of how to learn more about the rich possibilities embodied in einstein s field equations ? '' more recent reviews of vacuum fields can be found in bonnor @xcite .
for further details , see also the canonical monograph on exact solutions to einstein s field equations by stephani @xcite .
until now , no explicit exact solutions describing the system of a rotating centre with an additional axisymmetric ring , disc or torus have been found .
nevertheless , many stationary axisymmetric solutions of the ( electro-)vacuum einstein equations are known that do generalise those containing only isolated black holes .
these were mostly obtained by indirect methods known as `` generating techniques '' . two major approaches , developed by the end of the 1970 s the group - theoretic techniques and the soliton - theoretic ( or inverse - scattering ) techniques , work for spacetimes with two commuting symmetries .
( see kordas @xcite for a review and @xcite for more detailed analysis and interrelations between different formulations , e.g. those of kinnersley & chitre ; maison ; belinskii & zakharov ; harrison ; hoenselaers , kinnersley & xanthopoulos ; hauser & ernst ; neugebauer ; kramer & neugebauer ; or alekseev . )
other related methods have been proposed more recently , e.g. the simplification of the hauser ernst integral equation by sibgatullin @xcite , `` monodromy data transform '' by alekseev @xcite ( also @xcite ) , static gravitational multipoles @xcite ( cf .
@xcite ) and their superposition with stationary fields @xcite by gutsunaev & manko , linear transformation by quevedo @xcite or `` finite - gap '' ( algebraic - geometric ) solutions by korotkin & matveev @xcite .
most of them have been compared in reference @xcite .
( for other approaches , see e.g.@xcite or the results of nakamura and kyriakopoulos , referred to and worked out by @xcite ; cf . also @xcite . ) the crucial point of the soliton generating techniques is a solution of two linear differential equations ( lax pair ) the integrability conditions of which are exactly the einstein equations ( namely the ernst equation ) .
the linear problem can be tackled in order to generate new solutions from the known ones : given some ( `` seed '' ) metric with two killing vectors , it yields a different metric of the given type ( the procedure is often called bcklund transformation in analogy with the technique used for the kdv equation ) . in such a manner , many known spacetimes were reproduced , but also broad families of new solutions were provided characterised by arbitrarily large sets of free parameters . only a very restricted number of them have been given a clear physical interpretation , however .
though several of these results perhaps represent a rotating bh in an `` external '' gravitational field ( e.g. @xcite ) , none of the latter has yet been firmly linked with a concrete realistic body such as ring , disc or torus ( but cf .
section iv in letelier & oliveira @xcite ) .
namely , when performing the bcklund transformation , it is difficult to require specific physical properties of the spacetime being generated one rather tries what will come out .
see @xcite for a hint of how to overcome this .
more recently , letelier @xcite interpreted a solution obtained from a weyl seed by the inverse - scattering transformation with an even number @xmath166 of solitons as a superposition of this weyl solution i.e. of a given set of static multipoles with @xmath167 kerr(-nut ) solutions , represented by @xmath167 rotating axial bars in weyl - type coordinates . as an example , let us give the attempt @xcite which was `` cultivated '' from a general weyl metric by a real - two - soliton version of the belinskii - zakharov method .
the resulting metric was written as a generalisation of the kerr(nut ) solution and many of its properties were proven satisfactory .
choosing the inverted first morgan morgan disc ( see below ) as a seed , various characteristics of the superposition were plotted @xcite against the rotation parameter of the central hole , the mass of the disc and its inner radius .
the characteristics simplify very much in the limit when the bh becomes extreme .
one of the interesting features of this limit is the vanishing of the `` external '' gravitational field on the horizon @xcite .
this effect of expulsion of the external ( stationary axisymmetric ) fields from rotating ( and/or charged ) black holes , analogous to the meissner effect in ( super)conductors , was observed in magnetic fields before ( see the recent survey @xcite ) .
however , a serious problem has been found in the equatorial plane : the black - hole horizon has a sharp narrowing there .
this is caused by the appearance of a supporting surface between the hole and the source outside it @xcite .
indeed , the possible problems with physical interpretation of the solution ( e.g. negative density or pressure or matter moving at a superluminal speed ) also involve the occurrence of supporting singular surfaces or lines ( `` struts '' ) that indicate that a given system of bodies can not remain in stationary ( or static ) equilibrium .
the struts may equally well `` spoil '' a static superposition , though this case almost reduces just to adding the two potentials , @xmath168 .
the singularity may appear in calculating the second metric function , @xmath169.\ ] ] consequently , even the _ static _
axisymmetric case has only been afforded a few realistic superpositions .
let us refer to two of them involving a schwarzschild bh : the superposition with an infinite annular thin disc , obtained by lemos & letelier @xcite by inversion of the first counter - rotating finite disc of morgan & morgan @xcite and the one with an inverted isochrone thin disc , obtained by klein @xcite .
the morgan
morgan family of static axisymmetric finite thin counter - rotating discs @xcite is given by newtonian surface density @xmath170 since the zeroth member is clearly singular at the rim , most attention was devoted to the first ( @xmath171 ) member .
it appeared to be a prototype of a simple and physically meaningful weyl field and it was also considered as a limit case of a more general class of _ stationary _ discs @xcite .
the corresponding annular case , obtained by the inversion @xmath172 , @xmath173 , was shown to really allow for physically satisfactory situations @xcite .
there is nevertheless a problem at the rim of these discs , already suspected by checking the radial derivatives of ( [ density ] ) : the @xmath174-th member of the family has infinite @xmath174-th and higher derivatives of the density at the outer rim .
calculation of the kretschmann curvature invariant reveals that the rim is really singular and that the singularity is inherited by the inverted discs @xcite .
the superpositions of a schwarzschild bh with ( inverted ) `` higher '' morgan
morgan counter - rotating discs were studied in @xcite .
comparison of the results obtained for the first ten of them showed that the properties of the disc are very sensitive to the local density profile .
it will be desirable to also consider different classes of annular thin discs in superpositions , preferably of those not having any such drawback at the rim .
we have already explored one different family of annular disc solutions , namely , the case when density is a power - law in weyl radius , @xmath175 however , these discs also turned out not to be regular at the rim ; @xmath176-st radial derivative of their potential diverges , although all the radial derivatives of density are finite there , and the first @xmath177 derivatives even vanish . in a different context of galactic stellar discs ,
an interesting approach has been developed by sibgatullin , garcia & manko @xcite who examined a new method of reconstruction of the surface density distribution of a finite size disc with a central bh . allowing for self - gravitation of the matter around the horizon
, one has to reduce the generality of the problem elsewhere . whereas test discs are treated within stationary axisymmetric settings , analytical solutions with exact inclusion of the ambient material
have only been studied in more detail in static cases yet .
this is of course a serious limitation , because the collapsed cores of galaxies as well as ultra - compact components in binaries are likely to rotate , perhaps very rapidly , which means that inertial dragging should be important there in addition to the effects of pure scalar attraction .
moreover , the composite spacetimes presented in the literature are constructed with discs whose field can be expressed in a tractable form rather than with those which really follow from some model of accretion .
however , some superpositions such as e.g. those involving thin annular discs obtained by inversion of the morgan morgan solutions @xcite were shown to possess physically acceptable properties within an astrophysically plausible range of parameters .
as such , they can indicate some of the effects that could occur as a consequence of non - negligible gravitational contribution of the disc ( see @xcite for a more thorough survey ) .
let us return to the linear problem ( lax pair of linear equations ) associated with the ernst equation . in order to control
the physical properties of the solution being constructed , it is desirable to translate the physical boundary conditions of the ernst equation into the language of quantities which appear in the linear problem and then solve the latter ( rather than to apply a `` random '' bcklund transformation to some metric @xcite ) .
this leads to riemann
hilbert problems known from the theory of completely integrable differential equations @xcite . tackling the stationary axisymmetric boundary value problem with the help of the linear system ,
the field of two physically relevant types of sources has been discussed : that of black holes and that of finite thin discs ( @xcite and references therein ) .
the hope that the above methods could be used to describe superpositions of both ( e.g. @xcite ) has recently been supported by klein @xcite who presented an exact space - time of a regular black hole surrounded by an infinite annular disc as a subclass of korotkin s solutions to the ernst equation , obtained by riemann - surface techniques @xcite .
the resulting metric is given in terms of theta - functions , being asymptotically flat under suitable choice of parameters .
klein s solution seems to involve cases with physically acceptable source disc .
[ testfluid ] the limit case should be briefly mentioned when gravitation of the external matter is ignored and the central rotating bh fully determines the field
. then the spacetime geometry is described by kerr metric and the equilibrium configurations can be found analytically @xcite , at least if the fluid obeys a simple equation of state ( such as a barytropic relation ) and remains in dynamical equilibrium with purely rotational motion ( four - velocity @xmath178 ) .
this case is over - simplified from the astrophysical point of view , but it captures several crucial relativistic effects which do not occur in the newtonian limit namely , the possibility of fluid overflow across the inner edge , which acts , effectively , as the lagrange l@xmath6 point in a binary system .
the relativistic euler equation can be integrated in terms of specific enthalpy , @xmath179 , which can be further expressed as a function of @xmath180 , the profile of angular momentum density @xcite .
the only unknown function that needs to be specified is the rotation law , either in terms of @xmath181 ( which is a conserved quantity ) , or angular velocity @xmath182 .
rotation can be differential but the choice is not completely arbitrary because , according to rayleigh s criterion for linearly stable configurations , @xmath183 must be increasing monotonically outwards from the rotation axis ( a borderline case of constant @xmath183 was discussed in detail ) .
@xmath184 plays the role of an effective potential for inertial forces and determines equilibrium configurations in which the equipotential and isobaric surfaces coincide , so this way constant pressure surfaces can be found . for example , in a special case of marginally stable configuration with @xmath185 , one finds @xmath186 .
the fluid can establish a stationary configuration provided that the @xmath184-isosurface is closed ( see figure [ fig : w0 ] ) , otherwise overflow and subsequent accretion take place .
let us recall that equilibrium configurations with their elongated cross - section ( bounded by the critical , self - crossing lobe ) _ resemble _ the shape of roche lobes .
the possibility of fluid transport across the cusp occurs when the critical lobe is filled ; accretion can thus proceed without a necessary action of viscosity . however , the two situations are of very different nature . in the former case ,
it is the combined gravitational effect of the binary together with ( newtonian ) centrifugal force which define the lobes , while , in the latter case , the only source of gravity is the central body with relativistic effects playing a crucial role .
compared with the vacuum case ( with a bh but without a disc ) , modifications of the observable characteristics can be linked with the presence of self - gravitating matter .
various approaches have been adopted in order to construct solutions and discuss motion in the gravitational field of black - hole discs .
one can start , e.g. , by constructing equilibrium structures of self - gravitating polytropic thick discs @xcite .
formalism was developed also for perturbative solution of a ( slowly rotating , weakly gravitating , geometrically thin ) ring around a schwarzschild bh @xcite .
furthermore , equatorial motion was explored in pseudo - newtonian studies @xcite of a rotating bh surrounded by a massive thin equatorial ring , and in gr analysis of the corresponding static case @xcite .. this simple expression mimics the gravitational field of a non - rotating bh ; in particular , it reproduces the radii of marginally stable and marginally bound circular orbits of free test particles .
different variants of the pseudo - newtonian potential were proposed in order to include rotation of the bh and to obtain better agreement with relativity ( see @xcite for references ) .
the pseudo - newtonian potential does not aspire to represent any gravitational theory , and the corresponding potential is not required to satisfy any field equations .
this fact , however , does not diminish the practical value of the model , which applies even to regions with strong gravity and captures qualitative features of motion near the horizon . ]
the accretion - disc self - induced field was also discussed in connection with the disc s vertical structure and stability .
it turns out that certain modes of instability can be damped whereas others are amplified by the disc s gravity @xcite .
fully relativistic results employ numerical spacetimes @xcite . within these solutions ,
the computation of emission - line profiles were performed ; the case of moderately massive thin finite discs was treated by @xcite , and that of heavy toroids by @xcite . , but also in terms of physical measures circumferential radius @xmath187 and proper radial distance from the horizon . ]
another issue is the oscillation of the disc material excited by a small perturbation in the horizontal ( @xmath24 and @xmath16 ) or vertical ( @xmath5 ) direction @xcite . in the static case ,
the formulae for the frequencies of free horizontal and vertical oscillations give @xcite @xmath188 @xmath189.\ ] ] without the external source , @xmath190 reduces to the square of the schwarzschild orbital frequency , @xmath191 , whereas @xmath192 remains different due to the pericentre precession .
it was found that perturbations in the horizontal direction are more important for the stability of light discs , whereas the vertical perturbations are more dangerous for heavy discs : for superpositions of the given type , the curve of marginal vertical stability is more restrictive than that of marginal horizontal stability for @xmath193 .
it even fully excludes discs with @xmath194 @xcite .
disc self - gravity makes oscillations of the inner parts faster and the region of the horizontal mode trapping somewhat smaller
. when the black hole and the disc are allowed to rotate , manifestations of self - gravity could be even more pronounced , because in relativity the kinetic energy ( as well as any other form of energy ) also generates the field .
corresponding mass - energy currents give rise to frame dragging .
_ whereas the mass of the black hole is probably dominant in a real accretion system , the disc can bear much of the angular momentum , thus modifying the gravito - magnetic field of the central object significantly .
_ this could be decisive for the mechanisms of rotational energy extraction .
note , however , also the possibility of having accreting systems where the disc / torus mass can actually be comparable to that of the central black hole and which can represent a transient state of gravitational collapse . in this context ,
the most popular system is a close binary of ultra - compact objects .
late evolutionary stages of such systems ( in - spiral , merger and the ring - down phase ) and the fate of debris have been studied , numerically and perturbatively , as a tentative source of gravitational waves ( see @xcite and references therein ) and also as an engine behind gamma - ray bursts @xcite . in particular , at late stages of the neutron star
black hole binary in - spiral , the neutron star could tidally break up @xcite .
two - dimensional toroidal shapes similar to the one shown in figure [ fig : w0 ] were originally invented mainly as a model for large and very diluted flows which probably exist in quasars and similar objects .
these considerations have been continuously refined and discussed in various context ( recently , e.g. @xcite and references cited therein ) . as mentioned above , already in early works
it was recognised that self - gravity imposes limits on this scheme and should be taken into account in calculations .
self - gravity is also relevant for a recently proposed idea @xcite that non - stationary tori could be formed in accreting binaries .
a relatively small size and steep pressure gradients of these configurations put them in a corner of parameter space where they may show high - frequency quasi - periodic oscillations @xcite , which makes them very attractive to astronomers , but formation of such structures still needs to be addressed .
from the point of view of gravitational theory , and also for the nature of the accretion flow , the most important part is the inner region near to the bh .
however , the gravitational field of actual discs has substantial and even , possibly , dominant influence on their structure also at large distances ( exceeding several @xmath195 ; in physical units , @xmath196 cm ) where the self - gravity is comparable to the central field .
newtonian physics is sufficient at those distances . the importance of the disc self - gravity is often checked by toomre s parameter , @xmath197 : the local stability criterion with respect to axially symmetric perturbations requires @xmath198 @xcite ( @xmath199 is a local sound speed ) .
spiral structures may be produced already at @xmath200@xmath201 .
depending on the radial distribution of the thermodynamic quantities , global instabilities can indeed occur and play a key role in enhancing the accretion rate @xcite .
( of course , at this point one must recall the theories of galactic dynamics @xcite . )
this may facilitate angular momentum transport and provide a feeding mechanism for central black holes @xcite .
again , we concentrate on discs in galactic nuclei although similar mechanisms are relevant also for discs around stellar - mass black holes and , to a certain degree , even for circumstellar discs in star - forming regions .
in addition to the classical equations of mass , momentum and energy conservation , the description of newtonian gravitating fluids must include poisson s equation , @xmath202 which links the mass density distribution @xmath203 in the disc and the self - gravitating potential @xmath204 ( @xmath205 refers to the source and @xmath206 is the field point ) .
to understand the signatures of self - gravity expected in the fluid structure and motion , one can consider a purely rotating , steady - state system with zero viscosity and no significant radial pressure gradient .
the radial component of the euler equation ( in cylindrical coordinates ) is then : @xmath207 . in the absence of rotation ,
steady solutions are forbidden and the fluid falls towards the centre in a short time scale .
free fall can be avoided thanks to the centrifugal term which compensates the total gravitational attraction @xmath208 .
if radial self - gravity is unimportant , that is if @xmath209 , then rotation of the fluid is determined by the bh mass only , @xmath210 .
this assumption underlies most disc models in astrophysics .
in particular , newtonian geometrically thin disc models fall within this category @xcite . in the opposite situation ,
it is evident from the simple form of euler s equation that , as the disc carries some mass , angular velocity is expected to increase in order to prevent accretion .
this implies that ( radially ) self - gravitating discs are expected to rotate faster than non - self - gravitating ones . from the previous estimate
one can conclude that heavy ( fluid ) discs should exhibit a super - keplerian rotation curve ( in terms of @xmath211 ) .
keplerian rotation could still be maintained in a self - gravitating disc , but this would be a surprising result of some fine tuning among various quantities pressure gradients , advection of momentum , effects of turbulent transport , etc .
neither can the possibility of sub - keplerian rotation be ruled out .
such a case could occur in the outermost disc regions if @xmath212 , for instance due to the presence of a dense region near the edge or due to significant negative pressure gradients .
further , viscous stresses can switch the accretion .
let us now analyse the vertical component of the euler equation in the same idealized situation as described above ( again in the newtonian limit ) : @xmath213 .
this is hydrostatic equilibrium in @xmath5-direction which tells us that central gravity acts against pressure gradients and tends to gather material around the equatorial plane of the disc . for a bare central point mass ,
the magnitude of the vertical component of gravity decreases with radius . as an inevitable consequence
, the disc tends to ` flare ' ( its thickness being roughly proportional to the distance from the axis , or even steeper , and it can wildly oscillate in the vertical direction ) .
the atmosphere of the disc ( i.e. its less dense region away from the equatorial plane ) gets thicker where the gravity is weaker .
the geometrical semi - thickness of the disc then satisfies the relation @xmath214 .
models show such flaring in the non - self - gravitating regime .
regarding the vertical direction , the disc self - gravity is significant if @xmath215 . for any fluid element this results in enhancement of the vertical gravity ; thus , a vertically self - gravitating disc should be geometrically thinner and more dense than a non - self - gravitating one . as will be shown below
, some accretion disc models even yield @xmath216 in those regions , so in these circumstances the disc thickness can not exceed a certain limiting value .
this is important for the disc irradiation ( and self - irradiation ) .
some fundamental properties of discs and tori can be understood in the apparently simple framework of maclaurin spheroid theory .
early analytical investigations by poincar , dyson and , later , by chandrasekhar @xcite have been supplemented with recent numerical investigations ( hachisu @xcite ) .
these works shed light on possible shapes and dynamical characteristics of steady , self - gravitating polytropes in the inviscid limit .
both axially and non - axially symmetrical equilibrium patterns have been discussed .
it turns out that equilibrium sequences can describe any conceivable system ( depending on the polytropic index of the gas ) : these include multiple star systems , tri - axial ( jacobi ) ellipsoids , some very strange shapes ( ` pear ' , ` amonit ' , etc . ) , but also discs , tori and rings . in the case of global solid - body rotation , discs , tori , and rings
occupy the bottom - right part in the usual diagram of rotation frequency versus angular momentum . furthermore , in a recent paper by ansorg @xcite a sequence of compressible rings is investigated with high precision .
a bifurcation occurs from the maclaurin sequence to the multi - ring sequence .
what remains to be explored is the case of a full multi - ring sequence and its limit for the number of rings approaching infinity .
an interesting connection could appear with the theory of disc fragmentation into individual rings @xcite .
the theory of maclaurin spheroids and rings is built from simple arguments and is fundamental for understanding the problem of gravitating discs , but it still needs to be enriched with other physical ingredients .
disc modelling poses a difficult task even in the newtonian regime .
in addition to the technical embarrassment of solving the poisson equation with sufficient accuracy , difficulties are connected with strong coupling between magnetohydrodynamic terms and radiative transfer in the medium , and with our ignorance of boundary and initial conditions ( they can not be sufficiently constrained by observations but are decisive for the evolution of flows ) .
furthermore , huge values of the reynolds number ( typically @xmath217 ) are associated with accretion flows . in consequence
, simulations do not currently reach the numerical resolution required to investigate simultaneously both the large - scale streams and the smallest swirls where kinetic energy is dissipated .
this is another reason why ( magneto-)hydrodynamical simulations of accretion flows should be interpreted with caution @xcite .
despite the tremendous complexity of these problems , models and simulations have reached a fair level of credibility @xcite .
thermal equilibrium of the disc and its possible fragmentation have been also explored @xcite . given the technical obstacles inherent in the problem
, disc modellers often resort to a semi - phenomenological parameterisation of the turbulent transport by using a viscosity coefficient of the form @xmath218 , where @xmath219 is the disc half - thickness and @xmath220 is a parameter .
the @xmath221-parameterisation , originally advocated in the context of binary accretion discs by shakura & sunyaev @xcite , remains a recipe enabling study of the characteristics of a mean accretion flow at the lowest level of approximation .
however , it does not explain the process of triggering and maintaining the turbulence .
other viscosity laws have been proposed , for instance the @xmath2-prescription which is based on analogy with turbulence observed in laboratory sheared flows and gives @xmath222 @xcite . also , there is the prescription of marginally stable ( @xmath223 ) discs with @xmath224 , designed for weakly ( locally ) self - gravitating discs @xcite .
very probably turbulent transport and its effects can not be satisfactorily mimicked through simple prescriptions .
below we touch on several properties which can be derived from the classical @xmath221-prescription , as well as the properties of @xmath223 discs . given the current knowledge , one can not decide which of these approximations is better .
association of the density field @xmath225 with a prescribed newtonian potential is straightforward and always possible .
various classes of density potential pairs can be formed @xcite . unfortunately , the reverse procedure , i.e. the inversion of poisson s equation , is difficult in general .
no analytically tractable expression is available for @xmath204 that would correspond to a general three - dimensional system of finite size and mass .
exact solutions corresponding to non - spherical systems can not be written generally in a closed form if the matter distribution is two - dimensional , non - homogeneous or of non - trivial shape or topology .
astrophysical discs exhibit all three complications .
as an illustration we show first , in figure [ fig : map ] ( left panel ) , the density distribution in the standard disc model where self - gravity effects are _ not _ included @xcite .
the complexity of shape and density distribution is striking : discs appear to be complex objects for which the merit of special analytical ( and especially self - similar ) solutions is very limited or doubtful .
an alternative approach is to consider the integral definition for the potential , @xmath226 and to use expansions of green s function ( for instance in cylindrical harmonics or another kind of equivalent infinite series representation ) .
this approach is often employed in order to find self - similar solutions and to investigate the disc stability @xcite . however , as pointed out in @xcite , convergence of the series is not guaranteed in all points of matter distribution .
also , the expansion leads to uncertainties because the inferred series must be truncated . in order to be able to treat realistic systems
, one has to seek numerical solutions .
however , it is difficult to achieve sufficient computational accuracy using standard techniques . why is this ?
the first reason is that no boundary conditions are available in the vicinity of the disc ; instead , they can be imposed only at infinity ( @xmath227 for @xmath81 ) .
secondly , the physical boundary of the fluid is not known _ a priori _ and it can evolve with time .
one can not rely on the standard situation that is met within laboratory physics of a fluid confined within walls , in which case the boundary conditions are imposed by experimental settings .
the third difficulty is caused by the relatively large and variable aspect ratio : @xmath228 may be as large as unity and even greater at some parts of the disc , and as low as @xmath229 at other places ( cf . figure [ fig : map ] ) .
this fact hinders discussion of discs in active galactic nuclei , where the geometrical thickness of the disc can vary wildly over its diameter .
there are two practical ways of solving the problem numerically @xcite .
the first one employs straightforward determination of the potential from equation ( [ eq : poisson - integ ] ) everywhere in the disc .
this approach can be very accurate if one is able to cope with integrable singularities inside the source ( @xmath230 ) and it can account for very distorted disc shapes and large density gradients . however , the computations involved are generally very time consuming .
one can envisage some cell - based technique @xcite ( as done in particle simulations @xcite ) , but this is not expected to reach high precision ( which is necessary for stability analysis ) .
another possibility is to invert equation ( [ eq : poisson ] ) using a finite difference method or a finite element method @xcite .
this approach requires that boundary conditions can be obtained through equation ( [ eq : poisson - integ ] ) .
finally , it is worth noting that the inversion of poisson s equation for rotating stars can be performed via the mapping technique , e.g. transformation of space coordinates @xcite . to our knowledge , the mapping technique has never been used for discs .
since the efficient numerical computation of the potential is not trivial , people still seek approximate semi - analytical approaches that would complement purely numerical solutions and make them faster .
this is in particular the case for the infinite slab ( is ) approximation , which was introduced by paczyski and collaborators in the late 1970s in a series of papers devoted to marginally stable discs @xcite . in this approach
the disc is examined in @xmath231-plane , imposing axial symmetry and omitting the radial and azimuthal terms in equation ( [ eq : poisson ] ) .
the poisson equation then reads @xmath232 the main advantage of equation ( [ eq : psiisa ] ) is the possibility of determining the magnitude of the vertical gravity by direct integration along the vertical direction . in this way
even a full solution @xmath233 could apparently be found .
the vertical field is given by @xmath234 where @xmath235 is half of the local surface density .
this result is exact if the system is invariant under any translation perpendicular to @xmath5-axis , and for this the disc must be ( i ) of infinite size , ( ii ) strictly flat ( its geometrical thickness @xmath236 is constant ) and ( iii ) homogeneous in the radial direction ( vertical gradients can be non - zero in the direction of @xmath5-axis ) .
obviously , such properties are not realistic , but the validity of the is approximation can be extended to radially inhomogeneous , finite size discs with non - constant thickness , provided that the condition @xmath237 is satisfied .
it is not generally possible to check inequality ( [ eq : isaextended ] ) for a given configuration because the `` true '' potential @xmath204 is not known or is difficult to compute with sufficient accuracy ( for the technical reasons mentioned above ) .
also , the above - described way of getting @xmath233 can not be used to estimate the radial component of the field .
it is worth noting that equation ( [ eq : gzisa ] ) holds regardless of the relative masses of the disc and the central object , making the is approximation often very useful ( for example , it can be used to compute massive non - keplerian discs ) , but it is also important to know where this powerful approximation fails .
although the potential is a global function , we can determine this by considering its local properties by expressing @xmath238 .
inequality ( [ eq : isaextended ] ) is violated in regions where @xmath239 or , in terms of density gradients , when @xmath240 for a reasonable density stratification and attractive potential , we expect @xmath241 and @xmath242 .
this implies that the isapproximation may break down when ( i ) @xmath243 , which corresponds to discs and tori of moderate and large aspect ratios , and/or ( ii ) the radial and vertical density gradients are of the same order .
this applies if density waves or shocks with steep fronts are present , especially in boundary layers or close to the disc edges .
the flatter is the disc , the weaker are the edge effects . an interesting property of vertical stratification can be derived directly from the is approximation @xcite .
if the disc is able to reach hydrostatic equilibrium , then the total pressure gradient compensates total gravity .
differentiating equation of hydrostatic balance relative to @xmath5 gives @xmath244 + 4\pi { } \rho = 0 .
\label{eq : hydro}\ ] ] this equation becomes independent of the radius in the limit of a disc that stays in vertical equilibrium under its own gravity .
an analytical solution can be obtained assuming polytropic relation @xmath245 ( @xmath199 is a speed , @xmath246 is polytropic index ) .
the above - given equation is analogy of the famous hydrostatic equilibrium problem for gaseous spheres in equilibrium .
the resulting `` plane '' lane
emden equation writes @xcite @xmath247 for vertically isothermal structure with @xmath248 , the equation ( [ eq : ple - polytrope ] ) is the frank kamnetski equation @xcite @xmath249 and the solution is @xcite @xmath250 where @xmath251 is the isothermal sound speed and @xmath252 is the mid - plane density . from equation ( [ eq : sigmadef ] ) , the surface density of the self - gravitating disc is given by @xmath253 @xcite , and thus the disc mass depends solely on the radial distribution of central pressure @xmath254 .
discussion of solutions for various polytropic indices can be found in @xcite .
the issue is that , as the polytropic index @xmath246 runs from infinity ( in the isothermal case ) to smaller values , the vertical density profile tends to be flatter . in the limit @xmath255 ,
this profile becomes exactly rectangular in shape .
figure [ fig : rhoz ] shows the density profile in the direction perpendicular to the disc plane in a typical disc .
the isothermal case is also shown for comparison .
models of the disc vertical stratification are close to the models of stellar interiors and show that the vertical structure is not isothermal and can not be modelled by a polytrope ( with @xmath256 ) , at least not if the disc is optically thick .
discs are heated by turbulent motion and this results in a very complex vertical stratification of physical quantities . in many respects it would be useful for the disc modelling if the degree of self - gravity were indicated by an unmistakable criterion .
as a relevant quantity , one can intuitively suggest the mass ratio @xmath257 @xmath258 , where @xmath259 is the cumulative mass of the disc and @xmath260 is a half of the local surface density according to equation ( [ eq : sigmadef ] ) .
one might postulate that self - gravity is unimportant for @xmath261 , say .
note that , in a binary system , the @xmath257-value ( defined as the mass ratio of the two stars ) is critical in the definition of the roche lobes .
however , such a criterion is unreliable to explore an extended disc in its close neighbourhood , where it can not be reduced to a point mass .
we show below that the right criterion is based on the quantity @xmath262 which contains information about the disc mass ( relative to the central black hole ) as well as about the distribution of the mass density in space ( in terms of the aspect ratio ) . the ratio @xmath263 can be evaluated provided that the vertical component of the disc self - gravity is known . for @xmath264 ,
the disc and the central body contribute comparably to the field along @xmath5-axis ( this case is referred as the `` weak self - gravitating limit '' , wsg hereafter ) .
as soon as @xmath265 exceeds unity , the disc is _ vertically _ self - gravitating . within the is approximation , @xmath266 is given by equation ( [ eq : gzisa ] ) , so one obtains @xmath267 .
when the gas remains confined within altitudes @xmath268 , we can use the mid - plane value @xmath269 where @xmath270 is the mass density in @xmath74 , as defined in the absence of self - gravity .
the role of parameter @xmath271 was discussed in @xcite ; taylor expansion around @xmath74 shows that the mid - plane value of @xmath265 can be identified with @xmath61-parameter of paczyski @xcite .
one finds that the upper limit for the disc height is typically @xmath272 ; such discs stand on a borderline beyond which the approximation of geometrically thin discs becomes very crude and inadequate .
we notice that vertical self - gravity occurs preferentially ( i ) when the central bh has a small mass , ( ii ) if the local mass density in the disc is large , and ( iii ) at rather large distances from the centre .
if the surface density obeys a power - law in radius , for instance in the form @xmath273 , then @xmath274 relation ( [ eq : zetac ] ) shows that even low - mass discs can be self - gravitating if they are flat enough .
we can put numbers in this relation : given that standard discs in active galaxies have a power index @xmath275 and small , almost constant aspect ratio , @xmath276 , we find @xmath277 @xcite .
( the adopted value of @xmath278 corresponds to the outer , gas - pressure dominated region ; another value would not change qualitative conclusions at this point . ) in other words , a disc becomes vertically self - gravitating when its relative mass exceeds a small threshold value .
this value is model dependent ; typically it is order of a few @xmath279 .
note that the limit is much less than the crude estimate of @xmath280 suggested before . by introducing ratio @xmath281
one can follow the same kind of analysis for the radial component of gravity .
a disc is _ radially _ self - gravitating if @xmath282 , hence this regime can be referred to as a strongly self - gravitating ( ssg ) disc .
conclusions are less straightforward in this case because we lack sufficiently reliable expression for @xmath283 . in other words , there is no obvious analogy to the is approximation . some authors @xcite have proposed that the total potential can be approximated by the monopole term , @xmath284 , which is not fully satisfactory but gives at least the right order of magnitude in the ssg limit ( @xmath285 ) . considering , as above , a power law for the radial surface density profile ,
@xmath286 is given by @xmath287 in the equatorial plane .
using equation ( [ eq : zetac ] ) , we find @xmath288 ( the case of @xmath289 can be derived in similar manner ) .
the ssg regime corresponds to @xmath290 using the above values for @xmath278 and @xmath228 , @xmath291 corresponds to @xmath292 , well above the wsglimit . as @xmath293 is expected to be an increasing function of radius , the ssg limit involves more mass and
is reached at larger radii than the weak limit .
it therefore lies inside the vertically self - gravitating discs .
this value is in good agreement with the results established by lantian & xiaoci @xcite who have attempted to performed a direct , analytical integration of equation ( [ eq : poisson - integ ] ) .
although their investigation still contains some rough approximations ( for instance , singularities in the poisson kernel were artificially removed ) , these authors derived a formula for radial self - gravity which accounts for finite size effects . for @xmath294
, they found radial acceleration to be @xmath295 , \label{eq : lantian}\ ] ] which sets the ssg limit at @xmath296 ( assuming the eddington accretion rate ; see figure [ fig : map ] right ) .
note that this formula still remains singular at the disc edge @xcite . to circumvent the difficulty , a softening parameter is often introduced @xcite , however , correct behaviour can be obtained just by careful treatment of edges .
results discussed so far have not yet addressed the essential questions concerning the gravitational stability of self - gravitating , geometrically thin discs .
it is known for long , and confirmed by observations , that galaxies ( treated as discs of stars , gas and dust ) are regulated by self - gravity @xcite .
similar processes can form spirals and bar patterns in purely gaseous discs .
it was shown in a remarkable paper by goldreich & lynden - bell @xcite that instability sets in when @xmath297 where the lower limit , @xmath298 , depends on polytropic index @xmath246 ( the value of @xmath299 corresponds to a vertically isothermal disc , while @xmath300 for @xmath301 ) .
note that this is an analogy of toomre s original criterion , but with a slightly different threshold @xcite .
it is important to recall that goldreich & lynden - bell s criterion has been established from an analysis of growing unstable modes through a linear perturbation theory , within the framework of is approximation and assuming a non - viscous , uniformly rotating , flat disc .
it is therefore expected to be somewhat altered when relaxing some of the above assumptions .
it is easy to recognise that equation ( [ ineq : glb ] ) also gives @xmath302 .
we thus conclude that instabilities are not present in a disc if its size remains below the weak limit .
the region separating the weak , the unstable and the strong limit can be narrow or wide , depending on the surface density profile ; or , more precisely , on the function @xmath303 .
again , a very nice example is provided by saturn s rings which are vertically self - gravitating at their outer edge , but gravitationally stable , implying that the weak limit and the unstable limit do not coincide . according to equation ( [ eq : zeta ] )
, an increase of the @xmath265-parameter is expected as the radius increases .
this means that there is a boundary zone that separates the inner ( non - self - gravitating ) solution from the outer ( self - gravitating ) one .
the location of this zone can be found without accounting for self - gravity , simply by computing @xmath265 from a disc model @xcite .
this approach may underestimate the magnitude of self - gravity and overestimate the radius of such boundary , but it provides a useful order of magnitude estimate .
a more reliable result must include vertical self - gravity into consideration .
this can be achieved within the is approximation .
furthermore , in the context of @xmath221-prescription , the wsg limit is reached at @xmath304 . for accretion rates lower than a certain critical value , @xmath305
, one finds @xcite @xmath306 while @xmath307 for accretion rates larger than @xmath308 ; here , @xmath309 is expressed in schwarzschild radii , and the critical accretion rate is expressed in solar masses per year , @xmath310/yr ( @xmath311 is opacity ) .
noticeable deviations are observed when considering temperature- and density - dependent opacities and equation of state ( not yet specified self - consistently in equations ( [ eq : rsg1])([eq : rsg2 ] ) ) .
the existence of two different relations arises from the fact that bh discs generally possess an inner region ( dominated by radiation pressure ) and an outer ( gas pressure dominated ) region where the physical quantities ( namely , the density ) exhibit different variations with radius and the accretion rate .
the case ( [ eq : rsg2 ] ) corresponds approximately to accretion luminosity exceeding the eddington limit .
figure [ fig : weaklimit ] gives a theoretical relationship between the location of wsg limit and the accretion rate .
it was computed by different authors using the @xmath221-prescription and the same input parameters ( relevant for accretion flows onto super - massive bhs in galactic nuclei ) .
it is noticeable that no firm agreement has been achieved about the position of this limit , which largely fluctuates from one author to another simply because of different ingredients in models .
also , there are no direct observations available to confirm if this wsg - transition operates in nature and if it occurs where predicted .
however , in active galaxies there is a _ striking similarity _ between the location and size of the so - called broad - line region , and the value of @xmath312 .
this correspondence could indicate that the clouds are formed at the distance where self - gravity becomes important ( cf .
@xcite for a systematic discussion of the correlation ) . according to bi - dimensional calculations ,
a lower ( reliable ) limit appears to be @xmath313 for a @xmath314 m@xmath315 central black hole and @xmath316 .
this is only 5 milliparsecs from the centre , much less than previously estimated ( e.g. @xcite ) . of simple models that can be constructed analytically on the basis of @xmath221-parameterisation
, we pay a special attention to those where self - gravity is taken into account in the vertical direction , while keplerian rotation is still imposed .
this represents even a weaker category of wsg discs , because their rotation curve is assumed not to be altered by the disc mass . within this approach
, the above - discussed infinite slab approximation has been found very convenient , although some criticism can be raised at this point , because these disc models typically tend to be rather massive .
hence they become globally self - gravitating and no more keplerian ( this problem occurs especially if the surface density falls faster than the square of radius ) .
paczyski s original paper @xcite was devoted to the investigation of discs with constant ratio of the vertical self - gravity to the vertical component of the central field , i.e. constant value of @xmath265-parameter ( i.e. @xmath317 , which corresponds to @xmath61 in paczyski s notation ) .
the assumption of vertical self - gravity is justified by the fact that strongly self - gravitating discs are prone to gravitational instabilities @xcite which should make them evolve in a couple of dynamical time scales ( @xmath318 ) .
the endpoint of this evolution is the state of constant toomre s @xmath319-parameter , which is typical for differentially rotating discs of stars ( @xmath319 can be related to @xmath265 via equation ( [ eq : zeta ] ) ) .
soon it was realized that holding @xmath320 might produce some substantial bias @xcite ; a series of papers followed to elucidate the local response of the disc to vertical self - gravity and to study @xmath321 profile .
keplerian discs were investigated in which @xmath265 varies as a function of radius in a self - consistent fashion ( another motivation for this research was the chance of discovering new branches of solutions ) .
most explorations were carried out for steady discs , using the so - called `` vertically averaged approximation '' , where one seeks only for radial profiles of density , temperature and other physical quantities ( without considering vertical stratification in detail ) . with additional assumptions about opacity , equation of state , pressure regime , etc .
, one is naturally lead to self - similar solutions @xcite . as already mentioned above
, vertical self - gravity makes the accreted gas more concentrated around the mid - plane . as a result
, the disc tends to collapse in the vertical direction beyond some critical radius where @xmath322 ( cf .
figures [ fig : map ] and [ fig : rhoz ] ) .
this property is shared by all computations of geometrically thin discs , independently of the actual viscosity prescription ( in other words , if one assumes @xmath221-parameterisation for viscosity , then the disc height is almost independent of @xmath221 ) .
the opposite is true for geometrically thick discs which tend to be thicker when self - gravity is taken into account @xcite .
if prescription for viscosity is specified , the variation with radius can be found in explicit form .
examples are given in table [ tab : r - variations ] ; given the viscosity law and the heating mode of the disc that is on verge of self - gravitating regime , one finds the @xmath323-variation of the disc thickness , temperature , surface density , etc .
note that @xmath321 increases with @xmath323 in all models .
this leads inevitably to the onset of self - gravity , subsequent reduction of the disc thickness and further consequences on many aspects , such as self - irradiation , dynamical stability and the emitted spectrum of the disc .
it may also affect the disc warping @xcite in regions that are shadowed from the central source and can not reprocess its radiation ; although weak at those distances , the resulting lack of heating is in favour of gravitational instability .
lccccc model & @xmath324 & @xmath325 & @xmath219 & @xmath293 + @xmath221-disc ( no self - gravity ) ' '' '' & @xmath326 & @xmath327 & @xmath328 & @xmath329 + self - gravitating @xmath221-disc & 0 & @xmath330 & @xmath331 & @xmath332 + @xmath2-disc ( no self - gravity ) & @xmath333 & @xmath334 & @xmath335 & @xmath336 + self - gravitating @xmath2-disc & @xmath333 & @xmath334 & @xmath331 & @xmath337 + @xmath319-constant disc & @xmath338 & @xmath339 & @xmath340 & 0 + irradiated @xmath319-constant disc ' '' '' & @xmath341 & @xmath342 & @xmath343 & 0 + properties derived through the @xmath221-prescription deserve a comment .
the @xmath323-dependence of the @xmath265 comes out rather steep and much more pronounced than what is obtained for a bare standard disc ( neglecting self - gravity ) , in which case @xmath344 .
this means that the transition from non - self - gravitating regime to the fully self - gravitating one should occur in a very narrow region ; the weak , unstable and strong limits are expected to almost coincide with each other . as a consequence of the steep rise of self - gravity , the disc can suffer from a runaway - type increase of density ; one is even led to densities typical for stellar interiors and conditions suitable for deuterium and hydrogen burning @xcite .
temperature reaches a plateau in the outer region of the disc plane with value @xcite @xmath345}\ , , \label{eq_tsurmu}\ ] ] where @xmath346 is the mean mass per particle , @xmath347 is the accretion rate normalized to @xmath348 and @xmath349 . some authors have argued that such uniform mid - plane temperatures are unphysical and they have suggested the use of a different prescription for viscosity @xcite .
interestingly , this may lead back to the original idea of paczyski that viscosity is connected with the @xmath319-parameter .
we conclude this section by showing several typical scales of the problem in figure [ fig : scales ] .
these graphs are derived from basic arguments and they can help to identify physical mechanisms which are important from the observational point of view .
time - scales relevant for the gravitationally unstable regions of accretion discs are shown as a function of radius ( in the left frame ) .
the possibility of star formation was taken into account together with continuous migration and evolution in the medium ( here , @xmath350 was assumed for the central black hole ) .
if the stars form a dense ( quasi - spherical ) nuclear cluster in which the central bh resides , one can define the radius where stars are destroyed tidally or by their mutual collisions .
the resulting value depends on @xmath1 as well as details of the stellar structure ( as shown in the right frame ) .
significant effort has been devoted to searching for black holes during the last few decades .
traditionally , most approaches of present - day astronomy rely on the electromagnetic radiation coming from cosmic sources , but there are other possibilities which will be surely exploited in near future ( neutrinos , high - energy cosmic particles , and gravitational waves ) .
the idea of detecting black holes via ( electromagnetic ) radiation of nearby luminous matter can be traced back to the celebrated lecture , which rev .
john michell presented in front of the royal society as early as in 1783 .
understanding local physics of the medium is crucial if one aims to interpret complex and variable spectra that are observed .
x rays are particularly pertinent because they are produced near bh , as discussed by fabian and others @xcite .
indeed , given the intrinsic emissivity of the medium , there are only a few factors determining the source spectrum on the earth .
the main factor is the black hole mass @xmath1 ; the angular momentum of the black hole and the disc self - gravity play secondary roles .
the geometrical optics approximation is thus adequate and one can hope to infer actual values of bh parameters from the motion of test matter around it .
however , one has to be well aware of various environmental influences that hinder real observations . under conditions that are thought to be typical in active galactic nuclei ,
the inner regions of accretion flows contain diluted plasma with a corona threaded by magnetic fields @xcite .
high - energy photons are first generated in the corona .
these primary photons irradiate substantially colder gas of the accretion disc , where they are reprocessed further ( figure [ fig : collin ] ) .
a fraction of photons is reflected .
the temperature of the reflecting medium can be @xmath351k in galactic nuclei and about 23 orders of magnitude higher in the case of galactic bh candidates @xcite .
the quality of photon reprocessing depends critically on the ionization state and other local properties of the medium . in this way the resulting spectrum is formed with spectral features ( lines )
superposed on top of the background continuum .
x - ray spectra contain precious information about rapid flows from which one can infer properties of the central source of gravity and the state of the nearby gas .
manifestation of strong gravity has been searched in radiation of accretion discs with the main aim of finding the evidence for the central black hole @xcite .
majority of these papers deal with non - self - gravitating discs and assume that integrated light of the whole source is only very weakly affected by the disc gravitation ( however , see @xcite ) . in the case of our galactic centre ,
the presence of standard - type ( thin ) accretion disc is rather improbable ( e.g. narayan @xcite ) .
direct effect of the disc gravity on spectral line profiles is below current observational resolution for any present - day spectra of active galactic nuclei , so it can be safely neglected in the first approximation ( we briefly discuss this effect in the following paragraph ) .
it is noteworthy , however , that self - gravity of the disc has indirect impact on spectral features .
first , as discussed in section [ sec : astrophysical ] , the disc geometrical shape is affected by self - gravity with consequences for obscuration of some parts of the disc and the resulting variability , especially at large values of inclination ( this is also the case when gr effects of the central bh are maximal , so the two influences are mixed ) .
self - gravity also controls the amount of disc self - irradiation , which produces indirect signs in reflected light .
finally , self - gravity induces fragmentation in the disc medium ( depending in particular on the disc cooling properties @xcite ) with fundamental consequences on the disc structure .
fragments can even lead to star formation or contribute to star - bursts @xcite .
the clumpy material may also result in much more pronounced variability observed from the source @xcite .
let us now discuss direct effects of the combined bh
disc gravity on observed light rays . in the early works of cunningham & bardeen @xcite ,
a standard non - self - gravitating disc was assumed .
these authors initiated a series of investigations of model spectra with the aim of revealing the imprints of gr .
since then , many people have advanced various modifications of the original semi - analytical approach @xcite .
very prominent role has been played by studies of spectral features ( lines ) of iron in the x - ray band ; in particular , the so called disc - line scheme has been widely applied , as it captures the main spectral characteristics attributed to strong gravity @xcite .
the line is produced by reprocessing the primary continuum radiation near above the inner edge of the disc , within several tens of @xmath4 ( see @xcite for recent exposition of emission mechanisms ) .
the line intrinsic width is small , typically much less than the final observed profile that results from integration over the whole emitting surface .
standard picture of the disc thermal spectrum @xcite was reconsidered to account also for the disc gravity @xcite . for detailed examination it would be useful to produce templates of expected line profiles or even spectral models that are consistent with the formation of non - axially symmetric structures in accretion discs as well as the presence of a black hole in the centre @xcite .
there is an exciting possibility of detecting gravitational effects on polarized light of a source orbiting a black hole , but this idea still awaits technology usable for its practical implementation @xcite .
effects of the black hole itself on the light propagation are blurred in real sources in which the signal comes from different , insufficiently resolved regions .
so far , only a few cases have been discovered of an accreting black hole with a line that attains the characteristic shape , width and variability and could be understood as the interplay of doppler and gravitational shifts .
the best example of such spectral feature originating near a bh has been identified in the nucleus of seyfert galaxy mcg6 - 30 - 15 where the iron line was caught at the moment of high redshift @xcite .
large gravitational redshift can be connected with rapid bh rotation , because the radius of the disc inner edge ( where the line presumably originates ) decreases with bh angular momentum increasing @xcite .
magnetic fields can change this straightforward conclusion . also , the line is strongly variable , so its final ( continuum subtracted ) shape can not be resolved reliably with current knowledge ; but there is a good prospect of overpassing these deficiencies with more detailed models and using time - resolved spectroscopy that will be able to track rapid features @xcite .
reconstruction of accretion disc images will then be doable using reverberation techniques @xcite or doppler mapping @xcite .
this will also require good understanding of non - standard effects which are usually neglected , e.g.the effect of disc eccentricity @xcite .
as far as bh candidates in the milky way are concerned , the relativistic iron line was detected in a well - known micro - quasar grs 1915 + 105 @xcite ; rapid rotation of the central black hole is not necessary in order to explain the data in this case .
similar evidence for the broad and skewed iron line has been discovered with different degree of credibility in about eight other sources @xcite . in spite of a limited similarity of the relativistic iron line in these sources , we recall that recall that accretion discs have negligible gravity in bh candidates , and so we turn back to accretion discs galactic centres .
moving to a region upstream the accretion flow to the distance of hundreds @xmath4 and farther out from the core of a galaxy , the interaction between gaseous flows and other astronomical bodies ( stars forming a nuclear cluster ) can substantially change the structure , accretion rate and chemical composition of the flow , so we briefly touch this point , too .
that way , the medium feeding the bh can be enriched above the level of usual solar metalicity @xcite . on the other hand ,
moving satellites feel the presence of the interstellar gas , especially in nuclei of active galaxies with high concentrations of stars and the relatively dense interstellar environment @xcite . accreting medium
thus influences the motion and evolution of these bodies . even though self - gravitational effects of the medium become progressively less important near the centre , a relatively accurate knowledge of their impact is desired in order to derive , reliably , bh parameters from rotational motion of the surrounding luminous matter .
the dissipative interaction of stellar - mass satellites and the resulting orbital decay proceed with a time - scale that is defined by properties of both components of the system : in the inner disc , hydrodynamical drag dominates the orbit evolution , while in the self - gravitating outer region , the disc structure becomes clumpy and direct collisional interaction is reduced . in even more remote parts ,
gravitational relaxation is important for the cluster evolution , so the outer cluster is dynamically distinct from the core .
syer , clarke & rees @xcite proposed that a stellar - mass body passing the galactic core can be captured on a quasi - stable bound orbit around a central super - massive bh . in the simplest version of the model , stars have been considered as bullets crashing on the disc surface and passing through at hyper - sonic speed .
zurek , siemiginowska & colgate @xcite , armitage , zurek & davies @xcite and ivanov , igumenshchev & novikov @xcite examined various effects of passages ( of individual stars ) through diluted gaseous environment ( slab - disc geometry was assumed ; other people explored different contexts of this problem @xcite ) .
pineault & landry @xcite discussed the statistical rate and distribution of impacts over the disc surface .
self - gravity of the disc medium was also considered @xcite , as well as gr effects in stellar motion @xcite , effects of the orbital decay of satellites by gravitational radiation and close encounters @xcite , or stellar coalescences @xcite .
gradual changes of the physical and chemical state of satellites and of the disc itself were discussed @xcite .
radiation torques @xcite as well as star
gas hydrodynamic interaction @xcite induce different channels of angular momentum transport in the medium .
the clumpy structure @xcite of the outer , self - gravitating disc reduces the efficiency of direct star disc collisions , while turbulence in the medium tends to strengthen dissipation @xcite .
repetitive transitions bring the satellites into the plane of the disc at almost circular orbits @xcite ( though , one can speculate of several effects that slow the process of circularisation down , so that eccentric or non - equatorial orbits may be maintained even at late stages @xcite ) .
subsequent orbit evolution is reduced to a situation that was addressed in connection with the formation and migration of bodies inside proto - planetary discs @xcite .
the inner cluster becomes flattened and its structural dimension is related to the size of the disc .
one can immediately recognise plethora of consequences connected with non - sphericity , anisotropy and satellite segregation in the cluster : namely , the initial mass function is modified , as well as the form of spectral - line profiles and stellar velocity dispersions .
the onset of star formation leads to gradual conversion of gas into stars in the disc @xcite .
star - burst phenomenon has been indeed identified in some active galactic nuclei on @xmath352pc scales , which could represent outskirts of a dusty torus surrounding the central black hole @xcite . indeed
, violent star formation could coincide with the outer parts of a massive nuclear cluster that is embedded within the dusty torus .
even if @xmath353 , the effect of the disc gravity on the long - term evolution of stellar orbits should be taken into account together with dissipation via star
disc ( hydrodynamical ) interactions .
this stands in contrast with the fact that gravitation of accretion discs has been neglected in most works exploring the nuclear cluster structure .
the two influences are linked through their dependence on the disc density .
a remark is worth making regarding the mass ratio @xmath257 ( which was criticised as unsuitable self - gravity criterion in section [ sec : weak ] ) : it is a useful measure to assess the importance of stellar transitions across a self - gravitating disc in galactic nuclei , because attraction of the disc and the effects of transitions are both increasing , roughly proportionally , with @xmath7 .
circularisation of the orbits , evolution of their inclination and the resulting capture rate are visibly affected by the disc gravity .
an important point is the range of time - scales : ( i ) dynamical period ( of satellite stars revolution around the central mass ; the shortest time - scale ) , ( ii ) period of oscillations in orbital elements ( medium time - scale ; these oscillations are due to disc gravity ) and ( iii ) time for bringing the orbital plane into the disc ( caused by successive interactions with the disc ; long time - scale ) .
the orbital decay continues also when satellites are embedded in the disc plane ; its rate varies depending on the ability of the body to clear a gap along its trajectory through the medium .
different modes depend on the satellite masses , sizes and on the disc thickness and surface density at corresponding radius .
the evolution of orbital parameters is an issue in extra - solar planetary research @xcite , which tackles similar systems on much small scales cf .
the discussion of planetary formation and migration @xcite .
these topics have not yet been pursued to much detail in the context of galactic nuclear clusters embedded in the gaseous environment .
one can understand the origin of orbital parameters oscillations in terms of kozai s theory @xcite , taking into account the gradual adiabatic diffusion of quasi - integrals of motion . in order to see the reason for orbital oscillations , one borrows from classical mechanics traditional approach to the disc gravitational potential @xmath354 as a perturbation acting on the satellite motion @xcite .
again , the total potential is supposed to be dominated by the central field of the core , while the disc contribution @xmath355 can be transformed by a canonical transformation to the form which depends on mean orbital elements ( eccentricity @xmath356 , inclination @xmath357 , and argument of pericentre @xmath358 ; the bar denotes mean values ) and does _ not _ depend on the fast - changing variable ( mean anomaly ) .
see @xcite for the explicit form of this new @xmath359 in terms of integration ( over the mean anomaly ) along an unperturbed keplerian ellipse ; in that paper , kuzmin s potential was employed as a simple model for the disc gravitational field ( it would be quite interesting to examine a generalization in terms of miyamoto & nagai s potential @xcite , which can be still performed analytically ) . in the first order of perturbation theory
one finds that @xmath360 ; the mean semi - major axis @xmath361 is also constant and differs from the actual osculating semi - major axis by short - period terms only .
the mentioned averaging procedure allows us to disregard short - term effects and concentrate on the long - term evolution .
furthermore , owing to the axial symmetry of @xmath233 exists an additional constant of motion @xcite : @xmath362 ( figure [ fig : vk ] left ) . this way the problem of orbital evolution is reduced to the evolution of @xmath363 and @xmath364 which are further constrained by condition @xmath365 .
contours of the perturbation potential @xmath366 provide a convenient representation of mean orbital elements in the @xmath367-plane .
finally , _ with the inclusion of orbital dissipation in the gaseous medium , one finds that an adiabatic drift occurs and drives the orbit across the contours of constant @xmath359 ; subsequently , the drift provokes excursions of the orbital elements when a separatrix is traversed_. a representative example of a trajectory is shown in figure [ fig : vk ] ( right ) .
the interplay between kozai s mechanism and gradual changes of the perturbation potential can give occasional ( but large ) eccentricity jumps ; one of them is clearly visible in the figure .
next step would be to consider a consistent scheme for the whole cluster of satellites . in their classical paper , bahcall & wolf
@xcite anticipated a way to reveal a hypothetical central massive hole ( embedded in a collisional environment of a globular cluster ) through stellar velocity dispersion and broadened wings of line profiles integrated over the bh domain of influence . in order to calculate the form of central cusps , the authors employed time - dependent diffusion equation . of other assumptions ,
the most important was the spherical symmetry and isotropy of the cluster itself .
probability density of satellites can be expressed in terms of their semi - major axes : @xmath368 , where @xmath369 denotes the number of satellites in a shell with radius @xmath13 and width @xmath370 .
corresponding spatial density @xmath371 is thus proportional to @xmath372 . modifying the model assumptions ( namely , that about the importance of collisions ) leads to different but similar density profiles ( power - law at large radii ) .
for example , the well - known plummer @xcite sphere has a finite - density core with density falling off as @xmath373 , while jaffe @xcite and hernquist @xcite models both decline @xmath374 at large radii .
the latter case has a sound theoretical basis in the mechanics of violent relaxation .
see also @xcite and , for more recent discussion and references , see @xcite .
further , the collisional steady - state solutions have been reconsidered as a model for a cusp in the center of a galaxy @xcite .
conclusions of these papers can be compared with more advanced models that relax some of the assumptions and include new effects , e.g. resonances , enhanced loss - cone depletion or rotation of the cluster @xcite .
nevertheless , self - consistent treatment of non - spherical stellar systems with the effects of self - gravity _ and _ the presence of an accretion disc are still an open problem .
the disc induced dissipative orbital decay is distinct in its tendency to flatten the stellar component and even form a ring of stars that could be resolved observationally .
figure [ fig : cluster ] illustrates a stationary density profile of the cluster in a model where its members revolve around a bh and their orbits slowly decay due to dissipative effects .
one can distinguish different populations of satellites ( see the left panel ) : those which form almost spherical cluster far from the centre , outside the disc outer radius , and those which have been dragged into the disc plane near the centre ( the latter form a ring ) .
hence it is a certain generalisation of the bahcall
wolf @xcite cluster .
a complementary representation is provided in the right panel of figure [ fig : cluster ] which shows an azimuthal section across the cluster .
again , the inner cluster is evidently flattened and its dimension is tightly related to the size of the disc . it should be recalled that , with this approach , a relatively small volume is explored near bh , where the central mass has direct dynamical influence on the embedded stars and gas ( in other words , rotational support of the stellar motion is not a negligible fraction compared to their random motions ) . the linear size of this region is a function of @xmath1 and @xmath375 , and is typically less than @xmath376pc .
note that the @xmath377 relation resembles the empirical correlation which has been determined in a large sample of galactic bulges @xcite ( although the constant of proportionality can be different in the innermost region around the central bh , which is unresolved by current observations ) .
hence , in the inner cluster of stars , the obtained velocity dispersion should be determined by the joint action of gravity and of other agents , including dissipation in the interstellar environment .
it was proposed @xcite that the effect of interaction between stars and the accretion flow shapes the observed @xmath1@xmath375 relation in the innermost region of some active galaxies ( those with sufficiently dense accretion discs in the core ) .
the model of star
disc interaction leads to an axially symmetric , but non - spherical central cluster , and so the exact value of the power - law index of the @xmath1@xmath375 relation depends also on the observer line of sight ( the expected central velocity dispersion comes out somewhat larger for systems seen edge - on than for pole - on view ) .
the model therefore represents rather extreme case , however , it is obvious that the interaction between stars and the intergalactic environment should be taken into account also at larger distances from the core @xcite .
miralda - escud & kollmeier @xcite applied a similar idea to a much wider region ( and hence , applicable to actual present - day observations ) .
these authors propose star - disc collisions are a self - regulating process that helps to feed the central bh and control its growth .
the influence of the collisions remains imprinted in the bulge mass stellar velocity dispersion relation , which concerns the region in galactic nuclei greatly exceeding the domain of direct gravitational influence of the super - massive bh .
naively , one could expect that dissipation of satellite s orbital energy and momentum should always lead to a very dense nuclear cluster .
however , the outcome may be more complicated because of non - sphericity and satellite segregation ( caused by differing pace of orbital evolution , according to the mass function and the distribution of sizes of stellar bodies ) .
for example , recently , zier & biermann have showed @xcite , on the basis of numerical simulations , that a flattened shape of the cluster may arise due purely to gravitational effect of a massive binary bh in the core .
they found that a stellar ring is formed on a parsec scale in the orbital plane of the central binary , while the cluster is gradually depleted .
another approach to ( self - gravitating , eccentric ) galactic stellar discs was applied to the double nucleus of m31 @xcite and even the case of milky way was discussed recently @xcite . indeed ,
as indicated also in this paper , there is a broad variety of reasons why the compact nuclear cluster should acquire a non - spherical structure in the course of evolution @xcite .
the effects exhibited by self - gravitating accretion flows are merely one ingredient of a complex problem .
gravitational field of black - hole discs induces great diversity of effects and gives ample scope for new ideas , of which we could touch only very limited section .
first of all , as noted in the first part of this paper , even some very basic questions remain still open within the framework of exact general relativistic solutions for the black hole disc gravitational field .
our interest in global gravitational effects connected with astrophysical discs has lead us to focus on systems around super - massive bhs that reside in nuclei of galaxies , in which case the onset of self - gravity appears inevitable .
recent ideas of heavy tori being formed in certain moments of neutron - star binary coalescence make these topics fitting also to stellar mass objects , although their `` discs '' must be very different in many respects . because of expected instabilities , accreting bhs are pertinent also for forthcoming gravitational wave experiments .
astrophysical black holes are supposed to be embedded in complex cosmic environment which consists of multiple sub - systems spanning enormous range in the parameter space of masses , lengths and time - scales .
gaseous discs represent just one of the components in this formation . under such circumstances ,
analytical means can offer insight into physical mechanisms which are connected with the disc , but specific analytical solutions lack generality and must be inevitably simplified . in recent decades
, much effort has been devoted to clarify , which processes dominate the evolution and what are the applicable ranges of model parameters .
it is quite possible that some regimes of the disc operation have been missed in previous works . in this respect ,
an attractive recent speculation suggests that intermediate mass black holes ( with @xmath378 ) might also exist , possibly in cores of some globular clusters .
if this new family of black holes is revealed and if they also accrete gas from their close environs ( current observational evidence is rather scarce ) , it would provide a fascinating opportunity to study accreting black hole sources for which the dynamical time is of the order of @xmath379 second .
in other words , the expected variability of such object could span over intervals that can be readily sampled over many periods with available technology .
we concentrated our attention on a fairly conventional case of gas discs around super - massive black holes at small , sub - parsec length - scales where the motion of stellar - mass orbiters can serve as an accurate tool to probe the system .
this idea requires various intervening influences ( such as the perturbation of the central gravitational field caused by the disc / torus own gravity , or the orbital dissipation in the gaseous interstellar environment ) to be understood with sufficient accuracy .
interesting and fundamental questions , namely , whether and where exactly the onset of self - gravity occurs in real accretion discs , still remain open to further research .
we are grateful to j miller for very helpful discussions and criticism of the first version of the manuscript .
we thank j bik for his support , in various ways , of this work .
we also very much appreciate comments and advice on different points from m abramowicz , s collin , b czerny , p harmanec , a martocchia , g matt , j palou , l rezzolla , l ubr , d vokrouhlick and participants of the prague relativity seminar .
we acknowledge support under grants gacr 202/02/0735 , gacr 205/03/0902 and gauk 188/2001
. part of this paper was written during vk s visit to the ukaff at the university of leicester .
999 abramowicz m a , arkuszewski w and muchotrzeb b 1976 pressureless disk can not be the source of the kerr metric _ lett .
nuovo cim . _ * 15 * 476 abramowicz m a , calvani m and nobili l 1983 runaway instability in accretion disks orbiting black holes _ nature _ * 302 * 597 abramowicz m a , curir a , schwarzenberg - czerny a and wilson r e 1984 self - gravity and global structure of accretion discs _ mon . not .
r. astron .
soc . _ * 208 * 279 abramowicz a , henderson f and ghosh p 1983 evolution of accretion tori orbiting black holes .
mon . not .
r. astron .
* 203 * 323 abramowicz m , jaroszynski m and sikora m 1978 relativistic accreting disks _ astron .
astrophys . _ * 63 * 221 abramowicz m a , karas v and lanza a 1998 on the runaway instability of relativistic tori _ astron .
astrophys . _ * 331 * 1143 abramowicz m a and percival m j 1997 thermal equilibria of accretion discs around black holes : a review of recent progress _ class .
quantum grav . _
* 14 * 2003 abrassart a and czerny b 2000 toy model of obscurational variability in active galatic nuclei _ astron .
_ * 356 * 475 alekseev g a 2001 new integral equation form of integrable reductions of einstein equations _ theor . math .
phys . _ * 129 * 1466 alekseev g a and garcia a a 1996 schwarzschild black hole immersed in a homogeneous electromagnetic field _
d _ * 53 * 1853 alexander t 1999 the distribution of stars near the supermassive black hole in the galactic center _ astrophys .
j. _ * 527 * 835 amendt p , lanza a and abramowicz m a 1989 baroclinic toroidal configurations of equilibrium _ astrophys .
j. _ * 343 * 437 andalib s w , tohline j e and christodoulou d m 1997 a survey of the principal modes of nonaxisymmetric instability in self - gravitating accretion disk models _ astrophys .
j. suppl . _
* 108 * 471 ansorg m 2001 differentially rotating disks of dust : arbitrary rotation law _ gen .
grav . _ * 33 * 309 ansorg m. , kleinwchter a and meinel r 2003 uniformly rotating axisymmetric fluid configurations bifurcating from highly flattened maclaurin spheroids _ mon . not .
r. astron .
soc . _ * 339 * 515 ansorg m , kleinwchter a and meinel r 2003 relativistic dyson rings and their black hole limit _ astrophys . j. _ * 582 * l87 ansorg m , kleinwchter a , meinel r and neugebauer g 2002 dirichlet boundary value problems of the ernst equation _ phys . rev .
d _ * 65 * 044006 aretxaga i 2003 line formation in the inner starburst regions of agn , in _ winds , bubbles and explosions _ ed arthur j and henney w _ rev .
astrophys . _
* 15 * 329 armitage p j , livio m , lubow s h and pringle j e 2002 predictions for the frequency and orbital radii of massive extrasolar planets _ mon . not .
r. astron .
* 334 * 248 armitage p j , zurek w h and davies m b 1996 red giant
disk encounters : food for quasars ?
_ astrophys .
j. _ * 470 * 237 artymowicz p , lin d n c and wampler e j 1993 star trapping and metallicity enrichment in quasars and active galactic nuclei _ astrophys . j. _ * 409 * 592 bahcall j n and wolf r a 1976 star distribution around a massive black hole in a globular cluster _ astrophys .
j. _ * 209 * 214 bao g , hadrava p and stgaard e 1996 line profiles from relativistic eccentric rings _ astrophys .
j. _ * 464 * 684 bao g , wiita p j and hadrava p 1998 general relativistic effects on the spectrum reflected by accretion disks around black holes _ astrophys
. j. _ * 504 * 58 barbour j and pfister h ( ed ) 1995 _ mach s principle from newton s bucket to quantum gravity _
( boston : birkhauser ) bardeen j m 1973 timelike and null geodesics in the kerr metric , in _ black holes _ ed dewitt c and dewitt b s ( new york : gordon and breach ) p 215 bardeen j m 1973 rapidly rotating stars , disks , and black holes , in _ black holes _
ed dewitt c and dewitt b s ( new york : gordon and breach ) p 241 bardou a , heyvaerts j and duschl w 1998 influence of viscosity laws on the transition to the self - gravitating part of accretion disks _ astron .
astrophys . _
* 337 * 966 barrabs c and israel w 1991 thin shells in general relativity and cosmology : the lightlike limit _ phys .
d _ * 43 * 1129 baxandall f e , newall h f and stratton f j m 1930 the spectrum of @xmath2 lyrae _ ann .
solar phys .
cambridge _ * 2 * 1 begelman m c and rees m j 1978 the fate of dense stellar systems _ mon . not .
r. astron .
soc . _ * 185 * 847 beig r and schmidt b 2000 time - independent gravitational fields , in _
einstein s field equations and their physical implications _ ,
notes phys .
540 , ed schmidt b g ( berlin : springer ) p 325 bik j 1979 on the theories of the interacting perturbations of the reissner - nordstrm black hole
b _ * 29 * 945 bik j 2000 selected solutions of einstein s field equations : their role in general relativity and astrophysics , in _ einstein s field equations and their physical implications _ , lect .
notes phys .
540 , ed schmidt b g ( berlin : springer ) p 1 bik j and ledvinka t 1993 relativistic disks as sources of the kerr metric _ phys .
* 71 * 1669 bik j and ledvinka t 2000 electromagnetic fields around black holes and meissner effect _
nuovo cim .
b _ * 115 * 739 bik j , lynden - bell d and katz j 1993 relativistic disks as sources of static vacuum spacetimes _ phys
d _ * 47 * 4334 binney j and tremaine s 1987 _ galactic dynamics _
( princeton : princeton univ .
press ) blaes o m 1987 stabilization of non - axisymmetric instabilities in a rotating flow by accretion on to a central black hole _ mon . not .
r. astron .
soc . _ * 227 * 975 bodo g and curir a 1992 models of self - gravitating accretion disks _ astron .
astrophys . _ * 253 * 318 bonazzola s , gourgoulhon e and marck j - a 1998 numerical approach for high precision 3d relativistic star models _ phys .
d _ * 58 * 104020 bondi h 1952 on spherically symmetrical accretion _ mon . not .
r. astron .
soc . _ * 112 * 195 bonnor w b 1992 physical interpretation of vacuum solutions of einstein s equations .
part i. time - independent solutions _ gen .
* 24 * 551 bonnor w b , griffiths j b and maccallum m a h 1994 physical interpretation of vacuum solutions of einstein s equations .
time - dependent solutions _ gen .
* 26 * 687 boyarchuk a a , bisikalo d v , kuznetsov o a and chechetkin v m 2002 _ mass transfer in close binary stars _ , advances in astronomy and astrophysics 6 ( london : taylor & francis ) boyer r h 1965 rotating fluid masses in general relativity _ proc .
cambridge phil .
soc . _ * 61 * 527 bretn n , garca a a , manko v s and denisova t e 1998 arbitrarily deformed kerr - newman black hole in an external gravitational field _ phys .
d _ * 57 * 3382 bromley b c , chen k and miller w a 1997 line emission from an accretion disk around a rotating black hole : toward a measurement of frame dragging _ astrophys
. j. _ * 475 * 57 brower d and clemence g 1961 _ methods of celestial mechanics _ ( new york : academic press ) brown g e , lee c - h , wijers r a m j , lee h k , israelian g and bethe h a 2000 a theory of gamma - ray bursts _ new astronomy _ * 5 * 191 butterworth e m and ipser j r 1976 on the structure and stability of rapidly rotating fluid bodies in general relativity .
i. the numerical method for computing structure and its application to uniformly rotating homogeneous bodies _ astrophys . j. _ * 204 * 200 campanelli m and lousto c o 1999 second order invariant gravitational perturbations of a kerr black hole _ phys .
d _ * 59 * 124022 cannizzo j
k and reiff c m 1992 accretion disks in active galactic nuclei vertically averaged models _
. j. _ * 385 * 87 carter b 1973 black hole equilibrium states , in _ black holes _ ed dewitt c and dewitt b s ( new york : gordon and breach ) p 57 carter b 1979 the general theory of the mechanical , electromagnetic and thermodynamic properties of black holes , in _ general relativity , an einstein centenary survey _ ed hawking s w and israel w ( cambridge : cambridge univ . press ) p 294 castro - tirado a j , greiner j and paredes j m ( ed ) 2001 _ microquasars _ , astrophys .
space sci .
276 ( dordrecht : kluwer ) chakrabarti s k 1988 spacetime with self - gravitating thick disc _ j. astrophys .
_ * 9 * 49 chandrasekhar s 1969 _ ellipsoidal figures of equilibrium _
( new haven : yale university press ) chandrasekhar s 1978 the kerr metric and stationary axisymmetric gravitational fields _ proc . r. soc .
london a _ * 358 * 405 chandrasekhar s 1979 an introduction to the theory of the kerr metric and its perturbations , in _ general relativity , an einstein centenary survey _ ed hawking s w and israel w ( cambridge : cambridge univ . press ) p 370 chandrasekhar s 2000 _ the mathematical theory of black holes _ ( new york : oxford univ .
press ) chaudhuri s and das k c 1997 two - soliton solutions of axially symmetric metrics _ gen .
* 29 * 75 chaudhuri s and das k c 1997 axially symmetric metrics from laplace s seed by inverse scattering method _ j. math
. phys . _ * 38 * 5792 cheng k s and wang j - m 1999 _ astrophys .
j. _ * 521 * 502 chinea f j and gonzles - romero l m ( ed ) 1993 _ rotating objects and relativistic physics _ , lect .
notes phys .
423 ( berlin : springer ) chrzanowski p l 1975 vector potential and metric perturbations of a rotating black hole _
d _ * 11 * 2042 chrzanowski p l 1976 applications of metric perturbations of a rotating black hole : distortion of the event horizon _ phys .
* 13 * 806 ciufolini i and wheeler j a 1995 _ gravitation and inertia _ ( new jersey : princeton univ . press ) clarke c j s 1993 _ the analysis of space - time singularities _ , cambridge lect .
notes phys . 1 ( cambridge : cambridge univ .
press ) clment m j 1974 on the solution of the poisson equation for rapidly rotating stars _ astrophys .
j. _ * 194 * 709 clment g 2000 generating rotating einstein - maxwell fields _ annalen phys . _ * 9 * si-42 colgate s a 1967 stellar coalescence and the multiple supernova interpretation of quasi - stellar sources _ astrophys .
j. _ * 150 * 163 collin s , boisson c , mouchet m , dumont a - m , coup s _ et al _ 2002 are quasars accreting at super - eddington rates ?
. astrophys .
* 388 * 771 collin s and hur j - m 1999 influence of the metallicity and of the irradiation on the structure of accretion disks around massive black holes _ astron .
astrophys . _
* 341 * 385 collin s and hur j - m 2001 size - mass - luminosity relations in agn and the role of the accretion disc _ astron .
astrophys . _ * 372 * 50 collin s and zahn j 1999 star formation and evolution in accretion disks around massive black holes _ astron .
astrophys . _
* 344 * 433 collin s and zahn j 1999 star formation in accretion disks around massive black holes and pregalactic enrichment _ astrophys .
space sci . _ * 265 * 501 collin - souffrin s and dumont a - m 1990 line and continuum emission from the outer regions of accretion discs in active galactic nuclei .
ii radial structure of the disc _ astron .
astrophys . _ * 229 * 292 connors p a , stark r f and piran t 1980 polarization features of x - ray radiation emitted near black holes _ astrophys . j. _ * 235 * 224 cosgrove c m 1980 relations between the group - theoretic and soliton - theoretic techniques for generating stationary axisymmetric gravitational solutions
_ j. math .
* 21 * 2417 cosgrove c m 1981 bcklund transformations in the hauser - ernst formalism for stationary axisymmetric spacetimes _ j. math .
phys . _ * 22 * 2624 cosgrove c m 1982 relationship between the inverse scattering techniques of belinskii - zakharov and hauser - ernst in general relativity _ j. math .
* 23 * 615 cunningham c t 1975 the effects of redshifts and focusing on the spectrum of an accretion disk around a kerr black hole
_ astrophys .
j. _ * 202 * 788 cunningham c t and bardeen j m 1972 the optical appearance of a star orbiting an extreme kerr black hole _ astrophys .
j. _ * 173 * l137 czerny b 2003 comparison of gbh and agn accretion flows , in _ active galactic nuclei : from central engine to host galaxy _ , asp conf .
series 290 , ed collin s , combes f and shlosman i ( san francisco : astronomical society of the pacific ) p 59 daigne f and mochkovitch r 1997 gamma - ray bursts and the runaway instability of thick discs around black holes _ mon . not .
r. astron .
* 285 * l15 damour t , soffel m and xu c 1991 general - relativistic celestial mechanics .
i. method and definition of reference systems _ phys .
d _ * 43 * 3273 damour t , soffel m and xu c 1992 general - relativistic celestial mechanics .
translational equations of motion _ phys .
d _ * 45 * 1017 de andrade v c , blanchet l and faye g 2001 third post - newtonian dynamics of compact binaries : noetherian conserved quantities and equivalence between the harmonic - coordinate and adm - hamiltonian formalisms _ class .
quantum grav . _ * 18 * 753 demianski m 1976 stationary axially symmetric perturbations of a rotating black hole _ gen . rel . grav . _ * 7 * 551 de villiers j - p and hawley j f 2003 global general relativistic magnetohydrodynamic simulations of accretion tori _ astrophys
. j. _ * 592 * 1060 done c 2002 accretion flows in x - ray binaries _ phil .
a ) _ * 360 * 1967 duschl w j , stittmatter p
a and biermann p l 2000 a note on hydrodynamic viscosity and selfgravitation in accretion disks _ astron .
astrophys . _ * 357 * 1123 dyson f w 1893 the potential of an anchor ring _
a ) _ * 184 * 43 ellis g f
r and schmidt b g 1977 singular space - times _ gen .
* 8 * 915 elmegreen b 2002 star formation from galaxies to globules _ astrophys . j. _ * 577 * 206 embacher f 1984 the analog of electric and magnetic fields in stationary gravitational systems _ found .
* 14 * 721 ernst f j 1968 new formulation of the axially symmetric gravitational field problem _ phys .
d _ * 167 * 1175 ; 1968 _ ibid _ * 168 * 1415 evans n w and collett j l 1997 on the origin of cusps in stellar systems _ astrophys . j. _ * 480 * l103 evans n w and de zeeuw p t 1992 potential - density pairs for flat galaxies _ mon . not .
r. astron .
soc . _ * 257 * 152 fabian a c 1979 theories of the nuclei of active galaxies _ proc .
london a _ * 366 * 449 fabian a c , iwasawa k , reynolds c s and young a j 2000 broad iron lines in active galactic nuclei _ publ .
pacific _ * 112 * 1145 fabian a c , rees m j , stella l and white n e 1989 x - ray fluorescence from the inner disc in cygnus x-1 _ mon . not .
r. astron .
soc . _ * 238 * 729 fabian a c , vaughan s , nandra k , iwasawa k , ballantyne d r _ et al _ 2002 a long hard look at mcg6 - 30 - 15 with xmm - newton _ mon . not .
r. astron .
soc . _ * 335 * l1 fanton c , calvani m , de felice f and cadez a 1997 detecting accretion disks in active galactic nuclei _ publ .
* 49 * 159 fernandes j f q and lun a w c 1996 gauge invariant perturbations of black holes . i. schwarzschild space - time _ j. math .
phys . _ * 37 * 836 fernandes j f q and lun a w c 1997 gauge invariant perturbations of black holes . i. kerr space - time _ j. math .
phys . _ * 38 * 330 ferrarese l and merritt d 2000 a fundamental relation between supermassive black holes and their host galaxies _ astrophys .
j. _ * 539 * l9 fishbone l g and moncrief v 1976 relativistic fluid disks in orbit around kerr black holes _ astrophys .
j. _ * 207 * 962 font j a and daigne f 2002 the runaway instability of thick discs around black holes i. the constant angular momentum case _ mon . not .
r. astron .
soc . _ * 334 * 383 font j a and daigne f 2003 on the stability of thick accretion disks around black holes _ astrophys .
j. _ * 581 * l23 font j a and ibez j m 1998 non - axisymmetric relativistic bondi - hoyle accretion onto a schwarzschild black hole _ mon . not .
r. astron .
_ * 298 * 835 frank j , king a and raine d 2002 _ accretion power in astrophysics _ ( cambridge : cambridge univ . press ) frank j and rees m j 1976 effects of massive central black holes on dense stellar systems _ mon . not .
r. astron .
soc . _ * 176 * 633 frauendiener j and klein c 2001 exact relativistic treatment of stationary counterrotating dust disks : physical properties _ phys
d _ * 63 * 084025 freitag m and benz w 2002 a new monte carlo code for star cluster simulations .
central black hole and stellar collisions _ astron .
astrophys . _ * 394 * 345 frolov v p and novikov i d 1998 _ physics of black holes . basic concepts and new developments _ ( dordrecht : kluwer ) fukuda h , habe a and wada k 2000 the effect of the self - gravity of gas on gas fueling in a barred galaxy with a supermassive black hole
_ astrophys .
j. _ * 529 * 109 fukue j and sakamoto c 1992 vertical structures of self - gravitating gaseous disks around a central object _ publ .
. japan _ * 44 * 553 gammie c f 2001 nonlinear outcome of gravitational instability in cooling , gaseous disks _ astrophys .
j. _ * 553 * 174 gebhardt k , bender r , bower g , dressler a , faber s m _ et al _ 2000 a relationship between nuclear black hole mass and galaxy velocity dispersion _ astrophys .
j. _ * 539 * l13 ; 2001 _ ibid _ * 555 * l75 gleiser r j , nicasio c o , price r h and pullin j 2000 gravitational radiation from schwarzschild black holes : the second - order perturbation problem _ phys .
rep . _ * 325 * 41 goldreich p and lynden - bell d 1965 gravitational stability of uniformly rotating disks _ mon . not .
r. astron .
soc . _ * 130 * 97 goldreich p and sari r 2002 eccentricity evolution for planets in gaseous disks _ astrophys .
j. _ * 585 * 1024 goldreich p and tremaine s 1980 disk - satellite interactions _ astrophys . j. _ * 241 * 425 gonzlez g a and letelier p s 2000 rotating relativistic thin disks _ phys .
* 62 * 064025 gonzles g a and letelier p s 2004 exact general relativistic thick disks _ phys .
d _ in print goodman j 2003 self - gravity and quasi - stellar object discs _ mon . not . r. astron .
soc . _ * 339 * 937 goodman j and narayan r 1988 the stability of accretion tori .
iii the effect of self - gravity _ mon . not .
r. astron .
soc . _ * 231 * 97 greenberg r and brahic a 1984 _ planetary rings _
( tuscon : univ .
arizona press ) gutsunayev ts i and manko v s 1985 on the gravitational field of a mass possessing a multipole moment _ gen .
* 17 * 1025 gutsunaev ts i and manko v s 1988 on a family of solutions of the einstein - maxwell equations _ gen .
grav . _ * 20 * 327 hachisu i 1986 a versatile method for obtaining structures of rapidly rotating stars _ astrophys
. j. supp . _
* 61 * 479 ; 1986 _ ibid _ * 62 * 461 hachisu i and eriguchi y 1983 bifurcations and phase transitions of self - gravitating and uniformly rotating fluid _ mon . not .
r. astron .
soc . _ * 204 * 583 hachisu i , eriguchi y and nomoto k 1986 fate of merging double white dwarfs _ astrophys .
j. _ * 308 * 161 hagio f 1987 angular momentum transport by star - gas interaction and structure of accretion disks _ publ .
. japan _ * 39 * 887 hameury j - m , king a r , lasota j - p and auvergne m 1994 black - hole stellar accretion in active galactic nuclei _ astron .
astrophys . _ * 292 * 404 hartnoll s a and blackman e g 2002 iron line profiles from black hole accretion discs with spiral velocity structure _ mon . not .
r. astron .
soc . _ * 332 * l1 hashimoto m , eriguchi y and mller e 1995 equilibrium structure of self - gravitating keplerian disks _ astron . astrophys . _
* 297 * 135 hawley j f , balbus s a and winters w f 1999 local hydrodynamic stability of accretion disks _ astrophys .
j. _ * 518 * 394 heller c h and shlosman i 1994 fueling nuclear activity in disk galaxies : starbursts and monsters _ astrophys .
j. _ * 424 * 84 hernquist l 1990 an analytical model for spherical galaxies and bulges _ astrophys .
j. _ * 356 * 359 hockney r w and eastwood j w 1988 _ computer simulation using particles _
( bristol : hilger ed . )
hoenselaers c and dietz w ( ed ) 1984 _ solutions of einstein s equations : techniques and results _ , lect .
notes phys 205 ( berlin : springer ) holman m , touma j and tremaine s 1997 chaotic variations in the eccentricity of the planet orbiting 16 cyg b _ nature _ * 386 * 254 hoyle f and lyttleton r a 1939 the evolution of stars _ proc .
soc . _ * 35 * 592 hoyle f and lyttleton r a 1941 on the accretion theory of stellar evolution _ mon . not .
r. astron .
* 101 * 227 * 35 * 592 huben i , harmanec p and shore s n 1994 on the strategy of future observations and modeling of the beta lyr system _ astron .
astrophys . _ * 289 * 411 hughes s a 2000 evolution of circular , nonequatorial orbits of kerr black holes due to gravitational radiation _ phys .
d _ * 61 * 084004 ; 2001 _ ibid _ * 63 * 064016 ; 2001 _ ibid _ * 64 * 064004 hunter c 1963 the structure and stability of self - gravitating disks _ astrophys . j. _ * 126 * 299 hur j - m 1998 properties of self - gravitating @xmath221-discs revisited .
general scaling laws _ astron .
astrophys . _ * 337 * 649 hur j - m 2000 on the transition to self - gravity in low mass agn and yso accretion discs _ astron .
astrophys . _ * 358 * 378 hur j - m 2002 origin of non - keplerian motions of masers in ngc1068 _ astron .
astrophys . _
* 395 * l2 hur j m , collin - souffrin s , le bourlot j and pineau des forts g 1994 structure of the outer regions of accretion discs in active galactic nuclei _ astron . astrophys . _ * 290 * 19 ibez m h and sigalotti l 1984 the equilibrium of polytropic self - gravitating sheetlike masses _ astrophys .
j. _ * 285 * 784 igumenshchev i v , chen xingming and abramowicz m a 1996 accretion discs around black holes : two - dimensional , advection - cooled flows _ mon . not .
r. astron .
soc . _ * 278 * 236 igumenshchev i v , narayan r and abramowicz m a 2003 three - dimensional magnetohydrodynamic simulations of radiatively inefficient accretion flows _ astrophys .
j. _ * 592 *
1042 islam j n 1985 _ rotating fields in general relativity _ ( cambridge : cambridge univ . press ) israel w 1966 hypersurfaces and thin shells in general relativity _
nuovo cim .
b _ * 44 * 1 ivanov p b , igumenshchev i v and novikov i d 1998 hydrodynamics of black hole - accretion disk collision _ astrophys . j. _ * 507 * 131 iwasawa k , fabian a c , reynolds c s , nandra k , otani c _ et al _ 1996 the variable iron k emission line in mcg6 - 30 - 15 _ mon . not .
r. astron .
soc . _ * 282 * 1038 jaffe w 1983 a simple model for the distribution of light in spherical galaxies _ mon . not .
r. astron .
soc . _ * 202 * 995 jantzen r t , carini p and bini d 1992 the many faces of gravitoelectromagnetism _ annals phys . _
* 215 * 1 karas v , lanza a and vokrouhlick d 1995 emission - line profiles from self - gravitating thin disks _ astrophys .
j. _ * 440 * 108 karas v , martocchia a and ubr l 2001 variable line profiles due to non - axisymmetric patterns in an accretion disc around a rotating black hole _
. japan _ * 53 * 189 karas v and ubr l 2001 orbital decay of satellites crossing an accretion disc _ astron .
astrophys . _ * 376 * 686 karas v , ubr l and lechta m 2002 stellar dynamics in a galactic centre surrounded by an accretion disc , in _ gravitation : following the prague inspiration _ ed semerk o , podolsk j and ofka m ( singapore : world scientific ) p 85 karas v and vokrouhlick d 1994 relativistic precession of the orbit or a star near a supermassive black hole _ astrophys
. j. _ * 422 * 208 kato s 2001 basic properties of thin - disk oscillations _ publ .
. japan _ * 53 * 1 kato s , fukue j and mineshige s 1998 _ black - hole accretion disks _ ( kyoto : kyoto univ . press ) kegeles l s and cohen j m 1979 constructive procedure for perturbations of spacetimes _ phys . rev
. d _ * 19 * 1641 khanna r and chakrabarti s k 1992 effects of a self - gravitating disc on test particle motion around a kerr black hole _ mon . not .
r. astron .
soc . _ * 259 * 1 kim e einsel c lee h m spurzem r and lee m g 2002 dynamical evolution of rotating stellar systems
post - collapse , equal - mass system _ mon . not .
r. astron .
soc . _ * 334 * 310 king a r 2003 accretion in compact binaries , in _ compact stellar x - ray sources _ ed lewin w h g and van der klis m ( cambridge : cambridge univ .
press ) in print king i r 1966 the structure of star clusters .
iii . some simple dynamical models _ astron . j. _ * 71 * 64 king m and pfister h 2001 electromagnetic thirring problems _ phys .
d _ * 63 * 104004 klein c 1997 counter - rotating dust rings around a static black hole _ class .
quantum grav . _
* 14 * 2267 klein c 2003 exact relativistic treatment of stationary black - hole disk systems _
d _ * 68 * 027501 klein c and richter o 1998 physically realistic solutions to the ernst equation on hyperelliptic riemann surfaces _ phys .
d _ * 58 * 124018 kopal z 1978 _ dynamics of close binary systems _ astrophys .
space sci .
library 68 ( dordrecht : d reidel ) kordas p 1999 aspects of solution - generating techniques for space - times with two commuting killing vectors _ gen .
grav . _ * 31 * 1941 korotkin d 1997 some remarks on finite - gap solutions of the ernst equation _ phys .
a _ * 229 * 195 korotkin d and matveev v 2000 _ funct .
appl . _ * 34 * 1 kozai y 1962 secular perturbations of asteroids with high inclination and eccentricity _ astron .
j. _ * 67 * 591 kozowski m , jaroszynski m and abramowicz m a 1978 the analytic theory of fluid disks orbiting the kerr black hole _ astron .
astrophys . _ * 63 * 209 kozowski m , wiita p j and paczyski b 1979 self - gravitating accretion disk models with realistic equations of state and opacities _ acta astron . _ * 29 * 157 krivan w and price r h 1998 initial data for superposed rotating black holes _ phys . rev
. d _ * 58 * 104003 krolik j h 1999 _ active galactic nuclei _
( princeton : princeton univ . press )
kumar p 1999 the structure of the central disk of ngc 1068 : a clumpy disk model _ astrophys .
j. _ * 519 * 599 lantian y and xiaoci l 1990 radial self - gravity of accretion disc around supermassive black holes _ astrophys .
space sci . _ * 172 * 293 lanza a 1992 self - gravitating thin disks around rapidly rotating black holes _ astrophys .
j. _ * 389 * 141 laor a and netzer h 1989 massive thin accretion discs .
i calculated spectra _ mon . not .
r. astron .
soc . _ * 238 * 897 larwood j d 1997 _ the tidally induced warping , precession and truncation of accretion discs _ , phd thesis ( queen mary and westfield college , london ) lauer t r , gebhardt k , richstone d , tremaine s , bender r _ et al _ 2002 galaxies with a central minimum in stellar luminosity density _ astron j. _ * 124 * 1975 laughlin g , korchagin v and adams f c 1998 the dynamics of heavy gaseous disks _ astrophys .
j. _ * 504 * 945 laughlin g and ryczka m 1996 the effect of gravitational instabilities on protostellar disks _ astrophys .
j. _ * 456 * 279 ledvinka t 1998 _ thin disks as sources of stationary axisymmetric electrovacuum spacetimes _ , phd thesis ( charles univ . , prague )
lehner l 2001 numerical relativity : a review _ class .
quantum grav . _ * 18 * r25 leighly k m 1999 a comprehensive spectral and variability study of narrow - line seyfert 1 galaxies observed by asca _ astrophys .
j. supp . _
* 125 * 297 ; 1999 _ ibid _ 125 317 lemaitre a and dubru p 1991 secular resonances in the primitive solar nebula _ celestial mechanics and dynamical astronomy _ * 52 * 57 lemos j p s and letelier p s 1994 exact general relativistic thin disks around black holes _ phys .
d _ * 49 * 5135 letelier p s 1989 on soliton solutions to the vacuum einstein equations obtained from a general seed solution _ class .
quantum grav . _ * 6 * 875 letelier p s 1999 on the gravitational field of static and stationary axial symmetric bodies with multi - polar structure _ class . quantum grav . _ * 16 * 1207 letelier p s and oliveira s r 1987 exact self - gravitating disks and rings : a solitonic approach _ j. math
_ * 28 * 165 levin y and beloborodov a m 2003 stellar disk in the galactic center : a remnant of a dense accretion disk ? _ astrophys . j. _ * 590 * l33 levy h 1968 classification of stationary axisymmetric gravitational fields _
nuovo cim .
b _ * 56 * 253 lin d n c , artymowicz p and wampler e j 1994 star - disk interaction in quasars and agns , in _ theory of accretion disks 2 _ ed duschl w j , frank j , meyer f , meyer - hofmeister e and tscharnuter w m ( dordrecht : kluwer ) lin d n c and ida s 1997 on the origin of massive eccentric planets _ astrophys . j. _ * 477 * 781 lin d n c and papaloizou j c b 1996 theory of accretion disks ii : application to observed systems _ ann . rev . astron . astrophys . _ * 34 * 703 linet b 1977 stationary axisymmetric test fields on a kerr metric _ phys .
a _ * 60 * 395 liu b f , xie g z and ji k f 1994 theoretical restrictions on accretion disks _ astrophys .
space sci .
_ * 220 * 75 ; 1995 _ ibid _ * 225 * 299 longaretti p y 2002 on the phenomenology of hydrodynamic shear turbulence _
j. _ * 576 * 587 lu y , cheng k s , yang l t and zhang l 2000 a massive thick disc around a massive black hole and its runaway instability _ mon . not .
r. astron .
* 314 * 453 lynden - bell d 1969 galactic nuclei as collapsed old quasars _ nature _ * 223 * 690 lynden - bell d , bik j and katz j 1999 on accelerated inertial frames in gravity and electromagnetism _ annals phys . _ * 271 * 1 lynden - bell d , katz j and bik j 1995 mach s principle from the relativistic constraint equations _ mon . not .
r. astron .
soc . _ * 272 * 150 lynden - bell d and pringle j 1974 the evolution of viscous discs and the origin of the nebular variables _ mon . not .
r. astron .
soc . _ * 168 * 603 macfadyen a i and woosley s e 1999 collapsars : gamma - ray bursts and explosions in `` failed supernovae '' _ astrophys .
j. _ * 524 * 262 maison d 2000 duality and hidden symmetries in gravitational theories , in _
einstein s field equations and their physical implications _ , lect .
notes phys .
540 , ed schmidt b g ( berlin : springer ) p 273 manko v s 1992 the exterior gravitational field of a static and stationary mass with an arbitrary set of multipole moments _ gen .
* 24 * 35 manko v s and novikov i d 1992 generalizations of the kerr and kerr - newman metrics possessing an arbitrary set of mass - multipole moments _ class . quantum grav . _
* 9 * 2477 marsh t r and horne k 1988 images of accretion discs .
ii doppler tomography _ mon . not . r. astron .
soc . _ * 235 * 269 martocchia a 2000 _ x - ray spectral signatures of accreting black holes _ ,
phd thesis ( sissa , trieste ) martocchia a , karas v and matt g 2000 effects of kerr space - time on spectral features from x - ray illuminated accretion discs _ mon . not .
r. astron .
soc . _ * 312 * 817 martocchia a , matt g , karas v , belloni t and feroci m 2002 evidence for a relativistic iron line in grs 1915 + 105 _ astron .
astrophys . _ * 387 * 215 martocchia a , matt g , karas v , belloni t and feroci m 2003 iron line diagnostics for the grs 1915 + 105 black hole , in _ the restless high energy universe _ , proc . of the 2nd bepposax meeting , ed van den heuvel e p j , in t zand j j m and wijers r a m j , in press mashhoon b 1985 gravitational effects of rotating masses _ found .
* 15 * 497 masuda n and eriguchi y 1997 three - dimensional simulations of runaway instability of self - gravitating accretion disks _ astrophys . j. _ * 489 * 804 masuda n , nishida s and eriguchi y 1998 the runaway instability of self - gravitating tori with non - constant specific angular momentum around black holes _ mon . not .
r. astron .
soc . _ * 297 * 1139 matt g 1993 x - ray polarization properties of a centrally illuminated accretion disc _ mon . not .
r. astron .
soc . _ * 260 * 663 medvedev m v and narayan r 2000 axisymmetric self - similar equilibria of self - gravitating isothermal systems _ astrophys .
j. _ * 541 * 579 meinel r 2000 gravitational field of rotating disks and black holes _
annalen phys .
_ * 9 * 335 merloni a and fabian a c 2003 iron k@xmath221 line profiles and the inner boundary condition of accretion flows _ mon . not .
r. astron .
soc . _ * 342 * 951 mestel l 1963 on the galactic law of rotation _ astrophys .
j. _ * 126 * 553 mszros p 2001 gamma - ray bursts : accumulating afterglow implications , progenitor clues , and prospects _ science _ * 291 * 79 mineshige s and umemura m 1996 self - similar , self - gravitating viscous disks _ astrophys .
j. _ * 469 * l49 ; 1997 _ ibid _ * 480 * 167 mirabel i f 2003 jets in supermassive and stellar - mass black holes , in _ proc .
xxi symp . on relativistic astrophysics _ ed bandiera r , maiolino r and mannucci f ( singapore : world scientific ) miralda - escud j and kollmeier j a 2003 feeding accretion disks in active galactic nuclei with stars crashing on the disk _ astrophys .
j. _ submitted ( astro - ph/0310717 ) miyamoto m and nagai r 1975 three - dimensional models for the distribution of mass in galaxies _ publ .
. japan _ * 27 * 533 morbidelli a 2002 _ modern celestial mechanics : aspects of solar system dynamics _ ( london : taylor & francis ) morgan t and morgan l 1969 the gravitational field of a disk _ phys .
rev . _ * 183 * 1097 ( errata : 1969 _ ibid _ * 188 * 2544 ; 1970 _ ibid d _ * 1 * 3522 ) mller e and steinmetz m 1995 simulating self - gravitating fluids _ computer phys . commun . _ * 89 * 45 nakamura t , oohara k and kojima y 1987 general relativistic collapse to black holes and gravitational waves from black holes _ prog . theor . phys . suppl . _
* 90 * 1 narayan r 2000 hydrodynamic drag on a compact star orbiting a supermassive black hole _
j. _ * 536 * 663 narayan r 2002 why do agn lighthouses switch off ? in _ lighthouses of the universe : the most luminous celestial objects and their use for cosmology _ ed gilfanov m , sunyaev r _
et al _ ( berlin : springer ) p 405 narayan r , paczynski b and piran t 1992 gamma - ray bursts as the death throes of massive binary stars _ astrophys .
j. _ * 395 * l83 nayakshin s and kallman t r 2001 accretion disk models and their x - ray reflection signatures .
i. local spectra _ astrophys .
j. _ * 546 * 406 nelson r p , papaloizou j c b , masset f and kley w 2000 the migration and growth of protoplanets in protostellar discs _ mon . not .
r. astron .
soc . _ * 318 * 18 neugebauer g 1996 gravitostatics and rotating bodies , in _ general relativity , 46th scottish universities summer school in physics _
ed hall g s and pulham j r ( london : iop ) p 61 neugebauer g 2000 rotating bodies as boundary value problems _ annalen phys . _ * 9 * 342 neugebauer g and meinel r 2003 progress in relativistic gravitational theory using the inverse scattering method _ j. math .
phys . _ * 44 * 3407 nishida s and eriguchi y 1994 a general relativistic toroid around a black hole _ astrophys
. j. _ * 427 * 429 nishida s and eriguchi y 1996 runaway instability of black hole neutron torus systems _ astrophys . j. _ * 461 * 320 nishida s , lanza a , eriguchi y and abramowicz m a 1996 runaway instability and gamma - ray bursts _ mon . not .
r. astron .
soc . _ * 278 * l41 nolthenius r a and katz j i 1982 the passage of a star by a massive black hole _ astrophys .
j. _ * 263 * 377 novikov i d , pethick c j and polnarev a g 1992 tidal capture of stars by a massive black hole _ mon . not .
r. astron .
soc . _ * 255 * 276 novikov i d and thorne k s 1973 astrophysics of black holes , in _ black holes _ ed dewitt c and dewitt b s ( new york : gordon and breach ) p 343 ostriker j 1964 the equilibrium of self - gravitating rings _ astrophys . j. _ * 140 * 1067 ostriker j p 1983 viscous drag on an accretion disk due to an embedded stellar system _ astrophys .
j. _ * 273 * 99 paczyski b 1978 a model of selfgravitating accretion disk _ acta astron .
_ * 28 * 91 paczyski b 1978 a model of selfgravitating accretion disk with a hot corona _ acta astron . _ * 28 * 241 paczyski b 1998 are gamma - ray bursts in star - forming regions ? _ astrophys .
j. _ * 494 * l45 paczyski b and wiita p j 1980 thick accretion disks and supercritical luminosities _ astron .
astrophys . _ * 88 * 23 page d n and thorne k s 1974 disk - accretion onto a black hole .
i. time - averaged structure of accretion disks _ astrophys .
j. _ * 191 * 499 pandey u s and gupta a c 1998 equilibria of self - gravitating gaseous disks _ bull .
india _ * 26 * 731 papaloizou j c b and lin d n c 1995 theory of accretion disks i : angular momentum transport processes _ ann .
astrophys . _ * 33 * 505 papaloizou j c b and pringle j e 1984 the dynamical stability of differentially rotating discs with constant specific angular momentum _
mon . not .
r. astron .
soc . _ * 208 * 721 papaloizou j c b and savonije g j 1991 instabilities in self - gravitating gaseous discs _ mon . not .
r. astron .
soc . _ * 248 * 353 papaloizou j c b , terquem c and lin d n c 1998 on the global warping of a thin self - gravitating near - keplerian gaseous disk with application to the disk in ngc4258 _ astrophys .
j. _ * 497 * 212 peebles p j e 1972 star distribution near a collapsed object _ astrophys
. j. _ * 178 * 371 peng h 1983 on calculation of magnetic - type gravitation and experiments _ gen .
grav . _ * 15 * 725 perryman m a c 2000 extra - solar planets _ rep .
phys . _ * 63 * 1209 peterson b m 1997 _ introduction to active galactic nuclei _
( cambridge : cambridge univ . press ) peterson b m 2001 variability of active galactic nuclei , in _
advanced lectures on the starburst - agn connection _ ed aretxaga l , kunth d and mjica r ( singapore : world scientific ) p 3 pichon c and lynden - bell d 1996 new sources for kerr and other metrics : rotating relativistic discs with pressure support _ mon . not .
r. astron .
soc . _ * 280 * 1007 pineault s and landry s 1994 impact of small bodies onto disks around compact objects i. impact satellites _ mon . not .
r. astron .
* 267 * 557 plummer h c 1991 on the problem of distribution in globular star clusters _ mon . not .
r. astron .
soc . _ * 71 * 460 polyachenko v l , polyachenko e v and strelnikov a v 1997 stability criteria for gaseous self - gravitating disks _ astronomy lett . _
* 23 * 483 poutanen j 2001 understanding spectral variability and time lags in accreting black holes _ advances in space research _ * 28 * 267 pringle j e 1981 accretion discs in astrophysics _ ann .
astrophys . _
* 19 * 137 pringle j e 1996 self - induced warping of accretion discs _ mon . not .
r. astron .
soc . _ * 281 * 357 ; 1997 _ ibid _ * 292 * 136 pruet j , woosley s e and hoffman r d 2002 nucleosynthesis in gamma ray burst accretion disks _ astrophys .
j. _ * 586 * 1254 punsly b 2001 _ black hole gravitohydrodynamics _
( berlin : springer ) quevedo h 1992 generating solutions of the einstein - maxwell equations with prescribed physical properties _ phys .
d _ * 45 * 1174 quevedo h and mashhoon b 1991 generalizations of kerr spacetime _ phys .
d _ * 43 * 3902 quinlan g d , hernquist l and sigurdsson s 1995 models of galaxies with central black holes : adiabatic growth in spherical galaxies _ astrophys . j. _ * 440 * 554 quinlan g d and shapiro s l 1987 the collapse of dense star clusters to supermassive black holes binaries and gravitational radiation _ astrophys . j. _ * 321 * 199 quinlan g d and shapiro s l 1990 the dynamical evolution of dense star clusters in galactic nuclei _ astrophys . j. _ * 356 * 483 rauch k p 1995 dynamical evolution of star clusters around a rotating black hole with an accretion disc _ mon . not .
r. astron .
soc . _ * 275 * 628 rauch k p 1999 collisional stellar dynamics around massive black holes in active galactic nuclei _ astrophys .
j. _ * 514 * 725 rauch k p and tremaine s 1996 resonant relaxation in stellar systems _ new astronomy _ * 1 * 149 reynolds c s 1996 _ x - ray emission and absorption in active galaxies _ , phd thesis ( cambridge univ . ,
cambridge ) reynolds c s and nowak m a 2003 fluorescent iron lines as a probe of astrophysical black hole systems _ phys .
_ * 377 * 389 rice w m , armitage j , bate r and bonnell a 2003 the effect of cooling on the global stability of self - gravitating protoplanetary discs _ mon . not .
r. astron .
soc . _ * 339 * 1025 richard d and zahn j - p 1999 turbulence in differentially rotating flows . what can be learned from the couette - taylor experiment _ astron .
astrophys . _ * 347 * 734 rosswog s , davies m b , thielemann f - k and piran t 2000 merging neutron stars : asymmetric systems _ astron .
astrophys . _ * 360 * 171 raska a , dumont a - m , czerny b and collin s 2002 the structure and radiation spectra of illuminated accretion discs in active galactic nuclei i. moderate illumination _ mon . not .
r. astron .
soc . _ * 332 * 799 ruffert m and janka h - t 1999 gamma - ray bursts from accreting black holes in neutron star mergers _ astron .
astrophys . _ * 344 * 573 sakimoto p j and coroniti f v 1981 accretion disk models for qsos and active galactic nuclei the role of magnetic viscosity _ astrophys .
j. _ * 247 * 19 salow r m and statler t s 2001 eccentric nuclear disks with self - gravity : predictions for the double nucleus of m31 _ astrophys . j. _ * 551 * l49 sarbach o , heusler m and brodbeck o 2001 self - adjoint wave equations for dynamical perturbations of self - gravitating fields _ phys .
d _ * 63 * 104015 schinnerer e , eckart a and tacconi l j 2001 the nuclear stellar cluster in the seyfert 1 galaxy ngc 3227 : high angular resolution near - infrared imaging and spectroscopy _ astrophys .
j. _ * 549 * 254 schmid - burgk j 1967 finite amplitude density variations on a self - gravitating isothermal gas layer _ astrophys .
j. _ * 149 * 727 semerk o 2001 curvature singularity around first morgan morgan disc _ class .
quantum grav . _ * 18 * 3589 semerk o 2002 expulsion of external fields from extreme horizons : example of an external gravitational field _ czech . j. phys . _ * 52 * 11 semerk o 2002 towards gravitating discs around rotating black holes , in _ gravitation : following the prague inspiration _ ed semerk o , podolsk j and ofka m ( singapore : world scientific ) p 111 semerk o 2002 thin disc around a rotating black hole , but with support in - between _ class .
quantum grav . _
* 19 * 3829 semerk o 2003 static discs around a schwarzschild black hole : iii _ class
. quantum grav _ * 20 * 1613 semerk o and karas v 1999 pseudo - newtonian models of a rotating black hole field _ astron .
astrophys . _ * 343 * 325 semerk o and ek m 2000 static discs around a schwarzschild black hole : i _ class .
quantum grav . _
* 17 * 1613 semerk o and ek m 2000 oscillations of static disks around schwarzschild black holes : effect of self - gravitation _ publ .
. japan _ * 52 * 1067 semerk o , zellerin t and ek m 1999 the structure of superposed weyl fields _ mon . not .
r. astron .
soc . _ * 308 * 691 ( erratum : 2001 _ ibid _ * 322 * 207 ) semerk o , ek m and zellerin t 1999 test - particle motion in superposed weyl fields _ mon . not .
r. astron .
soc . _ * 308 * 705 shakura
n i and sunyaev r a 1973 black holes in binary systems . observational appearance _ astron
. astrophys . _ * 24 * 337 shapiro s l and marchant a b 1978 star clusters containing massive , central black holes monte carlo simulations in two - dimensional phase space _ astrophys .
j. _ * 225 * 603 shlosman i and begelman m c 1987 self - gravitating accretion disks in active galactic nuclei _ nature _ * 329 * 810 shlosman i and begelman m c 1989 evolution of self - gravitating accretion disks in active galactic nuclei _ astrophys .
j. _ * 341 * 685 shlosman i , begelman m c and frank j 1990 the fuelling of active galactic nuclei _ nature _ * 345 * 679 shore s
n and white r l 1982 self - gravitating accretion disk models for active galactic nuclei : self - consistent @xmath221-models for the broad emission - line region _ astrophys .
j. _ * 256 * 390 sibgatullin n r 1991 _ oscillations and waves in strong gravitational and electromagnetic fields _ ( berlin : springer ) sibgatullin n r , garcia a a and manko v s 2003 density distribution in massive galactic disks with a central black hole _
* 29 * 70 ; 2003 _ ibid _ * 29 * 825 sirko e and goodman j 2003 spectral energy distributions of marginally self - gravitating quasi - stellar object disks _ mon . not .
r. astron .
soc . _ * 341 * 501 smarr l 1973 surface geometry of charged rotating black holes _ phys .
d _ * 7 * 289 smarr l 1973 mass formula for kerr black holes _ phys . rev
. lett . _ * 30 * 71 stahler s w 1983 the equilibria of rotating isothermal clouds .
ii . structure and dynamical stability _ astrophys .
j. _ * 268 * 155 statler t s 2001 a simple family of models for eccentric keplerian fluid disks _ astron .
j. _ * 122 * 2257 stella l 1990 measuring black hole mass through variable line profiles from accretion disks _ nature _ * 344 * 747 stephani h , kramer d , maccallum m , hoenselaers c and herlt e 2003 _ exact solutions to einstein s field equations , second edition _
( cambridge : cambridge univ . press ) strzer h 1993 structure and spectra of accretion disks in the innermost parts of active galaxies _ astron .
astrophys . _
* 271 * 25 ubr l 2001 _ interaction between stars and the environment of galactic nuclei _ , phd thesis ( charles univ . , prague ) ubr l and karas v 1999 an orbiter crossing an accretion disc _ astron .
astrophys . _
* 352 * 452 syer d , clarke c j and rees m j 1991 star disc interactions near a massive black hole _ mon . not .
r. astron .
_ * 250 * 505 takahashi k and lee h m 2000 evolution of multimass globular clusters in the galactic tidal field with the effects of velocity anisotropy _ mon . not .
r. astron .
* 316 * 671 tanabe y 1979 generation of the lewis solutions from the weyl solutions _ j. math .
phys . _ * 20 * 1486 tassoul j - l 1978 _ theory of rotating stars _
( princeton : princeton univ .
press ) tertychniy s i 1990 some representations of the general solution of the linearised vacuum einstein equations for axially symmetric stationary metrics _ class .
quantum grav . _ * 7 * 1345 teukolsky s a 1973 perturbations of a rotating black hole .
i. fundamental equations for gravitational , electromagnetic , and neutrino - field perturbations _ astrophys . j. _ * 185 * 635 teukolsky s a 1998 black hole collisions , toroidal black holes , and numerical relativity , in _ black holes and relativistic stars _ ed wald r m ( chicago : univ .
chicago press ) p 123 thorne k s 1998 probing black holes and relativistic stars with gravitational waves , in _ black holes and relativistic stars _ ed wald r m ( chicago : univ .
chicago press ) p 41 thorne k s , price r h and macdonald d a ( ed ) 1986 _ black holes : the membrane paradigm _
( new haven : yale univ . press ) toomre a 1964 on the gravitational stability of a disk of stars _ astrophys .
j. _ * 139 * 1217 toomre a 1977 theories of spiral structure _ ann .
astrophys . _
* 15 * 437 torres del castillo g f 1990 generation of perturbations by means of decoupled equations and their adjoints _ gen .
* 22 * 1085 torres del castillo g f and cartas - fuentevilla r 1996 scattering by a reissner - nordstrm black hole _ phys .
d _ * 54 * 4886 torres del castillo g f and sols - rodrguez h g 1999 self - duality and gravitational perturbations _ j. math .
_ * 40 * 4099 tsuribe t 1999 self - similar viscous accretion and growth of a central core in self - gravitating disks _ astrophys .
j. _ * 527 * 102 turakulov z ia 1984 equilibrium of collisionless gravitating sphere and disk _ astrophysics _ * 19 * 448 ; also 1989 _ int . j. mod .
a _ * 4 * 3653 ; 1990 _ ibid _ * 5 * 725 usui f , nishida s and eriguchi y 1998 emission - line profiles from self - gravitating toroids around black holes _ mon . not .
r. astron .
soc . _ * 301 * 721 van der klis m 2000 millisecond oscillations in x - ray binaries _ ann . rev
astrophys . _ * 38 * 717 van putten m h p m 2001 gamma - ray bursts : ligo / virgo sources of gravitational radiation _ phys .
rep . _ * 345 * 1 van putten m h p m and levinson a 2003 theory and astrophysical consequences of a magnetized torus around a rapidly rotating black hole _
j. _ * 584 * 937 vein p r 1985 2 related families of determinantal solutions of the stationary axially - symmetric vacuum einstein equations _ class .
quantum grav . _
* 2 * 899 viergutz s u 1993 image generation in kerr geometry .
i. analytical investigations on the stationary emitter - observer problem _ astron .
astrophys . _ * 272 * 355 vilkoviskij e y and czerny b 2002 the role of the central stellar cluster in active galactic nuclei .
i. semi - analytical model _ astron .
astrophys . _
* 387 * 804 vogt d and letelier p s 2003 exact general relativistic perfect fluid disks with halos _ phys .
d _ * 68 * 084010 vokrouhlick d and karas v 1993 a star orbiting around a supermassive rotating black - hole _ mon . not .
r. astron .
* 265 * 365 vokrouhlick d and karas v 1998 stellar dynamics in a galactic centre surrounded by a massive accretion disc _ mon . not .
r. astron .
* 298 * 53 wald r m 1984 _ general relativity _
( chicago : univ .
chicago press ) ward w r 1997 protoplanet migration by nebula tides _ icarus _ * 126 * 261 will c m 1974 perturbation of a slowly rotating black hole by a stationary axisymmetric ring of matter .
i. equilibrium configurations _ astrophys .
j. _ * 191 * 521 will c m 1975 perturbation of a slowly rotating black hole by a stationary axisymmetric ring of matter .
penrose processes , circular orbits , and differential mass formulae _ astrophys .
j. _ * 196 * 41 wilms j , reynolds c s , begelman m c , reeves j , molendi s _ et al _ 2001 xmm - epic observation of mcg-6 - 30 - 15 : direct evidence for the extraction of energy from a spinning black hole ? _ mon . not .
r. astron .
* 328 * l27 wilson r e 1981 equilibrium figures for beta lyrae type disks _ astrophys .
j. _ * 251 * 246 wilson d b 1984 a runaway instability in thick accretion disks ? _ nature _ * 312 * 620 woodward j w , sankaran s and tohline j e 1992 tidal disruption of a star by a massive disk ( the axisymmetric roche problem ) _ astrophys .
j. _ * 394 * 248 wozniak h , combes f , emsellem e and friedli d 2003 numerical simulations of central stellar velocity dispersion drops in disc galaxies _ astron .
astrophys . _ * 409 * 469 yorke h w , bodenheimer p and laughlin g 1993 the formation of protostellar disks .
i. @xmath380 _ astrophys .
j. _ * 411 * 274 young p 1980 numerical models of star clusters with a central black hole .
i adiabatic models _ astrophys . j. _ * 242 * 1232 young a j , fabian a c , ross r r and tanaka y 2001 a complete relativistic ionized accretion disk in cygnus x-1 _ mon . not .
r. astron .
soc . _ * 325 * 1045 ek m and semerk o 2002 static discs around a schwarzschild black hole : ii _ czech .
_ * 52 * 19 zanotti o , rezzolla l and font j a 2003 quasi - periodic accretion and gravitational waves from oscillating ` toroidal neutron stars ' around a schwarzschild black hole _ mon . not .
r. astron .
* 341 * 832 zellerin t and semerk o 2000 two - soliton stationary axisymmetric sprouts from weyl seeds _ class .
quantum grav . _
* 17 * 5103 zhao hongsheng , haehnelt m g and rees m j 2002 feeding black holes at galactic centres by capture from isothermal cusps _ new astronomy _ * 7 * 385 zier c and biermann p l 2001 binary black holes and tori in agn . i. ejection of stars and merging of the binary _ astron .
astrophys . _ * 77 * 23 ; 2002 _ ibid _ * 396 * 91 zurek w h , siemiginowska a and colgate s a 1994 star - disk collisions and the origin of the broad lines in quasars _ astrophys .
j. _ * 434 * 46 ; 1996 _ ibid _ * 470 * 652 | fluid discs and tori around black holes are discussed within different approaches and with the emphasis on the role of disc gravity . first reviewed are the prospects of investigating the gravitational field of a black hole
disc system by analytical solutions of stationary , axially symmetric einstein s equations .
then , more detailed considerations are focused to middle and outer parts of extended disc - like configurations where relativistic effects are small and the newtonian description is adequate . within general relativity , only a static case has been analysed in detail .
results are often very inspiring , however , simplifying assumptions must be imposed : ad hoc profiles of the disc density are commonly assumed and the effects of frame - dragging and completely lacking .
astrophysical discs ( e.g. accretion discs in active galactic nuclei ) typically extend far beyond the relativistic domain and are fairly diluted . however , self - gravity is still essential for their structure and evolution , as well as for their radiation emission and the impact on the environment around .
for example , a nuclear star cluster in a galactic centre may bear various imprints of mutual star
disc interactions , which can be recognised in observational properties , such as the relation between the central mass and stellar velocity dispersion . |
einstein s recognition early last century that gravity can be interpreted as the curvature of space and time represented an enormous step forward in the way we think about fundamental physics . besides its obvious impact for understanding gravity over astrophysical distances complete with resolutions of earlier puzzles ( like the detailed properties of mercury s orbit ) and novel predictions for new phenomena ( like the bending of light and the slowing of clocks by gravitational fields ) its implications for other branches of physics have been equally profound .
these implications include many ideas we nowadays take for granted .
one such is the universal association of fundamental degrees of freedom with fields ( first identified for electromagnetism , but then cemented with its extension to gravity , together with the universal relativistic rejection of action at a distance ) .
another is the recognition of the power of symmetries in the framing of physical law , and the ubiquity in particular of gauge symmetries in their description ( again reinforcing the earlier discovery in electromagnetism ) .
a third is the systematization of the belief that the physical content nature s laws should be independent of the variables used in their description , and the consequent widespread penetration of geometrical methods throughout physics .
but the study of general relativity ( gr ) and other interactions ( like electromagnetism , and its later - discovered relatives : the weak and strong forces ) have since drifted apart
. like ex - lovers who remain friends , for most of the last century practitioners in either area have known little of the nitty gritty of each other s day - to - day struggles , even as they read approvingly of their occasional triumphs in the popular press . over the years
the study of both gravity and the other interactions has matured into precision science , with many impressive theoretical developments and observational tests . for gravity
this includes remarkably accurate accounts of motion within the solar system , to the point that gr through its use within the global positioning system ( gps ) is now an indispensable tool for engineers [ will 2001 ] . for the other interactions the successes include the development and testing of the standard model ( sm ) , a unified framework for all known non - gravitational physics , building on the earlier successes of quantum electrodynamics ( qed ) .
there is nevertheless a mounting chorus of calls for modifying general relativity , both at very short and very long distances .
these arise due to perceived failures of the theory when applied over distances much different from those over which it is well - tested .
the failures at short distances are conceptual , to do with combining gravity with quantum effects .
those at long distances are instead observational , and usually arise as ways to avoid the necessity for introducing the dark matter or dark energy that seem to be required when general relativity is applied to describe the properties of the universe as a whole .
the remainder of this chapter argues that when searching for replacements for gr over short and long distances there is much to be learned from other branches of physics , where similar searches have revealed general constraints on how physics at different scales can relate to one another .
the hard - won lessons learned there also have implications for gravitational physics , and this recognition is beginning to re - establish the connections between the gravitational and non - gravitational research communities . in a nutshell , the lessons distilled from other areas of physics make it likely that it is much more difficult to modify gravity over very long distances than over very tiny ones .
this is because very broad principles ( like unitarity and stability ) strongly restrict what is possible .
the difficulty of modifying gravity over long distances is a very useful ( but often neglected ) clue when interpreting cosmological data , because it strongly constrains the theoretical options that are available .
we ignore such clues at our peril .
the demand to replace general relativity at short distances arises because quantum mechanics should make it impossible to have a spacetime description of geometry for arbitrarily small scales .
for example , an accurate measurement of a geometry s curvature , @xmath0 , requires positions to be measured with an accuracy , @xmath1 , smaller than the radius of curvature : @xmath2 but for position measurements with resolution , @xmath1 , the uncertainty principle requires a momentum uncertainty , @xmath3 , which implies an associated energy uncertainty , @xmath4 , or equivalently a mass @xmath5 .
but the curvature associated with having this much energy within a distance of order @xmath1 is then @xmath6 , where @xmath7 defines the planck length , @xmath8 , and @xmath9 is newton s constant .
requiring eq .
( [ deltacriterion ] ) , then shows that there is a lower bound on the resolution with which spacetime can be measured : @xmath10 although this is an extremely short distance ( present experiments only reach down to about @xmath11 m ) , it is also only a lower bound .
depending on how gravity really works over short distances , quantum gravity effects could arise at much longer scales .
notice how crucial it is to this argument that the interaction strength , @xmath9 , has dimensions of length ( in fundamental units , for which @xmath12 ) .
imagine performing a similar estimate for an electrostatic field .
the coulomb interaction energy between two electrons separated by a distance @xmath1 is @xmath13 , where @xmath14 denotes the electron s electric charge . but
the energy required by the uncertainty principle to localize electrons this close to one another is @xmath15 , so the condition that this be smaller than @xmath16 is @xmath17 where the fine - structure constant , @xmath18 , is dimensionless .
this condition does nt depend on @xmath1 because the relative strength of quantum fluctuations to electrostatic interactions does not change with distance .
the observation that quantum fluctuations do not get worse at shorter distances in electrodynamics , and so is much less severe than the power - law competition found above for gravity . ] but do for gravity can be more technically expressed as the statement that qed is a _
renormalizable _ quantum field theory ( qft ) while gr is not . in
qft small - distance quantum fluctuations appear ( within perturbation theory ) as divergences at small distances ( or high momenta ) when summing over all possible quantum intermediate states .
for instance , given a hamiltonian , @xmath19 , the second - order shift in the energy of a state @xmath20 is @xmath21 where the approximate equality focusses on the sum over a basis of free single - particle states having energies @xmath22 when performing the sum over @xmath23 . because the combination @xmath24 $ ] typically falls with large @xmath25 , limit .
( relativistic calculations organize these sums differently to preserve manifest lorentz invariance at each step , but the upshot is the same . ) renormalizability means that these divergences can all be absorbed into the unknown parameters of the theory like the electron s charge and mass , for instance whose values must in any case be inferred by comparison with experiments . as the above estimates suggest , the hallmark of a nonrenormalizable theory is the appearance of couplings ( like newton s constant ) having dimensions of length to a positive power ( in fundamental units ) .
couplings like this ruin perturbative renormalizability because the more powers of them that appear in a result , the more divergent that result typically is .
for instance , a contribution that arises at @xmath26th order in newton s constant usually depends on @xmath9 through the dimensionless combination @xmath27 , where @xmath28 is the uv cutoff in momentum space ( equivalently , @xmath1 is the small - distance cutoff in position space ) .
by contrast , having more powers of dimensionless couplings , or those having dimensions of inverse powers of length , do not worsen uv divergences .
ever - worsening divergences ruin the arguments that show for renormalizable theories that all calculations are finite once a basic set of couplings are appropriately redefined .
removal of divergences can be accomplished , but only by introducing an infinite number of coupling parameters to be renormalized .
lack of renormalizability was for a long time regarded as a fundamental obstacle to performing any quantum calculations within gravity .
after all , if every calculation is associated with a new parameter that absorbs the new divergences , whose value must be inferred experimentally , then there are as many parameters as observables and no predictions are possible .
if this were really true , it would mean that any classical prediction of gr would come with incalculable theoretical errors due to the uncontrolled size of the quantum corrections . and
the presence of such errors would render meaningless any detailed comparisons between classical predictions and observations , potentially ruining gr s observational successes .
how can meaningful calculations be made ?
as it happens , tools for making meaningful quantum calculations using non - renormalizable theories exist , having been developed for situations where quantum effects are more important than they usually are for gravity [ weinberg 1979 , gasser 1984 ] . the key to understanding how to work with non - renormalizable theories is to recognize that they can arise as approximations to more fundamental , renormalizable physics , for which explicit calculations are possible .
the way non - renormalizable theories arise in this case is as a low - energy / long - distance approximation in situations for which short - distance physics is unimportant , and so is coarse - grained or integrated out [ gell - mann 1954 , wilson 1974 ] . for instance , consider the lagrangian density for the quantum electrodynamics of electrons and muons @xmath29 where @xmath30 and @xmath31 ( or @xmath32 and @xmath33 ) are the electron ( or muon ) mass and field . here
@xmath34 and @xmath35 , as usual , and @xmath36 represents the dirac matrices that satisfy @xmath37 .
this is a renormalizable theory because all parameters , @xmath38 , @xmath39 and @xmath40 , have non - positive dimension when regarded as a power of length in fundamental units .
suppose now we choose to examine observables only involving the electromagnetic interactions of electrons at energies @xmath41 ( such as the energy levels of atoms , for instance ) .
muons should be largely irrelevant for these kinds of observables , but not completely so .
muons are not completely irrelevant because they can contribute to electron - photon processes at higher orders in perturbation theory as virtual states .
it happens that any such effects due to virtual muons can be described at low energies by the following _ effective field theory _ of electrons and photons only : @xmath42 where the second line is obtained from the first by performing the field redefinition @xmath43 + \cdots \,.\ ] ] in both equations the ellipses describe terms suppressed by more than two powers of @xmath44 .
the lagrangian densities of eqs .
( [ lqed ] ) and ( [ leffqed ] ) are precisely equivalent in that they give precisely the same results for _ all _ low - energy electron / photon observables , provided one works only to leading order in @xmath45 . if the accuracy of the agreement is to be at the one - loop level , then equivalence requires the choice @xmath46 , and the effective interaction captures the leading effects of a muon loop in the vacuum polarization .
if agreement is to be at the two - loop level , then @xmath47 captures effects coming from higher loops as well , and so on .
this example ( and many many others ) shows that it must be possible to make sensible predictions using non - renormalizable theories .
this must be so because the lagrangian of eq .
( [ leffqed ] ) is not renormalizable its coupling has dimensions ( length)@xmath48 yet it agrees precisely with the ( very sensible ) predictions of qed , eq .
( [ lqed ] ) .
but it is important that this agreement only works up to order @xmath45 .
if we work beyond order @xmath45 in this expansion , we can still find a lagrangian , @xmath49 , that captures all of the effects of qed to the desired order .
the corresponding lagrangian requires more terms than in eq .
( [ leffqed ] ) , however , also including terms like @xmath50 that arise at order @xmath51 .
agreement with qed in this case requires @xmath52 .
sensible predictions can be extracted from non - renormalizable theories , but only if one is careful to work only to a fixed order in the @xmath44 expansion . what is useful about this process is that an effective theory like ( [ leffqed ] ) is much easier to use than is the full theory ( [ lqed ] ) . and any observable whatsoever
may be computed once the coefficients ( @xmath53 and @xmath54 in the above examples ) of the various non - renormalizable interactions are identified .
this can be done by comparing its implications with those of the full theory for a few specific observables .
what about the uv divergences associated with these new effective interactions ?
they must be renormalized , and the many couplings required to perform this renormalization correspond to the many couplings that arise within the effective theory at successive orders in @xmath44 .
but predictiveness is not lost because working to fixed order in @xmath44 means that only a fixed number of effective couplings are required in any given application . at present
this is the _ only _ known way to make sense of perturbatively non - renormalizable theories . in particular
it means that there is a hidden approximation involved in the use of a non - renormalizable theory the low - energy , @xmath44 , expansion that may not have been obvious from the get - go .
what would this picture mean if applied to gr ?
first , it would mean that gr must be regarded as the leading term in the low - energy / long - distance approximation to some more fundamental theory .
working beyond leading order would mean extending the einstein - hilbert action to include higher powers of curvatures and their derivatives , with the terms with the fewest derivatives being expected to dominate at low energies [ for a review see burgess 2004 ] .
since we do not know what the underlying theory is , we can not hope to compute the couplings in this effective theory from first principles as was done above for qed .
instead we treat these couplings as phenomenological , ultimately to be determined from experiment .
the most general interactions involving the fewest curvatures and derivatives , that are consistent with general covariance are @xmath55 where @xmath56 is the metric s riemann tensor , @xmath57 is its ricci tensor , and @xmath58 is the ricci scalar , each of which involves precisely two derivatives of the metric . the first term in eq . ( [ gravaction ] )
is the cosmological constant , which we drop because observations imply @xmath59 is ( for some unknown reason , see below ) extremely small .
once this is done the leading term in the derivative expansion is the einstein - hilbert action whose coefficient , @xmath60 gev , has dimensions of mass ( when @xmath12 ) , and is set by the value of newton s constant .
this is followed by curvature - squared terms having dimensionless effective couplings , @xmath61 , and curvature - cubed terms with couplings inversely proportional to a mass , @xmath62 , ( not all of which are written in eq .
( [ gravaction ] ) ) .
although the numerical value of @xmath63 is known , the mass scale @xmath39 appearing in the curvature - cubed ( and higher ) terms is not .
but since it appears in the denominator it is the lowest mass scale to have been integrated out that should be expected to dominate .
what its value should be depends on the scale of the applications one has in mind . for applications to the solar system or to astrophysics @xmath39
might reasonably be taken to be the electron mass , @xmath64 .
but for applications to inflation , where the scales of interest are much larger than @xmath64 , @xmath39 would instead be taken to be the lightest particle that is heavier than the scales of inflationary interest .
the einstein - hilbert term should dominate at low energies ( since it involves the fewest derivatives ) , and this expectation can be made more precise by systematically identifying which interactions contribute to a particular order in the semiclassical expansion .
to do so we expand the metric about an asymptotically static background spacetime : @xmath65 , and compute ( say ) the scattering amplitudes for asymptotic graviton states that impinge onto the geometry from afar .
if the energy , @xmath66 , of the incoming states are all comparable and similar to the curvatures scales of the background spacetime , dimensional analysis can be used to give an estimate for the energy - dependence of an @xmath67-loop contribution to a scattering amplitude , @xmath68 .
consider a contribution to this amplitude that involves @xmath69 external lines and @xmath70 vertices involving @xmath71 derivatives and @xmath72 attached graviton lines .
dimensional analysis leads to the estimate : @xmath73^{v_{id } } \,.\ ] ] notice that no negative powers of @xmath66 appear here because general covariance requires derivatives come in pairs , so the index @xmath71 in the product runs over @xmath74 , with @xmath75 .
this last expression displays the low - energy approximation alluded to above because it shows that the small quantities controlling the perturbative expansion are @xmath76 and @xmath77 .
use of this expansion ( and in particular its leading , classical limit see below ) presupposes both of these quantities to be small .
notice also that because @xmath78 , factors of @xmath77 are much larger than factors of @xmath76 , but because they do not arise until curvature - cubed interactions are important , the perturbative expansion always starts off with powers of @xmath76 .
( [ grcount1a ] ) answers a question that is not asked often enough : what is the theoretical error made when treating gravitational physics in the classical approximation ?
what makes it so useful in this regard is that it quantifies the size of the contribution to @xmath68 ( or other observables ) arising both from quantum effects ( _ i.e. _ loops , with @xmath79 ) , and from terms normally not included in the lagrangian ( such as higher - curvature terms ) .
this allows an estimate of the size of the error that is made when such terms are not considered ( as is often the case ) . in particular , eq .
( [ grcount1a ] ) justifies why classical calculations using gr work so well , and quantifies just how accurate their quantum corrections are expected to be . to see this ,
we ask which graphs dominate in the small-@xmath66 limit . for any fixed process ( _ i.e. _ fixed @xmath69 ) eq .
( [ grcount1a ] ) shows the dominant contributions are those for which @xmath80 that is , the dominant contribution comes from arbitrary tree graphs constructed purely from the einstein - hilbert @xmath81 action .
this is precisely the prediction of classical general relativity .
for instance , for the scattering of two gravitons about flat space , @xmath82 , we have @xmath83 , and eq .
( [ grcount1a ] ) predicts the dominant energy - dependence to be @xmath84 .
this is borne out by explicit tree - level calculations [ dewitt 1967 ] which give @xmath85 for an appropriate choice of graviton polarizations .
here @xmath86 , @xmath87 and @xmath88 are the usual lorentz - invariant mandelstam variables built from the initial and final particle four - momenta , all of which are proportional to @xmath89 .
this shows both that @xmath90 to leading order , and that it is the physical , invariant , centre - of - mass energy , @xmath91 , that is the relevant scale against which @xmath39 and @xmath63 should be compared .
the next - to - leading contributions , according to eq .
( [ grcount1a ] ) , arise in one of two ways : either @xmath92 these correspond to one - loop ( quantum ) corrections computed only using einstein gravity ; plus a tree - level contribution including precisely one vertex from one of the curvature - squared interactions ( in addition to any number of interactions from the einstein - hilbert term ) .
the uv divergences arising in the first type of contribution are absorbed into the coefficients of the interactions appearing in the second type .
both are suppressed compared to the leading , classical , term by a factor of @xmath93 .
this estimate ( plus logarithmic complications due to infrared divergences ) is also borne out by explicit one - loop calculations about flat space [ weinberg 1965 , dunbar 1995 , donoghue 1999 ] .
this is the reasoning that shows why it makes sense to compute quantum effects , like hawking radiation or inflationary fluctuations , within a gravitational context . for observables
located a distance @xmath94 away from a gravitating mass @xmath40 , the leading quantum corrections are predicted to be of order @xmath95 . for comparison , the size of classical relativistic corrections is set by @xmath96 , where @xmath97 denotes the schwarzschild radius . at the surface of the sun this makes relativistic corrections of order @xmath98 , while quantum corrections are @xmath99 .
clearly the classical approximation to gr is _ extremely _ good within the solar system . on the other hand , although relativistic effects can not be neglected near a black hole , since @xmath100 , the relative size of quantum corrections near the event horizon is @xmath101 which is negligible provided @xmath102 . since @xmath63 is of order tens of micrograms , this shows why quantum effects represent small perturbations for any astrophysical black holes , but would not be under control for any attempt to interpret the gravitational field of an elementary particle ( like an electron ) as giving rise to a black hole .
the good news is that it says that the observational successes of gr are remarkably robust against the details of whatever small - distance physics ultimately describes gravity over very small distances .
this is because _ any _ microscopic physics that predicts the same symmetries ( like lorentz invariance ) and particle content ( a massless spin-2 particle , or equivalently a long - range force coupled to stress - energy ) as gr , must be described by a generally covariant effective action like eq .
( [ gravaction ] ) .
because this is dominated at low energies by the einstein - hilbert action , it suffices to get the low - energy particle content and symmetries right to get gr right in all of its glorious detail [ deser 1970 ] .
the bad news applies to those who think they know what the fundamental theory of quantum gravity really is at small scales , since whatever it is will be very hard to test experimentally .
this is because all theories that get the bare minimum right ( like a massless graviton ) , are likely to correctly capture all of the successes of gr in one fell swoop . at low energies the only difference between the predictions of _
any _ such theory is the value of the coefficients , @xmath61 and @xmath103 _ etc _ , appearing in the low - energy lagrangian ( [ gravaction ] ) , none of which are yet observable .
there are two kinds of proposals that allow tests at low energies : those that change the low - energy degrees of freedom ( such as by adding new light particles in addition to the graviton more about these proposals below ) ; and those that change the symmetries predicted for the low - energy theory .
prominent amongst this latter category are theories that postulate that gravity at short distances breaks lorentz or rotational invariance , perhaps because spacetime becomes discrete at these scales . at first sight ,
breaking lorentz invariance at short distances seems batty , due to the high accuracy with which experimental tests verify the lorentz - invariance of the vacuum within which we live .
how could the world we see appear so lorentz invariant if it is really not so deeper down ?
surprisingly , experience with other areas of physics suggests this may not be so crazy an idea ; we know of other , emergent , symmetries that can appear to be very accurate at long distances even though they are badly broken at short distances .
most notable among these is the symmetry responsible for conservation of baryon number , which has long been known to be an ` accidental ' symmetry of the standard model .
this means that for _ any _ microscopic theory whose low - energy particle content is that of the sm , any violations of baryon number must necessarily be described by a non - renormalizable effective interaction [ weinberg 1979a , wilczek 1979 ] , and so be suppressed by a power of a large inverse mass , @xmath44
. this suppression can be enough to agree with observations ( like the absence of proton decay ) if @xmath40 is as large as @xmath104 gev .
could lorentz invariance be similarly emergent ?
if so , it should be possible to find effective field theories for which lorentz violation first arises suppressed by some power of a heavy scale , @xmath44 , even if lorentz invariance is not imposed from the outset as a symmetry of the theory .
unfortunately this seems hard to achieve , since in the absence of lorentz invariance it is difficult transformations , at least for the kinetic terms . ] in an effective theory to explain why the effective terms @xmath105 should have precisely the same coefficient in the low - energy theory .
( see however [ groot nebbelink 2005 ] for some attempts . )
the problem is that the coefficients of these terms are dimensionless in fundamental units , and so are unsuppressed by powers of @xmath44 .
but the relative normalization of these two terms governs the maximal speed of propagation of the corresponding particle , and there are extremely good bounds ( for some particles better than a part in @xmath106 ) on how much this can differ from the speed of light [ see , for instance , mattingly 2005 for a recent review ] .
this underlines why proponents of any particular quantum gravity proposal must work hard to provide the effective field theory ( eft ) that describes their low - energy limit [ see kostelecky 2004 , mattingly 2005 for some gravitational examples ] . since all of the observational implications are contained within the effective theory , it is impossible to know without it whether or not the proposal satisfies all of the existing experimental tests
this is particularly true for proposals that claim to predict a few specific low - energy effects that are potentially observable ( such as small violations of lorentz invariance in cosmology ) .
even if the predicted effects should be observed , the theory must also be shown not to be in conflict with other relevant observations ( such as the absence of lorentz invariance elsewhere ) , and this usually requires an eft formulation .
there also has been considerable activity over recent years investigating the possibility that gr might fail , but over very long distances rather than short ones .
this possibility is driven most persuasively from cosmology , where the hot big bang paradigm has survived a host of detailed observational tests , but only if the universe is pervaded by no less than _ two _ kinds of new exotic forms of matter : dark matter ( at present making up @xmath107 of the universal energy density ) and dark energy ( comprising @xmath108 of the cosmic energy density ) . because all of the evidence for the existence of these comes from their gravitational interactions ,
inferred using gr , the suspicion is that it might be more economical to interpret instead the cosmological tests as evidence that gr is failing over long distances .
but since the required modifications occur over long distances , their discussion is performed most efficiently within an effective lagrangian framework .
these next paragraphs summarize my personal take on what has been learnt to this point .
an important consideration when trying to modify gravity over long distances is the great difficulty in doing so in a consistent way .
almost all modifications so far proposed run into trouble with stability or unitarity , in that they predict unstable degrees of freedom like ` ghosts , ' particles having negative kinetic energy .
the presence of ghosts in a low energy theory is generally regarded as poison because it implies there are instabilities . at the quantum level
these instabilities usually undermine our understanding of particle physics and the very stability of the vacuum [ see cline 2004 for a calculation showing what can go wrong ] , but even at the classical level they typically ruin the agreement between the observed orbital decay of binary pulsars and gr predictions for their energy loss into gravitational waves .
the origin of these difficulties seems to be the strong consistency requirements that quantum mechanics and lorentz invariance impose on theories of massless particles having spin - one or higher [ weinberg 1964 , deser 1970 , weinberg 1980 ] , with static ( non - derivative ) interactions . a variety of studies indicate that a consistent description of particles with spins @xmath109 always requires a local invariance , which in the cases of spins 1 , 3/2 and 2 corresponds to gauge invariance , supersymmetry or general covariance , and this local symmetry strongly limits the kinds of interactions that are possible .
a remarkable equivalence between asymptotically anti - de sitter gravitational theories and non - gravitational systems in one lower dimensions may provide a loophole to some of these arguments , although its ultimate impact is not yet known . ]
although it remains an area of active research [ dvali 2000 ] , at present the only systems known to satisfy these consistency constraints consist of relativistic theories of spins 0 through 1 coupled either to gravity or supergravity ( possibly in more than 4 spacetime dimensions ) . as might be expected , widespread acceptance of the existence of a hitherto - unknown form of matter requires the concordance of several independent lines of evidence , and this constrains one s options when formulating a theory for dark matter .
it is useful to review this evidence when deciding whether it indicates a failure of gr or a new form of matter .
the evidence for dark matter comes from measuring the amount of matter in a region as indicated by how things gravitate towards it , and comparing the result with the amount of matter that is directly visible .
several types of independent comparisons consistently point to there being more than 10 times as much dark , gravitating material in space than is visible : * _ galaxies : _ the total mass in a galaxy may be inferred from the orbital motion of stars and gas measured as a function of distance from the galactic center .
the results , for large galaxies like the milky way , point to several times more matter than is directly visible . * _ galaxy clusters : _ similar measurements using the motion of galaxies and temperature of hot gas in large galaxy clusters also indicate the presence of much more mass than is visible . * _ structure formation : _ present - day galaxies and galaxy clusters formed through the gravitational amplification of initially - small primordial density fluctuations . in this case the evidence for dark matter arises from the interplay of two facts : first
, the initial density fluctuations are known to be very small , @xmath110 , at the time when the cmb was emitted .
second , small initial fluctuations can not be amplified by gravity until the epoch where non - relativistic matter begins to dominate the total energy density . but this does not give enough time for the initially - small fluctuations to form galaxies unless there is much more matter present than can be accounted for by baryons .
the amount required agrees with the amount inferred from the previous measures described above . * _ primordial nucleosynthesis : _ the total mass density of ordinary matter ( baryons ) in the universe can be inferred from the predicted relative abundance of primordial nuclei created within the hot big bang .
this predicted abundance agrees well with observations , and relies on the competition between nuclear reaction rates and the rate with which the universe cools . but both of these rates themselves depend on the net abundance of baryons in the universe : the nuclear reaction rates depend on the number of baryons present ; and the cooling rate depends on how fast the universe expands , and so at least , in gr on its total energy density .
the success of the predictions of big bang nucleosynthesis ( bbn ) therefore fixes the fraction of the universal energy density which can consist of baryons , and implies that there can at most be a few times more baryons than what would be inferred by counting those that are directly visible .
* _ the cosmic microwave background ( cmb ) : _ cmb photons provide an independent measure of the total baryon abundance .
they do so because sound waves in the baryon density that are present when these photons were radiated are observable as small temperature fluctuations .
since the sound - wave properties depend on the density of baryons , a detailed understanding of the cmb temperature spectrum allows the total baryon density to be reconstructed .
the result agrees with the bbn measure described above .
there are two main options for explaining these observations .
since dark matter is inferred gravitationally , perhaps the laws of gravity differ on extra - galactic scales than in the solar system .
alternatively , there could exists a cosmic abundance of a new type of hitherto - undiscovered particle . at present
there are several reasons that make it more likely that dark matter is explained by the presence of a new type of particle than by changing gr on long distances .
first , as mentioned above , sensible modifications are difficult to make at long distances that lack ghosts and other inconsistencies .
second , no phenomenological modification of gravity has yet been proposed that accounts for all the independent lines of evidence given above ( although there is a proposal that can explain the rotation of galaxies [ milgrom 1983 , sanders 2002 ] ) . on the other hand ,
all that is required to obtain dark matter as a new form of matter is the existence of a new type of stable elementary particle having a mass and couplings similar to those of the @xmath111 boson , which is already known to exist .
@xmath111 bosons would be excellent dark matter candidates if only they did not decay . a particle with mass and couplings like the @xmath111 boson , but which is stable
called a weakly interacting massive particle ( wimp ) would naturally have a relic thermal abundance in the hot big bang that lies in the range observed for dark matter [ for a review , see eidelman 2004 ] .
new particles with these properties are actually predicted by many current proposals for the new physics that is likely to replace the standard model at energies to be explored by the large hadron collider ( lhc ) . at the present juncture the preponderance of evidence
the simplicity of the particle option and the difficulty of making a modification to gr that works favours the interpretation of cosmological evidence as pointing to the existence of a new type of matter rather than a modification to the laws of gravity . * _ universal acceleration : _ since gravity is attractive , one expects an expanding universe containing only ordinary ( and dark ) matter and radiation to have a decelerating expansion rate .
evidence for dark energy comes from measurements indicating the universal expansion is _ accelerating _ rather than decelerating , obtained by measuring the brightness of distant supernovae [ perlmutter 1997 , riess 1997 , bahcall 1999 ] .
according to gr , accelerated expansion implies the universe is dominated by something with an equation of state satisfying @xmath112 , which is not true for ordinary matter , radiation or dark matter . * _ flatness of the universe : _ an independent measure of the dark energy comes from the observed temperature fluctuations in the cmb . because the cmb photons traverse the entire observable universe before reaching us , their properties on arrival depend on the geometry of the universe as a whole ( and so also , according to gr , on its total energy density ) .
agreement with observations implies the total energy density is larger than the ordinary and dark matter abundances , which fall short by an amount consistent with the amount of dark energy required by the acceleration of the universe s expansion [ komatsu 2009 ] .
again the theoretical options are the existence of a new form of energy density , or a modification of gr at long distances .
although there are phenomenological proposals for modifications that can cause the universe to accelerate ( such as [ dvali 2000 ] ) , all of the previously described problems with long - distance modifications to gr also apply here . by contrast
, there is a very simple energy density that does the job , consisting simply of a cosmological constant _
i.e. _ a constant @xmath113 ev@xmath114 in eq .
( [ gravaction ] ) , for which @xmath115 .
this is phenomenologically just what the doctor ordered , and agrees very well with the observations .
the theoretical difficulty here is that a cosmological constant is indistinguishable from the energy density of a lorentz - invariant vacuum , . whether this , together with supersymmetry , can solve the problem is under active study [ burgess 2005 ] . ] since both contribute to the stress tensor an amount @xmath116 . in principle
, this should be a good thing because we believe we can compute the vacuum energy .
the problem is that ordinary particles ( like the electron ) contribute such an enormous amount the electron gives @xmath117 ev@xmath118 that agreement with the observed value requires a cancellation [ weinberg 1989 ] to better than one part in @xmath119 .
dark matter and dark energy are two forms of exotic matter , whose existence is inferred purely from their gravitational influence on visible objects .
it is tempting to replace the need for two new things with a single modification to gravity over very large distances . yet
the preponderance of evidence again argues against this point of view .
first , it is difficult to modify gr at long distances without introducing pathologies .
second , it is difficult to find modifications that account for more than one of the several independent lines of evidence ( particularly for dark matter ) .
by contrast , it is not difficult to make models of dark matter ( wimps ) or dark energy ( a cosmological constant ) . for dark energy
this point of view runs up against the cosmological constant problem , which might indicate the presence of observably large extra dimensions , but for which no consensus yet exists . in summary ,
modifications to general relativity are widely mooted over both large and small distances .
this chapter argues that modifications at small distances are indeed very likely , and well worth seeking .
but unless the modification takes place just beyond our present experimental reach ( @xmath120 m ) [ arkani - hamed 1998 , antoniadis 1998 , burgess 2005 ] , it is also likely to be very difficult to test experimentally .
the basic obstruction is the decoupling from long distances of short - distance physics , a property most efficiently expressed using effective field theory methods .
the good news is that this means that the many observational successes of gr are insensitive to the details of whatever the modification proves to be .
modifications to gr over very long distances are also possible , and have been argued as more economical than requiring the existence of two types of unknown forms of matter ( dark matter and dark energy ) .
if so , consistency constraints seem to restrict the possibilities to supplementing gr by other very light spin-0 or spin-1 bosons ( possibly in higher dimensions ) . the experimental implications of such modifications are themselves best explored using effective field theories .
unfortunately , no such a modification has yet been found that accounts for all of the evidence for dark matter or energy in a way that is both consistent with other tests of gr and is more economical than the proposals for dark matter or energy themselves . to the extent that the utility of effective field theory relies on decoupling
, one might ask : what evidence do we have that planck - scale physics decouples ? there are two lines of argument that bear on this question .
first , once specific modifications to gravity are proposed it becomes possible to test whether decoupling takes place .
perhaps the best example of a consistent modification to gravity at short distances is string theory , and all the present evidence points to decoupling holding in this case .
but more generally , if sub - planckian scales do _ not _ decouple , one must ask : why has science made progress at all ?
after all , although nature comes to us with many scales , decoupling is what ensures we do nt need to understand them all at once .
if sub - planckian physics does not decouple , what keeps it from appearing everywhere , and destroying our hard - won understanding of nature ?
i thank the editors for their kind invitation to contribute to this volume , and for their patience in awaiting my contribution .
my understanding of this topic was learned from steven weinberg , who pioneered effective field theory techniques , and was among the first to connect the dots explicitly about gravity s interpretation as an effective field theory .
my research is funded by the natural sciences and engineering research council of canada , as well as by funds from mcmaster university and perimeter institute .
aghababaie , y. , burgess , c.p . ,
parameswaran , s.l . and
quevedo , f. ( 2004 ) nucl .
b * 680 * , 389 [ arxiv : hep - th/0304256 ] .
antoniadis , i. , arkani - hamed , n. , dimopoulos , s. and dvali , g. ( 1998 ) phys .
b * 436 * 257 [ arxiv : hep - ph/9804398 ] . arkani - hamed , n. , dimopoulos , s. and dvali , g. ( 1998 ) phys .
b * 429 * 263 [ arxiv : hep - ph/9803315 ] . arkani - hamed , n. , dimopoulos , s. , kaloper , n. and sundrum , r. ( 2000 ) phys . lett .
b * 480 * 193 , [ hep - th/0001197 ] .
burgess , c.p .
( 2005 ) aip conf .
proc . * 743 * , 417 [ arxiv : hep - th/0411140 ] .
carroll , s.m . and m. m. guica , m.m . , [ arxiv : hep - th/0302067 ] .
cline , j.m .
, jeon , s. and moore , g.d .
( 2004 ) phys .
d * 70 * 043543 [ arxiv : hep - ph/0311312 ] .
deser , s. ( 1970 ) gen .
* 1 * 9 [ arxiv : gr - qc/0411023 ] .
dewitt , b.s .
( 1967 ) _ phys .
rev . _ * 162 * 1239 .
gell - mann , m. and low , f.e .
( 1954 ) phys .
rev . * 95 * 1300 .
groot nibbelink , s. and pospelov , m. ( 2005 ) phys .
lett . * 94 * 081601 [ arxiv : hep - ph/0404271 ] .
kachru , s. , schulz , m.b . and
silverstein , e. ( 2000 ) phys . rev .
d * 62 * 045021 , [ hep - th/0001206 ] .
kostelecky , v. a. ( 2004 ) phys .
d * 69 * 105009 [ arxiv : hep - th/0312310 ] .
maldacena , j.m .
( 1998 ) adv .
* 2 * ( 1998 )
231 [ int . j. theor .
* 38 * ( 1999 ) 1113 ] [ arxiv : hep - th/9711200 ] .
mattingly , d. ( 2005 ) living rev .
* 8 * 5 [ arxiv : gr - qc/0502097 ] .
milgrom , m. ( 1983 ) _ ap .
j. _ * 270 * 365 - 370 ; 371 - 283 ; 384 - 389 . | we live at a time of contradictory messages about how successfully we understand gravity .
general relativity seems to work very well in the earth s immediate neighborhood , but arguments abound that it needs modification at very small and/or very large distances .
this essay tries to put this discussion into the broader context of similar situations in other areas of physics , and summarizes some of the lessons which our good understanding of gravity in the solar system has for proponents for its modification over very long and very short distances .
the main message is that effective theories , in the technical sense of ` effective ' , provide the natural language for testing proposals , and so are also effective in the colloquial sense . |
early surveys of the milky way , such as those of @xcite , photographed several regions of interstellar gas ionized by the ultraviolet radiation of nearby hot stars .
many other surveys were done over the ensuing decades ( see the list of references given by @xcite ) . however , it was not until after the discovery of pulsars that it became generally recognized that ionized hydrogen is a pervasive component of the interstellar medium . soon thereafter @xcite
conducted a photographic survey in search of this general diffuse medium , reaching a limiting brightness of 15 rayleigh ( @xmath8 ) , corresponding to an emission measure of about @xmath9 , and covering the band of latitude @xmath10 at a resolution of approximately 10@xmath11 .
somewhat later , @xcite , using an image tube camera , published an atlas of images in four emission lines ( [ sll ] , h@xmath12+[nii ] , o[iii ] , h@xmath13 ) and the blue continuum ( 422 nm ) , covering the region @xmath14 , reaching a surface brightness ( @xmath15 r ) about the same as that of sivan s survey , at an angular resolution of @xmath16 .
of course the h@xmath12 line is included in the pass band for the red plates of the national geographic society / palomar observatory sky survey ( poss 1 ) , and these show many prominent emission nebulae , those brighter than approximately 100 r. ( the eso / src southern sky atlas is similar , but goes a bit deeper . ) the second generation plates ( poss 2 ) go about a factor of four fainter , reaching a level of approximately 25 r. two difficulties with these photographic surveys are that their sensitivity is low , and because of the non - linear response of photographic emulsions and contributions from continuum radiation it is difficult to derive quantitative intensity information from them .
modern ccd detectors , on the other hand , are intrinsically linear , have high quantum efficiency , and allow for digital subtraction of continuum images .
although photographic surveys still have their place @xcite , most modern studies are done via narrow - band ccd imaging or fabry - perot spectroscopy .
an early example of the former technique is the work of , who surveyed the region @xmath17 to @xmath18 , @xmath19 to @xmath20 to a level of 4 r ( @xmath21 ) at 2 arcminute resolution .
an early example of the fabry - perot technique is the work of @xcite who observed several regions with a fabry - perot spectrometer .
reynolds and his collaborators have recently completed a survey of the northern sky with a fabry - perot instrument called the wisconsin h - alpha mapper ( wham ) @xcite .
a group at virginia polytechnic institute @xcite have begun a narrow - band survey of the northern hemisphere with an imaging camera similar to ours .
we present a comparison of these other surveys with our own in section [ others ] .
both of these narrow - band imaging surveys are complementary to reynolds fabry- perot survey , not competitive with it .
our angular resolution is 75 times better than that of wham .
the wham spectrometer , on the other hand , goes ten times fainter .
it also has high spectral resolution ( 0.026 nm , @xmath22 ) , and thus provides velocity as well as intensity information that the narrow- band imaging surveys can not .
perhaps as important , the high spectral resolution allows the wham instrument to separate interstellar h@xmath12 emission from that coming from the earth s geocorona and thus determine absolute intensities , whereas the images provided by the narrow - band surveys all contain an offset due to geocoronal emission ( as well as telluric oh emission ) , an offset which is variable both temporally and spatially . in section 5
we describe a technique for correcting our images for geocoronal emission that uses wham observations as fiducial points in those parts of the sky where our surveys overlap .
we present this survey to the astronomical community as a resource which we hope will find many applications .
we anticipate its use in two major areas of study : first , the structure of the interstellar medium , and second , the characterization of the galactic microwave foreground . the survey will provide detailed information on the structure of the diffuse , warm , ionized component of the interstellar medium , information necessary for understanding the dynamics and evolutionary history of the interstellar gas . since the survey covers the entire sky south of @xmath23 declination , has higher angular resolution and reaches a fainter intensity level than does any previous large scale survey , we expect a detailed examination will show several new structures , such as stellar wind bubbles , supernova remnants , bow shock nebulae , planetary nebula shells , and other interesting objects .
some examples are noted in section [ examples ] . even where specific h@xmath2 emission features are not seen , the survey images should be scientifically useful . since the production of both kinds of emission results from the interaction of a proton with an electron
, the ratio of intensities of free - free and h@xmath2 emission depends primarily on quantum - mechanical factors , with only a slight dependence on temperature and no dependence on density : @xmath24}\over{i_{h \alpha } [ r ] } } = 10.4 \nu^{-2.14 } t_4^{0.527 } 10^{0.029/t_4 } ( 1 + 0.08)\ ] ] where @xmath25 is the observing frequency in ghz , @xmath26 is the temperature of the gas in units of @xmath27 k , and the second term in the last factor ( 1 + 0.08 ) is the relative abundance of helium .
thus our h@xmath12 images will allow accurate measurement of anisotropies in the galactic free - free emission at microwave wavelengths ( or proof that these are negligible ) , emission which must be subtracted from satellite or ground - based measurements to obtain the spatial spectrum of the true cosmic background radiation . in an early pilot study for this project we showed that no more than 7% of the 44 @xmath28 anisotropy in the cmbr measured in the saskatoon experiment could be due to contamination by galactic free - free emission .
the high angular resolution of our survey may be important given the discovery of a third acoustic peak in the microwave background spectrum at angular scales of about @xmath29 , well within the range of spatial frequencies we sample .
a general review of the implications of h@xmath12 observations for studies of the cosmic microwave background can be found in .
our ccd camera , from spectrasource , inc . , contains a 1024 x 1024 texas instruments chip with the following characteristics : 12-@xmath30 pixels , full - well capacity 80 ke , quantum efficiency 62% , gain 1.4 e adu@xmath31 , and readout noise 18 e. it is cooled thermoelectrically to about @xmath32c , yielding a dark current of 0.04 e s@xmath31 . in a 20-minute exposure ,
the maximum we use , the dark level amounts to about 34 adu ( 48 e ) , about 10% the level of our typical exposures .
laboratory measurements showed that the chip response was linear up to the full - well capacity ( 80,000 e @xmath33 60,000 adu ) , but the variance increased non - linearly above 20,000 adu ( 28,000 e ) .
except for some very bright objects , our exposures yield values well below this limit .
the sky is imaged onto the chip with a 52-mm focal length canon lens operated at f/1.6 , yielding a field of view of @xmath34 and a scale of @xmath35 pixel@xmath31 .
the camera is attached to a byers series 2 german mount with microstepping motors on each axis .
the pointing accuracy is better than @xmath36 up to a zenith distance of 75 .
the unguided tracking accuracy is better than @xmath37 hr@xmath31 .
thus , with @xmath38 pixels , no external guider is needed .
the system is housed in a 10-foot dome ( from technical innovations , inc . ) on a concrete pad located near the 1-meter telescope building at cerro tololo inter- american observatory ( ctio ) .
the small size of the instrument and dome has caused the ctio staff to give our instrument the nickname of el enano the dwarf .
we added magnetic microswitch sensors , power relays , and a position encoder to the dome system so that it could be controlled by a computer .
we also added a third dome rotation motor , replaced some plastic parts with metal , and modifed the shutter cable system in order to improve reliability of the dome operation .
a texas weather instruments weather station completes the system .
two pcs running scheduling and control programs written in visual basic under windows 95 control the camera , mount , dome , weather station , and light box .
they communicate with each other and with the outside world via ethernet using standard tcp / ip protocol .
we are grateful to las cumbres observatory , inc .
, for loaning us much of this equipment for the duration of the project . because the twilight sky is not uniform over our wide field of view @xcite , we do not use sky flats for calibration .
instead we use a lightbox containing a translucent screen illuminated by incandescent light bulbs and by hydrogen emission lamps .
a filter wheel mounted between the lens and the sky contains an h@xmath2 filter of 3.2-nm bandwidth and a dual - band notch filter , which excludes h@xmath2 but transmits two 6.1-nm bands of continuum radiation on either side of h@xmath2 , centered at 644 nm and 677 nm ( figure [ filtertrans ] ) .
both filters were obtained from custom scientific , inc .
the choice of the h@xmath12 filter bandwidth was governed by two considerations .
we want a narrow bandpass to reduce the foreground sky continuum , but large enough that transmission losses near the edge of the field are small . because our filter is placed in front of the camera lens , light imaged at any point on the detector enters the filter as a parallel beam at an angle of incidence equal to the angular distance of the point from the optical axis . for a simple interference filter ( multi - layer filters are somewhat more complicated ) the central wavelength at angle of incidence @xmath39 is given by the formula @xcite @xmath40 where @xmath41 is the index of refraction .
the filter band shifts towards the blue with increasing angle of incidence ; equivalently , a line shifts towards the red with respect to the filter central wavelength , and moves to a different part of the filter transmission curve . for @xmath42 , near the corners of our images , the wavelength shift at 656.3 nm is approximately 1.0 nm . we have therefore chosen a filter with 1.6 nm half - width . the central wavelength of an interference filter shifts to longer wavelengths with increasing temperature , in our case about 0.03 nm c@xmath31 . the filter wheel is encased in styrofoam and maintained at a constant 25 c temperature to minimize such temperature - related shifts and to inhibit moisture from condensing on the filters or the lens . the temperature is deliberately set higher than the design temperature of 20 c , and higher than the night- time ambient temperature , so that a ) only active heating , not cooling , is needed to maintain the temperature ; b ) the bandpass is shifted slightly to the red of h@xmath12 at the center of the field , thus reducing the variation in transmission with angle of incidence as the central wavelength shifts blueward towards the edge of the field ; and c ) the possibility of condensation on the lens is eliminated . the transmission of 656.3 nm light as a function of angle of incidence , as measured on our lightbox while illuminated by a hydrogen emission lamp , is shown in figure [ hfiltertrans ] .
figure [ filtertrans ] also shows the position of atmospheric oh emission lines and interstellar emission lines . the h@xmath12 band
( 653.3 - 659.3 nm for t / t@xmath43 ) contains only five lines of the 6 - 1 band of oh . but some of these , particularly the 655.4 nm line , may make a significant contribution to the general foreground level , comparable to that from geocoronal h@xmath12 . to the extent that this foreground emission can be modelled as a 2-dimensional plane , it is removed from our final images by the methods discussed in sections [ corrections ] and [ georemoval ] .
in addition , the interstellar emission lines of [ nii ] at 654.8 and 658.3 nm lie within the h@xmath12 filter band .
the filter transmission is approximately 39% and 26% respectively at these wavelengths
( at the operating temperature of 25 c ) compared to the transmission at h@xmath12 of 78% . in some astronomical objects , the [ nii ] lines are comparable in strength to h@xmath12 , so this contamination by [ nii ] lines should be taken into account in any careful quantitative measurements of the atlas images .
for example , if the 658.3 line has the same intensity as h@xmath2 and the 654.8 nm line has 1/3 that strength ( the ratio of the two [ nii ] lines is always 1/3 ) , about 33% of the brightness measured in the h@xmath12 pass band is really [ nii ] emission , not h@xmath12 .
( see also the discussion of the [ nii ] lines in section [ intensityscale ] . )
the continuum filter bands were chosen primarily to avoid emission lines in the night sky @xcite .
the 644 band falls longward of the stronger lines of the 9 - 3 band of oh and excludes most of the short wavelength lines of the 6 - 1 band of oh , as well as excluding the oi lines at 630.0 and 636.4 nm .
the 677 band falls between the 6 - 1 and 7 - 2 bands of oh , and excludes the interstellar lines of hei at 667.8 nm and of [ sii ] at 671.6 nm .
the transmission at the wavelength of the [ sii ] line at 673.1 nm is only 16% , though it will be somewhat higher , up to 70% , at high angles of incidence . in many astronomical objects
this line is less than 10% the strength of h@xmath44 , though it can be of comparable strength in shocked regions , such as in supernova remnants , and in the very faint emission from the warm interstellar medium @xcite .
the exposure times used for the continuum exposures are four times shorter than for the h@xmath12 images , so the contamination from the [ sii ] 671.6-nm line is probably no more than a few per cent in most regions , though it may be as high as 25% in strongly shocked regions which appear near the edges of our fields .
one of the robot s computers positions the telescope and dome while the other operates the camera , schedules the observations , and runs the ftp and email communication links . a log of the night s operations and sample images are sent to the pi each morning via email and ftp , but the full data set is recorded on disk and then transferred to dat tape during the day .
we can inquire of the status of the robot and modify its observing schedule via email , and make changes in the control programs via ftp .
but we do not operate the camera in real time
it is a true robot , in the sense that its real - time operations are autonomous .
the robot monitors the weather station continuously , and will not attempt observations if humidity or wind conditions exceed safe limits , but as an extra precaution it also asks permission ( via email ) of the 4-meter telescope operator before opening the dome . the operator can also close the dome via email command .
the only other human intervention , during normal operations , is for weekly tape - changing and preventive maintenance .
all observations are taken in a dark sky , after the end of evening twilight and before the beginning of morning twilight , and with the moon below the horizon
. a normal set of science observations consists of five 20-minute exposures through the h@xmath2 filter interspersed among six 5-minute exposures through the dual- band continuum filter .
use of multiple exposures allows removal of cosmic ray events and other transients , such as meteor , satellite , and airplane trails .
interspersing the h@xmath2 and continuum exposures helps compensate for changes in sky foreground levels with time or zenith distance .
reflective ghosts of bright stars appear in the images .
these are caused by light focused at a point on the ccd surface reflecting back through the lens to the filter , where it is again reflected and focused at a different point on the detector . to assist in identifying and removing the reflective ghosts , the camera pointing is dithered in a raster pattern of 1/4@xmath45 steps between each exposure . the ghosts move in the opposite direction to the stars themselves and so can be readily identified in the reduction procedure .
( transmissive ghosts are produced by small angle scattering off of imperfections in the filter .
these move in the same direction as the stars during dithering , and so can not be simply identified and removed during processing . ) between observations of each science field , we take a few exposures pointed at the pole to obtain data on changes in extinction during the night .
we also take occasional short exposures of bright standard objects .
the program that schedules the observations gives priority to fields at large hour angle and high ( more northerly ) declinations , those for which the observing window is smallest . before observing more easterly fields it attempts to observe those fields which are about to become unavailable in the west .
the program does not schedule an observation if the beginning or ending zenith distance would be greater than 65@xmath46 . because the camera is on a german mount , which must be flipped 180at the meridian ,
observations are scheduled so they can be completed without crossing the meridian .
overall , 45% of the fields were observed in the western part of the sky and 55% in the east .
biases , darks , and flats are taken during twilight at the beginning and end of each night .
each time we normally record 15 biases , five darks each of 20 minute duration ( to match the maximum exposure times of the science images ) , and five flats for each of the two filters with the light box illuminated by two sets of incandescent lamps and for the h@xmath12 filter with the light box illuminated by two sets of hydrogen emission tubes .
the exposure times for the flats were adjusted to give images with a mean brightness several thousand times the readout noise , but below the level where the variance of the detector noise becomes non - linear .
the system noise is a combination of photon noise from the sky foreground ( with a small contribution from h@xmath12 line emission ) , dark current noise , and readout noise .
we can calculate the expected noise on a single exposure from the following formula : @xmath47 where n is the noise level ( adu ) ,
g is the gain ( e adu@xmath31 ) , @xmath13 is the conversion factor of rayleighs to photon intensity at 656.3 nm ( @xmath48 ) , @xmath49 is the intensity in the h@xmath12 line ( r ) , @xmath50 is the specific intensity of the sky foreground ( r nm@xmath51 ) , @xmath52 is the bandwidth of the filter ( nm ) , a is the area of the lens ( @xmath53 ) , @xmath54 is the solid angle of the sky focused on one pixel ( arcminute@xmath55 ) , t is the transmission of the optical system , q is the quantum efficiency , d is the dark current ( e s@xmath31 ) , @xmath56 is the integration time ( s ) , and @xmath57 is the readout noise ( electrons ) . using the parameters given in section [ equipment ] , a value of 10 r for the brightness of the geocoronal h@xmath12 line @xcite , @xmath58 = 13 r / nm @xcite for the specific intensity of the sky foreground , an assumed value of t = 0.6 ( for the combined transmission of lens , filter , and ccd window ) , filter bandwidths of @xmath52 = 3.2 nm and 12.2 nm for the h@xmath12 and continuum filters respectively ( section [ filters ] ) , and @xmath59 = 1200 s and 300 s for the integration times of the h@xmath12 and continuum exposures ( section [ observationalprocedure ] ) , we get expected noise values of 25 adu for a single h@xmath12 exposure and 22 adu for a single continuum exposure .
typical noise values actually achieved are 30 adu and 24 adu respectively , in reasonable agreeement with expectation .
note that for the above parameters the contribution to the noise from photons is larger than the detector readout noise , as desired , but only by a factor of two .
the noise contributed by the dark current is completely negligible .
the expected signal , after subtraction of the dark current , can be calculated from the formula : @xmath60 it follows that the intensity scale factor @xmath61 , the number of adu corresponding to an intensity of 1 r , is given by @xmath62 with the same parameters as used above for calculating the expected noise , we would predict an intensity scale factor of 11.2 adu r@xmath31 for a 20-minute exposure .
the actual value , determined from observation of planetary nebulae ( section [ intensityscale ] ) is 6.8 adu r@xmath31 , a difference of 40% .
one or more of the factors @xmath63 , @xmath54 , @xmath64 , or @xmath65 may be overestimated , but if these estimates are reduced in order to generate better agreement between the expected and measured intensity scale factors , then the discrepancy discussed above between the expected and measured noise would increase .
perfect agreement can not be expected for such _ ab initio _ calculations , and we conclude that the system in the field performs reasonably close to expectations .
after co - adding the images ( section [ dataprocessing ] ) , the noise is less .
a typical value on the continuum - subtracted image ( section [ continuumsubtraction ] ) is 14 adu ( 2 r ) , and on the images smoothed with a 5 pixel median filter ( section [ smoothing ] ) it is 5 adu ( 0.7 r ) .
large scale features of sub - rayleigh intensity are visible on several images .
we estimate that the faintest feature visible on our images has a brightness of approximately 0.5 rayleigh .
although the aim of this survey is the discovery and study of objects emitting at the h@xmath12 line , there may be some applications in which the continuum point sources ( stars ) are of interest .
the limiting r - magnitude on the co - added images is about 14.8 at b = @xmath1 , deteriorating ( because of confusion ) to r = 12.5 at b = @xmath66 . as mentioned in section [ observationalprocedure ] , we took short exposures of the south celestial pole region before and after each set of science images .
aperture photometry of several stars on these polar images showed that variations in extinction during one night or from night to night ( for those nights for which data appears in the atlas ) rarely exceeded 0.1 mag ( 10% ) . similarly ( section [ observationalprocedure ] ) , we observed regions containing bright standard objects on most nights on which data appearing in the atlas were obtained .
we measured the brightness of carefully selected standard regions on these images . for those standard objects which were observed over a large range of elevations , we derived an average extinction coefficient of 0.082 mags airmass@xmath31 at 656.3 nm , compared to the standard extinction for ctio at this wavelength of 0.096 mags airmass@xmath31 @xcite .
the difference is within the seasonal and night - to - night variations of extinction at ctio @xcite .
none of our data were obtained at zenith distances greater than @xmath67 , an air mass of 2.4 and extinction of 0.2 mag ( 20% ) , and most images were observed much closer to the zenith .
again , individual intensities obtained on particular nights never differed by more than than 0.1 mag ( 10% ) from that expected for the zenith distance at the time of observation .
both the studies of the polar images and the standard objects indicate that variations in extinction do not affect the brightness of our images by more than 10% .
therefore we have made no attempt to correct our data for atmospheric extinction .
co - adding of the calibration exposures ( biases , darks , and flats ) is done on site automatically , with the medians of evening and morning sets recorded separately .
these , and all science images , are recorded on tape and shipped to swarthmore for later processing . no significant differences between morning and evening calibration images were detected , and these were simply averaged , if both existed .
the science images were corrected for bias and dark current in the usual way and then divided by the flat field image appropriate to each filter , normalized to the mean flat value of the central 100 pixel @xmath68 100 pixel area . early in the project we assumed that the lightbox was uniformly illuminated and emitted light at a constant brightness in all directions .
we then attributed an observed decrease in brightness of the flats near their edges to vignetting in the camera system .
however , using the flats as observed caused a rise in brightness of both the h@xmath12 and continuum images near their edges , leading us to conclude that the lightbox emission was not truly isotropic , but decreased in intensity with angle to the normal .
we do not know the origin of this effect , but suspect it may be caused by nonsymmetric scattering properties of the plastic diffuser screens .
we corrected for this effect by determining the average of each flat image at constant radial distances from the center , and dividing the flat by this circularly symmetric function .
such a procedure preserves the features in the flat image due to pixel - to - pixel gain differences and to such things as dust specs on the camera window , but any symmetric features ( such as true vignetting or gain variations which are a function of distance from chip center ) are not corrected for and may propagate into the final images .
however , examination of the continuum images at high galactic latitude ( @xmath69 ) , where there are no bright features , shows that on average they are flat to within @xmath70 ( 1.3 r ) out to @xmath71 from the center ( see figure [ avgduplot ] ) , indicating that such vignetting - type effects , if they are present at all , are fairly small .
the sets of six continuum and five h@xmath2 images are aligned and coadded separately , pixel by pixel , after removing transients produced by cosmic rays , meteors , airplanes , and satellites .
we characterize a value as an outlier , to be excluded from the co - addition , if it lies more than 3.5 times the standard deviation from the mean .
given the nature of our data , determining the mean and standard deviation is not completely straightforward .
first , since we have a small number of values contributing to each pixel ( five or six ) , one of which ( if influenced by a transient event ) may differ by a large amount from the others , a straight average may not give a good estimate of the mean .
likewise , the rms average of the deviations from this mean may not give a good estimate of the standard deviation .
therefore we choose to take the median rather than the straight average as an estimate of the mean , and the median of the absolute deviation from this median as an estimate the standard deviation .
a further complication is that each image is taken at a different zenith distance and therefore will contain a different airglow and geocoronal contribution to the foreground level .
the offsets due to these differences are much larger than the noise
. therefore the median of pixels from the original set of images would give the value of the middle image , with no averaging of the noise .
likewise , the median of the absolute differences from this median would overestimate the standard deviation , for it would include the offsets as well as the variations due to noise . the first step ,
therefore , is to make corrections for the differences in foreground level .
we assume that the continuum images consist of a galactic ( and extragalactic ) component @xmath72 plus a foreground component @xmath73 which varies with time : @xmath74 because the spatial transmission function of the h@xmath12 filter ( figure [ hfiltertrans ] ) is still included in the line contribution to the h@xmath12 images at this stage ( it is removed later in the processing
see section [ continuumsubtraction ] ) , the corresponding relation for the h@xmath12 images is a little different : @xmath75 + t_h [ h_{for } ( t ) + h_{g } ] \label{hobs}\ ] ] where @xmath76 is the galactic contribution to the h@xmath12 line , @xmath77 is the time - variable foreground ( geocoronal ) contribution to the h@xmath12 line , and @xmath78 is the spatial transmission function for the h@xmath12 line .
the factor 1.2 accounts for the different filter bandwidth and exposure times for the h@xmath12 images compared to the continuum images , and is the average of the factors needed to match star intensities during the continuum subtraction process ( section [ continuumsubtraction ] ) . note : equation [ hobs ] does not account for the contribution of telluric oh line emission ( section [ filters ] ) to the h@xmath12 images , for which a different spatial transmission function should be used .
to the extent that these lines are important , the separation of the line and continuum components of the h@xmath12 image is only approximate .
we first determine the average brightness level , @xmath79 or @xmath80 , of the central 800 @xmath68 800 pixel area of each image [ using the `` sky '' procedure of the idl daophot package @xcite ] . for the @xmath81 continuum image
, we simply add a constant @xmath82 to bring it to the brightness level of the first image of the set .
for the h@xmath2 images , we use the function @xmath83 determined from equation [ cobs ] and from the @xmath79 values to interpolate the first term of equation [ hobs ] to the time of each h@xmath12 observation .
subtracting this term from the observed brightness level @xmath80 gives the line contribution to the brightness level , the second term of equation [ hobs ] , and thus the function @xmath84 .
it is then a simple matter , using equation [ hobs ] , to correct each image to the brightness level at the time of the first image .
this procedure should compensate for linear changes in airglow and geocoronal level with observation time , at least within the time frame of a particular set of images .
( in section [ georemoval ] we discuss a procedure for matching levels of _ all _ images . ) since all the images of the set now have the same average value , medians of the image values at each pixel and medians of the absolute deviation at each pixel should give a good estimate of the mean and standard deviation .
next , reflective ghosts are identified and marked on each image . this procedure
could not be completely automated because of a random variation by several pixels of the position of the axis of reflection from image to image .
this is probably due to an imperfect constraint of the plane of the filter wheel or a slight looseness of the filters in their mounts .
an empirical study of a few images showed that the ghosts are about 0.05% of the brightness of the stars , so that only the very brightest stars , those with central brightness exceeding the full - well capacity of the chip ( 80,000 e @xmath33 60,000 adu ) , produce ghosts on an image greater than the noise . in the ghost - removal procedure ,
an approximate axis position is assumed , the positions of the ghosts of the brightest stars calculated and marked on the image display , and the axis position adjusted interactively until the marked positions coincide with the real ghost positions . a circle of radius 12 pixels ( @xmath85 ) around each ghost is marked and not used when the images are co- added .
next the images are approximately aligned by shifting and rotating to remove the positional dither .
they are then aligned more exactly by finding the positions of all the intermediate brightness stars , those with central intensity between 10 times the sky foreground and a level of 14000 adu , and then finding the shifts and rotation which minimize the differences of position between the stars on each image compared to the first one of the set .
the aligned images are then co - added after removal of cosmic ray events and other transients . as described above , this is done by first finding at each pixel position the median of the set of images , and the median of the absolute difference from this median . for small numbers of points ,
the mean can be significantly different from the median , so now the true mean at each pixel position is calculated , excluding those points whose difference from the median is more than 3.5 times the median absolute difference .
regions containing ghosts are also excluded when calculating the mean . at this point
we have co - added means of the images taken through the h@xmath12 and continuum filters , with transients and ghosts removed .
next the continuum image is aligned to the h@xmath12 image , using the same procedure as described above .
aperture photometry on the same intermediate brightness stars on each image establishes an intensity scale factor between the two mean images . the continuum image , multiplied by this scale factor ,
is then subtracted from the h@xmath12 image .
the continuous airglow foreground and most stars are removed by this procedure , but the geocoronal h@xmath12 emission and residuals of the bright stars remain .
these bright star residuals are probably due to slight differences of the point spread functions on the two images , small non- linearities of the ccd response , or saturation of the detector ( for the very brightest stars ) .
a sample of the continuum , h@xmath12 , and star - subtracted images for one field is shown in figure [ orionquad ] .
the image should in principle now contain only line emission , primarily h@xmath44 , both interstellar and geocoronal , and stellar h@xmath12 emission and absorption , as modified by the transmission function of the h@xmath12 filter .
( there remains also some residual contamination by telluric oh lines section [ filters ] . ) the basic transmission function is determined by dividing the h@xmath12 flat ( produced by observing the lightbox illuminated by the hydrogen emission lamps ) by the continuum flat ( produced by observing the lightbox illuminated by the incandescent lamps ) , both with the h@xmath12 filter in place , and both normalized to the mean of the central 100 pixel @xmath68 100 pixel area . in principle the h@xmath12 component of the original images should be corrected for this transmission factor before they are shifted and co - added , but it is difficult to separate accurately the line and continuum fractions of the image before the co - addition .
therefore we constructed an effective h@xmath44 transmission function by de - dithering and co - adding the original function in precisely the same way as was done for the original images obtained with the h@xmath12 filter .
the star - subtracted image was then divided by this function to obtain a corrected image , with the transmission characteristics of the h@xmath12 filter removed .
for many fields , this procedure produces a good image , planar except for real interstellar h@xmath12 features . for some fields
, however , obvious instrumental effects remain .
one of these manifests itself as an increase of brightness with radius , in shape like the inverse of the h@xmath12 filter transmission function ( figure [ hfiltertrans ] ) .
this probably arises from imperfect subtraction of the continuum , perhaps due to the foreground brightness changing non - linearly with time .
if some continuum radiation remains in the image , then division by the h@xmath12 filter transmission function will cause such an upturn near the edges . to correct for this phenomenon , we iteratively subtracted from the images showing this effect a constant divided by the h@xmath12 filter transmission function until we found the value which minimized the upturn near the edges . even with this effect removed , some images showed a residual pattern at low amplitude ( a few rayleigh ) of a circularly symmetric brightness , rising and then falling with radius .
such an effect could arise from one or more emission lines in the night sky spectrum shifting across the filter band with increasing angle of incidence .
but the position and shape of the maximum does not seem to correspond with that expected for any known line .
it could also be some artifact of the flat - fielding procedure .
the cause of this effect is not understood .
we determined this function individually where we could , for those images which contained no bright interstellar h@xmath12 features , by finding the average of each image at constant radial distances from the center , after normalizing the image to have zero mean , and subtracting this circularly symmetric function from the image . for those fields , mainly at low galactic latitudes , where the genuine h@xmath12 features were too bright for this procedure to work , we merely subtracted the average of all the functions determined at high latitudes .
an example showing graphs of the average brightness as a function of radius for an original continuum - subtracted image and for the same image after these two stages of correction is shown in figure [ curlsequin ] . these procedures produced images which are fairly planar , but they do contain geocoronal h@xmath12 emission ( and possibly some residual oh emission ) as well as h@xmath12 emission from the interstellar medium . the geocoronal emission is variable in time and with position on the sky .
we have assumed that it changes fairly slowly in both dimensions , and have modelled it as a plane with a certain level ( which we call a piston ) , slope in declination ( tip ) , and slope in right ascension ( tilt ) .
we determined these parameters ( piston , tip , and tilt ) for each image so that neighboring fields fit together into a smooth mosaic , as described in section [ georemoval ] .
finally , as mentioned above , the star - subtraction procedure does not work perfectly , so that the images contain some high spatial frequency residuals of imperfectly subtracted bright stars .
we removed these as best we could , but with some loss of resolution , by applying a spatial median filter of width 5 pixels ( 40 ) to the images .
the smoothed image for the orion region is shown in figure [ orionquad]d . by automated pattern - matching of stars in our images to those in the sao catalog , we determine astrometric solutions for each mean and mean continuum image .
the solution , a gnomonic projection , is stored in the image header according to standard practice @xcite . here
we detail the method of solution and describe its accuracy .
the method is similar to others that we might have used ( e.g. , @xcite ) if they had existed when we first needed them .
after each observation , the telescope control software annotates the image header with the celestial coordinates of the intended center of the image . from the image ,
our algorithm generates a list of stars centroids and instrumental magnitudes . from the sao catalog ,
our algorithm culls a list of j2000 positions and apparent visual magnitudes of stars within 9 of the intended center .
the stars celestial positions are converted to angles @xmath86 of a gnomonic projection centered on the image s intended center ( e.g. , @xcite ) .
the pattern - matching algorithm compares the list of instrumental centroids and magnitudes to the list of gnomonic angles @xmath86 and sao magnitudes .
it uses valdes method of similar triangles @xcite as implemented by @xcite to determine a tentative match between the 50 brightest stars in one list and the 50 brightest stars in the other list .
it then iteratively refines the solution by extending the star lists , typically to 200 stars in common .
the custom keyword n_astrom in the image header logs the number of stars that contributed to the astrometric solution .
the instrumental positions match the catalog positions to @xmath87 pixel ( 8) r.m.s . if first - order plate equations are applied .
the residuals of the first - order fit exhibit a third - order polynomial with maximum amplitude of @xmath88 pixels ( @xmath89 ) . fitting with third - order polynomials reduces the r.m.s . of the residuals by a factor of two , but for simplicity we chose to retain only the first - order fit in the image header .
astrometric solutions were computed separately for the co - added h@xmath12 and continuum images . because the continuum image was shifted to match the h@xmath12 image prior to subtraction , the continuum - subtracted images and the smoothed images have the same astrometric solution as the h@xmath12 image .
to calibrate the intensity scale of our images , we compared aperture photometry of many compact ( diameter @xmath90 ) planetary nebulae ( pne ) on our images with the spectrophotometric fluxes reported by @xcite ( see figure [ dopitaplot ] ) .
dopita & hua report fluxes for ( 656.3 nm ) and the forbidden line of [ n ii ] at 658.34 nm .
the latter spectral line and its companion at 654.81 nm contribute to the flux in our `` '' images .
the intensity of [ n ii ] 654.81 nm is 1/3 that of [ n ii ] 658.34 nm @xcite .
the transmission of the `` '' filter , at our 25 c operating temperature and at normal incidence ) .
the net effect is that the intensity - weighted average transmission of the two [ n ii ] lines is a weakly varying function of angle of incidence that we have not attempted to allow for in determining the overall intensity calibration .
] , is 39% , 78% , and 26% at 654.81 nm , 656.3 nm , and 658.34 nm respectively . by accounting for the expected relative contributions to the flux measured through our filter for each pn according to the fluxes given by dopita & hua , we determined the single scalar that best matched our instrumental fluxes ( in analog to digital units , or adu ) to theirs ( in ) .
we chose to use in the comparison group only those pne that had an flux larger than @xmath91 , because many of the fainter ones had large uncertainties in our aperture photometry .
if the aperture photometry routine [ aper in the idl daophot package @xcite ] reported an error of less than 6% for any particular measurement , we replaced that error estimate with 6% ; three such replacements occurred in each of the two groups ( open or filled circles ) plotted in figure [ dopitaplot ] . for those 18 pne selected for the intensity calibration , the r.m.s .
differences between our fluxes , determined from continuum - subtracted images and corrected for [ n ii ] emission , and the fluxes of dopita & hua , is 8% . because dopita & hua s spectrophotometry is accurate to better than @xmath92%
, we consider the 8% r.m.s .
scatter between our measurements and theirs as indicative of the limit of precision of fluxes of unresolved sources measured from our continuum - subtracted images .
we attribute this 8% limit ( @xmath93 ) on precision to uncorrected extinction ( section [ performance ] ) and the undersampling of the optical system s point spread function by our pixels .
the transmission of our `` continuum '' filter at the wavelength of the [ sii ] line at 673.1 nm is 16% at normal incidence , and as large as 70% at the corners of our images ( [ filters ] ) , but because our continuum images are exposed for 25% of the time of the images , and because in all of these pne the [ s ii ] @xmath94673.1 line is less than 10% the brightness of , we have not corrected our fluxes for [ sii ] as we did for [ n ii ] . as a worst - case example , a single measurement of a particular pn that happened to be near the corner of an image , the leakage of [ s ii ] @xmath94673.1 would amount to less than 2% of the flux . in an image in which that same pn appears near the center , the effect would be less than 0.4% .
using the pne calibration just described , and converting from flux to surface brightness units , we find that in a single pixel of our images , 1 r = @xmath95 adu .
this factor has been applied to all the continuum and smoothed images available for public use ( section [ access ] ) ; their intensities are expressed in units of decirayleighs ( 1 dr = 0.1 r ) . as an independent check of the calibration using the 18 pne described previously , we compared our measurement of the total flux of the very bright rosette nebula to the flux measurement of @xcite . integrating the flux within a 1 radius circle , as celnik did , and setting our image s zero point to agree with the dashed contours of celnik s figure 2 , and correcting our measurements for [ n ii ] leakage with the spectral line ratio measured for the rosette , i(6583)/i ( ) = 0.37 ( @xcite ; position 58 ) we find 1 r = @xmath96 adu , where the uncertainty is set by the 7% uncertainty reported by celnik for the integrated flux . because the pne calibration is based upon many objects measured on many images rather than a single object ( the rosette ) measured on a single image , the rosette calibration is secondary to the pne calibration that we actually applied , 1 r = @xmath95 adu .
the ratio of the two ( 7.4/6.8 = 1.09 ) indicates that our intensity calibration is consistent with others to approximately @xmath97% .
( see section [ others ] for comparisons with other surveys . )
each individual image has an unknown offset due primarily to geocoronal emission , which is spatially and temporally variable . to account for variations in geocoronal emission ,
we have adopted a model in which the extra foreground emission is a 2-dimensional ( 2-d ) plane , the three parameters of which we call `` piston '' for the mean value of the intensity of the foreground ( ) and `` tip '' and `` tilt '' ( collectively ) for the foreground s slopes .
thus , the galactic emission , @xmath98 , that we seek can be determined from the emission that we observe , @xmath99 , as follows : @xmath100 where @xmath101 is a linear transform of the x- and y - coordinates of each pixel : @xmath102 $ ] .
the challenge is to find the specific and for each individual image in a physically justified manner .
our approach to this challenge extends the method of in a few ways : 1 ) theirs solves for , while ours solves for and ; 2 ) theirs matches pixels in one image with pixels in an adjacent image using a simple shift in @xmath101 , while ours uses the astrometry stored in the fits headers ; 3 ) theirs uses a median of differences of matched pixels in each overlapping region , while ours fits a 2-d plane to those differences ; 4 ) the overall zero - point(s ) for the pistons ( and tips and tilts in our case ) are determined by different methods ; 5 ) their algorithm for solving the system of equations is a simplex method ; ours is a singular value decomposition .
the large angular size of our images required us to implement item 1 because the foreground sky is not constant across an image , item 2 because of image distortion and rotation , especially near the celestial pole , and item 3 because according to our model the difference between two adjacent images is a 2-d plane whereas in their model the difference is a single scalar .
our method of determining the specific and for each individual image is best described by an example .
figure [ 4images ] shows four overlapping images .
there are five pairs of overlapping images : images 1 and 2 , 1 and 4 , 2 and 3 , 2 and 4 , plus 3 and 4 overlap .
we begin with the overlap between image 1 and 2 .
we subsample the `` good '' part of image 2 on a 20 pixel by 20 pixel grid and compute the position on the sky ( @xmath103 ) of each pixel from image 2 s astrometric solution .
for those ( @xmath103 ) we compute the ( x , y ) in image 1 s astrometric frame in order to pair the pixels in image 2 with pixels in image 1 at the same ( @xmath103 ) .
each such pair of pixels provides an equation such as @xmath104 where the subscripts refer to the image ( 1 or 2 ) and the superscripts ( @xmath105 or @xmath106 ) refer to observed or galactic intensity .
because the galactic emission is constant with time , i.e. , @xmath107 , it cancels out in the difference and we have @xmath108 in this equation we have observed @xmath109 at location @xmath110 of image 1 and have observed @xmath111 at @xmath112 of image 2 .
the unknown parameters for images 1 and 2 are @xmath113 and @xmath114 .
likewise , each pixel in image 3 that overlaps image 2 provides an equation @xmath115 and similarly for the three remaining overlapping image pairs .
we can define a matrix equation @xmath116 where vector @xmath117 contains the intensity differences between all pixel pairs of image i and image j , i.e. , @xmath118 ; vector @xmath119 contains the model parameters , i.e. , @xmath120 ; and describes the coupling of @xmath117 and @xmath119 by equations like those above . in the illustrated example
, @xmath119 would be 12 elements long ( 4 images and 3 parameters per image ) .
each pair of pixels from adjacent images adds an element to @xmath117 and a row to .
we then solve @xmath116 for @xmath119 by a standard method of linear algebra , singular value decomposition .
however , even if we have more equations than unknowns , the equations are degenerate such that the overall piston , tip , and tilt are arbitrary .
in many circumstances it may suffice to append three rows to to force the mean values @xmath121 and @xmath122 to zero .
another option is to refer the data to intensity values known by other means . in our case
we get additional equations by comparing our data with wham data .
if all calibrations are done correctly , if we average our data over a 1-diameter circle of sky and integrate wham s spectra over velocity while eliminating the geocoronal emission in the spectra , and if the contamination of our data by telluric oh emission is negligible , then our measurement of the galactic intensity and wham s should be equal , i.e. , @xmath123 .
three wham data points are shown in figure [ 4images ] as circles . in our hypothetical example , we append to equations like the following : @xmath124 in this equation , which applies to point c in figure [ 4images ] , @xmath125 is the observed intensity of image 3 centered on point c and averaged over a 1 diameter circle that corresponds to the wham beam size ; @xmath126 is the intensity at point c as measured by wham ; and @xmath127 is the location of point c in image 3 s coordinate system . in principle , only three such fiducial points ( e.g. , points a , b , and c ) are needed to constrain the overall piston , tip , and tilt .
we append to equations similar to the one above for image 2 and point a , image 3 and point b , and image 4 and point c , but not image 2 and point b because the circle around point b does not lie wholely within image 2 s `` good '' region , making @xmath128 inaccurate . for small numbers of images ,
it is practical to simply use a very sparsely subsampled grid of pixels , and let every pair of pixels lengthen @xmath117 by one element and by one row . because the time to solve the equation @xmath116 for @xmath119 grows in proportion to the product of the number of pixel pairs times the square of the number of free parameters of the model ( in our case , three times the number of images ) , a more efficient approach
is required for large numbers of images .
we chose to subsample our images , and then in each region of overlap , we inscribed a rectangle and within the rectangle we fit a 2-d plane to the differences in intensities of paired pixels .
we then evaluated the plane at the corners of the rectangle and used those values as representative of all the paired - pixel differences and locations in the region of overlap . choosing the rectangle s four corners of course over - determines the plane , but doing so is a simple and approximate method of representatively weighting the data as it would be if all pixels within the irregularly - shaped region of overlap were given equal weight .
an iterative method is used to fit the plane robustly , i.e. , such that outliers do not affect the solution .
simulations in which a large image is carved into a set of overlapping images and then random pistons , tips , and tilts are added to those images , and subsequently solved for , show that our algorithm can recover the pistons , tips , and tilts very accurately and reliably under optimum conditions .
in practice , with our actual images , our experience with the algorithm has been much less satisfying . with our images ,
if we solve for and simultaneously the solution tends to introduce large scale undulations in the mosaic of the ( galactic ) sky . wherever the solution is not tied to fiducial points , such as in the southern sky ( @xmath129 ) where no wham data exists , the solution tends to be unreliable .
slightly different input parameters give rather different results , even physically impossible results , such as negative brightness in some regions .
the solution tends to err in such a way that in extrapolating beyond the set of images containing wham fiducial points to those without any , erroneous gradients trade off with erroneous pistons , so that the solution is `` smooth '' from one image to another , but has the wrong overall `` figure '' or shape .
we tested the reliability and accuracy by using a subset of the wham data to determine a solution and a different subset to compare with our data after the solution has been applied .
for example , to test our algorithm s ability to extrapolate from the north to the south , i.e. , from where there is wham data to where there is not , we included as input ( cf .
equation [ eq - iwham ] ) wham data with declination @xmath130 , determined a solution , applied it to our images , and then compared our resulting intensities with wham data in the region @xmath131
this empirical test convinced us that in order to prevent the solution from diverging in the southern sky where no wham data exists , i.e. , @xmath133 , we had to provide additional , artificial fiducial points in that region that would be treated the same as actual wham observations . in order to create these artificial fiducial points , which are shown in figure [ figwhamaitoff ] , we fit a function of the form @xmath134 to wham data with @xmath135 and @xmath136 , but excluding the orion / eridanus region , @xmath137 .
the parameters @xmath138 and @xmath139 were fit by eye in order to ignore the somewhat - brighter regions remaining within those limits of @xmath140 , @xmath141 , and @xmath142 . the function
so selected , @xmath143 } , \label{cscblaw}\ ] ] is plotted in figure [ figcscblaw ] . that the sky is fainter at high latitudes than the extrapolation of a @xmath144 law determined at lower latitudes
is believed to be due to the local bubble , plus a small , increasing contribution for dust - scattered h@xmath12 at the lower latitudes @xcite .
if we assign all of this effect to the local bubble , then the isotropic component in equation [ cscblaw ] , -0.75 r , corresponds to the emission missing from that part of the wim that the local bubble has `` carved out . ''
for a mid - plane r.m.s .
electron density @xmath145 @xmath146 @xcite , the `` missing '' intensity of 0.75 r requires a 240-pc path length of fully ionized gas at 8000 k. that length is approximately equal to the 200-pc radius of the local bubble as modeled by @xcite , so our paramter @xmath147 rayleigh is consistent with prior models for the local wim . in the end , we chose to include data from all 542 images to determine simultaneously the and for each of the images . with 3166 pairs of overlapping images , and each region of overlap represented by 4 points as previously described , there were 12664 equations like equation [ i1mi2 ] .
we used 802 wham reference points spaced 4 in right ascension between @xmath148 , and spaced 5 in declination between @xmath149 .
we did not use any wham - measured surface brightnesses less than zero ( because those are probably erroneous ) or larger than 30 r because we wanted to avoid bright regions for which multiplicative errors might affect our estimate of the additive terms .
we also used 527 artificial points created as described previously ( cf . equation [ cscblaw ] ) .
the 802 wham fiducial points and 527 artificial fiducial points have 4803 separate measurements in our set of 542 images , resulting in 4803 equations like equation [ eq - iwham ] .
in order to give approximately equal weight to the two types of equations , i.e. , those that compared overlapping images ( equation [ i1mi2 ] ) and those that compared our data to wham data ( equation [ eq - iwham ] ) , we replicated the latter 4803 equations to produce 9606 such equations , or nearly as many as the 12664 equations of the other type .
we then solved the equations for the 542 images and by singular value decomposition .
we solved the equations twice ; prior to the second iteration , we discarded 52 ( 0.4% ) of the 12664 equations that were found to deviate by more than 10 times the average residual difference after applying the initial solution s and .
as expected , because only a very small fraction of the equations were excised , the second set differed only very slightly from the first set .
the pistons , tips , and tilts determined in the second set were applied to the continuum - subtracted ( and corrected ) images and the smoothed images ( sections [ continuumsubtraction][smoothing ] ) , and the latter were used to make the mosaic map mentioned in section [ description ] .
the survey consists of 268 fields covering the sky from @xmath1 to @xmath20 declination , with the same centers as those in the iras sky survey atlas ( issa)@xcite , plus , instead of issa region 001 , three fields centered at @xmath150 declination and 0 , 8 , and 16 hours right ascension to cover the polar region . in order to reduce the noise in mosaic maps and to allow confirmation of objects discovered at the faintest brightness levels , we repeated all fields with offsets of @xmath151 in both coordinates .
( the three polar fields were repeated at the same declination , @xmath150 , but displaced 4 hours in right ascension . )
the field centers are listed in table [ centerstable ] .
images of the individual fields are available from our web site ( section [ access ] ) .
some examples of interesting fields are shown in figure [ favorites ] , and a mosaic of the entire region of the sky included in the atlas is shown in figure [ mosaicfig ] .
all observations used in the survey were obtained in the period november 1997 to october 2000 , a time of increasing solar activity .
the virginia tech spectral - line survey ( vtss ) is an imaging survey similar to the one described here but for the northern hemisphere @xcite .
when the vtss observations of the northern hemisphere are complete , combining them with our southern hemisphere data will provide an all - sky mosaic with approximately 3 resolution .
comparison of vtss data with our own shows they are very similar , but with some differences , as the remainder of this section describes .
the vtss includes much of the milky way ( @xmath152 ) clearly visible ( @xmath153 ) from the martin observatory in southwestern virginia .
the vtss s cryogenically - cooled ccd camera has a 58-mm lens , 27 pixels , a scale of 16 pixel@xmath31 , and covers a 136 @xmath68136 field of view .
the 1.75 nm bandwidth ( fwhm ) of the vtss s filter limits the usable field of view ( cf .
equation [ cwlshift ] ) to a 10 diameter circle centered on the optical axis .
as soon as they are processed to final form , individual images from the vtss are posted on the vtss web site ( http://www.phys.vt.edu/~{}halpha/[http://www.phys.vt.edu/~\{}halpha/ ] ) . in order to make the comparison , we selected four continuum - subtracted images from the vtss : aql04 , aql12 , cnc02 , and ori11 .
[ the first three letters of each image s name refer to the constellation ( aquila , cancer , and orion ) and the number is simply a counter .
] each of the first three vtss images is an average of six 360-second exposures , and the last image ( ori11 ) is an average of twelve 600-second exposures .
we mapped , pixel - by - pixel , all of our final smoothed images that overlapped each vtss image onto that vtss image .
because our pixel size is 08 and the vtss images pixel size is 16 , each of our images has four pixels within the area of one vtss pixel .
between seven and ten of our images overlap each vtss image , but most of them do so only partially , such that typically two to four of our images cover a particular point on the sky .
thus , in mapping our images upon a vtss image , we average between 8 and 16 pixels data for each vtss pixel .
the resulting mosaic of our images , custom - made to match the pixel scale and orientation of the vtss image , is then nearly ready to be compared with the vtss image .
because a constant has been subtracted from each vtss image to force its median to be zero , we need to correct the zero - point of intensity scale of each vtss image .
rather than calculate a simple offset , we fit an arbitrary 2-d plane of emission to the difference between the vtss image and our custom - made mosaic . by subtracting that 2-d plane from the vtss image
, we effectively corrected the vtss image for whatever geocoronal or telluric oh foreground emission might be included in it .
( our images have been corrected already for this emission as described in section [ georemoval ] . ) at the top of figure [ fig - vtss ] , we show the vtss image aql04 and our custom - made mosaic , for side - by - side visual comparison .
notably , the structure of the diffuse emission is nearly identical , as it should be , because both of us are measuring the emission .
the vtss image has more noticeable residuals around stars , but aside from that the two images are nearly identical .
the width of the point spread function , measured from the same few unresolved emission knots , is @xmath154 pixels ( 32 ) fwhm in either image . in the vtss images
, this fwhm represents an intrinsic resolution limit , while in our case it represents a kind of effective width of the native resolution of our images and the @xmath155 kernel of the median filtering of our images , combined with the binning and averaging of multiple images that make up each custom - made mosaic .
also in figure [ fig - vtss ] , in the lower left panel , we show a pixel - by - pixel comparison of the vtss image aql04 and our mosaic image , after each has been convolved with a @xmath156 pixel ( @xmath157 ) kernel .
the convolution serves only to reduce noise and to reduce effects of slightly dissimilar point spread functions . rather than plot
_ every _ pixel , which would make the density of points in the plot too high , we randomly selected 1% of the pixels across the image for the plot . from this plot , and especially the one next to it , from the ori11 region , we determine that the intensity calibration of our data and the vtss data are not the same . for an object within the ori11 region ,
the surface brightness measured from a vtss image will be 80% that measured from our images .
for the aql04 region the measured value is 85% , but with greater uncertainty due to the smaller dynamic range of surface brightness in aql04 compared to ori11 . for the other two images , aql12 and cnc02 , the emission is too faint for the intensity ratio to be determined accurately .
for display purposes , we have corrected the vtss image aql04 in figure [ fig - vtss ] by dividing it by 0.80 , but we have not attempted to determine the cause of this effect or whether it affects other vtss images or perhaps only these two images that we have examined . in order to compare the sensitivity of our survey to that of the vtss , we further processed the images of the two regions aql12 and cnc02 , which have some faint emission in them , by median filtering each with a @xmath158 kernel and subtracting the result from each image in order to create residual images .
the purpose of creating residual images is to prevent genuine emission from affecting our measurement of the noise in the image .
the r.m.s .
surface brightness in those residual images were 0.23 r and 0.15 r for vtss images aql12 and cnc02 , respectively , and 0.22 r and 0.22 r respectively for their associated custom - made mosaics from our data . in determining the r.m.s .
, we used only the cores of the distributions , not the wings that are created by the residuals around stars .
thus , except for the possible difference in overall intensity calibration , when convolved to a similar resolution , our data and the vtss data give similar results in those regions of sky in which they overlap ( i.e. , for @xmath159 , and @xmath160 ) . as noted in section
[ intro ] , the wham survey is a velocity - resolved survey of h@xmath2 emission over 3/4 of the entire sky ( @xmath161 ) .
the wham survey consists of more than 37,000 individual spectra , each with velocity resolution of 12 km s@xmath31 over a 200 km s@xmath31 ( 0.44 nm ) spectral window .
each velocity - integrated spectrum has a 3@xmath162 sensitivity of 0.15 r over its 1 diameter beam .
geocoronal emission is separated from galactic emission spectrally in the wham data .
because of the wham survey s sensitivity and accuracy , it is an ideal data set to compare our own to , after appropriately integrating the wham spectra and smoothing our data to match wham s 1 angular resolution .
figure [ hassmwham ] compares wham data with our own . in figure
[ hassmwham ] , the wham values are velocity - integrated and have had geocoronal emission removed @xcite , while this work s values are the average within the 1 diameter beam of wham of our final smoothed images after removal of geocoronal emission as described in section [ georemoval ] .
the locations on the sky of the 834 points plotted in figure [ hassmwham ] are shifted 2 in right ascension with respect to the locations of the points used to fit and remove the geocoronal emission , in order to assure that the set of points used for the comparison contains none of the points used for the fitting . the r.m.s . of the residuals of points with wham - measured surface brightness
less than 2 r is 0.4 r. because the wham sensitivity is 0.05 r ( @xmath93 ) , the 0.4 r r.m.s .
difference between wham s values and our own is almost entirely due to our data s noise and is an empirical limit of the sensitivity of our data after convolution with a 1-diameter beam . for faint sources ,
the dominant errors are additive , but for bright sources , multiplicative errors could be more important than additive ones . in fitting and removing foreground emissions ( section [ georemoval ] ) , we specifically excluded regions brighter than 30 r so that any multiplicative errors that might exist would not affect the fitting too badly .
consequently , for a fair comparison of our intensity scale to wham s , we should include only regions much brighter than 30 r in order to reduce the effect of the additive offsets determined by the fitting procedure .
generally , bright regions are better than faint regions for comparisons of two independent scalings of intensity . for the 15 points in figure [ hassmwham ] with wham - measured surface brightness greater than 60 r
, the ratio of our values to wham s has a mean of 1.00 and a @xmath93 dispersion of 0.13 , i.e. , our intensity scale matches wham s .
ten points in the log - log ( top ) plot of figure [ hassmwham ] , i.e. , 1.2% of the total , are off - scale in the linear - log ( bottom ) plot ; some of these outliers are caused by imprecise continuum - subtraction or bleed trails near bright stars , some represent small fractions of very bright intensities , and the remainder occur at regions with large variation in surface brightness across the 1 diameter wham beam and may be caused by the inaccuracy in our modeling it as a uniform circular disk even though it is slightly tapered near its edge .
the wham survey is more sensitive than ours , and because of the spectral information it contains , the foreground emission can be removed , providing absolute intensities , whereas we must resort to the method of fitting neighboring fields ( section [ georemoval ] ) to put all of our images on the same intensity scale . on the other hand , our survey has greater angular resolution , which is important in the study of structure at small angular scales , such as in filamentary nebulae . to illustrate this point , in figure [ orieri ] we show the orion / eridanus region as observed in both surveys . at the other end of the angular resolution scale
are the photographic survey of the southern galactic plane with the uk schmidt telescope of the anglo- australian observatory @xcite , and the deep h@xmath12 survey by the marseille observatory group with a scanning fabry - perot interferometer located at the european southern observatory in chile @xcite .
the former will provide images of @xmath163 resolution over a @xmath164 field at a sensitivity of 5 - 10 r , but will be limited to the galactic plane ( @xmath165 ) .
the latter has lower resolution ( @xmath166 pixels ) and a smaller field of view ( @xmath167 ) but higher sensitivity ( 0.2 r ) and of course gives velocity information lacking in an imaging survey .
but it covers a very limited region of the galactic plane ( @xmath168 and @xmath169 ) .
figure [ highres ] compares one of our images with a corresponding field from the aao / ukst survey and one of the fields from the marseille survey , both of the latter taken from @xcite ( figs .
1 and 6 ) . images of individual fields in the survey may be examined on the web at http://amundsen.swarthmore.edu/shassa/ or http://www.lco.org / shassa/. four images , similar to those shown in figure [ orionquad ] , are available for each location : h@xmath12 , continuum , continuum - subtracted , and smoothed ( see section [ dataprocessing ] for a description of the processing involved in generating these image types ) .
individual images in fits format may be downloaded via the above web page .
each full image file contains about 4 mbytes , but compressed versions ( in gzip format ) are also available , averaging about 1 mbyte each
. those who wish to download a large number of images may find it more convenient to do so via the anonymous ftp server amundsen.swarthmore.edu/shassa/ or the alternative server ftp.lco.org / shassa/. all of the survey images can also be obtained ( in compressed form ) as a set of three cd - roms .
contact wayne rosing ( ) for details .
those who use the survey images in a publication or publish research based on the survey images should include an acknowledgement to the `` southern h - alpha sky survey atlas , which was produced with support from the national science foundation '' and make reference to this paper .
* reminder : * several factors may affect the photometric accuracy of the images in this atlas : \a ) the removal of the geocoronal foreground emission ( section [ georemoval ] ) is intended to offset each image so as to make them all consistent with sparsely - sampled wham data north of declination -30 . the accuracy of that process is discussed in section 6.2.2 and displayed in figure 13 . south of declination -30 , the `` background '' intensity of each image is unverified and for some applications should be treated as a free parameter .
\b ) [ nii ] emission may contribute significantly to the `` h@xmath12 '' brightness and [ sii ] emission may distort the `` continuum '' brightness of some astronomical objects ( sections [ filters ] and [ intensityscale ] ) .
\c ) telluric oh emission may contaminate both the h@xmath12 and continuum images ( sections [ filters ] , [ continuumsubtraction ] , and [ corrections ] ) .
\d ) corrections for other instrumental effects ( sections [ calibrations ] and [ corrections ] ) may distort the relative brightness across some objects .
users are advised to consider carefully how the reduction procedures described in this paper may affect their particular photometric applications .
we have purposely excluded new scientific analysis from this paper .
however , papers already published illustrate some of the scientific applications for which the type of data in this survey is appropriate .
first , the large size of the images ( @xmath170 ) allows objects , such as the large magellanic cloud , to be seen in their entirety without the need for piecing together a mosaic @xcite .
second , objects of low surface brightness but large angular size will be found here which have been missed in other studies .
two examples are the shell around the planetary nebula abell 36 @xcite and the 20-diameter filamentary bubble centered on @xmath171 @xcite visible to the upper - left of the gum nebula in figure [ mosaicfig ] .
third , previous studies which were mainly qualitative can now be made more quantitative .
for example , @xcite calibrated the sharpless brightness classes of diffuse nebulae on an absolute scale of rayleighs .
[ hall also discovered faint outer extensions of several planetary nebulae with radii many times the radii listed in the catalog of galactic planetary nebulae @xcite . ]
fourth , the correlation of h@xmath12 emission with dust emission can be very informative .
dust clouds , bright at 100@xmath30 , often show a `` silver lining '' of h@xmath12 emission on the side facing strong ionizing flux coming from the uv - emitting stars in the galactic plane ( @xcite , figure 7 ) . in another example @xcite a comparison of 100@xmath30 and h@xmath12 emission helped delineate the three - dimensional structure of the complex region near the orion belt star @xmath172 ori .
finally , our survey has proved useful in setting limits on galactic emission which might contaminate the dasi measurements of the cosmic microwave background @xcite .
we expect that many other applications will be found for the images in this atlas .
in this paper , and on the accompanying web sites ( http://amundsen.swarthmore.edu/shassa/ or http://www.lco.org/shassa/ ) , we present a digital atlas of the southern sky ( @xmath0 to @xmath1 ) as seen at 656.3 nm , the h@xmath12 emission line of hydrogen .
the atlas consists of 542 fields , each field covering a @xmath173 region of the sky .
the basic angular resolution is @xmath174 pixel@xmath31 .
smoothing to a resolution of @xmath175 allows features as faint as 0.5 rayleigh to be seen .
we are grateful to the following students who have participated in various phases of this project : peter austin , gang chen , katherine hall , nini khosrowshahi , eun oh , yukhi tajima , and anteneh tesfaye ( swarthmore college ) ; nat farney , brett schneider , and dan seaton ( williams college ) ; lynne raschke and andrew voellmy ( haverford college ) ; and chad bender , ray chen , patrick hentges , dan logan , ian odwyer , and jim pulokas.(univ . of illinois ) .
gang chen also worked for one year as a practical trainee at las cumbres observatory , where he was responsible for development of the code governing the robotic observing protocol , for the design and construction of the light box , and for the initial development and testing of the camera system .
the project has been supported by grants or contracts from the national science foundation ( ast 9529057 & ast 9900622 to jeg , ast-9874670 to prm ) , nasa ( nas7 - 1260 to jpl ) , a cottrell scholarship from the research corporation to prm , las cumbres observatory , dudley observatory , the fund for astrophysical research , the university of illinois , and swarthmore college .
the observations were obtained at cerro tololo inter - american observatory , which is operated by the association of universities for research in astronomy , inc .
, under cooperative agreement with the national science foundation .
we are extremely grateful to the staff of ctio for their support of this project . without their hard work on our behalf
, the survey would have proceeded much more slowly , and might never have been completed .
acker , a. , marcout , j. , ochsenbein , f. , stenholm , b. , & tylenda , r. 1992 , strasbourg - eso catalogue of galactic planetary nebulae ( garching : european southern observatory ) barnard , e. e. 1927 , a photographic atlas of selected regions of the milky way ( washington dc : carnegie institution of washington ) broadfoot , a.l .
, & kendall , k.r .
1968 , j. geophys .
res , 73 , 426 calabretta , m. & greisen , e. w. 2000 , in astronomical data analysis software and systems ix , ( astronomical society of the pacific conference series , san francisco ) , 216 , 571 celnik , w. e. 1983 , , 53 , 403 chromey , f.r . , & hasselbacher , d. 1996 , , 108 , 944 dennison , b. , simonetti , j. h. , & topasna , g. a. 1998 , pub .
astron . soc .
australia , 15 , 147 dopita , m. a. & hua , c. t. 1997 , , 108 , 515 gaustad , j. e. , rosing , w. , mccullough , p. r. , & van buren , d. 1999 , iau symp . 190 : new views of the magellanic clouds , 190 , 99 gaustad , j. e. , mccullough , p. r. & van buren , d. 1996 , , 108 , 351 georgelin , y. m. , russeil , d. , amram , p. , georgelin , y. p. , marcelin , m. , parker , q. a. , & viale , a. 2000 , , 357 , 308 haffner , l. m. , reynolds , r. j. & tufte , s. l. 1999 , , 523 , 223 haffner , l. m. 2001 , private communication haffner , l. m. et al .
2001 , in preparation haffner , l. m. et al .
2001 , http://www.astro.wisc.edu/wham/ hall , k.j
. 1999 , senior thesis ( swarthmore pa : swarthmore college , dept . of phys . &
halverson , n. w. , leitch , e. m. , pryke , c. , kovac , j. , carlstrom , j. e. , holzapfel , w. l. , dragovan , m. , cartwright , j. k.,mason , b. s. , padin , s. , pearson , t. j. , shepherd , m. c. , & readhead , a. c. s. 2001 , astro - ph/0104489 , submitted to hamuy , m. , walker , a. r. , suntzeff , n. b. , gigoux , p. , heathcote , s. r. , & phillips , m. m. 1992 , , 104 , 533 hamuy , m. , suntzeff , n. b. , heathcote , s. r. , walker , a. r. , gigoux , p. , & phillips , m. m. 1994 , , 106 , 566 hewish , a. , bell , s. j. , pilkington , j. d. , scott , p. f. , and co11ings , r. a. 1968 , , 217 , 709 jenkins , e. b. , wo ' zniak , p. r. , sofia , u. j. , sonneborn , g. , & tripp , t. m. 2000 , , 538 , 275 johnson , h. m. 1953 , , 118 , 370 landsman , w. 2001 , http://idlastro.gsfc.nasa.gov/ le coarer , e. , amram , p. , boulesteix , j. , georgelin , y. m. , georgelin , y. p. , marcelin , m. , joulie , p. , & urios , j. 1992 , , 257 , 389 marsden , b. g. 1982 , , 64 , 284 mccullough , p. r. , bender , c , gaustad , j. e. , rosing , w. , & van buren , d. 2001a , , 121 , 1578 mccullough , p.r .
, gaustad , j.e . ,
rosing , w. & van buren , d. 1999 , in the sloan summit on microwave foregrounds , ( astronomical society of the pacific conference series , san francisco ) , 181 , 253 mccullough , p. r. , gaustad , j. e. , rosing , w. & van buren , d. 2001b , in tetons 4 : galactic structure , stars and the interstellar medium , ( astronomical society of the pacific conference series , san francisco ) , 231 , in press mccullough , p. r. , reach , w. t. , and treffers , r. r. 1990 , , 22 , 750 mink , d. j. 1997 , asp conf .
ser . 125 : astronomical data analysis software and systems vi , 6 , 249 netterfield , c.b . ,
jarosik , n. , page , l. , wilkinson , d. , and wo11ack , e. 1995 , , 445 , 69 nossal , s. , reynolds , r. j. , roesler , f. l. , scherb , f. , & harlander , j. 1993 , , 98 , 3669 osterbrock , d. e. 1989 , astrophysics of gaseous nebulae and active galactic nuclei ( mill valley , ca : university science books ) osterbrock , d. e. & martel , a. 1992 , , 104 , 76 osterbrock , d. e. , fulbright , j. p. , martel , a. r. , keane , m. j. , trager , s. c. , & basri , g. 1996 , , 108 , 277 parker , q. a. & phillipps , s. 1998 , pub .
astron . soc .
australia , 15 , 28 parker , r. a. r. , gull , t. r. , and kirschner , r. p. 1979
, an emission - line survey of the milky way , nasa sp- 434 , ( washington dc : nasa ) regan , m. w. & gruendl , r. a. 1995 , in astronomical data analysis software and systems iv , ( astronomical society of the pacific conference series , san francisco ) , 77 , 335 reynolds , r. j. , scherb , f. , & roesler , f. l. 1973 , , 185 , 869 reynolds , r. j. 1991 , , 372 , l17 reynolds , r. j. , tufte , s. l. , haffner , l. m. , jaehnig , k. , & percival , j. w. 1998 , pub .
astron . soc .
australia , 15 , 14 reynolds , r. j. 2001 , private communication richmond , m. w. 1996 , private communication ( source code and documentation are available at http://freshmeat.net/projects/match/ ) russeil , d. , georgelin , y. m. , amram , p. , georgelin , y. p. , laval , a. , & marcelin , m. 1998 , pub .
astron . soc .
australia , 15 , 9 seife , c. 2001 , science , 292 , 823 sivan , j. p. 1974
, suppl . , 16 , 163 smith , w. j. 1990 , modern optical engineering ( new york : mcgraw - hill ) , p. 195
stetson , p. b. 1987 , , 99 , 191 stone , r. p. s. & baldwin , j. a. 1983 , , 204 , 347 valdes , f. g. , campusano , l. e. , velasquez , j. d. , & stetson , p. b. 1995 , , 107 , 1119 valls - gabaud , d. 1998 , pub .
astron . soc .
australia , 15 , 111 wheelock , s. l. et al .
1994 , iras sky survey atlas : explanatory supplement , jpl publication 94 - 11 , ( pasadena : jpl ) , p. a-1 wood , k. & reynolds , r. j. 1999 , , 525 , 799 lllllrr 002 & @xmath176 & -8000 & @xmath177 & -79 43 & 305 6 & -37 1 + 003 & @xmath178 & -8000 & @xmath179 & -79 48 & 296 4 & -35 7 + 004 & @xmath180 & -8000 & @xmath181 & -79 60 & 291 8 & -29 1 + 005 & @xmath182 & -8000 & @xmath183 & -80 12 & 294 0 & -21 7 + 006 & @xmath184 & -8000 & @xmath185 & -80 17 & 300 8 & -17 6 + 007 & @xmath186 & -8000 & @xmath187 & -80 12 & 308 7 & -18 9 + 008 & @xmath188 & -8000 & @xmath189 & -79 60 & 313 8 & -24 9 + 009 & @xmath190 & -8000 & @xmath191 & -79 48 & 313 0 & -32 5 + 010 & @xmath176 & -7000 & @xmath177 & -69 43 & 309 1 & -46 8 + 011 & @xmath192 & -7000 & @xmath193 & -69 45 & 296 2 & -46 7 + 012 & @xmath194 & -7000 & @xmath194 & -69 50 & 286 0 & -41 8 + 013 & @xmath195 & -7000 & @xmath195 & -69 57 & 280 9 & -33 9 + 014 & @xmath196 & -7000 & @xmath196 & -70 04 & 280 8 & -25 0 + 015 & @xmath197 & -7000 & @xmath197 & -70 11 & 284 6 & -16 9 + 016 & @xmath198 & -7000 & @xmath199 & -70 15 & 291 1 & -10 8 + 017 & @xmath200 & -7000 & @xmath201 & -70 17 & 299 5 & -7 7 + 018 & @xmath202 & -7000 & @xmath203 & -70 15 & 308 4 & -8 1 + 019 & @xmath204 & -7000 & @xmath205 & -70 10 & 316 5 & -11 9 + 020 & @xmath206 & -7000 & @xmath207 & -70 03 & 322 5 & -18 5 + 021 & @xmath208 & -7000 & @xmath209 & -69 55 & 325 5 & -26 9 + 022 & @xmath210 & -7000 & @xmath211 & -69 49 & 324 4 & -35 7 + & & & & & & + figure 6comparison of fluxes of compact planetary nebulae measured by us to fluxes reported by @xcite .
our fluxes have been corrected for the two [ n ii ] emission lines adjacent to ( see section [ intensityscale ] ) .
the pne represented by the solid circles were used to determine the overall gain calibration of our images ( i.e. , the number of adus per unit flux ) , thus the mean difference between the fluxes in `` this work '' and those of dopita & hua is zero by definition of the gain factor : @xmath95 adu corresponds to 1 rayleigh of emission in 1 pixel .
fainter pne ( which were not used ) are shown as open circles .
the logarithms are base 10 , and the units of the fluxes are .
figure 7this hypothetical example of four overlapping images is used to explain the mosaicing method in section [ georemoval ] .
each image is numbered and outlined by a bold solid line .
a thin solid line outlines the `` good '' region of each image , shown as solid white ; the grey region between the thin solid line and the bold solid line indicates the part of each image that is flagged as `` bad '' due to edge effects .
each region of overlap between a pair of `` good '' regions is indicated by an inscribed dashed rectangle .
circles indicate locations and sizes of regions observed by the 1 diameter beam of wham .
data within either the dashed rectangles and/or the circles contribute to determining each image s three free parameters , piston , tip , and tilt ( and ) .
figure 8locations of wham surface brightness data are plotted on this hammer - aitoff projection as plus signs ( @xmath215 ) ; these correspond approximately to 4 spacing in right ascension and 5 spacings in declination from -25 to + 10 .
those wham points used to determine the parameters of the fit in equation [ cscblaw ] are enclosed in squares .
the dots ( @xmath216 ) are spaced 4 in right ascension and 5 in declination from @xmath217 to @xmath218 , and with galactic latitude @xmath219 , but excluding regions near the lmc , smc , and extended parts of the gum nebula .
equation [ cscblaw ] is evaluated at the locations indicated by dots in order to constrain the fit for each individual image s piston , tip , and tilt in the southern sky as described in section [ georemoval ] .
figure 9equation [ cscblaw ] is plotted here as a solid line and is a fit to the subset of wham data defined in the previous figure ( the plus signs enclosed by squares ) .
the data plotted here are from wham ( reynolds 2001 ) .
figure 10some interesting fields from the atlas : 108 l = @xmath220 , b = @xmath221 , filaments in antlia ; 204 l = @xmath222 , b = @xmath223 , part of the eridanus filament ; 640 l = @xmath224 , b = @xmath225 , a bright rim in pyxis ; 576 l = @xmath226 , b = @xmath227 , part of the gum nebula and the vela snr ; 150 l = @xmath228 , b = @xmath229 sco ; 187 l = @xmath230 , b = @xmath231 oph .
figure 12comparison of our continuum - subtracted images to those of the vtss survey shows them to be similar with a few differences ( see section [ others ] for details not in this caption ) .
our data ( left image above ) , rendered to match one of the vtss images ( right image above ) , illustrates the comparison .
the vtss image , aql04 , 10 in diameter , centered on @xmath232 [ 2000 ] , with north up and east to the left , was selected at random for the comparison , and from it an arbitrary 2-d plane of emission was subtracted in order to minimize the residuals between it and the image rendered from our data .
the lower - left plot shows a pixel - by - pixel comparison of the data from those two images after convolution with a @xmath156 pixel ( @xmath157 ) kernel .
if the intensity calibration of our data and the vtss data were the same , the data would scatter about the diagonal solid line .
the dashed line is a first - order fit ( with zero intercept ) to the data with slope 0.85 .
a similar plot made for the vtss region ori11 , centered on @xmath233 [ 2000 ] , is shown at lower - right .
it has a large enough dynamic range of surface brightness to allow a good comparison of the vtss intensity calibration with our own .
the slope of the dashed line in the lower - right plot is 0.80 .
figure 13the surface brightness measured by wham and by this work are compared .
this work s values are the average of our data within the 1-diameter wham beam . in the top plot
, the diagonal line indicates equality of the two surface brightness measurements and the two additional lines are @xmath234 r , where x is the abscissa and y is the ordinate of the plot .
the bottom plot shows the same data as the top plot , but in the form of residuals ( our values minus wham s ) compared to wham s values ; in the bottom plot , the two curved lines are @xmath235 r. the r.m.s . of the residuals of points with wham - measured surface brightness less than 2 r
is 0.4 r. figure 14our continuum - subtracted data ( upper image ) and the wham survey data ( lower image ) of the same region , orion / eridanus , illustrate the effect of improved angular resolution in filamentary regions .
the upper image has 2 resolution ; the lower image has 60 resolution and has been interpolated between wham beams ( data provided by @xcite ) .
figure 15our field 033 ( top ) , compared to field ha176 of the aao / ukst survey ( left ) and the region around gum 37 from the marseille observatory survey ( right ) .
the second image is scanned from @xcite and does not represent the full image quality of the aao / ukst survey . | we have completed a robotic wide - angle imaging survey of the southern sky ( @xmath0 to @xmath1 ) at 656.3 nm wavelength , the h@xmath2 emission line of hydrogen .
each image of the resulting southern h - alpha sky survey atlas ( shassa ) covers an area of the sky @xmath3 square at an angular resolution of approximately 0.8 arcminute , and reaches a sensitivity level of 2 rayleigh ( @xmath4 ) per pixel , corresponding to an emission measure of @xmath5 , and to a brightness temperature for microwave free - free emission of @xmath6 at @xmath7 .
smoothing over several pixels allows features as faint as 0.5 rayleigh to be detected . |
the international committee for weights and measures recommended several radiations for the practical realization of the metre . at the border of the visible and near - infrared ranges ,
in particular , the international bureau of weights and measures ( bipm ) recommends the 5s@xmath0(f=3 ) - 5d@xmath1(f=5 ) two - photon transition in @xmath2rb with a standard uncertainty of 5 khz ( the relative standard uncertainty of @xmath3 ) @xcite .
recent development in phase - stabilized optical frequency combs based on mode - locked femtosecond lasers allowed determination of the absolute frequency of a similar transition in rb , 5s@xmath0 - 7s@xmath0 , which is 100 times weaker than 5s-5d , yet less sensitive to stray magnetic fields . at the 5s-5d transition
the rubidium atoms must be carefully shielded against the magnetic field to avoid any linear zeeman shifts . on the other hand ,
the 5s and 7s levels have the same land @xmath4 factors which cancels the linear zeeman shifts in the 5s-7s transition .
the ac - stark effect in the 5s-7s transition is also smaller than in the 5s-5d case . because of the difference in the signal strengths , all previous measurements of the 5s-7s transition @xcite yielded absolute values of its frequency less accurate than for the 5s-5d transition @xcite . in this work
we report the measurement of the absolute frequency measurements of the 5s@xmath0 - 7s@xmath0 transitions in @xmath5rb and @xmath2rb with relative standard uncertainty smaller than @xmath6 which is better than measured previously and is comparable with the measurements of the 5s-5d transition @xcite .
the experimental setup is shown in fig .
[ fig : setup ] .
we have used a commercial ring - cavity titanium sapphire ( tisa ) laser to study the two - photon 5s@xmath0 - 7s@xmath0 transitions at 760 nm in a hot rubidium vapour cell at temperatures up to @xmath7c .
the tisa laser , pre - stabilised by a fabry - prot cavity , has the linewidth of 300 khz .
the two - photon spectroscopy signal is observed by a photomultiplier tube in the 7s-6p-5s radiative cascade , with the 6p-5s blue fluorescence around 421 nm . the laser is digitally locked to the measured transition .
the digital lock , a technique generally used in optical atomic clocks , allows for long interrogation of the non - modulated reference beam with an optical frequency comb .
that technique and good long - term stability of our system reduce the statistical uncertainties of measured frequencies below 1 khz , an order of magnitude better than in refs .
@xcite , thanks to the longer averaging times .
the low statistical uncertainty enabled studies of systematic shifts with accordingly higher precision .
the idea of the digital lock is depicted in fig .
[ fig : digilock ] . an acousto - optic modulator ( aom ) driven by a direct digital synthesizer ( dds ) square - wave modulates the light frequency with the step @xmath8 equal to the half - width of the line
the aom carrier frequency @xmath9 is chosen such that the aom efficiency with @xmath10 and @xmath11 is the same .
the microcontroller ( atmel at91sam7s ) , which controls the dds , counts the photomultiplier pulses from the fluorescence signal of the 6p-5s transition . the error signal for the laser lock is calculated from the difference of the counts corresponding to fluorescence for @xmath12 and @xmath13 . the software pi regulator in the microcontroller calculates the correction @xmath14 and applies it to the tisa laser .
switching between the @xmath15 frequencies in the dds is completed in 150 ns and the acoustic wave needs few @xmath16s to completely propagate the frequency changes through the aom pbmo0@xmath17 crystal .
the following acquisition of the photomultiplier pulses takes 38 ms which is long enough to ignore chirping effects from switching the @xmath15 frequencies .
the power of the light sent to the rubidium cell is stabilised by the software pi regulator on the embedded pc ( fox board g20 ) with a half - wave plate mounted on a piezo - driven mount and a polariser . to exclude the residual doppler effect , caused by the wave - front curvature , the counter - propagating beams in the two - photon spectroscopy are not focused and their relative positions are controlled by a ccd beam profiler .
the digital lock can be also used to measure the line profile ( fig .
[ fig : line ] ) by changing @xmath18 in the digital lock algorithm and recording accordingly the fluorescence intensity .
since the counter - propagating beams are not focused , the transition is characterized by a power - broadened lorentzian profile .
the measurement accuracy may be reduced at the line center where the digital lock is less precise because of small value of the fluorescence derivative with respect to the laser frequency leading to underestimation of measured fluorescence at given frequency .
rb f=2-f=2 line profile for different intensities of the probing light .
the lorentz profile is fitted to the measured data . ]
part of the tisa light is sent directly to the er - dopped fiber optical frequency comb ( menlo fc1500 - 250-wg ) .
the comb , the dds synthesizers and counters in the experiment are locked to a microwave rb frequency standard ( srs fs725 ) , disciplined by the gps ( connor winfield fts 375 ) with fractional freuqency uncertainty better than @xmath19 at our measurement times .
the fractional allan variance of the frequency of the tisa laser locked to the @xmath5rb f=2-f=2 transition measured with the optical frequency comb is presented in fig .
[ fig : allan ] . after 1000
s the rb frequency standard reaches its final stability , and our system is further disciplined by the gps which improves the stability , as seen in the plot , after @xmath20 s integration time .
rb f=2-f=2 transition measured with the optical frequency comb . ]
several systematic effects should be taken into account to deduce the transition frequency .
most pronounced is the systematic pressure shift of the measured transition .
the rubidium vapour pressure is determined by the cell temperature @xcite .
we measured the absolute frequencies at different temperatures ( accuracy of 0.2 k ) and estimated the coefficient of the pressure shift interaction rb - rb as -17.82(81 ) khz / mtorr ( fig .
[ fig : press]a ) .
this value differs from the value in ref .
@xcite since we used a more accurate expression @xcite for calculating the vapour pressure at given temperature .
finally , the shift was extrapolated to the zero pressure . in the same way we measured and calibrated the ac - stark shift .
surprisingly , the measured ac - stark shift is much lower than calculated in ref .
analysis of this effect is , however , beyond the scope of this report .
in addition to the rb - rb pressure shift , additional contribution comes from collisions with residual gas present in the cell ( mainly ar , since rubidium acts as a getter for other impurities ) . in recent work of wu et al .
@xcite the 6s-8s transition frequency measured in 10 different commercial cs vapour cells varied by hundreds of khz , but measurements of the 5s-5d transition @xcite proved that in rb this effect is not that big .
our rb cell has been filled under vacuum of the order of @xmath21 torr but we assume it could drop to about @xmath22 torr because of the cell ageing .
the rb - ar collisional shift was estimated assuming van der waals interaction @xcite and verified with data in ref .
[ fig : press]b depicts the determination of the quadratic zeeman shift coefficient measured by application of an external , calibrated magnetic field .
the actual magnetic field was measured by a precise magnetometer .
the black body radiation @xcite and second order doppler shifts were calculated for a given stabilised cell temperature .
the systematic shift caused by different efficiencies of the aom probing two sides of the transition is most likely the fundamental limit of the measured uncertainty in the digital lock scheme .
this shift is dependent on the laser frequency derivatives of fluorescence corresponding to @xmath12 and @xmath13 .
this effect is most clearly seen when the residual doppler effect caused by some misalignment of the probe beams broadens the line by few mhz . the resulting change of these derivatives shifts the transition by tens of khz .
this shift is quantified by varying the @xmath18 value in the dds lock and the aom diffraction order ( + @xmath9 or -@xmath9 ) .
these systematic shifts can be significantly reduced in future measurements .
for example , in the optical clocks this effect is eliminated by the injection locked slave laser .
rb f=2-f=2 transition . ]
the accuracy budget for typical experimental conditions , i.e. laser intensity of 11 w/@xmath23 in each beam , beams @xmath24 diameter of 1.4 mm and temperature of @xmath25c , is presented in table [ tab : acc ] .
.accuracy budget for typical experimental conditions ( m - measured , c - calculated ) . [ cols="^,^,^",options="header " , ] in table [ tab : res ] we present the measured frequencies of four 5s@xmath0 - 7s@xmath0 transitions in @xmath5rb and @xmath2rb .
comparison of the measured frequency of the @xmath5rb f=2-f=2 transition with the previously measured values is depicted in fig .
[ fig : zest ] . with the knowledge of the hyperfine splitting of the 5s@xmath0 state @xcite the hyperfine a constants of the 7s@xmath0 state were derived , as well as the 5s@xmath0 - 7s@xmath0 transition isotope shift .
the results are presented in table [ tab : hyp ] .
comparisons presented in fig .
[ fig : zest ] and table [ tab : hyp ] show that while there is a good agreement of all previous measurements of the a hyperfine constants and isotope shift of the 5s-7s transition , the absolute frequency measurements agree within the expanded uncertainties only with the values given in refs .
@xcite and all transition frequencies measurements in ref .
@xcite are systematically higher that those measured in ref .
@xcite , where the measurement scheme was similar to ours . since the measurements of the hyperfine and isotope structures are not absolute but relative , the discrepancy between ref .
@xcite and the present work seems to indicate a systematic shift of the 5s-7s transition that remains to be identified .
we have performed a series of measurements of the absolute frequency of the 5s@xmath0 - 7s@xmath0 two - photon transitions in rubidium vapour with an optical frequency comb .
we have also estimated the a constants of the hyperfine splitting of the 7s@xmath0 state and the isotope shift between @xmath2rb and @xmath5rb of the 5s@xmath0 - 7s@xmath0 transition .
thanks to low statistical uncertainty and thorough study of all systematic shifts , the accuracy of the 5s@xmath0 - 7s@xmath0 transition frequency obtained in the present work is higher than previously reported @xcite .
this work has been performed in the national laboratory famo in toru and supported by the polish national science centre project no .
2012/07/b / st2/00235 and the team projects of the fnp , co - financed by the eu within the european regional development fund .
bureau international des poids et mesures ( bipm ) , in _ report of the 86th meeting of the comit international des poids et mesures _ , ( bipm , svres , france , 1997 ) h. c. chui , m. s. ko , y. w. liu , j. t. shy , j. l. peng , and h. ahn , opt .
lett . , * 30 * , 842 ( 2005 ) a. marian , m. c. stowe , d. felinto , and j. ye , phys .
lett . * 95 * , 023001 ( 2005 ) k. pandey , p. v. kiran kumar , m. v. suryanarayana , and v. natarajan , opt .
* 33 * , 1675 ( 2008 ) i. barmes , s. witte , and k. s. e. eikema , phys .
lett . , * 111 * , 023007 ( 2013 ) d. touahri , o. acef , a. clairon , j .- j .
zondy , r. felder , l. hilico , b. de beauvoir , f. biraben , f. nez , opt . commun . * 133 * , 471 ( 1997 ) , d. j. jones , s. a. diddams , j. k. ranka , a. stentz , r. s. windeler , j. l. hall , s. t. cundiff , science * 288 * , 635 ( 2000 ) , j. e. bernard , a.a .
madej , k.j .
siemsen , l. marmet , c. latrasse , d. touahri , m. poulin , m. allard , and m. ttu , opt .
* 173 * , 357 ( 2000 ) , c. s. edwards , g. p. barwood , h. s. margolis , p. gill , and w. r. c. rowley , metrologia * 42 * , 464 ( 2005 ) c. b. alcock , v. p. itkin , and m. k. horrigan , canadian metallurgical quarterly * 23 * , 309 ( 1984 ) c.m .
wu , t. w. liu , m. h. wu , r. k. lee , w. y. cheng , and s. bergeson , opt .
lett . , * 38 * , 3186 ( 2013 ) r. s. trawiski , acta phys . pol . a , * 110 * , 51 ( 2006 ) k.h . weber and k. niemax ,
a atoms and nuclei , * 307 * , 13 ( 1982 ) j. w. farley and w. h. wing , phys .
a , * 23 * , 2397 ( 1981 ) s. bize , y. sortais , m. s. santos , c. mandache , a. clairon , and c. salomon , europhys .
, * 45 * , 558 ( 1999 ) , e. arimondo , m. inguscio , and p. violino , rev .
* 49 * , 31 ( 1977 ) | we report the absolute frequency measurements of rubidium 5s-7s two - photon transitions with a cw laser digitally locked to an atomic transition and referenced to an optical frequency comb .
the narrow , two - photon transition , 5s-7s ( 760 nm ) insensitive to first order in a magnetic field , is a promising candidate for frequency reference .
the performed tests yield the transition frequency with accuracy better than reported previously . |
as the nearest early - type system , the milky way bulge provides a unique opportunity to study the details of its resolved stellar populations .
five years ago , the presence of double red clump ( rc ) was discovered in the higher latitude ( @xmath0@xmath1 ) fields of the milky way bulge from wide - field photometric surveys ( @xcite ; @xcite ; ) .
these and follow - up observations have established that ( 1 ) the two rcs have almost identical mean colours , ( 2 ) the double rc feature is only evident among metal - rich stars , while metal - poor populations show only faint rc , ( 3 ) the separation between the two rcs is vanished at low galactic latitudes , and ( 4 ) the relative strength of the two rcs changes strongly with longitude .
these observations are widely accepted as evidence for the x - shaped bulge originated from the disk and bar instabilities , which led to the consensus that even higher latitude region of the milky way bulge has more characteristics of a `` pseudo bulge '' , rather than a classical bulge ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) . without the crucial distance information ,
however , an alternative interpretation , in which the brighter rc is representing an intrinsically brighter subpopulation rather than a distance effect , must be explored in detail to see whether it can equally reproduce the key observations described above .
it is now well established that most globular clusters ( gcs ) host multiple populations with helium and light - elements enhanced second and third generation stars , for which the origin can be traced back to the chemical enrichments and pollutions by asymptotic - giant - branch ( agb ) stars and/or fast - rotating massive - stars ( frms ) .
some massive gcs , such as @xmath2 cen and terzan 5 , also show evidence for supernovae enrichment .
chemical enrichments and pollutions similar to these might have affected stellar populations in the bulge as well .
furthermore , proposed that bulge populations might be helium - enriched due to their super - solar metallicity and the helium enrichment parameter ( @xmath3 ) between 2 and 3 , with an average helium content of 0.31 0.35 . in general ,
super - helium - rich subpopulations are placed on the bluer horizontal - branch ( hb ) during the core helium - burning phase .
however , when the metallicity is high enough as in the bulge , the strong metallicity effect overwhelms the helium effect , and the metal - rich counterpart would be placed on the brighter rc .
this was correctly pointed out by @xcite in their investigation of the double rc observed in the metal - rich bulge gc terzan 5 @xcite .
the double rc observed in terzan 5 is very analogous to that in the bulge in terms of the magnitude difference , metallicity , and population ratio .
both @xcite and @xcite noted this similarity , but they dismissed this possibility in the bulge mostly based on the fact that the two rcs in terzan 5 , unlike the double rc in the bulge , also show the difference in colour .
the colour of the rc stars , however , can vary with metallicity , helium abundance , or age , similarly to the case of more metal - poor hb stars ( * ? ? ?
* see figure [ f1 ] below ) .
spectroscopic observations of red - giant - branch ( rgb ) and rc stars in the bulge field show a large range in metallicity , @xmath4 } \leq 0.5}$ ] , with a mean metallicity at almost solar .
the metallicity distribution functions ( mdfs ) derived from these observations show bimodality with metal - rich and metal - poor populations .
a metallicity gradient is also observed in the bulge along the minor axis , in the sense that the fraction of the metal - rich population is decreased at high galactic latitudes .
further showed that the @xmath5-elements enhancement is decreased at higher metallicities .
many of these observations and recent theoretical studies have suggested a possibility of a composite bulge in the milky way , in which a classical bulge ( cb ) can coexist with a boxy pseudo bulge originated from the bar (; ; @xcite ; ; ; @xcite ; @xcite ) . in this composite bulge , low galactic latitude ( @xmath6 ) fields
would be dominated by the most metal - rich bar population , while relatively metal - poor cb population becomes more and more important at higher latitudes (;
; @xcite ; ) .
the purpose of this paper is to show that , in such a composite bulge , the multiple population models can equally explain the presence of the double rc and other key observations in the milky way bulge .
the rc stars are metal - rich counterpart of core helium - burning hb stars in metal - poor gcs . in order to study the rc populations in the bulge ,
we have therefore constructed a series of synthetic hb models for the metal - rich populations , following the techniques developed by @xcite and as updated by @xcite .
our models are based on the yonsei - yale ( @xmath7 ) hb evolutionary tracks and isochrones with enhanced helium abundance @xcite , all constructed under the assumption that @xmath8 = 0.3}$ ] .
following our recent investigations for the halo gcs with multiple populations @xcite , mass - loss parameter @xmath9 was adopted to be 0.40 , and the mass dispersion on the rc was adopted to be @xmath10 for each subpopulation . in our multiple population models for the cb component of the bulge ,
the faint rc ( frc ) is produced by first - generation stars ( g1 ) , which were assumed to follow the standard helium enrichment parameter ( @xmath11 , @xmath12 ) , while the brighter rc ( brc ) is populated by super - helium - rich ( y = 0.39 ) second - generation stars ( g2 ) .
the choice of y = 0.39 for g2 is based on the empirical constraint from super - helium - rich subpopulations in @xmath2 cen @xcite , which is also comparable with the theoretical predictions for the helium content in the ejecta of both super - agb stars and type ii supernovae @xcite .
note further that this choice of helium abundances for g1 and g2 ( y = 0.27 & 0.39 at z = 0.02 ) , in the mean ( y = 0.33 ) , is consistent with the suggestion from the analysis of the red giant - branch bump of the milky way bulge (;
spectroscopy of rc stars in the bulge indicates that stars in brc are somewhat more metal - rich ( 0.12 0.23 dex ) than those in frc , and therefore the metallicity difference between g2 and g1 was assumed to be @xmath13 = 0.2}$ ] dex in our models .
the ages for g1 and g2 were adopted to be 12 gyr and 10 gyr , respectively @xcite , and the two rcs are assumed to be equally populated @xcite .
figure [ f1 ] shows our models for the two rcs from g1 and g2 in the @xmath14 colour magnitude diagram ( cmd ) .
the models are presented at four different metallicity regimes relevant to the milky way bulge to illustrate the sensitivity of the double rc feature on metallicity .
it is clear from the models in panel ( b ) that the separation in magnitude between the two rcs ( @xmath150.5 mag ) is analogous to the observed difference , while the colour difference is negligible as is observed .
more metal - poor models in panel ( c ) demonstrate , however , that , when the metallicity is decreased , the helium - rich g2 start to leave the rc regime to the bluer hb , while helium - normal g1 are placed on the frc .
when the metallicity is decreased further as in panel ( d ) , all the core helium - burning stars produced by g2 are completely out of the brc and are placed on the blue hb , while those from g1 are still on the frc or red hb .
this case is qualitatively similar to the extended hb morphology observed in gcs with multiple populations such as ngc 2808 ( see * ? ? ?
consequently , no double rc would be observed in the relatively metal - poor population .
this is mainly because helium - rich stars evolve more rapidly , and hence have lower masses at given age ( see * ? ? ?
* ; * ? ? ?
interestingly , the most metal - rich models in panel ( a ) show that the separation between the two rcs would be diminished or vanished in super - metal - rich ( @xmath16 \geq 0.2}$ ] ) population .
this is because even g1 , following the standard @xmath11 , becomes very helium - rich ( @xmath17 ) at super - metal - rich regime , increasing the luminosity of frc . in a two - component composite bulge
, this most metal - rich regime would be more relevant to the bar component . the helium enhanced evolutionary tracks adopted in our modeling were constructed under the assumption that @xmath18 } = 0.3}$ ] for all @xmath19}}$ ] regimes . in the bulge fields ,
however , the @xmath20-enhancement is observed to be decreased with increasing @xmath19}}$ ] . in order to investigate the effect of @xmath18}}$ ] , our @xmath20-enhanced evolutionary tracks are rescaled back to @xmath18}=0.0}$ ] by adopting the relation suggested by @xcite .
panels ( e ) ( h ) of figure [ f1 ] show our rc models constructed for @xmath18}=0.0}$ ] . at solar metallicity ( @xmath19 } \approx 0.0}$ ] ) , the scaled - solar ( @xmath18}=0.0}$ ] ) models
are fainter by @xmath210.1 mag compared to the @xmath20-enhanced ( @xmath18}=0.3}$ ] ) models , but the magnitude difference between the two rcs is little affected .
furthermore , according to ( * ? ? ?
* see their figure 19 ) , the difference in @xmath18}}$ ] between the two dominant components ( at @xmath19 } \approx + 0.15}$ ] and @xmath22 ) in the bulge is observed to be only @xmath210.1 dex , and therefore the effect of @xmath18}}$ ] on the magnitude difference between the two rcs would be negligible . in panel ( c ) of figure [ f2 ]
, our single metallicity models in figure [ f1 ] ( for @xmath18}=0.3}$ ] ) are combined to generate a composite model for the cb component of the bulge . for this , models in the metallicity range of @xmath23 } \leq + 0.25 $ ] were used , weighted according to the skewed gaussian mdf truncated at both ends with a peak at @xmath16 = -0.1}$ ] . in these models ,
the magnitude difference between the two rcs is @xmath24 mag , while the colour difference is negligible , @xmath25 , which are in reasonable agreements with the observed differences ( see figure [ f3 ] ) .
similar results are obtained even if the peak metallicity and age are varied by @xmath15@xmath26 dex and @xmath15@xmath27 gyr . }
= -0.25 $ ] and @xmath28 is adopted roughly in the same metallicity range , the magnitude and colour differences are varied by only @xmath150.02 and 0.00 mag , respectively . ] for comparison , panel ( a ) shows our model for the gc terzan 5 , where the metallicities for g1 and g2 were adopted from @xcite , and the helium abundance for g2 was adopted to be y = 0.35 .
ages for g1 and g2 are identical to those adopted for the cb models ( i.e. , 12 & 10 gyr , respectively ) .
our models show that a larger difference in metal abundance ( @xmath13 } = 0.5 $ ] dex ) between g2 and g1 in terzan 5 is mainly responsible for the observed colour difference between the two rcs , while a smaller difference in helium abundance is also a contributing factor .
if no difference in age is adopted between g1 and g2 for terzan 5 , as was assumed by @xcite , the magnitude and colour differences are reduced by @xmath150.05 and @xmath150.01 mag , respectively .
the panel ( b ) of figure [ f2 ] also presents our model for the bar component of the bulge , in which we assume that only metal - rich g1 ( with an age of 10 gyr ) is populated considering the disk origin of the bar population .
following the observed mdf for the most metal - rich bar component , a gaussian mdf with the mean @xmath29 } = + 0.15 $ ] and @xmath30 } = 0.10 $ ] dex is adopted . as described above , this super - metal - rich g1 , following the standard @xmath31 , becomes relatively helium - rich ( @xmath32 0.31 ) and thus produces relatively bright ( @xmath33 mag ) rc stars that would be placed between the brc ( g2 ) and frc ( g1 ) of the cb model in panel ( c ) .
the panel ( d ) of figure [ f2 ] shows the cmd of the overlapping populations ( i.e. , cb + bar ) . in order to reproduce the low latitude fields of the bulge ,
the population ratio between the bar and cb is adopted to be 2:1 . in this case , the split in the rc is not clear , but the gaussian decomposition indicates two components separated by @xmath150.30 mag , which is in reasonable agreement with the observed smaller difference at low galactic latitudes , such as the baade s window . , @xmath34 mag )
are included in our simulations . ] ) field of the bulge ( 2mass data from @xcite ; figure 2 of @xcite ) .
note that the observed cmd is contaminated by the foreground disk stars at @xmath35 and 0.7 . ]
our models can reproduce all of the key observations described in section 1 .
firstly , as described above , the observed magnitude difference ( @xmath150.5 mag ) between the two rcs is reproduced while maintaining no or little colour difference between them ( see figure [ f2 ] ) .
secondly , the double rc feature is reproduced only for the relatively metal - rich ( @xmath16 } \geq -0.4 $ ] ) populations as suggested from the observations .
our models for the metal - poor populations in the bulge show only frc ( see figure [ f1 ] lower panel ) , because , unless very metal - rich ( @xmath16 } \geq -0.4 $ ] ) , super - helium - rich g2 can not be trapped on the brc .
thirdly , as illustrated by a schematic diagram in figure [ f4 ] ( three vertical panels ) , the latitude dependence of the double rc feature can be explained within a two - component composite bulge . at low latitudes ,
the bulge is dominated by the most metal - rich bar population with only g1 , which , on the rc , is placed between the two rcs from a minority cb component , completely filling the gap of the double rc . at higher latitudes ,
however , contribution from the bar component is decreased / vanished , while the cb component ( with g1 and g2 ) becomes more and more important , producing a larger magnitude difference between the two rcs with increasing @xmath36 .
finally , in a two - component bulge with a spheroidal cb embedded in a tilted bar ( pseudo bulge ) , the longitude dependence of the rc luminosity function can be reproduced in our model as well . as also illustrated in figure [ f4 ] ( three horizontal panels ) , the observed luminosity function , in our scenario , is the sum of a cb component ( with double rc ) and a metal - rich bar component ( with single rc ) . towards negative longitudes the bar component is placed at the far side of the bulge and therefore frc becomes gradually more populated , while the opposite effect is anticipated towards positive longitudes .
most of these observations were obtained in @xmath37 fields ( see figure 11 of * ? ? ?
* ) , where the cb component is expected to be comparable to the bar component in terms of population ratio .
more detailed population models that fully take into account the observed mdf and its dependence on galactic latitude and longitude will be presented in our forthcoming paper ( in preparation ) , where we will also compare the observed rc luminosity functions with the models constructed under various combinations of input parameters .
the kinematics of rc and giant stars in the bulge can provide further constraints on our models .
the radial velocity ( @xmath38 ) measurements show that the bulge rotates cylindrically , which is a strong evidence for the bar at low latitude fields ( @xmath39@xmath1 ) . for the higher latitudes , however , this is not necessarily inconsistent with our cb dominated scenario in this region of the bulge , because , for example , @xcite showed that an initially non - rotating cb could absorb a significant fraction of the angular momentum from the bar within a few gyr .
the observed kinematics of stars in the two rcs are also consistent with our scenario . at @xmath40 , found an excess of stars moving towards the sun in the brc , while that receding from the sun is observed in the frc . in our composite bulge scenario , this can be understood by the depth ( @xmath152 kpc ; @xmath41 0.5 mag ) effect of the bar . since the rcs from the bar and cb
are mixed in the cmd of this field , rc stars in near side of the bar would be superimposed on brc from cb , while those in far side of the bar is mixed with frc from cb . therefore , if the stars in the bar are in streaming motions , we would see some difference in @xmath38 , in the mean , between brc and frc . at higher latitudes ( @xmath42 & @xmath43 ) , however , there is no difference in @xmath38 between the two rcs , which is consistent with our cb ( with multiple populations ) dominated scenario in these regions of the bulge .
the distance measurements for the two rcs from soon to be released gaia trigonometric parallax data would provide the most direct test for the suggested model . if the two rcs are due to the x - shaped structure , @xmath150.5 mag difference in the distance modulus is expected between the brc and frc , while essentially no difference , in the mean , is predicted in our multiple population models .
this should settle the issue once and for all .
if confirmed , our result would suggest that the bar in the milky way bulge is not sufficiently buckled to form the x - shaped structure to be observed as double rc at higher latitudes ( @xmath44 ) .
furthermore , if the split rc is mostly due to the effect of multiple populations , rather than a distance effect , the previous studies on the structure of the milky way bulge and bar based on rc , high latitude fields in particular , should be reexamined in the new paradigm .
finally , our result for the cb component of the milky way bulge , the nearest early - type system , would also suggest that the early - type galaxies would be similarly prevailed by super - helium - rich subpopulation .
if @xmath1550% of stars in early - type galaxies are helium and light elements enhanced g2 , this would have profound impacts on the interpretation of integrated spectra by employing population synthesis models , such as the origins of the na enhanced giant elliptical galaxies @xcite and the uv upturn phenomenon @xcite .
we thank the referee for a number of helpful suggestions which led to several improvements in the manuscript .
we also thank sang - il han for his assistance in obtaining the observed data for figure 3 .
support for this work was provided by the national research foundation of korea to the center for galaxy evolution research .
babusiaux , c. , gmez , a. , hill , v. , et al .
2010 , , 519 , a77 bekki , k. 2012 , , 421 , l44 bureau , m. , aronica , g. , athanassoula , e. , et al .
2006 , , 370 , 753 calamida , a. , bono , g. , lagioia , e. p. , et al .
2014 , , 565 , a8 carretta , e. , bragaglia , a. , gratton , r. g. , et al .
2009 , , 505 , 117 chung , c. , yoon , s .- j .
, & lee , y .- w . 2011 , , 740 , l45 clarkson , w. , sahu , k. , anderson , j. , et al .
2008 , , 684 , 1110 dantona , f. , & caloi , v. 2004 , , 611 , 871 dantona , f. , ventura , p. , caloi , v. , et al .
2010 , , 715 , l63 decressin , t. , meynet , g. , charbonnel , c. , prantzos , n. , & ekstrm , s. 2007 , , 464 , 1029 dercole , a. , vesperini , e. , dantona , f. , mcmillan , s. l. w. , & recchi , s. 2008 , , 391 , 825 de propris , r. , rich , r. m. , kunder , a. , et al .
2011 , , 732 , l36 erwin , p. , saglia , r. p. , fabricius , m. , et al .
2015 , , 446 , 4039 ferraro , f. r. , dalessandro , e. , mucciarelli , a. , et al .
2009 , , 462 , 483 gonzalez , o. a. , rejkuba , m. , zoccali , m. , et al .
2011 , , 530 , a54 gratton , r. g. , carretta , e. , & bragaglia , a. 2012 , , 20 , 50 han , s .- i . , kim , y .- c . , lee , y .- w . , et al . 2009
, globular clusters - guides to galaxies , 33 hill , v. , lecureur , a. , gmez , a. , et al .
2011 , , 534 , a80 jang , s. , lee , y .- w . ,
j . , & na , c. 2014 , , 443 , l15 jeong , h. , yi , s. k. , kyeong , j. , et al .
2013 , , 208 , 7 johnson , c. i. , rich , r. m. , kobayashi , c. , kunder , a. , & koch , a. 2014 , , 148 , 67 joo , s .-
j . , & lee , y .- w .
2013 , , 762 , 36 lee , j .- w . ,
kang , y .- w . ,
lee , j. , & lee , y .- w .
2009 , , 462 , 480 lee , y .- w . ,
demarque , p. , & zinn , r. 1994 , , 423 , 248 lee , y .- w . , joo , j .- m . , sohn , y .-
j . , et al . 1999 , , 402 , 55 lee , y .- w . ,
, han , s .-
, et al . 2005 , , 621 , l57 li , z .- y . ,
& shen , j. 2012 , , 757 , l7 lim , d. , han , s .-
i . , lee , y .- w . , et al .
2015 , , 216 , 19 maeder , a. 1992 , , 264 , 105 marino , a. f. , milone , a. p. , piotto , g. , et al . 2009 , , 505 , 1099 marino , a. f. , milone , a. p. , przybilla , n. , et al .
2014 , , 437 , 1609 mcwilliam , a. , & zoccali , m. 2010 , , 724 , 1491 nataf , d. m. , udalski , a. , gould , a. , fouqu , p. , & stanek , k. z. 2010 , , 721 , l28 nataf , d. m. , gould , a. p. , pinsonneault , m. h. , & udalski , a. 2013 , , 766 , 77 nataf , d. m. , udalski , a. , skowron , j. , et al .
2015 , , 447 , 1535 ness , m. , freeman , k. , athanassoula , e. , et al .
2013 , , 432 , 2092 ness , m. , freeman , k. , athanassoula , e. , et al .
2012 , , 756 , 22 obreja , a. , domnguez - tenreiro , r. , brook , c. , et al .
2013 , , 763 , 26 origlia , l. , rich , r. m. , ferraro , f. r. , et al .
2011 , , 726 , l20 pagel , b. e. j. , simonson , e. a. , terlevich , r. j. , & edmunds , m. g. 1992 , , 255 , 325 piotto , g. , villanova , s. , bedin , l. r. , et al .
2005 , , 621 , 777 reimers , d. 1977 , , 61 , 217 renzini , a. 1994 , , 285 , l5 rojas - arriagada , a. , recio - blanco , a. , hill , v. , et al .
2014 , , 569 , a103 saha , k. 2015 , arxiv:1505.07048 saha , k. , gerhard , o. , & martinez - valpuesta , i. 2015 , arxiv:1505.04508 saito , r. k. , zoccali , m. , mcwilliam , a. , et al .
2011 , , 142 , 76 saito , r. k. , minniti , d. , dias , b. , et al .
2012 , , 544 , a147 salaris , m. , chieffi , a. , & straniero , o. 1993 , , 414 , 580 skrutskie , m. f. , cutri , r. m. , stiening , r. , et al .
2006 , , 131 , 1163 timmes , f. x. , woosley , s. e. , & weaver , t. a. 1995 , , 98 , 617 uttenthaler , s. , schultheis , m. , nataf , d. m. , et al .
2012 , , 546 , a57 van dokkum , p. g. , & conroy , c. 2010 , , 468 , 940 vsquez , s. , zoccali , m. , hill , v. , et al .
2013 , , 555 , a91 ventura , p. , & dantona , f. 2009 , , 499 , 835 ventura , p. , di criscienzo , m. , carini , r. , & dantona , f. 2013 , , 431 , 3642 wegg , c. , & gerhard , o. 2013 , , 435 , 1874 woosley , s. e. , & weaver , t. a. 1995 , , 101 , 181 zoccali , m. , gonzalez , o. a. , vasquez , s. , et al .
2014 , , 562 , a66 zoccali , m. , hill , v. , lecureur , a. , et al .
2008 , , 486 , 177 | the presence of two red clumps ( rcs ) in high latitude fields of the milky way bulge is interpreted as evidence for an x - shaped structure originated from the bar instability . here
we show , however , that this double rc phenomenon is more likely to be another manifestation of multiple populations observed in globular clusters ( gcs ) in the metal - rich regime . as in the bulge gc terzan 5 ,
the helium enhanced second generation stars in the classical bulge component of the milky way are placed on the bright rc , which is about 0.5 mag brighter than the normal rc originated from the first generation stars , producing the observed double rc . in a composite bulge , where a classical bulge can coexist with a boxy pseudo bulge
, our models can also reproduce key observations , such as the dependence of the double rc feature on metallicity and galactic latitude and longitude .
if confirmed by gaia trigonometric parallax distances , this would indicate that the milky way bar is not sufficiently buckled to form the x - shaped structure in the bulge , and suggest that the early - type galaxies would be similarly prevailed by super - helium - rich subpopulation .
[ firstpage ] galaxy : bulge galaxy : structure galaxy : formation galaxies : elliptical and lenticular , cd globular clusters : general stars : horizontal - branch |
tuberculosis ( tb ) detection and treatment saved 22 million of lives , between 1995 and 2012 , following the 2013 report of the world health organization ( who ) @xcite .
however , in 2012 , there were 8.6 million of new tb cases and 1.3 million of tb deaths @xcite .
tb prevention , diagnosis and treatment , requires adequate funding , sustained over many years , which represents a worldwide scale challenge .
mathematical dynamic models are an important tool in analyzing the spread and control of infectious diseases . many tb mathematical models
have been developed see , e.g. , @xcite and the references cited therein .
the main differences of the models proposed in are the way they represent reinfection , since there is no consensus on wether a previous infection gives or not protection .
the way recently infected individuals progress to active disease is not the same in all models : they can be `` fast progressors '' or `` slow progressors '' . in some models
, it is assumed that only 5 to 10% of the infected individuals are fast progressors .
the remaining models consider that individuals are able to contain the infection asymptomatic and non infectiously ( latent individuals ) , having a much lower probability of developing active disease by endogenous reactivation .
more recent models also assume exogenous reinfection of latent and treated individuals , based on the fact that infection and/or disease do not confer full protection @xcite .
this assumption has an important impact on the efficacy of interventions @xcite . in this paper
, we consider a tb mathematical model from @xcite , where exogenous reinfection is considered . without treatment , tb mortality rates are hight @xcite .
different interventions are available for tb prevention and treatment : vaccination to prevent infection ; treatment to cure active tb ; treatment of latent tb to prevent endogenous reactivation . in this work
, we study the implementation of two post - exposure interventions that are not widely used : treatment of early latent individuals with anti - tb drugs ( e.g. , treatment of recent contacts of index cases ) and prophylactic treatment / vaccination of the persistent latent individuals .
we propose an optimal control problem that consists in analyzing how these two control measures should be implemented , for a certain time period , in order to reduce the number of active infected individuals , while controlling the interventions implementation costs .
optimal control is a branch of mathematics developed to find optimal ways to control a dynamic system @xcite .
other authors applied optimal control theory to tb models ( see , e.g. , @xcite ) .
this approach allows the study of the most cost - effective intervention design by generating an implementation design that minimizes an objective function .
the intensity of interventions can be relaxed along time , which is not the case considered in most models , for which interventions are modeled by constant rates @xcite .
the paper is organized as follows . in section [ sec : model ]
we present the mathematical model for tb that will be study in this paper .
two control functions @xmath0 and @xmath1 are then added to the original model from @xcite .
section [ sec : oc : problem ] is dedicated to the formulation of the optimal control problem .
we prove the existence of an unique solution and derive the expression for the optimal controls according to the pontryagin maximum principle @xcite .
section [ sec : numericalresults ] has four subsections dedicated to a numerical and cost - effectiveness analysis of the optimal control problem .
we start by illustrating the problem solutions for a particular case ( section [ subsec : example ] ) .
we then introduce some summary measures in section [ subsec : smeasures ] to describe how the results change when varying transmission intensity ( section [ subsec : beta ] ) and protection against reinfection ( section [ subsec : sigma ] ) . in section [ subsec :
oc : strategies ] , we analyze the cost - effectiveness of three intervention strategies : applying @xmath0 or @xmath1 separately and applying the two control measures simultaneously .
we end with section [ sec : discussion ] of discussion .
following the model proposed in @xcite , population is divided into five categories : susceptible ( @xmath2 ) ; early latent ( @xmath3 ) , , individuals recently infected ( less than two years ) but not infectious ; infected ( @xmath4 ) , , individuals who have active tb and are infectious ; persistent latent ( @xmath5 ) , , individuals who were infected and remain latent ; and recovered ( @xmath6 ) , , individuals who were previously infected and treated .
we assume that at birth all individuals are equally susceptible and differentiate as they experience infection and respective therapy . the rate of birth and death , @xmath7 , are equal ( corresponding to a mean life time of 70 years @xcite ) and no disease - related deaths are considered , keeping the total population , @xmath8 , constant with @xmath9 .
parameter @xmath10 denotes the rate at which individuals leave @xmath3 compartment ; @xmath11 is the proportion of infected individuals progressing directly to the active disease compartment @xmath4 ; @xmath12 and @xmath13 are the rates of endogenous reactivation for persistent latent infections ( untreated latent infections ) and for treated individuals ( for those who have undergone a therapeutic intervention ) , respectively .
parameters @xmath14 and @xmath15 are factors that reduce the risk of infection , as a result of acquired immunity to a previous infection , for persistent latent individuals and for treated patients , respectively .
these factors affect the rate of exogenous reinfection . as in @xcite , in our simulations
we consider three different cases for the protection against reinfection conferred by treatment : same protection as natural infection ( @xmath16 ) ; lower protection than conferred by infection ( @xmath17 ) ; and higher protection than conferred by infection ( @xmath18 ) , see section [ subsec : sigma ] .
parameter @xmath19 is the rate of recovery under standard treatment of active tb , assuming an average duration of infectiousness of six months .
the values of the rates @xmath10 , @xmath11 , @xmath12 , @xmath13 , @xmath14 and @xmath19 are taken from @xcite and the references cited therein ( see table [ parameters ] for the values of the parameters ) .
additional to standard treatment of infectious individuals , we consider two post - exposure interventions targeting different sub - populations : early detection and treatment of recently infected individuals ( @xmath3 ) and chemotherapy or post - exposure vaccine of persistent latent individuals ( @xmath5 ) .
these interventions are applied at rates @xmath20 and @xmath21 .
we consider , without loss of generality , that the rate of recovery of early latent individuals under post - exposure interventions is equal to the rate of recovery under treatment of active tb , @xmath22 , and greater than the rate of recovery of persistent latent individuals under post - exposure interventions , @xmath23 @xcite .
since we are interested in studying these interventions along time , we add to the original model two control functions , @xmath24 and @xmath25 , which represent the intensity at which these post - exposure interventions are applied at each time step . the dynamical control system that we propose is given by @xmath26 the assumption that the total population @xmath8 is constant , allows to reduce the control system from five to four state variables .
we decided to maintain the tb model in form , using relation @xmath27 as a test to confirm the numerical results .
.parameter values for the control system .
[ cols= " < , < , < " , ] we align the remaining alternative strategies by increasing effectiveness and recompute the icer : icer(b)=acer(b)=5.7 and icer(a)=@xmath28 .
hence , we conclude that strategy * b * has the least icer and therefore is more cost - effective than strategy * a*. for this illustration we have considered the same cost for both interventions ( @xmath29 ) .
results should depend strongly on the choice of these parameters , however this discussion is out of the scope of our present work .
in this work we study the potential of widespread of two post - exposure interventions that are not widely used : treatment of early latent individuals and prophylactic treatment / vaccination of persistent latent individuals .
we propose an optimal control problem that consists in analysing how these two control measures should be implemented , for a certain time period , in order to reduce the number of active infected individuals , while controlling the interventions implementation costs .
this approach differs from others @xcite since it allows intensity of intervention to be changed along time .
as previous suggested @xcite , interventions impact can be sensitive to transmission intensity and reinfection .
we choose a dimensionless measure of effectiveness to compare different scenarios : assuming different transmission intensity ( @xmath30 ) or assuming different assumptions on protection against reinfection conferred by treatment ( @xmath15 ) .
effectiveness of optimal intervention decreases with transmission .
there is a change in the intervention profile from low to high transmission . in high transmission settings ,
the intensity of treatment of persistent latent individuals @xmath31 for the optimal solution is reduced . since treatment of persistent latent individuals reduces the reactivation rate ( from @xmath12 to @xmath13 ) , when reinfection is very common and it overcomes reactivation impact , the advantage of treating this population group is less pronounced .
the susceptibility to reinfection after treatment is still an open question . in one hand
, treatment can reduce the risk of tb by reducing the amount of bacteria present in the lungs .
on the other hand , we can argue that latent infection boosts immunity by constant stimulation of the immune system , so treatment could reduce protection .
we vary parameter @xmath15 to explore these two possible scenarios : @xmath32 when treatment enhances protection and @xmath33 when treatment impairs protection .
results show that treatment of persistent latent individuals should be less intense or even absent for the case where treatment impairs protection .
similar results were obtained for the case of constant treatment rates in @xcite .
in fact , for the correspondent case with maximum intensity ( @xmath34 and @xmath35 ) , we can have an increase of the equilibrium proportion of infectious individuals ( @xmath36 ) .
we can conclude that reinfection has an important role in the determination of the optimal control strategy , by diminishing the intervention intensity on persistent latent individuals : first when transmission is very high corresponding to a very high reinfection rate and secondly when this population group has a lower susceptibility to reinfection ( @xmath37 ) .
interestingly , the reinfection threshold @xmath38 of the model with no controls still marks a change in the model behaviour . even though , we are comparing equilibrium results to transient short time interventions .
cost - effectiveness analysis of alternative combinations of the two interventions is conducted . for @xmath39 , treatment of only early latent individuals
is the more cost - effective strategy , despite of treatment of both early latent and persistent latent individuals having a higher effectiveness .
the total cost associated with treatment of persistent latent individuals is very high , especially because this population group can be very big in comparison to the others .
it is believed that about one third of world s population is latent infected with tb . here , for simplicity
, we have considered the cost parameters both equal to one .
however , this depends greatly on the type of intervention used and results can be changed . for example , if intervention on persistent latent individuals could be done by vaccination , then the per person unit cost could be significantly reduced . plus ,
treatment of early latent individuals implies contact tracing of index cases and prophylactic treatment , which can also be very expensive .
the hamiltonian @xmath40 associated to the problem in is given by @xmath41 where @xmath42 is the _ adjoint vector_. according to the pontryagin maximum principle @xcite , if @xmath43 is optimal for problem with the initial conditions given in table [ icbeta100 ] and fixed final time @xmath44 , then there exists a nontrivial absolutely continuous mapping @xmath45
\to \mathbb{r}^5 $ ] , @xmath42 , such that @xmath46 and @xmath47 the minimality condition @xmath48 holds almost everywhere on @xmath49 $ ] .
moreover , the transversality conditions @xmath50 hold .
[ lem : thm ] for problem with fixed initial conditions @xmath51 , @xmath52 , @xmath53 , @xmath54 and @xmath55 and fixed final time @xmath44 , there exists adjoint functions @xmath56 , @xmath57 , @xmath58 , @xmath59 and @xmath60 such that @xmath61 \dot{\lambda^*_2}(t ) = \lambda^*_2(t)\left(\delta + \tau_1 + \mu\right ) - \lambda^*_3(t ) \phi \delta - \lambda^*_4(t ) ( 1 - \phi ) \delta - \lambda^*_5(t)\tau_1 u^*_1(t ) \\[0.1 cm ] \dot{\lambda^*_3}(t ) = -w_0 + \lambda^*_1(t ) \frac{\beta}{n } s^*(t ) - \lambda^*_2(t ) \frac{\beta}{n}(s^*(t ) + \sigma l_2^*(t ) + \sigma_r r^*(t))\\ \qquad \quad + \lambda^*_3(t ) \left(\tau_0 + \mu\right ) + \lambda^*_4(t)\sigma \frac{\beta}{n } l_2^*(t ) - \lambda^*_5(t)\left(\tau_0 - \sigma_r \frac{\beta}{n } r^*(t ) \right ) \\[0.1 cm ] \dot{\lambda^*_4}(t ) = - \lambda^*_2(t ) \frac{\beta}{n}i^*(t ) \sigma - \lambda^*_3(t ) \omega + \lambda^*_4(t)\left(\sigma \frac{\beta}{n } i^*(t ) + \omega + \tau_2 u^*_2(t ) + \mu\right)\\ \qquad \quad - \lambda^*_5(t)\left ( \tau_2 u^*_2(t ) \right ) \\[0.1 cm ] \dot{\lambda^*_5}(t ) = -\lambda^*_2(t ) \sigma_r \frac{\beta}{n}i^*(t ) - \lambda^*_3(t ) \omega_r + \lambda^*_5(t)\left(\sigma_r \frac{\beta}{n } i^*(t ) + \omega_r + \mu\right ) \ , , \end{cases}\ ] ] with transversality conditions @xmath62 furthermore , @xmath63 system is derived from the pontryagin maximum principle ( see , @xcite ) and the optimal controls come from the minimality condition . for small final time @xmath44 ,
the optimal control pair given by is unique due to the boundedness of the state and adjoint functions and the lipschitz property of systems and ( see @xcite and references cited therein ) .
existence of an optimal solution @xmath64 associated to an optimal control pair @xmath65 comes from the convexity of the integrand of the cost function @xmath66 with respect to the controls @xmath67 and the lipschitz property of the state system with respect to state variables @xmath68 ( see , , @xcite ) . for small final time @xmath44
, the optimal control pair is given by that is unique by the lemma above . because the state system is autonomous , uniqueness is valid for any time @xmath44 and not only for small time @xmath44 .
we fix @xmath39 and @xmath69 and the remaining parameters according to table [ parameters ] and vary @xmath44 .
results for the proportion of infectious individuals are shown in the figure [ fig : i : beta100:sigual : tf ] . with @xmath70 .
parameters according to table [ parameters ] , @xmath39 and @xmath69 . ]
the general behaviour do not change significantly with @xmath44 .
the proportion of infected individuals slightly increases towards the end of the intervention for @xmath71 .
this tendency is more pronounced for higher @xmath44 .
figure [ fig : secanal : wi ] shows the results for different combination of the weight constants on the objective functional @xmath66 .
we fix @xmath39 and @xmath69 and the remaining parameters according to table [ parameters ] and vary @xmath72 , @xmath73 and @xmath74 .
efficacy decreases when the costs @xmath73 and @xmath74 increase , corresponding to an earlier relaxation of the intensity of treatment @xmath75 in the optimal solution due to cost restrictions .
+ the change in efficacy is more pronounced for the cases where the weight associated with infectious individuals @xmath72 change in comparison to the weights associated with the controls @xmath76 ( figures [ fig : secanal : w12 ] and [ fig : secanal : w0 ] ) .
results are less sensitive to the variation between the weight controls @xmath73 and @xmath74 ( figures [ fig : secanal : w1 ] and [ fig : secanal : w2 ] ) .
this work was partially supported by the portuguese foundation for science and technology ( fct ) through the : _ centro de matemtica e aplicaes _ , project pest - oe / mat / ui0297/2014 ( rodrigues ) ; _
center for research and development in mathematics and applications _ ( cidma ) , project pest - oe / mat / ui4106/2014 ( silva and torres ) ; post - doc fellowship sfrh / bpd/72061/2010 ( silva ) ; project ptdc / eei - aut/1450/2012 , co - financed by feder under pofc - qren with compete reference fcomp-01 - 0124-feder-028894 ( torres ) . c. castillo - chavez and z. feng , _ mathematical models for the disease dynamics of tuberculosis _ , in : advances in mathematical population dynamics - molecules , cells and man ( eds .
m. a. horn , g. simonett and g. f. webb ) , vanderbilt university press , 1998 , 117128 .
m. g. m. gomes , p. rodrigues , f. m. hilker , n. b. mantilla - beniers , m. muehlen , a. c. paulo and g. f. medley , _ implications of partial immunity on the prospects for tuberculosis control by post - exposure interventions _ , j. theoret .
( 2007 ) , 608617 .
a. v. rie , v. zhemkov , j. granskaya , l. steklova , l. shpakovskaya , a. wendelboe , a. kozlov , r. ryder and m. salfinger , _ tb and hiv in st petersburg , russia : a looming catastrophe ? _ , int .
j. tuberc .
lung dis . 9
( 2005 ) , 740745 .
h. s. rodrigues , m. t. t. monteiro and d. f. m. torres , _ optimal control and numerical software : an overview _ , in : systems theory : perspectives , applications and developments ( ed .
f. miranda ) , nova science publishers , new york , 2014 , 93110 .
arxiv:1401.7279 p. rodrigues , c. j. silva and d. f. m. torres , _ optimal control strategies for reducing the number of active infected individuals with tuberculosis _ ,
proceedings of the siam conference on control and its applications ( ct13 ) , san diego , california , usa , july 8 - 10 , 2013 , pp . 4450 .
s. verver , r. m. warren , n. beyers , m. richardson , g. d. van der spuy , m. w. borgdorff , d. a. enarson , m. a. behr and p. d. van helden , _ rate of reinfection tuberculosis after successful treatment is higher than rate of new tuberculosis _ , am .
j. respir .
171 ( 2005 ) , 14301435 .
r. m. warren , t. c. victor , e. m. streicher , m. richardson , n. beyers , n. c. g. pittius and p. d. helden , _ patients with active tuberculosis often have different strains in the same sputum specimen _ , am .
j. respir .
169 ( 2004 ) , 610614 . | we propose and analyse an optimal control problem where the control system is a mathematical model for tuberculosis that considers reinfection .
the control functions represent the fraction of early latent and persistent latent individuals that are treated .
our aim is to study how these control measures should be implemented , for a certain time period , in order to reduce the number of active infected individuals , while minimizing the interventions implementation costs .
the optimal intervention is compared along different epidemiological scenarios , by varying the transmission coefficient . the impact of variation of the risk of reinfection , as a result of acquired immunity to a previous infection for treated individuals on the optimal controls and associated solutions , is analysed .
a cost - effectiveness analysis is done , to compare the application of each one of the control measures , separately or in combination .
* keywords : * tuberculosis ; optimal control ; post - exposure interventions ; efficacy function ; cost effort .
* mathematics subject classification 2010 : * 92d30 ; 49m05 . |
although it is well known that the wavelike phenomena of classical physics are ruled by hyperbolic equations , there are at least two modern motivations for studying the scalar wave equation on curved lorentzian four - manifolds .
they are as follows . 0.3 cm ( i ) in the course of studying the einstein vacuum equations in four dimensions , i.e. @xmath0 it was conjectured in ref .
@xcite that they admit local cauchy developments for initial data satisfying eq .
( 1.1 ) such that the metric @xmath1 induced by @xmath2 on a given spacelike hypersurface @xmath3 and the extrinsic - curvature tensor @xmath4 of @xmath3 are prescribed . ]
sets @xmath5 with locally finite @xmath6 curvature and locally finite @xmath6 norm of the first covariant derivatives of @xmath4 .
this means that the spacetime constructed by evolution from smooth data can be smoothly continued , together with a time foliation , as long as the curvature of the foliation and the first covariant derivatives of its extrinsic curvature remain @xmath6-bounded on the leaves of the foliation @xcite .
the proof that this is indeed the case relies on a number of technical ingredients , including the construction of a parametrix ( an approximate green function of the wave operator , that provides a progressive wave representation @xcite for solutions of the wave equation ) for solutions of the homogeneous wave equation @xmath7 on a fixed einstein vacuum background .
one has then to obtain control of the parametrix and of its error term by using only the fact that the curvature tensor is bounded in @xmath6 @xcite .
note that , at a deeper level , the metric @xmath2 can be viewed to determine the elliptic or hyperbolic nature is a connected , four - dimensional , hausdorff four - manifold of class @xmath8 , a linear partial differential operator is a linear map @xmath9 with coefficients @xmath10 given by functions of class @xmath11 .
characteristic polynomial _ of the operator @xmath12 at a point @xmath13 is @xmath14 where @xmath15 is a cotangent vector at @xmath16 .
the cone in the cotangent plane @xmath17 at @xmath16 defined by @xmath18 is called the characteristic cone ( or conoid ) . by construction ,
such a cone is independent of the choice of coordinates , because the higher order terms ( also called leading or principal symbol ) of @xmath12 transform into higher - order terms by a change of coordinates .
the operator @xmath12 is said to be hyperbolic at @xmath16 if there exists a domain @xmath19 , a convex open cone in @xmath17 , such that every line through @xmath20 cuts the characteristic cone in @xmath21 real distinct points . in particular , second - order differential operators with higher - order terms @xmath22 are hyperbolic at @xmath16 if and only if the cone defined by @xmath23 is convex , i.e. , if the quadratic form @xmath24 has signature @xmath25 .
] of the operator @xmath26 , where @xmath27 can denote covariant differentiation with respect to the levi - civita connection on spacetime , or on a vector bundle over spacetime , depending on our needs . when @xmath2 is riemannian , i.e. positive - definite , this operator is minus the laplacian , whereas if @xmath2 is lorentzian , one gets the wave operator .
note also that , in four - dimensional manifolds , our lorentzian world lies in between two other options , i.e. a riemannian metric @xmath2 with signature @xmath28 and elliptic operator @xmath29 , and a ultrahyperbolic metric @xmath2 with signature @xmath30 and ultrahyperbolic operator @xmath29 . in the so - called euclidean ( or riemannian )
framework used by quantum field theorists in functional integration , where the metric is positive - definite , the most fundamental differential operator is however the dirac operator , obtained by composition of clifford multiplication with covariant differentiation .
its leading symbol is therefore clifford multiplication , and it generates all elliptic symbols on compact riemannian manifolds @xcite .
this reflects the better known property according to which , out of the dirac operator and its ( formal ) adjoint , one can define two operators of laplace type , as well as powers of these operators . 0.3 cm ( ii ) recent work on the self - dual road to noncommutative gravity with twist has found it useful to start from a classical , undeformed spacetime which is a self - dual solution of the vacuum einstein equation , e.g. a kasner spacetime @xcite . within that framework , it is of interest to solve first the scalar wave equation in such a kasner background .
since such a task was only outlined in ref .
@xcite , we find it appropriate to develop a systematic calculus in the present paper .
relying in part upon ref .
@xcite , we begin by considering the scalar wave equation ( 1.2 ) for a classical scalar field @xmath31 when the kasner parameters @xmath32 take the values @xmath33 , respectively , i.e. @xmath34 where @xmath31 admits the integral representation @xmath35 one can then set @xcite @xmath36 where @xmath37 has to solve , for consistency , the equation @xcite @xmath38w(\xi_{1},\xi_{2},\xi_{3},t)=0 , \label{(1.6)}\ ] ] and the @xmath39 term in the factorization ( 1.5 ) ensures that , in eq .
( 1.6 ) , the first derivative of @xmath40 is weighed by a vanishing coefficient .
this is a sort of canonical form of linear second - order ordinary differential equations with variable coefficients ( see section 10.2 of ref .
@xcite ) , and eq .
( 1.6 ) can be viewed as a @xmath41-parameter family of such equations , the parameters being the triplet @xmath42 .
section ii relates eq .
( 1.6 ) to the bessel functions , studies a specific choice of cauchy data and eventually solves eq .
( 1.3 ) through an integral representation that relies upon integration in the complex domain .
section iii evaluates the bicharacteristics in kasner spacetime , relating them to elliptic integrals , while sec .
iv builds the parametrix of our scalar wave equation through a pair of integral operators where the integrand consists of amplitude and phase functions . concluding remarks and open problems
are presented in sec .
v , while relevant background material is described in the appendices .
from now on , we therefore study until the end of next subsection the ordinary differential equation @xmath44w(t)=0 . \label{(2.1)}\ ] ] this is a particular case of the differential equation @xmath45f(t)=0 , \label{(2.2)}\ ] ] which is solved by the linear combination @xmath46 by comparison of eqs .
( 2.1 ) and ( 2.2 ) we find @xmath47 and hence , in light of what we pointed out at the end of sec .
i , our partial differential equation ( 1.4 ) is solved by replacing @xmath48 and @xmath49 in ( 2.3 ) by some functions @xmath50 and @xmath51 , whose form depends on the choice of cauchy data , i.e. ( see sec .
iii ) @xmath52 the bessel function @xmath53 is not regular at @xmath54 and hence , by using this representation , we are considering an initial time @xmath55 .
we use the linearly independent bessel functions @xmath56 and @xmath53 which describe accurately the time dependence of the integrand in eq .
note that the three choices @xmath57 are equivalent , since the three coordinates @xmath58 in the scalar wave equation @xcite are on equal footing . only the calculational details change .
more precisely , on choosing @xmath59 , one finds @xmath60 whereas , upon choosing @xmath61 , one finds @xmath62 the task of solving our wave equation ( 1.3 ) can be accomplished provided that one knows the cauchy data @xmath63 indeed , from our eqs .
( 1.4 ) , ( 1.5 ) and ( 2.5 ) , one finds ( denoting by an overdot the partial derivative with respect to @xmath64 ) @xmath65 @xmath66 .
\label{(2.8)}\end{aligned}\ ] ] equations ( 2.7 ) and ( 2.8 ) are a linear system of algebraic equations to be solved for @xmath67 and @xmath68 , and they can be studied for various choices of cauchy data . for example , inspired by the simpler case of scalar wave equation in two - dimensional minkowski spacetime , we may consider the cauchy data @xcite @xmath69 @xmath70 where @xmath12 has dimension of length .
thus , by virtue of the identity @xmath71 , \label{(2.11)}\ ] ] we obtain from ( 2.7 ) and ( 2.9 ) @xmath72 , \label{(2.12)}\ ] ] while ( 2.8 ) and ( 2.10 ) yield @xmath73 an interesting generalization of the cauchy data ( 2.9 ) and ( 2.10 ) might be taken to be @xmath74 @xmath75 since it reduces to ( 2.9 ) and ( 2.10 ) at @xmath76 , which is indeed the value of initial time assumed in the minkowski spacetime example considered in ref .
@xcite ( whereas in kasner spacetime we take so far @xmath77 to have enough equations to determine @xmath50 and @xmath51 ) .
hereafter , to avoid cumbersome formulas , we keep choosing the cauchy data ( 2.9 ) and ( 2.10 ) . at this stage ,
( 2.7 ) , ( 2.8 ) , ( 2.12 ) and ( 2.13 ) lead to @xmath78 @xmath79 where ( 2.12 ) should be used to express @xmath80 .
the integrand of eq .
( 1.4 ) is therefore expressed in factorized form through bessel functions , decaying exponentials and oscillating functions , but the evaluation of the integral is hard , even in this simple case .
note now that the original hyperbolic equation ( 1.3 ) is a particular case of the general form @xcite @xmath81 \equiv \left[{\partial^{2}\over \partial t^{2}}-\left ( \sum_{j , k=1}^{n}a^{jk } { \partial^{2}\over \partial x^{j } \partial x^{k}}+b { \partial \over \partial t } + \sum_{j=1}^{n}b^{j}{\partial \over \partial x^{j}}+c \right)\right]u=0 . \label{(2.18)}\ ] ] in the general theory , @xmath82 is a symmetric tensor , @xmath83 is a @xmath8 vector field and @xmath84 is a @xmath8 scalar field . in our case
, we have @xmath85 and @xmath86 thus , for all @xmath87 ( as we said before , we avoid @xmath54 , which is a singularity of the kasner coordinates ) , we can exploit the integral representation ( see appendix a ) of the solution of hyperbolic equations with variable coefficients @xcite , while remarking that eq .
( 1.3 ) is also of a type similar to other hyperbolic equations for which the mathematical literature ( see appendix a ) has proved that the cauchy problem is well posed @xcite . on referring the reader to chapters @xmath88 and @xmath89 of ref .
@xcite for the interesting details , we simply state here the main result when eqs .
( 2.18 ) and ( 2.19 ) hold . 0.3 cm * theorem 2.1 * the solution of the scalar wave equation ( 1.3 ) with cauchy data ( 2.9 ) and ( 2.10 ) at @xmath90 admits the integral representation @xmath91 , \label{(2.20)}\ ] ] where @xmath92 is a fundamental solution ( see appendix b ) of the adjoint equation @xmath93=0 , \label{(2.21)}\ ] ] @xmath94 being the adjoint operator acting , in our case , as @xmath95 while the integrand @xmath96 $ ] is the differential @xmath41-form @xmath97=\left[bu s- \left(s { \partial u \over
\partial \tau } -u { \partial s \over \partial \tau}\right)\right ] { \rm d}y^{1 } \wedge { \rm d}y^{2 } \wedge { \rm d}y^{3 } \nonumber \\ & + & \sum_{j=1}^{3}(-1)^{j } \sum_{k=1}^{3 } a^{jk } \left(s { \partial u \over
\partial y^{k}}-u { \partial s \over \partial y^{k}}\right ) { \rm d}\tau \wedge { \rm d}y^{1 } \wedge ... { \widehat { { \rm d } y^{j } } } \wedge ... { \rm d}y^{3}. \label{(2.23)}\end{aligned}\ ] ] with this notation , the hat upon @xmath98 denotes omission of integration with respect to that particular variable , and @xmath99 is the region of integration viewed as a cell in the complex domain , with boundary @xmath100 .
integration over @xmath100 should be therefore interpreted in the sense of the calculus of exterior differential forms .
our @xmath99 is a manifold defined by the conditions @xmath101 @xmath102 which describe a sphere of radius @xmath103 in the complex domain , centered at the real point @xmath104 .
moreover , the symbolic notation @xmath105 indicates the process of describing the boundary @xmath100 down around the domain of dependence on the space where the initial data ( 2.9 ) and ( 2.10 ) are assigned ( such a space is a two - dimensional plane when the kasner exponents @xmath43 are chosen , whereas , for more general exponents , it corresponds to a singular surface of infinite curvature ) .
a quite complicated evaluation of residues is involved in eq .
( 2.20 ) , because the fundamental solution @xmath92 of eq . ( 2.21 ) is singular where the hadamard - ruse - synge world function ( see appendix b ) vanishes .
equation ( 1.3 ) is just a particular case of the following general form of scalar wave equation in kasner spacetime : @xmath107 in light of the technical results in appendix a , it is rather important to study eq . (
3.1 ) with generic values of parameters , which is what we do now .
the leading symbol of the wave operator @xmath108 is the contravariant form @xmath109 of the metric , and hence the characteristic polynomial reads as @xmath110 an hypersurface @xmath111 is called a _
characteristic _ of @xmath108 if the restriction @xmath112 can be expressed by using only derivatives tangential to @xmath111 of the restrictions @xmath113 and @xmath114 .
this implies that the characteristics of wave equations on @xmath94 are the null hypersurfaces of @xmath94 @xcite .
hence there is , at each point of the characteristic surface @xmath111 , a unique null direction that is both normal and tangential to @xmath111 .
the curves on @xmath111 that are tangential to this null direction field form a congruence on @xmath111 and are called the _ bicharacteristics_.
suppose now that @xmath115 is a coordinate neighbourhood such that @xmath116 where @xmath117 and @xmath118 . if @xmath111 is a characteristic , one has @xcite
@xmath119 by introducing @xmath120 as a local coordinate , this implies , for some @xmath121 , the equation @xmath122 the differential equations @xmath123 where @xmath124 is a parameter , have a unique smooth solution for given initial values @xmath125 . since the gradient of @xmath120 is orthogonal to a hypersurface @xmath126 , the bicharacteristics of @xmath111 are obtained by integrating eq .
( 3.6 ) subject to the initial condition @xmath127 .
if the map @xmath128 is an integral curve of eq .
( 3.6 ) , and @xmath129 one finds @xmath130 on a bicharacteristic , one has @xmath131 , and hence it follows from eq .
( 3.5 ) that the equations @xmath132 hold on a bicharacteristic .
one can introduce an invertible parameter transformation by defining a function @xmath133 such that @xcite @xmath134 if one then defines @xmath135 the equations ( 3.9 ) are transformed into the equivalent hamilton - like set @xmath136 @xmath137 where @xmath138 plays the role of hamiltonian function .
interestingly , eqs .
( 3.12 ) and ( 3.13 ) are the equations of null geodesics , in canonical form .
the null nature of such geodesics follows immediately from the definition ( 3.11 ) and eq .
( 3.5 ) , i.e. @xmath139_{\sigma_{\omega } } = \left[\left({{\rm d}\tau \over { \rm d}\psi}\right)^{2}g^{\alpha \beta } ( \partial_{\alpha}\chi)(\partial_{\beta}\chi)\right]_{\sigma_{\omega } } = \left[\left({{\rm d}\tau \over { \rm d}\psi}\right)^{2 } a \chi \right]_{\sigma_{\omega}}=0 .
\label{(3.14)}\ ] ] thus , the bicharacteristics are null geodesics .
a characteristic contains a congruence of bicharacteristics , the latter being the null geodesics of spacetime @xmath140 .
the bicharacteristics of @xmath108 can also be viewed as the projections on the manifold @xmath94 of the curves in the cotangent bundle @xmath141 , called bicharacteristic strips , that are the integral curves of the hamiltonian system of ordinary differential equations ( 3.12 ) and ( 3.13 ) .
the equation to which eq .
( 3.5 ) reduces when @xmath142 ( which occurs on characteristics and bicharacteristics ) is the well known _ eikonal equation _
@xcite ) . if the parameter @xmath143 is set equal to @xmath144 , there is complete formal analogy between eqs .
( 3.12 ) , ( 3.13 ) and a set of hamilton equations .
as far as wave propagation is concerned , the _ wave fronts _ of our wave equation ( 3.1 ) are characteristic surfaces ( satisfying therefore eqs .
( 3.3 ) and ( 3.5 ) ) and propagate along bicharacteristics @xcite . with
our kasner metric and @xmath145 , the eqs .
( 3.12 ) and ( 3.13 ) for bicharacteristics , bearing in mind that @xmath146 in @xmath147 units , take the form ( there is no summation over @xmath148 in eq . ( 3.16 ) below ) @xmath149 @xmath150 @xmath151 @xmath152 equation ( 3.18 ) is solved by @xmath153 further differentiation with respect to @xmath154 of eq .
( 3.15 ) leads therefore , by virtue of ( 3.17 ) , to the equivalent decoupled system given by @xmath155 together with eqs .
( 3.16 ) and ( 3.17 ) . upon defining @xmath156
we find , more explicitly , the following nonlinear equation for @xmath157 : @xmath158 supplemented by the initial conditions @xmath159 since the right - hand side of eq .
( 3.21 ) is independent of @xmath154 we are dealing with an autonomous differential equation , which can be solved by separation of variables . for this purpose , setting @xmath160 one finds @xmath161 which implies , denoting by @xmath162 an integration constant , @xmath163 and hence @xmath164 where a further integration yields @xmath165 for some constant @xmath166 . in our case ,
the integral of @xmath167 is such that @xmath168 so that eq .
( 3.26 ) for the geodesic parameter @xmath154 in terms of the time variable @xmath64 reads as @xmath169 as far as we can see , such an integral can not be evaluated explicitly for generic values of kasner parameters , but in the particular case @xmath170 which is technically harder than the @xmath43 choice studied in ref .
@xcite , we have found an explicit formula in terms of elliptic integrals .
to be self - contained , we recall some basic definitions , as follows . 0.3 cm ( i ) if @xmath171-{\pi \over 2},{\pi \over 2 } \right[$ ] , the elliptic integral of the first kind , here denoted by @xmath172 $ ] , is defined by @xmath173 \equiv \int_{0}^{\omega}(1-m \sin^{2}\theta)^{-{1 \over 2}}{\rm d}\theta . \label{(3.30)}\ ] ] ( ii ) if @xmath171-{\pi \over 2},{\pi \over 2 } \right[$ ] , the elliptic integral of the second kind , @xmath174 $ ] , reads as @xmath175 \equiv \int_{0}^{\omega}(1-m \sin^{2}\theta)^{{1 \over 2}}{\rm d}\theta .
\label{(3.31)}\ ] ] ( iii ) if @xmath115 lies the same open interval as in ( i ) and ( ii ) , the incomplete elliptic integral of the third kind is given by @xmath176 \equiv \int_{0}^{\omega}(1-n \sin^{2}\theta)^{-1 } ( 1-m \sin^{2}\theta)^{-{1 \over 2}}{\rm d}\theta .
\label{(3.32)}\ ] ] moreover , in order to express our results , we have to consider the three roots @xmath177 of the algebraic equation @xmath178\over ( { \tilde \xi}_{3})^{2}}=0 . \label{(3.33)}\ ] ] if the three roots are all real , we follow the convention according to which @xmath179 . if @xmath180 is the real root while the remaining two roots are complex conjugate , we agree that @xmath181 and @xmath182 are such that @xmath183 . with this understanding , and defining @xmath184 @xmath185 @xmath186 \right .
& + & { 1 \over r_{2}}\bigr[(\gamma + 4 ( { \tilde \xi}_{3})^{2}r_{2 } ) e_{i}[{\rm arcsin}f(r , t),f_{3,1}(r)/f_{2,1}(r ) ] \nonumber \\ & - & \left . \gamma e_{iii}[f_{3,1}(r),{\rm arcsin}f(r , t),f_{3,1}(r)/f_{2,1}(r)]\bigr]\right \ } , \label{(3.36)}\end{aligned}\ ] ] the integral ( 3.28 ) reads as @xmath187 .
\label{(3.37)}\end{aligned}\ ] ] furthermore , since eq .
( 3.37 ) can not be explicitly inverted to express @xmath188 , it is more convenient to use the identities @xmath189 so that the remaining components of null geodesic equations are expressed , from ( 3.26 ) and ( 3.38 ) , in the form @xmath190 @xmath191 with the choice ( 3.29 ) of kasner parameters , also ( 3.39 ) can be expressed in terms of elliptic integrals , while ( 3.40 ) takes the remarkably simple form @xmath192+\gamma t^{4 \over 3 } + 4({\tilde \xi}_{3})^{2}t^{2 } \over t^{{4 \over 3}}}}. \label{(3.41)}\ ] ] in the formula ( 3.39 ) , when @xmath193 , we find @xmath194 where @xmath195 \sqrt{(t^{2/3}-r_{1})(t^{2/3}-r_{2})(t^{2/3}-r_{3 } ) { f_{3,1}(r)\over r_{2}r_{3 } } } , \label{(3.43)}\ ] ] while , if @xmath196 , we find @xmath197 where , upon defining @xmath198 we write @xmath199\biggr ) \nonumber \\ & + & 3 \gamma \biggr((t^{2/3}-r_{2})(t^{2/3}-r_{3})t^{2/3 } \nonumber \\ & + & r_{3 } ( t^{2/3}-r_{1})^{2 } { \sqrt{(t^{2/3}-r_{2})(t^{2/3}-r_{3})}r_{1 } \over ( t^{2/3}-r_{1})^{3/2}\sqrt{\rho(t)r_{2}r_{3 } } } \nonumber \\ & \times & \biggr({(r_{1}+r_{3})\over ( r_{1}-r_{3 } ) } e_{i}[{\rm arcsin}f(r , t),f_{3,1}(r)/f_{2,1}(r ) ] + { r_{2}\over r_{1 } } e_{ii}[{\rm arcsin}f(r , t),f_{3,1}(r)/f_{2,1}(r ) ] \nonumber \\ & + & { ( r_{1}+r_{2}+r_{3})\over ( r_{3}-r_{1 } ) } e_{iii}[f_{3,1}(r),{\rm arcsin}f(r , t),f_{3,1}(r)/f_{2,1}(r)]\biggr ) \biggr)\biggr ] \biggr \}. \label{(3.46)}\end{aligned}\ ] ] our original calculation of bicharacteristics in kasner spacetime is therefore completed .
since we are studying a wave equation , we may expect that the solution formula involves amplitude and phase functions , as well as the cauchy data ( here , unlike sec .
ii , we exploit techniques that do not need to avoid @xmath54 ) @xmath200 @xmath201 which are again assumed to be fourier transformable . however , the variable nature of the coefficients demands for a nontrivial generalization of the integral representation ( 1.4 ) .
this is indeed available , since a theorem guarantees that the solution of the cauchy problem ( 3.1 ) , ( 4.1 ) and ( 4.2 ) can be expressed in the form @xcite @xmath202 where , on denoting by @xmath203 the fourier transform of the cauchy data , the operators @xmath204 act according to ( hereafter , @xmath205 , with covariable @xmath206 ) @xmath207 where the @xmath208 are real - valued phase functions which satisfy the initial condition @xmath209 and @xmath210 is a regularizing operator which smoothes out the singularities acted upon by it @xcite . in other words ,
the cauchy problem is here solved by a pair of fourier - maslov integral operators of the form ( 4.4 ) , and such a construction ( leaving aside , for the moment , its global version , which can be built as shown in chapter vii of ref .
@xcite ) generalizes the monochromatic plane waves for the dalembert operator from minkowski spacetime to kasner spacetime .
strictly , we are dealing with the approximate green function for the wave equation , called the parametrix . in our case , since we know a priori that ( 4.3 ) and ( 4.4 ) yield an exact solution of the cauchy problem , we can insert them into eq .
( 3.1 ) , finding that , for all @xmath211 , @xmath212 \sim \sum_{k=1}^{2 } ( 2\pi)^{-3 } \int p[{\rm e}^{{\rm i}\varphi_{k}}\alpha_{jk } ] { \hat u}_{j}(\xi){\rm d}^{3}\xi , \label{(4.6)}\ ] ] where @xmath213 can be neglected with respect to the integral on the right - hand side of eq .
( 4.4 ) , because @xmath210 is a regularizing operator .
next , we find from eq .
( 3.1 ) that @xmath214={\rm e}^{{\rm i}\varphi_{k}}({\rm i}a_{jk}+b_{jk } ) , \label{(4.7)}\ ] ] where @xmath215 @xmath216 if the phase functions @xmath208 are real - valued , since the exponentials @xmath217 can be taken to be linearly independent , we can fulfill eq .
( 3.1 ) , up to the negligible contributions resulting from @xmath213 , by setting to zero in the integrand ( 4.6 ) both @xmath218 and @xmath219 .
this leads to a coupled system of partial differential equations .
our cauchy problem ( 3.1 ) , ( 4.1 ) and ( 4.2 ) is therefore equivalent to solving the equations @xmath220 equation ( 4.10 ) is the _ dispersion relation _ for the scalar wave equation in kasner spacetime .
such a dispersion relation takes a neater geometric form upon bearing in mind the form ( 3.1 ) of the wave ( or dalembert ) operator @xmath221 in kasner coordinates , i.e. @xmath222=0 , \label{(4.11)}\ ] ] @xmath223\alpha_{jk}=0 .
\label{(4.12)}\ ] ] let us bear in mind that the indices @xmath224 and @xmath4 are not tensorial , but they merely count the number of functions contributing to the fourier - maslov integral operator ( 4.4 ) .
we can therefore exploit the four - dimensional concept of gradient of a function @xcite as the four - dimensional covariant vector defined by the differential of the function , i.e. @xmath225 where @xmath226 is the levi - civita connection on four - dimensional spacetime , and we exploit the identity @xmath227 .
the consideration of @xmath228 is not mandatory at this stage , but it will be helpful in a moment , when we write ( see below ) in tensor language the equations expressing the dispersion relation .
we arrive therefore , upon multiplying eq .
( 4.11 ) by @xmath229 , while dividing eq . ( 4.12 ) by @xmath229 , at the following geometric form of dispersion relation in kasner spacetime ( with our notation we actually write it in the same way in any lorentzian spacetime ) : @xmath230 = { \rm div}\bigr[(\alpha_{jk})^{2}{\rm grad}\varphi_{k}\bigr]=0 , \label{(4.14)}\ ] ] @xmath231 where the four - dimensional divergence operator acts according to is a gradient , one gets therefore the wave operator on scalars , i.e. @xmath232 ] @xmath233 note that , if the ratio @xmath234 is much smaller than a suitable parameter having dimension @xmath235 , eq . ( 4.15 ) reduces to the eikonal equation and hence the phase function reduces to the hadamard - ruse - synge world function of appendices b and c. this property makes contact with the asymptotic expansion presented in appendix b. however , it is possible to devise a strategy to solve exactly eqs .
( 4.14 ) and ( 4.15 ) . for this purpose
we remark that , upon defining the covariant vector @xmath236 eq .
( 4.14 ) is equivalent to solving the first - order partial differential equation expressing the vanishing divergence condition for @xmath237 , i.e. @xmath238 of course , this equation is not enough to determine the four components of @xmath237 , but there are cases where further progress can be made ( see below ) . after doing that , we can express the ( covariant ) derivative of the phase function from the definition ( 4.17 ) , i.e. @xmath239 and the insertion of eq .
( 4.19 ) into eq .
( 4.15 ) yields @xmath240 interestingly , this is a tensorial generalization of a famous nonlinear ordinary differential equation , i.e. the ermakov - pinney equation @xcite @xmath241 if @xmath242 is replaced by @xmath243 , @xmath244 is set to zero and @xmath245 is promoted to a function of spacetime location , eq .
( 4.21 ) is mapped into eq .
( 4.20 ) . after solving this nonlinear equation for @xmath246 $ ]
, the task remains of finding the phase function @xmath208 by writing and solving the four components of eq .
( 4.19 ) . to sum up
, we have proved the following original result .
0.3 cm * theorem 4.1 * for any lorentzian spacetime manifold @xmath140 , the amplitude functions @xmath247 and phase functions @xmath248 in the parametrix ( 4.4 ) for the scalar wave equation can be obtained by solving , first , the linear condition ( 4.18 ) of vanishing divergence for a covariant vector @xmath237 .
all nonlinearities of the coupled system are then mapped into solving the nonlinear equation ( 4.20 ) for the amplitude function @xmath229 .
eventually , the phase function @xmath208 is found by solving the first - order linear equation ( 4.19 ) . in kasner spacetime ,
( 4.18 ) takes indeed the form @xmath249 this suggests considering @xmath250 and @xmath251 such that @xmath252 so that eq .
( 4.22 ) leads to the equation @xmath253 this is precisely the vanishing divergence condition satisfied by retarded potentials in minkowski spacetime in the coordinates @xmath254 .
their integral representation is well known to be of the form ( recall that we work in @xmath147 units ) @xmath255 @xmath256 where @xmath257 and hence eqs .
( 4.23 ) , ( 4.25 ) and ( 4.26 ) solve completely the problem of finding the auxiliary covariant vector @xmath237 in kasner spacetime .
we should now solve eq .
( 4.20 ) for @xmath229 .
the reader might wonder what has been gained by turning the task of solving the scalar wave equation into the task of solving eq .
( 4.20 ) . in this equation
, we can first get rid of the part linear in @xmath258 in the @xmath259 operator by setting @xmath260 which leads to @xmath261 { \tilde \alpha}_{jk}=t^{2}\psi_{\gamma}\psi^{\gamma}. \label{(4.29)}\ ] ] next , we can get rid of powers of @xmath262 by setting @xmath263 , which yields @xmath264 and the same formula holds with @xmath64 replaced by @xmath265 . thus , upon choosing @xmath266 , we obtain eventually the amplitude functions by solving the following nonlinear equation for @xmath267 : @xmath268 + 4t^{2}\bigr[(\psi_{0})^{2}-\sum_{l=1}^{3}t^{-2p_{l}}(\psi_{l})^{2}\bigr ] .
\label{(4.31)}\end{aligned}\ ] ] the form ( 4.31 ) of the equation for @xmath267 makes it possible to apply the powerful adomian method @xcite for the solution of nonlinear partial differential equations . for this purpose , inspired by ref .
@xcite , we define the four linear operators occurring in @xmath12 , i.e. @xmath269 the remainder ( i.e. , lower order part ) of the linear operator @xmath12 , i.e. @xmath270 the nonlinear term ( hereafter we omit the subscripts @xmath271 for simplicity of notation ) @xmath272 , \label{(4.34)}\ ] ] while the part of the right - hand side which is independent of @xmath167 is here denoted by @xmath273 , i.e. @xmath274 . \label{(4.35)}\ ] ] hence the nonlinear equation ( 4.31 ) can be re - expressed in the form @xmath275 the idea is now to apply the inverse of @xmath276 , or @xmath277 , or @xmath278 , or @xmath279 to this equation , which , upon bearing in mind the identities @xcite @xmath280 @xmath281 @xmath282 @xmath283 the @xmath284 s being constants fixed by the initial and boundary conditions , leads to the following four equations : @xmath285 @xmath286 @xmath287 @xmath288 now we add these four equations , and upon defining @xmath289 , \label{(4.45)}\end{aligned}\ ] ] @xmath290 @xmath291 + g\eta , \label{(4.47)}\ ] ] we arrive at the fundamental formula @xmath292 at this stage , if the function @xmath167 has a poincar asymptotic expansion @xcite , which can be convergent or divergent and is written in the form @xmath293 we point out that ( 4.49 ) leads in turn to a poincar asymptotic expansion of the nonlinear term @xmath294 defined in eq .
( 4.34 ) in the form @xmath295+a_{1}[f_{0},f_{1}]+ ... + a_{k}[f_{0}, ... ,f_{k}]+ ... , \label{(4.50)}\ ] ] where , by virtue of the formula @xmath296 we can evaluate the poincar asymptotic expansion of squared logarithmic derivatives according to @xmath297_{,l } \right \}^{2 } \nonumber \\ & \sim & \left[{(f_{{0},l}+f_{{1},l})\over f_{0 } } -{f_{1}f_{{0},l}\over ( f_{0})^{2 } } -{f_{1}f_{{1},l}\over ( f_{0})^{2 } } + { ( f_{1})^{2}f_{{0},l}\over ( f_{0})^{3}}+ ... \right]^{2}. \label{(4.52)}\end{aligned}\ ] ] in light of ( 4.34 ) and ( 4.52 ) we find @xmath298={3 \over 4f_{0}}\left[\sum_{l=1}^{3}t^{-2p_{l}}(f_{{0},l})^{2 } -(f_{{0},t})^{2}\right ] , \label{(4.53)}\ ] ] @xmath299={3 \over 4f_{0}}\left(f_{1}\left(1-{f_{1}\over f_{0}}\right)\right)^{2 } \left \ { \sum_{l=1}^{3}t^{-2p_{l } } \left[{f_{{1},l}\over f_{1 } } -{f_{{0},l}\over f_{0}}\right]^{2 } -\left[{f_{{1},t}\over f_{1 } } -{f_{{0},t}\over f_{0}}\right]^{2}\right \ } , \label{(4.54)}\ ] ] plus a countable infinity of other formulas for @xmath300 $ ] , ...
@xmath301 $ ] ... .
note that , unlike the case of simpler nonlinearities @xcite , the functionals @xmath302 involve division by @xmath303 .
the solution algorithm is now completely specified , because eq . ( 4.48 ) yields the recursive formulas @xcite @xmath304 \ ; \forall n=0,1, ...
,\infty , \label{(4.55)}\ ] ] and hence @xmath305 where , by exploiting the partial sum of the geometric series , we find @xmath306 . \label{(4.57)}\ ] ] since the operator @xmath307 is built from the inverses of differential operators , it is a pseudo - differential operator , and it remains to be seen whether , for sufficiently large values of @xmath308 , it only contributes to the terms @xmath210 in the parametrix ( 4.4 ) , so that we only need the limit @xmath309 .
\label{(4.58)}\ ] ] the adomian method we have used is well suited to go beyond weak nonlinearity and small perturbations , but of course the nontrivial technical problem is whether the series for the unknown function @xmath167 is convergent , and also how fast . if it were necessary to consider hundreds of terms , the algorithm would be of little practical utility .
an interesting alternative , which can not be ruled out at present , is instead the existence of an asymptotic expansion of @xmath167 involving only finitely many terms , whose rigorous theory is described in a monograph by dieudonn @xcite . in such a case we might write @xmath310 -ga_{1}[f_{0},f_{1}=kf_{0}-ga_{0}[f_{0 } ] ] , \label{(4.59)}\ ] ] which is fully computable by virtue of eqs .
( 4.45)-(4.47 ) and ( 4.53)-(4.55 ) .
we find it therefore encouraging that an exact solution algorithm has been obtained for the scalar parametrix in kasner spacetime .
last , but not least , eqs .
( 4.19 ) for the gradient of phase functions can be integrated to find @xmath311 bearing in mind that @xmath312 and eq .
( 4.23 ) , while the @xmath313 functions may be fixed by demanding consistency with eq . ( 4.5 ) .
this method leads to the following formulas for the complete evaluation of phase functions : @xmath314 @xmath315 where @xmath316 denotes the triplet @xmath317 deprived of the @xmath318-th coordinate , and no summation over @xmath318 is performed on the right - hand side .
the work in ref .
@xcite succeeded in the difficult task of setting up a solution algorithm for defining and solving self - dual gravity field equations to first order in the noncommutativity matrix .
however , precisely the first building block , i.e. the task of solving the scalar field equation in a classical self - dual background was only briefly described .
this incompleteness has been taken care of in the present paper for the case of kasner spacetime , first with a particular choice of kasner parameters : @xmath319 . the physics - oriented literature had devoted efforts to evaluating quantum propagators for a massive scalar field in the kasner universe @xcite , but the relevance for the classical wave equation of the mathematical work in refs .
@xcite had not been appreciated , to the best of our knowledge . as far as we know , our original results in secs .
iii and iv are substantially new .
we have indeed evaluated the bicharacteristics of kasner spacetime in terms of elliptic integrals of first , second and third kind , while the nonlinear system for obtaining amplitude and phase functions in the scalar parametrix has been first mapped into eqs .
( 4.18)-(4.20 ) , a set of equations that holds in any curved spacetime .
furthermore , the nonlinear equation ( 4.20 ) has been mapped into eq .
( 4.31 ) , and the latter has been solved with the help of the adomian method , arriving at eqs .
( 4.53)-(4.59 ) .
there is however still a lot of work to do , because the proof that the asymptotic expansion of @xmath320 is of the poincar type or , instead , only involves finitely many terms , might require new insight from asymptotic and functional analysis .
this adds evidence in favour of noncommutative gravity needing the whole apparatus of classical mathematical physics for a proper solution of its field equations ( see also the work in ref .
@xcite , where noether - symmetry methods have been used to evaluate the potential term for a wave - type operator in bianchi i spacetime ) .
g. e. and e. d. g. are grateful to the dipartimento di fisica of federico ii university , naples , for hospitality and support .
as we know from sec . v @xcite , eq .
( 1.3 ) is a particular case of the wave equation ( 3.1 ) . the operator @xmath108 in eq .
( 3.1 ) is an example of what is called , in the mathematical literature , a fuchsian hyperbolic operator with weight @xmath321 with respect to @xmath64 . in general , the weight is @xmath322 , and such fuchsian hyperbolic operators read as ( hereafter @xmath323 \times { \bf r}^{n}$ ] ) @xmath324 subject to @xmath325 conditions stated in ref .
@xcite which specify the relation between @xmath4 and @xmath21 , the form of the coefficients , hyperbolicity , quadratic form associated to the operator ( see below ) , estimates for principal part and lower order terms of the operator .
when all these @xmath325 conditions hold , one can prove the following theorem @xcite : 0.3 cm * theorem a1*. for any functions @xmath326 and @xmath327 \times { \bf r}^{n})$ ] , there exists a unique solution @xmath328 \times { \bf r}^{n})$ ] such that @xmath329 \times { \bf r}^{n } , \label{(a2)}\ ] ] @xmath330 and the solution has a finite propagation speed . for the operator in eq .
( 3.1 ) , the quadratic form of the general theory , obtained by replacing @xmath331 in all spatial derivatives of second order , reads as @xmath332 according to tahara , for fuchsian hyperbolic operators , the quadratic form @xmath333 leading to theorem a1 should be positive - definite as a function of @xmath334 for any @xmath335 , with symmetric coefficients of class @xmath336 on @xmath337 $ ] , and such that @xmath338 for the operator in eq . ( 3.1 ) one finds indeed @xmath339 thus , bearing in mind that , when the @xmath340 kasner parameters are all nonvanishing , one of them is negative and the other two are positive , one obtains ( on defining @xmath341 for all @xmath342 ) @xmath343 and hence condition ( a5 ) of the general theory is fulfilled .
this is also the case of the operator in eq .
( 1.3 ) , for which @xmath344 which implies that @xmath345 in other words , the hyperbolic equation studied in our paper can always rely upon the tahara theorem on the cauchy problem .
if instead we resort to the garabedian technique of integration in the complex domain , strictly speaking , we need to assume analytic coefficients @xcite , which is not fulfilled , for example , by @xmath346 in ( 2.19 ) if we replace @xmath64 by a complex @xmath347 and want to consider also the value @xmath348
. however , ref .
@xcite describes the way out of this nontrivial technical difficulty . for this purpose
, one considers first a more complicated , inhomogeneous equation @xmath81=f \label{(a10)}\ ] ] with analytic coefficients and analytic right - hand side , from which one can write down a direct analogue of the solution ( 2.20 ) in the form @xmath349 + \int_{d}(pf -u m[{\cal p}]){\rm d}\tau \wedge { \rm d}y^{1 } \wedge { \rm d } y^{2 } \wedge { \rm d } y^{3 } \right ] , \label{(a11)}\ ] ] where @xmath350 is called a
_ parametrix _
( i.e. a distribution @xcite that provides an approximate inverse ) and is given by @xmath351 in terms of the world function @xmath352 of appendix b. the notation @xmath105 means that the manifold of integration @xmath99 is supposed to approach the real domain in such a way that it folds around the characteristic conoid @xmath353 without intersecting it .
equation ( a11 ) defines a volterra integral equation for the solution of the cauchy problem .
it follows that @xmath354 varies continuously with the derivatives of the coefficients of eq .
similarly , the second partial derivatives of @xmath354 depend continuously on the derivatives of the coefficients of a high enough order .
thus , when they are _ no longer analytic , we may replace these coefficients by polynomials approximating an appropriate set of their derivatives _ in order to establish the validity of ( a11 ) in the general case by passage to the limit .
note also that the integral equation ( a11 ) has a meaning in the real domain even where the partial differential equation ( a10 ) is not analytic , since the construction of the parametrix @xmath350 and of the world function @xmath352 only requires differentiability of the coefficients of a sufficient order @xcite .
more precisely , for coefficients possessing partial derivatives of all orders , we introduce a polynomial approximation that includes enough of these derivatives to ensure that the solution of the corresponding approximate equation ( a11 ) converges together with its second derivatives .
the limit has therefore to be a solution of the cauchy problem associated with the more general coefficients , and must itself satisfy the volterra integral equation ( a11 ) .
in his analysis of partial differential equations , hadamard discovered the importance of the _ world function _ @xcite , which can be defined as the square of the geodesic distance between two points with respect to the metric @xmath355 in the analysis of second - order linear partial differential equations @xmath356=\left[\sum_{i , j=1}^{n}a^{ij}{\partial^{2}\over \partial x^{i } \partial x^{j } } + \sum_{i=1}^{n}b^{i}{\partial \over \partial x^{i}}+c \right]u=0 , \label{(b2)}\ ] ] the first - order nonlinear partial differential equation ( cf .
( 3.5 ) ) for the world function @xmath352 reads as @xcite @xmath357 where the coefficients @xmath358 are the same as those occurring in the definition of the operator @xmath359 ( this is naturally the case because the wave or laplace operator can be always defined through the metric , whose signature determines the hyperbolic or elliptic nature of the operator , as we stressed in sec .
the world function can be used provided that the points @xmath16 and @xmath360 are so close to each other that no caustics occur . a _ fundamental solution _
@xmath361 of eq .
( b2 ) is a distribution @xcite , and can be defined to be @xcite a solution of that equation in its dependence on @xmath362 possessing , at the parameter point @xmath363 , a singularity characterized by the representation @xmath364 where @xmath365 are supposed to be regular functions of @xmath16 in a neighbourhood of @xmath360 , with @xmath366 at @xmath367 , and where the exponent @xmath21 depends on the spacetime dimension @xmath308 according to @xmath368 .
the sources of nonvanishing @xmath369 are either a mass term in the operator @xmath359 @xcite or a nonvanishing spacetime curvature @xcite .
the term @xmath370 plays an important role in the evaluation of the integral ( 2.20 ) , as is stressed in sec .
6.4 of ref .
@xcite . in kasner spacetime
, the hadamard green function ( b4 ) has been evaluated explicitly only with the special choice of parameters @xmath371 in ref .
@xcite . in that case , direct integration of the geodesic equation ( appendix c ) yields eventually an exact formula for the hadamard - ruse - synge world function in the form @xcite @xmath372 having defined @xmath373 @xmath374 following our remarks at the end of sec .
ii , we expect that the choice of kasner parameters made in sec .
ii would still lead to a formula like ( b5 ) for the world function , but with @xmath375 however , as far as we know , the extension of these formulas to generic values of kasner parameters is an open problem .
the calculation in ref .
@xcite is so enlightening and relevant for our purposes that it deserves a brief summary .
to begin , the geodesic equation in a kasner spacetime with metric @xmath376 is the following coupled system of nonlinear differential equations : @xmath377 @xmath378 where @xmath379 is the affine parameter of the geodesic
. equation ( c3 ) can be solved for @xmath380 , because it yields @xmath381 which implies @xmath382 having denoted by @xmath383 three integration constants .
the constancy along the geodesic of the ( pseudo-)norm squared @xmath384 , where @xmath385 , yields ( @xmath103 being negative ( resp .
positive ) for timelike ( resp .
spacelike ) geodesics ) @xmath386 from which we obtain @xmath387 on the other hand , the world function is the square of the geodesic distance between the points @xmath388 and @xmath108 , say , i.e. @xmath389 moreover , following ref .
@xcite , one defines @xmath390 upon considering the particular choice @xmath371 , and defining @xmath391 these formulae make it possible to re - express the integration constants in the form @xmath392 @xmath393 where @xmath394 .
we can now square up the product @xmath395 from ( c12 ) , finding eventually @xmath396 on the other hand , the geodesic distance in ( c8 ) becomes in our case @xmath397 , \label{(c14)}\ ] ] and if we square it up and then exploit ( c13 ) we obtain @xmath398 because the terms involving products of square roots cancel each other . at this stage
, we can re - express the squares of @xmath399 and @xmath400 from ( c11 ) , i.e. @xmath401 by virtue of ( c15 ) and ( c16 ) , we find eventually the result ( b5 ) , where the role played by @xmath402 in the formulas has been made explicit .
s. kleinerman , geom .
special volume gafa2000 , 1 ( 2000 ) .
s. kleinerman , int . j. mod .
d * 22 * , 1330012 ( 2013 ) .
f. treves , _ introduction to pseudodifferential and fourier integral operators .
volume 2 : fourier integral operators _ ( plenum press , new york , 1980 ) .
j. szeftel ,
arxiv:1204.1769 [ math.ap ] . g. esposito , _ dirac operators and spectral geometry _
, cambridge lecture notes in physics vol .
* 12 * ( cambridge university press , cambridge , 1998 ) .
e. di grezia , g. esposito , and p. vitale , phys .
d * 89 * , 064039 ( 2014 ) ; phys .
d * 90 * , 129901 ( 2014 ) .
e. t. whittaker and g. n. watson , _ modern analysis _ ( cambridge university press , cambridge , 1927 ) .
j. d. jackson , _ classical electrodynamics _ ( wiley , new york , 1999 ) .
p. r. garabedian , _ partial differential equations _
( chelsea , new york , 1964 ) .
o. a. oleinik , comm .
pure appl . math .
* 23 * , 569 ( 1970 ) .
a. menikoff , amer .
j. math . * 97 * , 548 ( 1975 ) .
f. g. friedlander , _ the wave equation in curved space - time _ ( cambridge university press , cambridge , 1975 ) .
l. vitagliano , int . j. geom .
11 * , 1460039 ( 2014 ) . y.
choquet - bruhat , in _ battelle rencontres _ , edited by c. m. dewitt and j. a. wheeler
( benjamin , new york , 1968 ) .
t. levi civita , _ caratteristiche dei sistemi differenziali e propagazione ondosa _
( zanichelli , bologna , 1931 ) . h. nariai , nuovo cimento b * 35 * , 259 ( 1976 ) . v. p. ermakov , univ .
kiev , series * iii * 9 , 1 ( 1880 ) .
e. pinney , proc .
* 1 * , 681 ( 1950 ) .
h. r. lewis and w. b. riesenfeld , j. math .
* 10 * , 1458 ( 1969 ) .
g. adomian , j. math .
appl . * 135 * , 501 ( 1988 ) .
h. poincar , acta math .
* 8 * , 295 ( 1886 ) .
j. dieudonn , _ calcul infinitesimal _
( hermann , paris , 1980 ) .
h. tahara , proc .
japan acad .
a * 54 * , 92 ( 1978 ) .
p. r. garabedian , j. math . mec . * 9 * , 241 ( 1960 ) .
b. s. dewitt , phys .
lett . * 4 * , 317 ( 1960 ) .
a. paliathanasis , m. tsamparlis , and m. t. mustafa , int . j. geom .
* 12 * , 1550033 ( 2015 ) , arxiv:1411.0398 [ math - ph ] .
j. hadamard , _ lectures on cauchy s problem in linear partial differential equations _
( dover , new york , 1952 ) .
h. s. ruse , proc .
lond . math .
32 * , 87 ( 1931 ) .
j. l. synge , proc .
soc . * 32 * , 241 ( 1931 ) .
b. s. dewitt , _ dynamical theory of groups and fields _ ( gordon & breach , new york , 1965 )
. g. bimonte , e. calloni , l. di fiore , g. esposito , l. milano , and l. rosa , class .
quantum grav .
* 21 * , 647 ( 2004 ) . | the scalar wave equation in kasner spacetime is solved , first for a particular choice of kasner parameters , by relating the integrand in the wave packet to the bessel functions .
an alternative integral representation is also displayed , which relies upon the method of integration in the complex domain for the solution of hyperbolic equations with variable coefficients . in order to study the propagation of wave fronts ,
we integrate the equations of bicharacteristics which are null geodesics , and we are able to express them , for the first time in the literature , with the help of elliptic integrals for another choice of kasner parameters . for generic values of the three kasner parameters , the solution of the cauchy problem is built through a pair of integral operators , where the amplitude and phase functions in the integrand solve a coupled system of partial differential equations .
the first is the so - called transport equation , whereas the second is a nonlinear equation that reduces to the eikonal equation if the amplitude is a slowly varying function .
remarkably , the analysis of such a coupled system is proved to be equivalent to building first an auxiliary covariant vector having vanishing divergence , while all nonlinearities are mapped into solving a covariant generalization of the ermakov - pinney equation for the amplitude function .
last , from a linear set of equations for the gradient of the phase one recovers the phase itself .
this is the parametrix construction that relies upon fourier - maslov integral operators , but with a novel perspective on the nonlinearities in the dispersion relation .
furthermore , the adomian method for nonlinear partial differential equations is applied to generate a recursive scheme for the evaluation of the amplitude function in the parametrix .
the resulting formulas can be used to build self - dual solutions to the field equations of noncommutative gravity , as has been shown in the recent literature . |
the recent development in electronics , combining the magnetic and semiconducting materials ( so - called magnetoelectronis or spintronics ) , has brought half - metallic ferromagnets , initially predicted by de groot and collaborators in 1983 @xcite , to the centre of scientific research .
these magnetic metals have the peculiarity that the minority - spin band - structure is semiconducting while the majority - spin band - structure is metallic . such half - metals exhibit ,
ideally , a 100% spin polarization at the fermi level and therefore these compounds should have a fully spin - polarized current and should be ideal spin injectors into a semiconductor , thus maximizing the efficiency of spintronic devices @xcite .
an important part of the scientific effort has been focused on the study and fabrication of half - metallic heusler alloys .
the first family of heusler alloys studied were of the form x@xmath0yz , crystallizing in the @xmath1 structure which consists of four fcc sublattices , where x a high valent transition or noble metal atom , y a low - valent transition metal atom and z an sp element @xcite .
such heusler compounds have attracted a lot of interest due to the possibility to study in the same family of alloys a series of interesting diverse magnetic phenomena like itinerant and localized magnetism , antiferromagnetism , helimagnetism , pauli paramagnetism or heavy - fermionic behaviour @xcite .
the heusler alloys of the second class are of the form xyz , crystallizing in the @xmath2 structure , and consisting of three fcc sublattices ; they are often called half- or semi - heusler alloys in literature , while the @xmath1 compounds are referred to as full heusler alloys .
the interest in these types of intermetallic alloys was revived after the prediction @xcite , using first - principles calculations , of half - metallicity in nimnsb , a half - heusler compound .
the main advantages of heusler alloys with respect to other half - metallic systems ( e.g. some oxides like cro@xmath0 and fe@xmath3o@xmath4 and some manganites like la@xmath5sr@xmath6mno@xmath3 ) @xcite ) are their relatively high curie temperatures @xcite . while for the other compounds the curie temperature is near the room temperature , _ e.g. _ for nimnsb it is 730 k and for co@xmath0mnsi it reaches the 985 k @xcite . and
their structural similarity to the zinc - blende structure , adopted by binary semiconductors widely used in industry ( such as gaas on zns ) .
heusler alloys have been already incorporated in spin - filters @xcite , tunnel junctions @xcite , and gmr devices @xcite .
the most successful recent applications in spintronics concern the half - metallic full heusler alloys .
the group of westerholt has incorporated co@xmath0mnge in the case of spin - valves and multilayer structures @xcite .
the group of reiss managed to create magnetic tunnel junctions based on co@xmath0mnsi @xcite . a similar study by sakuraba and collaborators resulted in the fabrication of magnetic tunnel junctions using co@xmath0mnsi as magnetic electrodes and al - o as the barrier and their results are consistent with the presence of half - metallicity for co@xmath0mnsi @xcite . dong and
collaborators recently managed to inject spin - polarized current from co@xmath0mnge into a semiconducting structure @xcite . in a recent article
@xcite we have reviewed theoretical results of the basic electronic and magnetic properties of the half - metallic heusler alloys .
the gap has its origin for both the half- and full - heusler alloys in the @xmath7-@xmath7 hybridisation , which is fundamental for understanding their electronic and magnetic properties . for both families of compounds
the total spin magnetic moment scales with the number of valence electrons and can be described by a slater - pauling rule . in the present contribution
we review our most recent results on the electronic properties of these alloys obtained from first - principles electronic structure calculations . in section [ sec2 ]
we summarize our older results ( see references @xcite and @xcite for an extended review ) .
the following sections are dedicated to phenomena which can destroy half - metallicity . in particular section [ sec3 ]
discuss the effect of the lattice parameter , section [ sec4 ] is dedicated to the effect of spin - orbit coupling , section [ sec6 ] to the doping and disorder effects and the formation of defects . in sections
[ sec9 ] and [ sec10 ] we discuss two special cases : the quaternary heusler compounds and the case of half - metallic ferrimagnets , respectively . in section [ sec11 ]
we review the theoretical studies on the behaviour of the half - metallicity and of the magnetization as a function of the temperature . finally in section [ sec12 ] we summarize and conclude our review .
note that the problems of surface / interface states @xcite are reviewed in other papers in this volume . to obtain these results
we have employed a variety of electronic structure methods each one suitable for the property under investigation .
the results in section [ sec2 ] , [ sec3 ] and [ sec4 ] have been obtained using the full - potential version of the screened korringa - kohn - rostoker ( kkr ) method @xcite .
the relativistic calculations to produce the orbital moments in [ sec4 ] are performed with the fully - relativistic version of the kkr method where the dirac equation is solved @xcite .
results in sections [ sec6 ] and [ sec10 ] come from the full potential nonorthogonal local orbital minimum basis band structure scheme ( fplo ) @xcite .
finally for the study of the quaternary heusler alloys the kkr method is employed within the coherent potential approximation ( cpa ) to simulate the disorder in the crystal @xcite .
the electronic , magnetic and gap properties of half - heusler alloys have been extensively studied in reference @xcite and of full - heusler compounds in reference @xcite .
these results have been extensively also reviewed in reference @xcite and in the introductory chapter of @xcite and the reader is directed to them for an extended discussion . in this section we will only briefly overview these properties .
several heusler alloys have been predicted to be half - metals @xcite ; perhaps the most widely studied among them is nimnsb , the first to be predicted as a half - metal .
the electronic band - structure calculations of nimnsb reveal that for both spin directions there is one band at around @xmath8 ev with reference to the fermi level which arises from the @xmath9 states of the sb atom .
the next three lowest bands are due to the @xmath10 states of sb and accommodate also electrons from the transition metal atoms @xcite .
the ni and mn majority - spin @xmath7 states are energetically close and form a common band .
the situation is different for the minority - spin states : due to the strong exchange splitting of mn , the minority mn @xmath7-states are unoccupied , while the ones of ni are occupied .
the minority - spin gap arises between the bonding and antibonding @xmath7-hybrids created among the mn and ni states ( in this respect , the origin of the gap is similar to the one in compound semiconductors like gaas or znse ) .
the minority - spin bonding states have most of their weight at the ni atom and are occupied , while the antibonding states have their weight mostly at the mn atom and are unoccupied .
this leads to very strong localized spin moments at the mn atoms ( see table [ tableorb ] for the values of the moments ) @xcite .
a simple band counting shows that the total spin moment is exactly 4 @xmath11 for nimnsb ; this is a secial case of the slater - pauling behaviour , leading to integer moments for half - metallic systems .
the total moment of the half - metallic @xmath2 heusler alloys follows a simple rule .
the total number of electrons , @xmath12 , is given by the sum of the number of spin - up and spin - down electrons , while the total moment @xmath13 per unit cell is given by the difference .
since 9 minority bands are fully occupied , we obtain the simple `` rule of 18 '' for half - metallicity in @xmath2 heusler alloys @xcite @xmath14 this behaviour of the total moment resembles the slater - pauling behaviour of the binary transition metal alloys with the difference that in the latter ones the total spin moment decreases with the number of valence electrons @xcite .
the half - heusler alloys with 18 valence electrons ( e.g. , cotisb ) are nonmagnetic semiconductors , where the gap is again created by the @xmath7-@xmath7 hybridization . in figure [ fig1 ] (
left panel ) we have gathered the calculated spin magnetic moments per formula unit for the half - heusler alloys which we have plotted as a function of the total number of valence electrons .
the dashed line represents the rule @xmath15 obeyed by these compounds .
the total moment @xmath13 is an integer ( in @xmath16 ) , assuming the values 0 , 1 , 2 , 3 , 4 and 5 if @xmath17 .
the value @xmath18 corresponds to the semiconducting phase and the value @xmath19 to the maximal moment when all 5 majority @xmath7-states are filled .
firstly we varied the valence of the lower - valent ( _ i.e. _ magnetic ) transition metal atom .
thus we substitute v , cr and fe for mn in the nimnsb and comnsb compounds using the experimental lattice constants .
for all these compounds we find that the total spin moment scales accurately with the total charge and that they all present half - metallicity . of valence electrons per unit cell for all the studied half ( left panel ) and full ( right panel ) heusler alloys .
the dashed line represents the slater - pauling behaviour.,title="fig : " ] of valence electrons per unit cell for all the studied half ( left panel ) and full ( right panel ) heusler alloys .
the dashed line represents the slater - pauling behaviour.,title="fig : " ] in the second part of this section we discuss the properties of the full - heusler alloys like co@xmath0mnsi and co@xmath0cral .
their electronic and magnetic properties are similar to the half - heusler compounds with the additional complication of the presence of 2 co atoms per unit cell .
in addition to the bands the half - heusler alloys , there are five states exclusively located at the co sites and near the fermi level @xcite .
now , the fermi level is located among these states so that three out of five are occupied and two of them unoccupied leading to smaller gaps .
now there are 12 minority - spin occupied states and the compounds with 24 valence electrons like fe@xmath0vsb are semiconductors .
if v is substituted by a higher - valent atom , spontaneous spin polarization occurs , and the exchange splitting shifts the majority states to lower energies .
thus the extra electrons fill in only majority states : co@xmath0cral ( 27 valence electrons ) has a total spin moment of 3 @xmath11 and co@xmath0mnsi ( 29 valence electrons ) a spin moment of 5 @xmath11 as can be seen in table [ tableorb ] .
the presence of the second co atom has also another important consequence .
now each mn atom has an octehedral environment of co atoms as first neighbours and not four as was the case for comnsb .
this leads to an increased hybridization between the @xmath7 orbitals of the mn and co atoms and the spin moment of co in co@xmath0mnsi is slightly lower than 1 @xmath11 , considerably larger than the co moment in comnsb . following the above discussion we investigate the slater - pauling behaviour with the `` rule of 24 '' now instead of 18 in the half - heuslers : @xmath20 in figure [ fig1 ] ( right panel ) we have plotted the total spin moments for all the studied compounds as a function of the total number of valence electrons .
the dashed line represents the slater - pauling rule of half - metallic full heusler alloys .
the `` magical number 24 '' arises from the fact that the minority band contains 12 electrons per unit cell : 4 are occupying the low lying @xmath9 and @xmath10 bands of the @xmath21 element and 8 the co - like minority @xmath7 bands ( the mn - co bonding hybrids : @xmath22 and @xmath23 , and the non - bonding co states localized only at the co sites : @xmath24 ) . since 7 minority bands ( 2@xmath25co @xmath26 , 5@xmath25mn @xmath7 )
are unoccupied , the largest possible moment is 7 @xmath16 and would occur if all majority @xmath7-states were occupied .
of course it would be impossible to get a compound with a total spin moment of 7 @xmath16 but even @xmath27 is difficult to obtain . as it was shown by wurmehl et al .
@xcite , the on - site correlations in co@xmath0fesi play a critical role for this compound and calculations within the lda+u scheme , rather than the lda , give a spin moment of 6 @xmath16 .
the exchange interactions , stabilizing ferromagnetism , have been studied in reference @xcite using the adiabatic ( frozen - magnon ) approximation to calculate the inter - atomic exchange parameters for the half - heusler alloys nimnsb and comnsb .
the dominant interaction is between the mn atoms .
the magnetic interactions are more complex in full - heusler alloys co@xmath0mnsi and co@xmath0cral @xcite . in both cases ,
ferromagnetism is stabilized by the inter - sublattice interactions between the mn(cr ) and co atoms and between co atoms belonging to different sublattices . before closing this section
we discuss also the role of the @xmath21-elements in half - metallic heusler alloys .
while the @xmath21-elements are not responsible for the appearance of the minority gap , they are nevertheless very important for the physical properties of the heusler alloys and the structural stability of the @xmath2 structure . there are three important features : \(i ) while an sb atom has 5 valence electrons ( 5@xmath28 , 5@xmath29 ) , in the nimnsb compound each sb atom introduces a deep lying @xmath9-band , at about @xmath8 ev , and three @xmath10-bands below the centre of the @xmath7-bands .
these bands accommodate a total of 8 electrons per unit cell , so that formally sb acts as a triply charged sb@xmath30 ion .
analogously , a te - atom behaves in these compounds as a te@xmath31 ion and a sn - atom as a sn@xmath32 ion .
this does not mean , that locally such a large charge transfer exists .
in fact , the @xmath9- and @xmath10-states strongly hybridize with the transition metal ( tm ) @xmath7-states and the charge in these bands is delocalized and locally sb even loses about one electron , if one counts the charge in the wigner - seitz cells .
what matters here is that the @xmath9- and @xmath10-bands accommodate 8 electrons per unit cell , thus reducing the electrons which have to be accommodated by the @xmath7-bands of the tm atoms .
\(ii ) the @xmath21-atom is very important for the structural stability of the heusler alloys .
for instance , it is difficult to imagine that the calculated half - metallic nimn and ptmn alloys with zinc - blende structure actually exist , since metallic alloys prefer highly coordinated structures like fcc , bcc , hcp etc
. therefore the @xmath21-elements are decisive for the stability of the @xmath2 compounds .
a careful discussion of the bonding in these compounds has been recently published by nanda and dasgupta @xcite using the crystal orbital hamiltonian population ( cohp ) method . for the semiconductor fevsb
they find that , while the largest contribution to the bonding arises from the v-@xmath7
fe-@xmath7 hybridization , contributions of similar size arise also from the fe-@xmath7 sb-@xmath10 and the v-@xmath7
sb-@xmath10 hybridization .
similar results are also valid for the semiconductors cotisb and nitisn and in particular for the half - metal nimnsb .
since the majority @xmath7-band is completely filled , the major part of the bonding arises from the minority band , so that similar arguments as for the semiconductors apply .
\(iii ) another property of the @xmath21-elements is worthwhile mentioning : substituting the sb atom in nimnsb by sn , in or te destroys the half - metallicity @xcite .
this is in contrast to the substitution of ni by co or fe , which is documented in table [ tableorb ] .
the total moment of 4 @xmath16 for nimnsb is reduced to 3 @xmath16 in comnsb and 2 @xmath16 in femnsb , thus preserving half - metallicity . in nimnsn the total moment is reduced only to 3.3 @xmath16 ( instead of 3 ) and in nimnte the total moment increases only to 4.7 @xmath16 ( instead of 5 ) .
thus by changing only the @xmath21-element it is rather difficult to preserve the half - metallicity , since the density of states changes more like in a rigid band model @xcite .
changes of the lattice parameter can influence the electronic and magnetic properties of the @xmath2 and @xmath1 heusler alloys . to the best of our knowledge
no relevant experimental study exists . in figure [ fig2 ]
we show the dos of nimnsb and comnsb for the experimental lattice parameter and the ones compressed and expanded by 2 % .
first one sees that , upon compression , the fermi level moves in the direction of the conduction band , upon expansion towards the valence band . in both cases , however , the half - metallic character is conserved . in order to explain this behaviour
, we first note that the position of fermi level is determined by the metallic dos in the majority band .
as we believe , the shift of @xmath33 is determined by the behaviour of the sb @xmath10-states , in particular by the large extension of these states as compared to the @xmath7 states . upon compression
the @xmath10-states are squeezed and hybridize stronger , thus pushing the @xmath7-states and the fermi level to higher energies , i.e. , towards the minority conduction band .
in addition the mn @xmath7 and ni or co @xmath7 states hybridize stronger , which tends to increase the size of the gap while the mn(ni ) @xmath7-bandwidth increases , which tends to shrink the gap .
as shown in figure [ fig2 ] the first of the two effects is stronger leading finally to an increase of the gap width . upon expansion
the opposite effects are observed . in the case of nimnsb and for the experimental lattice
constant the gap - width is @xmath34 0.4 ev . when the lattice is expanded by 2%the gap
shrinks to 0.25 ev and when compressed by 2% the gap - width is increased to 0.5 ev .
similarly in the case of comnsb , the gap is 0.8 ev for the experimental lattice constant , 0.65 ev for the 2% expansion and 0.9 ev for the case of the 2%compression . for the full - heusler alloys
the pressure dependence has been recently studied by picozzi et al .
@xcite for co@xmath0mnsi , co@xmath0mnge and co@xmath0mnsn , using both the lda and the somewhat more accurate gga .
the general trends are similar : the minority gap increases with compression , and the fermi level moves in the direction of the conduction band .
for example in the case of co@xmath0mnsi the gap - width is 0.81 ev for the theoretical equilibrium lattice constant of 10.565 . when the lattice constant is compressed to @xmath34 10.15 , the gap - width increases to about 1 ev .
the calculations show that for the considered changes of the lattice constant of @xmath35 2 % , half - metallicity is preserved
. there can be sizeable changes of the local moments , but the total moment remains constant , since @xmath33 stays within the gap .
the results presented in this section up to now concern the homogeneous change of the lattice constant under hydrostatic pressure .
block and collaborators have studied the effect of tetragonalization on the properties of both nimnsb and co@xmath0cral @xcite .
they found that although tetragonalization largely affected the shape and width of the gap , the systems remained half - metallic for moderate degrees of strain .
.[tableorb ] calculated spin ( @xmath36 ) and orbital ( @xmath37 ) magnetic moments in @xmath16 for the xmnsb half - heusler and x@xmath0yz full - heusler compounds .
the last three columns are the total spin and orbital magnetic moment and their sum , respectively .
the indexes x , y , z refer to the corresponding atoms in the compound .
[ cols=">,>,>,>,>,>,>,>,>,>",options="header " , ] [ tablehferri ] the ideal case for realistic applications would be a half - metallic antiferromagnet rather than the half - metallic ferromagnets .
in such a compound the majority- and minority - spin densities of states are not connected by a symmetry transformation , as for common antiferromagnets and the material is better described as a fully compensated ferrimagnet , having a net magnetic moment that is , due to the half - metallic character , precisely equal to zero .
such a material would have the advantage that due to the absence of a total magnetic moment it would not give rise to stray fields .
van leuken and de groot have suggested a possible route towards a half - metallic antiferromagnet starting from the semiconducting @xmath2-type compound fevsb @xcite . due to the difficulties to construct such a material
, it would be useful to study also the half - metallic ferrimagnetic materials , like femnsb @xcite or the mn@xmath0vz @xcite compounds where the mn and fe or v spin moments are anti - parallel . in reference
@xcite we studied the possibility of appearance of half - metallicity in the case of the full - heusler compounds mn@xmath0vz where z is an @xmath21 atom belonging to the iiib or ivb column of the periodic table .
total energy calculations have shown that , when z is al , ga , in or sn , the compounds are ferrimagnetic , while the compounds containing si and ge are non - magnetic at the equilibrium lattice constant . a small expansion of the lattice leads to the emergence of ferrimagnetism also in these compounds .
although all compounds are not half - metallic at their equilibrium lattice constant , small expansion of the lattice pushes the fermi level within the gap which is now situated in the majority - spin band contrary to all other full - heusler alloys . in table
[ tablehferri ] we have gathered the atom - resolved and total spin moments for some of the compounds under study , for the largest and smallest lattice parameters for which half - metallicity is present .
the total spin moment is @xmath38 @xmath16 for the mn@xmath0val(ga or in ) compounds which have 22 valence electrons per unit cell and @xmath39 @xmath16 for the mn@xmath0vsi(ge or sn ) compounds with 23 valence electrons ( the negative sign of @xmath13 is according to the convention set by equation [ eq : sp2 ] ) . thus these compounds follow the slater - pauling behaviour and the `` rule of 24 '' .
the lighter the element and the smaller the number of valence electrons , the wider is the gap and the more stable is the half - metallicity with respect to the variation of the lattice constant .
the discussion so - far has been focused on the magnetic ground state of half - metallic heusler alloys . at elevated temperatures @xmath40 the spin polarization @xmath41
must clearly be reduced , and at the curie temperature @xmath42 it must vanish together with the average magnetization .
an qualitative and quantitative understanding of this effect is interesting from the point of view of basic science , but is also required in order to predict the applicability of half - metals in realistic devices .
a theoretical study of the temperature dependence of @xmath43 requires knowledge of the response of the spectral function ( density of states ) to an increase of the temperature .
this is a two - step process .
firstly , one needs to find the response of the magnetic excitations to temperature .
secondly , the spectral function must be found in the presence of the excitations .
both steps are not trivial , thus suitable approximations are needed . to our knowledge , there have been three different approaches to the subject for the case of half - metallic ferromagnets : a tight - binding approach @xcite , an approach based on constrained density - functional theory combined with the concept of a disordered local moment @xcite , and an approach based on dynamical mean - field theory ( dmft ) @xcite .
skomski and dowben @xcite consider a tight - binding model of a half - metallic ferromagnet .
magnetic excitations , in the form of magnons , will cause a rotation of the spin quantization axis , in general different at each atom @xmath44 , given by a unitary transformation @xmath45 .
this causes a projection of spin - up states of one atom to spin - down states of another .
they arrive at a crude estimate of this spin - mixing contribution to the spin - down dos : @xmath46 , in terms of the spin - up dos @xmath47 , the magnetization @xmath48 at @xmath49 , and at temperature @xmath40 , @xmath50 .
this gives an estimate for the polarization @xmath51 .
this picture is appealing because of its simplicity , and should be correct when the magnetic excitations are predominantly of the heisenberg type and have a long wavelength , i.e. , at low temperatures .
then , the local spin quantization axis is unchanged over large distances ( of the order of a fraction of the magnon wavelength ) , and , at each instant , the material can be considered locally half - metallic ; the reduction of @xmath52 comes about from the angle between the local spin quantization axes and the global axis corresponding to the average magnetic moment .
however , as is argued by leai et al .
@xcite , at higher temperatures the wavelength of the magnetic excitations can become smaller and affect the local electronic structure .
the hybridization between atomic wavefunctions is crucial for the gap , and the hybridization strength can be altered if the local spin axes of neighbouring atoms are different .
therefore , the spectral function should be in principle re - calculated self - consistently in the presence of the excitations . since the characteristic times of magnetic excitations are much longer than the electron hopping times , the adiabatic approximation within density - functional theory can be employed : an external constraint is used to force a non - collinear state , representing the magnon ; under this constraint , the electronic structure is calculated self - consistently within density - functional theory .
such an approach was applied in reference @xcite , with the conclusion that a non - collinear configuration can affect the hybridization strongly enough that the polarization collapses . in particular , it was found for nimnsb that , under the constraint of a spin spiral , the gap does not close easily , but the minority - spin @xmath7 bands around the gap move in such a way that they cross the fermi level .
because the @xmath7 bands have a high spectral weight , @xmath53 increases abruptly when these reach @xmath33 , and the polarization collapses . the effect was quantified by using the disordered local moment approach . within this method , the magnetic state at @xmath54 is approximated by a state of the form ni(mn@xmath55mn@xmath56 , i.e.
, the spin is flipped a fraction @xmath57 of the mn atoms in a random way , representing a disordered moment at the mn site . for @xmath58 the ferromagnetic ground state
is recovered , while @xmath59 corresponds to @xmath42 .
the electronic structure of this disordered state was found within the cpa and the kkr green function method . in this way
, @xmath52 can be calculated as a function of @xmath60 , and , if @xmath50 is independentlu known , one can find @xmath61 .
the results show a collapse of the polarization at about 0.4@xmath42 , which corresponds to room temperature .
chioncel et al .
@xcite have considered the effect of the dynamic electron - magnon interaction , resulting from electron correlations . according to this picture
, there exist excitations within the ( spin - down ) half - metallic gap as superpositions of spin - up electron states and spin - down electron states with magnons @xcite .
these excitations , also called non - quasiparticle states , produce at @xmath49 a spectral weight at energies just above @xmath33 ( or just below @xmath33 if the gap is among the majority - spin states ) . for @xmath54 , however , spin - down spectral
weight appears also at @xmath33 , and the polarization is reduced .
this correlation - induced effect depends on the value of the on - site coulomb interaction @xmath62 , and can be captured within the dmft . for the heusler alloy femnsb ,
calculations within the lda+dmft reported in @xcite show that , at @xmath63 , @xmath64 if @xmath65 ev and @xmath66 if @xmath67 ev .
unfortunately , there are not many experimental results concerning the temperature dependence of the polarization in heusler alloys .
one indirect result is provided in reference @xcite , where tunneling magnetoresistance ( tmr ) data are compared from junctions containing a co@xmath0mnsi heusler alloy as a first electrode with data from junctions with co@xmath68fe@xmath69 or ni@xmath70fe@xmath71 ( the second electrode is in all cases co@xmath68fe@xmath69 , and the barrier is alo@xmath72 ) .
it is shown that , at low temperatures , the co@xmath0mnsi - based junction exhibits an 85% tmr ratio ( relative change of resistance due to a change of alignment of the magnetic moments of the leads ) , while the co@xmath68fe@xmath69- and ni@xmath70fe@xmath71-based junctions exhibit 65% and 70% tmr ratio respectively . at higher temperatures ( above 150 k )
the situation is reversed , with the co@xmath0mnsi - based junction showing a lower tmr ratio ( 30% at 300 k ) than the other two ( 45% at 300 k ) .
this evidence suggests that , at low @xmath40 , co@xmath0mnsi has a higher spin polarization than co@xmath68fe@xmath69 or ni@xmath70fe@xmath71 , but at high @xmath40 , the polarization of the heusler alloy drops more rapidly .
we note here that co@xmath0mnsi has a high curie temperature of 985 k. more recently , sakuraba and collaborators @xcite managed to grow a co@xmath0mnsi / al - o / co@xmath0mnsi tunnel junction . at low temperatures ( 2 k ) ,
an extremely high tmr ratio of 570% was reported , while with increasing temperature the tmr dropped rapidly , reaching about 67% at 300 k. on the other hand , a co@xmath0mnsi / al - o / co@xmath73fe@xmath74 junction showed a less pronounced decrease of tmr from 160% at 2 k to 70% at 300 k , while a junction co@xmath73fe@xmath74/al - o / co@xmath73fe@xmath74 showed a less drastic decrease of tmr , approximately from 60% to 30% .
the authors of reference @xcite propose an explanation in terms of spin - flip processes due to magnetic impurities at the interface co@xmath0mnsi / al - o .
alternatively , it is possible that there is a rapid loss of half - metallicity with increasing temperature , causing a decrease of @xmath52 much faster than @xmath60 .
in this review we have given an introduction to the electronic structure and the resulting magnetic properties of half - metallic heusler alloys , which represent interesting hybrids between metallic ferromagnets and semiconductors .
first , we reviewed a few recent results on the electronic properties of these alloys .
the origin of the gap in these half - metallic alloys and its connection to the magnetic properties are well understood .
the gap in half - heuslers arises between the bonding and antibonding @xmath7-hybrids created by the transition metal atoms ( e.g. ni and mn in nimnsb ) . in full - heusler compounds , like co@xmath0mnsi
, there is the extra complication of the appearance of co @xmath7-states which do not hybridize with the mn states .
there are exactly 9 occupied minority - spin states for the half - heuslers and 12 for the full - heuslers ; thus the total spin moment follows a slater - pauling behaviour for both families .
moderate changes of the lattice parameter shift the fermi level , but overall do not affect half - metallicity .
spin - orbit coupling also induces states within the gap but the alloys keep a very high degree of spin - polarization at the fermi level .
orbital moments are small as expected for typical intermetallic ferromagnets .
we discussed the effect of substitutional doping , disorder and defects on the properties of these alloys . in many cases moderate degrees of disorder and doping do not alter half - metallicity . also defects with low formation energies , which are the most likely to occur , keep the half - metallic character of the compounds .
quaternary heusler alloys are a special case of full - heusler alloys where one site is occupied randomly by two chemical elements .
these alloys , when the two extreme normal compounds are half - metallic , follow the slater - pauling behaviour keeping the half - metallicity of the normal heusler alloys .
finally we discuss the appearance of half - metallic ferrimagnetism in mn@xmath0vz ( z and @xmath21 element ) .
the results summarized above contribute in the idea that half - metallic ferromagnets are feasible experimentally since all studied effects keep an almost 100% degree of spin - polarization at the fermi .
however as discussed in the last section temperature effects quickly destroy the half - metallicity and the alloys transit to a normal metal .
moreover , surfaces / interfaces of heusler alloys are known from first principles calculations to show strong surface / interface states which can severely reduce the degree of spin - polarization @xcite .
a theoretical analysis @xcite shows that interface states at a half - metal / semiconductor interface can reduce the tunneling magnetoresistance ratio significantly .
99 webster p j and ziebeck k r a , in _ alloys and compounds of d - elements with main group elements .
_ , edited by h.r.j .
wijn , landolt - brnstein , new series , group iii , vol .
19,pt.c ( springer - verlag , berlin ) , 1988 , pp .
75 - 184 bergmann a , grabis j , toperverg b p , leiner v , wolff m , zabel h , and westerholt k 2005 b * 72 * 214403 ; grabis j , bergmann a , nefedov a , westerholt k , and zabel j 2005 b * 72 * 024437 ; _ idem _ 2005 b * 72 * 024438 sakuraba y , hattori m , oogane m , ando y , kato h , sakuma a , miyazaki t , and kubota h 2006 _ appl .
_ * 88 * 192508 ; sakuraba y , nakata j , oogane m , kubota h , ando y , sakuma a , and miyazaki t 2005 * 44 * l1100 galanakis i , leai m , bihlmayer g , and blgel s 2005 b * 71 * 71 214431 ; leai m , galanakis i , bihlmayer g , and blgel s 2005 * 17 * 3121 ; galanakis i 2005 * 288 * 411 ; galanakis i 2004 * 16 * 8007 ; galanakis i 2002 * 14 *
6329 ishida s , fujii s , kashiwagi s and asano s 1995 _ j. phys . soc .
jpn . _ * 64 * 2152 ; fujii s , ishida s and asano s 1995 _ j. phys .
jpn . _ * 64 * 185 ; zhang m , liu z h , hu h n , liu g d , cui y t , wu g h , brck r , de boer f r , and li y x 2004 _ j. appl . phys . _ * 95 * 7219 ; zhang m , dai x , hu h , liu g , cui y , liu z , chen j , wang j , and wu g 2003 * 15 * 7891 | half - metallic heusler alloys are amongst the most promising materials for future magnetoelectronic applications .
we review some recent results on the electronic properties of these compounds .
the origin of the gap in these half - metallic alloys and its connection to the magnetic properties are well understood . changing the lattice parameter shifts slightly the fermi level .
spin - orbit coupling induces states within the gap but the alloys keep a very high degree of spin - polarization at the fermi level .
small degrees of doping and disorder as well as defects with low formation energy have little effect on the properties of the gap , while temperature effects can lead to a quick loss of half - metallicity .
finally we discuss two special issues ; the case of quaternary heusler alloys and the half - metallic ferrimagnets . |
we do not know what mechanism has been chosen by nature to break the electroweak symmetry .
there is a general agreement in the sense that the higgs sector of the standard model is incomplete and new physics must exist at the tev scale in order to explain the origin of electroweak symmetry breaking .
an appealing possibility is that a new strong interaction breaks the electroweak symmetry through the formation of a condensate of new fermions which is not a singlet of the electroweak symmetry group .
this hypothetical new interaction is generally called technicolor@xcite . in general ,
technicolor models have a rich spectrum of composite pseudoscalar and vector states .
some non - minimal models predict the existence of colored technihadrons such as color octet technirhos and color triplet technipions .
they are the focus of this work .
because technicolor is a strongly coupled theory , it is necessary to use effective methods in order to study its phenomenology . in general , the methods that have been proved useful in the description of the properties and interactions of usual hadrons are also used in the study of technihadrons .
one of such methods is vector meson dominance ( vmd)@xcite . a related , but perhaps more elegant , method for describing such interactions is the so called hidden symmetry .
a realization of this technique applied to a strong electroweak sector is known as the bess model@xcite .
vmd was invented in order to describe the interaction of electrons with hadrons ( the rho meson and pions ) . in this scheme
the electron interacts directly only with the ( proto-)photon while the pion interacts directly only with the ( proto-)rho meson .
the interaction between electrons and pions is made possible by a mixing of the ( proto-)photon with the ( proto-)rho . in a similar way , it is assumed that the normal fermions interacts with technihadrons through a mixing of the standard model gauge bosons with the vector technimesons .
nevertheless , some deviations from complete vmd occur in its original context@xcite .
for example , it has been claimed that the decay of the usual ( qcd ) @xmath0 into a pair of leptons is better described if a direct coupling of the leptons to the @xmath1 is introduced@xcite . on the other hand , in @xcite
it is shown that the introduction of a direct coupling of the photon to the charged pions , allows the abandon of the universality condition in order to obtain a correct pion form factor .
moreover , we know that technicolor dynamics is not like qcd .
all these circumstances motivate us to ask about the meaning of the limits on technirho production at the tevatron when departure from traditional vmd are consired . in this work
we investigate the consequences of deviations form complete vmd ( which we will call partial vmd ) for the production of a color octet technirho and a pair of color triplet technipions .
the first studies of the color octet technirho phenomenology @xcite were done using a version of vmd ( generalized to color interaction ) based on non - diagonal propagators . in this work ,
we start with a version of vmd which is equivalent to the previous one but is formulated in terms of mass - mixing terms .
as it was previously shown @xcite , the mixing of the color octet technirho with the gluon can be described by the lagrangian : @xmath2 where @xmath3 this lagrangian is invariant under @xmath4 provided that the fields transform as : @xmath5 notice that the fields @xmath6 and @xmath7 are not physical fields because they are not mass eigenstates .
we call them the proto - gluon and the proto - technirho , respectively .
the physical fields can be written as : @xmath8 where @xmath9 is given by @xmath10 as usual , we estimate the value of @xmath9 scaling up the mixing between the photon and the usual rho meson obtained in normal vmd : @xmath11 where @xmath12 is the number of technicolors . as usual , we set @xmath13 . when the lagrangian ( [ eq : l1 ] ) is written in terms of the physical fields , we find that there is no coupling between a color octet technirho and two gluons @xcite .
nevertheless , it has been shown @xcite that an operator with dimension six exits that restores this coupling .
this operator may be written as : @xmath14 of course we do not know the value of the constant @xmath15 and hence an important uncertainty exists about the contribution of this operator .
for simplicity , we work in the pessimistic limit assuming that the constant @xmath15 is small enough to make this contribution negligible .
in fact , at the tevatron this term is not important due to the low gluon luminosity at large partonic center - of - mass energies @xcite .
however , this term is crucial for the study of the color octet technirho phenomenology at the lhc . in this case
, a reliable estimation of @xmath15 based on dynamical calculations would be very valuable , but this analysis is beyond the scope of this work .
now we want to couple these fields to quarks . in normal vmd , this is done assuming that the quarks only interacts with the proto - gluon . nevertheless , because both , the proto - gluon and the proto - technirho , transform like gauge fields , it is possible to write a covariant derivative with both fields : @xmath16\psi.\ ] ] in this way
, we introduce a direct coupling between the quarks and the proto - technirho which strength is measured by the parameter @xmath17 .
this kind of direct coupling can be produced by extended technicolor although in this case it must be proportional to the quark mass and , hence , small .
nevertheless because we do not know exactly all the properties of the underlying ( extended ) technicolor theory , we can not ignore the possibility that a more important coupling can be generated .
when lagrangian ( [ eq : lag_q_gauge ] ) is written in terms of physical fields we obtain : @xmath18 where @xmath19 is the usual qcd coupling constant . notice that the coupling of the gluon to quarks is the usual one and it is independent of @xmath17 . on the other hand , the coupling of the physical color octet technirho not only depends on @xmath17 , it vanishes for @xmath20 .
that means that a relative small deviation from normal vmd can produce the decoupling of the physical technirho from quarks .
of course this effect has important consequences for the present mass limits obtained at the tevatron .
in order to evaluate the effects , on the technirho phenomenology , of such a modification of the technirho - quark interaction , we compute ( with the help of lanhep @xcite and calchep @xcite ) the cross section for the resonant production of dijets at the tevatron and we compare it with the experimental limits obtained by the cdf collaboration for a luminosity of @xmath21 pb@xmath22 ( run i)@xcite .
we use the cteq6l @xcite parton distribution function , @xmath23 gev and we implement the following kinematical cuts : @xmath24 and @xmath25 where @xmath26 is the jet pseudo - rapidity .
as we have already said , we work in the limit where the color octet technirho does not couple to two gluons , hence it is produced only by quark fusion .
for this reason , we do not interpret the experimental results as limits on the color octet technirho mass , but as limits on the @xmath17 parameter .
the results are shown in figure [ fig : x ] .
the curves represent the @xmath27 c.l .
limits on @xmath28 and the region between them is allowed .
the region between the lines is allowed ] the color - octet technirho remains invisible in this channel if @xmath29 .
for the value of @xmath30 given by equation ( [ eq : sina ] ) , we obtain that the value of @xmath17 needed in order to explain the non - observabilty of the color - octet technirho at the tevatron is @xmath31 .
this value of @xmath17 seems too high because it implies that etc would produce unacceptable big masses for the @xmath32 and @xmath33 quarks .
nevertheless , this numerical result depends of the expression of @xmath30 shown in equation ( [ eq : sina ] ) that was scaled up from qcd .
this procedure is highly dangerous because we know that technicolor s dynamics is not qcd - like .
of coure , we can ask whether , in a realistic model , the mixing angle can be small enough in order to have an @xmath34 compatible with the @xmath32 and @xmath33 quarks . to our knowledge , this is an open question . on the other hand , in the context of the top quark production in the so called extended bess model @xcite
it has been considered a direct coupling of a color - octet vector resonances ( equivalent to our color - octet technirho ) to quarks of the same order of magnitude of the lower bound of our @xmath17 .
let us turn again our attention to the color octet technirho production by gluon fusion , governed by the dimension 6 operator showed in equation ( [ eq : op6 ] ) .
we have already said that this process is subdominant at the tevatron due to the low gluon luminosity at large partonic center - of - mass energies . nevertheless , it necessary to study its contribution
when the coupling of the technirho to quarks is suppressed . in this case
we compute the production cross using @xmath35 tev and we compare our results with the experimental limits on the production of a color octet technirho , in order to obtain limits on the parameter @xmath15 .
the results are shown in figure [ fig : c_2 ] .
the region below the curve is allowed parameter .
the region below the curve is allowed ] notice that even for large values of @xmath15 the color octet technirho remains invisible in the dijet spectrum .
we recall that , on the ground of dimensional analysis , the natural value of @xmath15 is of ordrer @xmath36
another interesting process is the production of colored technipions . in this section
we focus on the production of the color triplet ones . in the one family technicolor model , there are four color triplet technipions which we denote generically by @xmath37 .
we assume that all of them have the same mass . in an exact analogy with the previous case
, the lagrangian may be written in terms of our extended covariant derivative : @xmath38 where @xmath39 in this case , the parameter @xmath40 measures the strength of the direct coupling between the proto - gluon and the color triplet technipion .
the relevant feynman rules for the technipion pair production , written in terms of the physical fields , are shown in table [ tab : fr ] .
.feynman rules relevant for the technipion pair production .
[ cols="<,<",options="header " , ] [ tab : gr ] we estimate the pair production of color octet technirhos using only the diagrams a , b and d of fig . [ fig : lhc_diag ] because we expect that the dominant contribution to this process , at the lhc , comes from gluon fusion .
the result is shown in figure [ fig : lhc ] .
it is expected that lhc will have a luminosity of @xmath41 pb@xmath22/yr
so the color octet technirhos must be abundantly produced ( between @xmath42 and @xmath43 events per year in the whole possible mass range ) at the lhc .
of course , a more detailed analysis taking into account realistic detector effects and background must be done , but such a work is beyond the scope of this paper . ]
we have written down the more general lagrangian ( using operators with dimension up to four ) that describes the interaction of the the color octet technirho with quarks , gluons and colored technipions .
our lagrangian deviate from the usual implementation of vector meson dominance because it includes a direct coupling of the technirho to quarks .
we found that the coupling of the physical technirho to quarks is modified in such a way that it can be significatively suppressed with respect to the usual expectations of normal vmd .
in fact , we compared our predictions to experimental data from the cdf collaboration and we found that a relative small direct coupling can render the color octet technirho invisible .
this effect has consequences also for the pair production of colored technipions , which becomes non - resonant .
nevertheless , the data obtained at the tevatron exclude the existence of color triplet technipions for @xmath44 gev .
finally , we propose to pay attention to the pair production of color octet technirhos as a viable discovery channel at the lhc .
the author received support from universidad austral de chile ( did grant s-2006 - 28 ) . for a review see : d. dominici , riv .
nuovo cim . *
20 * , 1 ( 1997 ) [ arxiv : hep - ph/9711385 ] .
j. schechter , phys .
d * 34 * , 868 ( 1986 ) .
a. r. zerwekh , arxiv : hep - ph/0603096 k. d. lane and m. v. ramana , phys .
d * 44 * , 2678 ( 1991 ) .
e. eichten and k. d. lane , phys .
b * 327 * , 129 ( 1994 ) [ arxiv : hep - ph/9401236 ] .
a. r. zerwekh and r. rosenfeld , phys .
b * 503 * , 325 ( 2001 ) [ arxiv : hep - ph/0103159 ] .
r. sekhar chivukula , a. grant and e. h. simmons , phys .
b * 521 * , 239 ( 2001 ) [ arxiv : hep - ph/0109029 ] .
a. r. zerwekh , int .
j. mod .
a * 19 * ( 2004 ) 4387 .
a. v. semenov , arxiv : hep - ph/0208011 .
+ a. semenov .
lanhep - a package for automatic generation of feynman rules .
user s manual .
inp msu preprint 96 - 24/431 , moscow , 1996 ; hep - ph/9608488 + a. semenov . nucl.inst . & meth
. a393 ( 1997 ) p. 293 .
+ a. semenov .
lanhep - a package for automatic generation of feynman rules from the lagrangian .
updated version 1.3 .
inp msu preprint 98 - 2/503 . + home page : http://theory.sinp.msu.ru/ semenov / lanhep.html a. pukhov , arxiv : hep - ph/0412191 . f. abe _ et al . _
[ cdf collaboration ] , phys .
d * 55 * , 5263 ( 1997 ) [ arxiv : hep - ex/9702004 ] .
j. pumplin , d. r. stump , j. huston , h. l. lai , p. nadolsky and w. k. tung , jhep * 0207 * , 012 ( 2002 ) [ arxiv : hep - ph/0201195 ] .
r. casalbuoni , p. chiappetta , d. dominici , a. fiandrino and r. gatto , z. phys .
c * 69 * , 519 ( 1996 ) [ arxiv : hep - ph/9505212 ] .
t. affolder _ et al . _
[ cdf collaboration ] , 1.8-tev , phys .
* 85 * , 2056 ( 2000 ) [ arxiv : hep - ex/0004003 ] . | in this paper we address the question : `` do the limits on technirho production at the tevatron mean what we think they do ? ''
these limits are based on calculations that rely on vector meson dominance ( vmd ) .
vmd was invented in order to describe the interaction of electrons with hadrons ( the rho meson and pions ) .
the method has been used also as a tool in the study of technicolor phenomenology .
nevertheless there are evidences in the sense that , even in its original , context vmd is not completely realized . in this work
we investigate the consequences of a deviation from complete vmd for the phenomenology of colored technihadrons .
we focus specially on the production of the color octet technirho and color triplet technipions .
we found that a relative small direct coupling of the proto - technirho to quarks is enough to suppress or even eliminate the interaction among quarks and the physical technirho . on the other hand , it is possible to suppress the coupling of the physical technirho to technipions but in this case a large interaction among the technipions and the proto - gluon must be introduced .
the consequences for the limits on the mass of the color octet technirho and colored triplets technipions are also investigated . |
in classical mechanics , lagrangian and hamiltonian formulations are completely the same description of a dynamical system . usually more attention to the hamiltonian formulation is paid because it has properties of a canonical system . in post - newtonian
( pn ) mechanics of general relativity , the two formulations are still adopted . are they completely equivalent ?
ten years ago two independent groups [ 1,2 ] answered this question .
they proved the complete physical equivalence of the third - order post - newtonian ( 3pn ) arnowitt - deser - misner ( adm ) coordinate hamiltonian approach to and the 3pn harmonic coordinate lagrangian approach to the dynamics of spinless compact binaries .
this result was recently extended to the inclusion of the next - to - next - to - leading order ( 4pn ) spin - spin coupling [ 3 ] . however , there are two different claims on the chaotic behavior of compact binaries with one body spinning and spin effects restricted to spin - orbit ( 1.5pn ) coupling .
that is , the 2pn harmonic coordinate lagrangian dynamics allow the onset of chaos [ 4 ] , but the 2pn adm hamiltonian dynamics are integrable , regular and non - chaotic [ 5,6 ] .
an explanation to the opposite results was given in [ 7 ] .
in fact , the 2pn hamiltonian and lagrangian formulations are not exactly equal but are only approximately related . as its detailed account , the equations of motion for the lagrangian formulation use lower - order terms as approximations to higher - order acceleration terms in the euler - lagrange equations , while these approximations do not occur in the equations of motion for the hamiltonian formulation . it is natural that the lagrangian has approximate constants of motion but the hamiltonian contains exact ones .
these facts were regarded as the essential point for the two formulations having different dynamics . in this sense , the two claims that seem to be explicitly conflicting were thought to be correct .
recently , the authors of [ 8 ] revisited the equivalence between the hamiltonian and lagrangian formulations at pn approximations .
they found that the two formulations at the same pn order are nonequivalent in general and have differences .
three simple examples of pn lagrangian formulations , including a relativistic restricted three - body problem with the 1pn contribution from the circular motion of two primary objects , a spinning compact binary system with the newtonian term and the leading - order spin - orbit coupling [ 8 ] and a binary system of the newtonian term and the leading - order spin - orbit and spin - spin couplings [ 9 ] , were used to show that the differences are not mainly due to the lagrangian having the approximate euler - lagrange equations and the approximate constants of motion but come from truncation of higher - order pn terms between the two formulations transformed . an important result from the logic
is that an equivalent hamiltonian of a lower - order lagrangian is usually at an infinite order from a theoretical point of view or at a higher enough order from numerical computations .
based on this , the integrability or non - integrability of the lagrangian can be known by that of the hamiltonian .
more recently , chaos in comparable mass compact binary systems with one body spinning was completely ruled out [ 10 ] .
the reason is that a completely canonical higher - order hamiltonian , which is equivalent to a lower - order conservative lagrangian and holds four integrals of the total energy and the total angular momentum in an eight - dimensional phase space , is typically integrable [ 11 ] .
this result is useful to clarify the doubt on the absence of chaos in the 2pn adm hamiltonian approach [ 5,6 ] and the presence of chaos in the 2pn harmonic coordinate lagrangian formulation [ 4 ] . as a point to illustrate ,
two other doubts about different chaotic indicators resulting in different dynamical behaviours of spinning compact binaries among references [ 12 - 15 ] and different descriptions of chaotic parameter spaces and chaotic regions between two articles [ 4,16 ] have been clarified in [ 17 - 19 ] .
it is worth noting that the logic result on the equivalence of the pn hamiltonian and lagrangian approaches at different orders is not easy to check because the exactly equivalent hamiltonian of the lagrangian is generally expressed as an infinite series whose convergence is unknown clearly in most cases . to provide enough evidence for supporting this result , we select a part of the 1pn lagrangian formulation of relativistic circular restricted three - body problem [ 20 ] , where the euler - lagrange equations can be described by a converged taylor series and the equivalent hamiltonian
can also be written as another converged taylor series . for our purpose ,
the hamiltonian is derived from the lagrangian in sect .
2 . then in sect .
3 numerical methods are used to evaluate whether various pn order hamiltonians and the 1pn lagrangian with various pn order euler - lagrange equations are equivalent . finally , the main results are concluded in sect .
as in classical mechanics , a lagrangian formulation @xmath0 and its hamiltonian formulation @xmath1 satisfy the legendre transformation in pn mechanics .
this transformation is written as @xmath2 here @xmath3 and @xmath4 are coordinate and velocity , respectively .
canonical momentum is @xmath5 taking a special pn circular restricted three - body problem as an example , now we derive the hamiltonian from the lagrangian in detail .
the circular restricted three - body problem means the motion of a third body ( i.e. a small particle of negligible mass ) moving around two masses @xmath6 and @xmath7 ( @xmath8 ) .
the two masses move in circular , coplanar orbits about their common center of mass , and have a constant separation @xmath9 and the same angular velocity .
they exert a gravitational force on the particle but the third body does not affect the motion of the two massive bodies . taking the unit of mass @xmath10 , we have the two masses @xmath11 and @xmath12 . the unit of length requires that the constant separation of the two bodies should be unity . the common mean motion , the newtonian angular velocity @xmath13 , of the two primaries is also unity . in these unit systems ,
the two bodies are stationary at points @xmath14 and @xmath15 with @xmath16 and @xmath17 in the rotating reference frame .
state variables @xmath18 of the third body satisfy the following lagrangian formulation @xmath19 @xmath20.\end{aligned}\ ] ] in the above equations , the related notations are specified as follows .
@xmath21 is of the form @xmath22 where the distances from body 3 to bodies 1 and 2 are @xmath23 @xmath24 stands for the newtonian circular restricted three - body problem .
@xmath25 is a 1pn contribution due to the relativistic effect to the circular motions of the two primaries .
@xmath26 is also a 1pn contribution from the relativistic effect to the third body , and is only a part of that in [ 20 ] for our purpose .
@xmath27 is the 1pn effect with respect to the angular velocity @xmath13 of the primaries and is given by @xmath28 in fact , the separation @xmath9 is a mark of @xmath25 and @xmath26 as the 1pn effects when the velocity of light , @xmath29 , is taken as one geometric unit in later numerical computations . the lagrangian ( 3 ) is a function of velocities and coordinates , therefore , its equations of motion are the ordinary euler - lagrange equations : @xmath30 since the momenta @xmath31 and @xmath32 of the forms @xmath33 are linear functions of velocities @xmath34 and @xmath35 , accelerations can be solved exactly from eq .
they have detailed expressions : @xmath36 the newtonian terms @xmath37 and @xmath38 and the 1pn terms @xmath39 and @xmath40 are @xmath41 @xmath42 , \\
y_1 & = & 2\omega_1(y-\dot{x})+\frac{u_y}{u}l_2+\frac{3}{a}[u(y-2\dot{x } ) \nonumber \\ & & -(\dot{x}u_x+\dot{y}u_y)(x+\dot{y})],\end{aligned}\ ] ] where @xmath43 and @xmath44 .
considering that @xmath45 is at the 1pn level , eqs .
( 12 ) and ( 13 ) have the taylor expansions @xmath46 + \frac{x_1}{c^2}[\sum\limits_{j=0}^{k-1}(-1)^{j}(\frac{\delta}{c^2})^{j}],\\ \ddot{y } & \approx & y_0[\sum\limits_{i=0}^{k}(-1)^{i}(\frac{\delta}{c^2})^{i } ] + \frac{y_1}{c^2}[\sum\limits_{j=0}^{k-1}(-1)^{j}(\frac{\delta}{c^2})^{j}].\end{aligned}\ ] ] they are the euler - lagrange equations with pn approximations to an order @xmath47 , labeled as @xmath48 . as a point to illustrate ,
the case of @xmath49 with @xmath50 corresponds to the newtonian euler - lagrange equations , marked as @xmath51 . from a theoretical viewpoint , as @xmath52
, @xmath48 is strictly equivalent to @xmath53 given by eqs .
( 12 ) and ( 13 ) , namely , @xmath54 .
note that for the generic case in [ 8 ] , the momenta are highly nonlinear functions of velocities , so no exact equations of motion similar to eqs .
( 12 ) and ( 13 ) but approximate equations of motion can be obtained from the euler - lagrange equations ( 9 ) .
this means that we do not know what the pn approximations like eqs .
( 18 ) and ( 19 ) are converged as @xmath52 .
the velocities @xmath34 and @xmath35 obtained from eqs .
( 10 ) and ( 11 ) are expressed as @xmath55 of course , they can be expanded to the @xmath56th order @xmath57+(1+\frac{\omega_1}{c^2}){y},\\ \dot{y } & \approx & p_{y}[\sum\limits_{i=0}^{k}(-1)^{i}(\frac{\delta}{c^2})^{i}]-(1+\frac{\omega_1}{c^2}){x}.\end{aligned}\ ] ] as mentioned above , eqs .
( 22 ) and ( 23 ) are exactly identical to eqs .
( 20 ) and ( 21 ) when @xmath52 . in light of eqs .
( 1 ) , ( 20 ) and ( 21 ) , we have the following hamiltonian @xmath58 its taylor series at the @xmath56th order is of the form @xmath59 it is clear that @xmath60 with @xmath61 is the newtonian hamiltonian formulation , and can be expressed in terms of the jacobian constant @xmath62 as @xmath63 .
additionally , @xmath64 is closer and closer to @xmath65 as @xmath56 gets larger . without doubt
, the exact equivalence between @xmath65 and @xmath64 should be @xmath66 .
of course , what @xmath64 is converged as @xmath52 is still unknown for the general case in [ 8 ] .
it should be emphasized that @xmath67 is the @xmath56th order pn approximation to the euler - lagrange equations @xmath53 that is exactly derived from the 1pn lagrangian @xmath68 , and @xmath69 is the @xmath56th order pn approximation to the hamiltonian @xmath65 .
because of the exact equivalence between @xmath53 and @xmath65 , @xmath67 is the @xmath56th order pn approximation to the hamiltonian @xmath65 , and @xmath69 is the @xmath56th order pn approximation to the euler - lagrange equations @xmath53 . additionally , @xmath70 and @xmath71 are exactly equivalent , i.e. , @xmath72 .
however , it would be up to a certain higher enough finite order @xmath56 rather than up to the infinite order @xmath56 that the equivalence @xmath73 can be checked by numerical methods . see the following numerical investigations for more details
besides the above analytical method , a numerical method is used to estimate whether these pn approaches have constants of motion and what the accuracy of the constants is . above all
, we are interested in knowing whether these pn approaches are equivalent . an eighth- and ninth - order runge kutta fehlberg algorithm of variable time - steps is used to solve each of the above euler - lagrange equations @xmath48 and hamiltonians @xmath64 .
parameters and initial conditions are @xmath74 , @xmath75 , @xmath76 and @xmath77 . note that the initial positive value of @xmath35 is given by the jacobian constant .
this orbit in the newtonian problem @xmath24 is a kolmogorov - arnold - moser ( kam ) torus on the poincar section @xmath78 with @xmath79 in fig .
1(a ) , therefore , it is regular and non - chaotic
. the integrator can give errors of the energy @xmath60 for the lagrangian @xmath24 in the magnitude of order @xmath80 or so .
the long - term accumulation of energy errors is explicitly present in fig .
1(b ) because the integration scheme itself yields an artificial excitation or damping .
if this accumulation is neglected , the energy should be constant .
this shows that the energy @xmath60 is actually an integral of the lagrangian @xmath24 .
however , the existence of this excitation or damping does not make the numerical results unreliable during the integration time of @xmath81 due to such a high numerical accuracy . in this sense , not only the integrator does not necessarily use manifold correction methods [ 21 - 23 ] , but also it gives true qualitative results as a symplectic integration algorithm [ 24 - 27 ] does . when the pn terms @xmath25 and @xmath26 are included , what about the accuracy of energy integrals given by the related pn approximations ?
let us answer this question .
taking the separation between the primaries , @xmath82 , we plot fig .
2(a ) in which the errors of energies of the 1pn euler - lagrange equations @xmath83 and hamiltonian @xmath84 are shown .
it is worth noting that the error of energy is estimated by means of @xmath85 , where @xmath84 is regarded as the energy of @xmath83 at time @xmath86 and @xmath87 is the initial energy .
obviously , the error for @xmath83 is larger in about 10 orders of magnitude than that for @xmath84 .
this result should be very reasonable because differences between @xmath83 and @xmath84 exist explicitly but the canonical equations are exactly given by the 1pn hamiltonian @xmath84 , as shown in the above analytical discussions .
in other words , the difference between @xmath83 and @xmath84 is at 1pn level .
of course , the higher the order @xmath56 gets , the smaller the difference between @xmath48 and @xmath64 becomes .
this is why we can see from figs .
2(a ) and 2(b ) that the error of the 8pn euler - lagrange equations @xmath88 and hamiltonian @xmath89 is typically smaller than that of the 1pn euler - lagrange equations @xmath83 and hamiltonian @xmath84 . without doubt , @xmath53 and @xmath65 should be the same in the energy accuracy if no roundoff errors exist in fig .
in addition to evaluating the accuracy of energy integrals of these pn approaches , evaluating the quality of these pn approaches to the euler - lagrange equations @xmath53 or the hamiltonian @xmath65 is also necessary from qualitative and quantitative numerical comparisons . see the following demonstrations for more information .
besides the method of poincar sections , the method of lyapunov exponents is often used to detect chaos from order .
it relates to the description of average exponential deviation of two nearby orbits .
based on the two - particle method [ 28 ] , the largest lyapunov exponent is calculated by @xmath90 where @xmath91 and @xmath92 are distances between the two nearby trajectories at times 0 and @xmath86 , respectively . a globally stable orbit is said to be regular if @xmath93 but chaotic if @xmath94 . generally speaking
, it costs a long enough time to obtain a stabilizing value of @xmath95 from the limit . instead , a quicker method to find chaos is a fast lyapunov indicator [ 29,30 ] , defined as @xmath96 the globally stable orbit is chaotic if this indicator increases exponentially with time @xmath97 but ordered if this indicator grows polynomially .
it can be seen clearly from the poincar section of fig .
3(a ) that the dynamics of @xmath53 or @xmath65 in fig .
2(c ) is chaotic .
this result is supported by the lyapunov exponents in figs .
3(b ) and 3(c ) and the flis in fig .
3(d ) and 3(e ) . what about the dynamics of these various pn approximations ?
the key to this question can be found in figs .
3(b)-3(e ) . here
are the related details . as shown in fig .
3(b ) , lower order pn approximations to the euler - lagrange equations @xmath53 , such as the 1pn euler - lagrange equations @xmath83 and the 4pn euler - lagrange equations @xmath98 , are so poorer that their dynamics are regular , and are completely unlike the chaotic dynamics of @xmath53 . with increase of the pn order @xmath56 , higher order pn approximations to the euler - lagrange equations
@xmath53 become better and better .
for example , the 8pn euler - lagrange equations @xmath88 allows the onset of chaos , as @xmath53 does .
seen particularly from the evolution curve on the lyapunov exponent and time , the 12pn euler - lagrange equations @xmath99 seems to be very closer to @xmath53 .
these results are also suitable for the pn hamiltonian approximations to the hamiltonian @xmath65 in fig .
when the lyapunov exponents in figs .
3(b ) and 3(c ) are replaced with the flis in figs .
3(d ) and 3(e ) , similar results can be given .
when the separation @xmath100 is instead of @xmath82 in fig .
3(a ) , an ordered kam torus occurs .
that means that the @xmath53 dynamics is regular and non - chaotic . in figs 3(f)-3(i ) , lower order pn approximations such as @xmath88 ( or @xmath89 ) have chaotic behaviors , but higher order pn approximations such as @xmath99 ( or @xmath101 ) have regular behaviors . in short , the above numerical simulations seem to tell us that the euler - lagrange equations ( or the hamiltonian approaches ) at higher enough pn orders have the same dynamics as the euler - lagrange equations @xmath53 ( or the hamiltonian @xmath65 ) .
there is a question of whether these results depend on the separation @xmath9 . to answer it
, we fix the above - mentioned orbit but let @xmath9 begin at 10 and end at 250 in increments of 1 . for each given value of @xmath9 ,
the fli is obtained after integration time @xmath102 . in this way
, we have dependence of flis on the separations @xmath9 in several pn lagrangian and hamiltonian approaches , plotted in fig .
5.5 is referred as a threshold value of fli for distinguishing between the regular and chaotic cases at this time .
that is to say , an orbit is chaotic when its fli is larger than threshold but ordered when its fli is smaller than threshold . in light of this
, we do not find that there are dramatic dynamical differences between the euler - lagrange equations @xmath53 ( or the hamiltonian @xmath65 ) and the various pn approximations such as the 1pn hamiltonian @xmath84 and the 1pn euler - lagrange equations @xmath83 .
however , it is clearly shown in table 1 that regular and chaotic domains of smaller separations @xmath9 in the lowest pn approaches @xmath83 and @xmath84 are explicitly different from those in @xmath53 or @xmath65 . as claimed above
, this result is of course expected .
when the order @xmath56 gets higher and higher , @xmath48 and @xmath64 have smaller and smaller dynamical differences compared with @xmath53 or @xmath65 .
two points are worth noting .
first , the same order pn approaches like @xmath99 and @xmath101 ( but unlike @xmath53 and @xmath65 ) are incompletely equivalent in the dynamical behaviors for smaller values of @xmath9 .
second , all the pn approaches @xmath83 , @xmath84 , @xmath99 , @xmath101 , @xmath103 , @xmath53 and @xmath65 can still have the same dynamics when @xmath9 is larger enough .
the two points are due to the differences among these approaches from the relativistic effects depending on @xmath9 ; smaller values of @xmath9 result in larger relativistic effects but larger values of @xmath9 lead to smaller relativistic effects .
now we are interested in quantitative studies on the various pn approximations @xmath67 to the hamiltonian @xmath65 and the various pn approximations @xmath69 to the euler - lagrange equations @xmath53 . in other words , we want to know how the deviation @xmath104 between the position coordinate @xmath105 for @xmath67 ( or @xmath69 ) and the position coordinate @xmath106 for @xmath65 ( or @xmath53 ) varies with time . to provide some insight into the rule on the deviation with time
, we should consider the regular dynamics in various pn approximations because the chaotic case gives rise to exponentially sensitive dependence on initial conditions .
for the sake of this purpose , the parameters and initial conditions unlike the aforementioned ones are @xmath107 , @xmath108 and @xmath78 .
when @xmath109 is given in fig .
5(a ) , the curve @xmath53 is used to estimate the accuracy of numerical solutions between @xmath65 and @xmath53 , which begins in about the magnitude of @xmath110 and is in about the magnitude of @xmath111 at time @xmath112 .
the difference numerical solutions between @xmath65 and @xmath83 is rather large . with increase of @xmath56
, @xmath67 is soon closer to @xmath65 .
for instance , @xmath113 is basically consistent with @xmath65 after time @xmath114 , and @xmath99 is almost the same as @xmath65 .
similarly , this rule is suitable for the approximations @xmath69 to the euler - lagrange equations @xmath53 in fig .
after the integration time reaches 10000 for each @xmath115 $ ] in figs .
5(c ) and 5(d ) , appropriately larger separation @xmath9 and higher enough order @xmath56 are present such that @xmath67 and @xmath69 are identical to @xmath65 or @xmath53 . in a word ,
it can be seen clearly from fig .
5 that @xmath67 and @xmath69 are equivalent as @xmath56 is sufficiently large .
in general , pn lagrangian and hamiltonian formulations at the same order are nonequivalent due to higher order terms truncated .
a lower order lagrangian is possibly identical to a higher enough order hamiltonian .
it is difficult to check this equivalence because the euler - lagrange equations are not exactly but approximately derived from the lagrangian . to cope with this difficulty , we take a simple relativistic circular restricted three - body problem as an example and investigate the equivalence of pn lagrangian and hamiltonian formulations . this dynamical problem is described by a 1pn lagrangian formulation , in which the euler - lagrange equations not only are exactly given but also can be expressed as a converged infinite pn order taylor series .
the lagrangian has an exactly equivalent hamiltonian , expanded to another converged infinite pn order taylor series .
numerical results support the equivalence of the 1pn lagrangian with the euler - lagrange equations at a certain specific higher order and the pn hamiltonian approach to a higher enough order . in this way
, we support indirectly the general result of [ 8,10 ] that a lower order lagrangian approach with the euler - lagrange equations at some sufficiently higher order can be equivalent to a higher enough order hamiltonian approach . | it was claimed recently that a low order post - newtonian ( pn ) lagrangian formulation , which corresponds to the euler - lagrange equations up to an infinite pn order , can be identical to a pn hamiltonian formulation at the infinite order from a theoretical point of view .
this result is difficult to check because in most cases one does not know what both the euler - lagrange equations and the equivalent hamiltonian are at the infinite order .
however , no difficulty exists for a special 1pn lagrangian formulation of relativistic circular restricted three - body problem , where both the euler - lagrange equations and the equivalent hamiltonian not only are expanded to all pn orders but also have converged functions .
consequently , the analytical evidence supports this claim .
as far as numerical evidences are concerned , the hamiltonian equivalent to the euler - lagrange equations for the lower order lagrangian requires that they both be only at higher enough finite orders . |
the metal - insulator transition has been at the forefront of physics research since anderson s seminal paper @xcite . as a result of quantum interference
, the wavefunctions of a disordered noninteracting system can become exponentially localized in configuration space .
the phenomenon has been experimentally observed in different setups , more recently with bose - einstein condensates @xcite .
a proposal for an experiment with ultracold atoms in a two - dimensional geometry also exists @xcite .
lattice models , such as the anderson tight - binding and the powerlaw random banded matrix ( prbm ) models @xcite , have been extensively employed in the analysis of the anderson metal - insulator transition . at criticality , it was found that the eigenstates exhibit multifractal features @xcite .
the inverse participation ratios , @xmath0 , contain information about the structure of the eigenstates @xmath1 written in the basis vectors @xmath2 of the configuration space .
in particular , @xmath3 measures the level of delocalization of the eigenstates in the chosen basis @xcite . at the anderson transition , the probability amplitudes @xmath4
display large fluctuations and @xmath5 shows anomalous multifractal scaling with respect to the system size @xcite , @xmath6 where @xmath7 denotes the average over an ensemble of realizations and eigenstates , @xmath8 is the dimension of the hamiltonian matrix @xcite , and @xmath9 represents the generalized dimension .
multifractality is reflected by the nonlinear dependence of the generalized dimension on @xmath10 .
in contrast , @xmath11 in the metallic phase , where @xmath12 is the system dimension , and @xmath13 in the insulating phase . experimentally , multifractality has been observed in disordered conductors @xcite and in systems with cold atoms @xcite .
recently , new studies have led to the conclusion that multifractal correlations are not exclusive to the critical point of the anderson - transition . in disordered systems , they are present away from criticality @xcite and even in extended states @xcite .
they are also found in the ground states of clean systems @xcite .
studies of the dynamics of noninteracting systems at the metal - insulator transition have shown that the loschmidt echo @xcite , the survival probability @xcite , and the spreading of wavepackets @xcite at the mobility edge exhibit a powerlaw behavior , where the exponent coincides with the generalized dimension for @xmath14 .
the generalized dimension of the eigenstates , @xmath15 , is extracted from eq .
( [ eq : generalized ] ) by performing a scaling analysis of @xmath16 . however , in studies of dynamics the main interest is on the generalized dimension associated with the initial state @xmath17 and denoted by @xmath18 .
the latter is obtained from a scaling analysis of the level of delocalization of the initial state with respect to the energy eigenbasis , that is the analysis of @xmath19 where @xmath20 and @xmath21 .
when investigating localization in real space , the initial state usually corresponds to a basis vector of the configuration space .
we also note that the two generalized dimensions above have been shown @xcite to be related through the expression @xmath22 .
a natural question following this brief summary of the anderson localization is what happens to the above findings when interaction is included .
it had been conjectured already in @xcite and then confirmed with perturbative arguments @xcite and rigorously @xcite that localization may persist .
studies about many - body localization ( mbl ) have recently boomed @xcite .
the interest in the subject is in part motivated by the access to new experimental tools , such as cold atoms in optical lattices @xcite , that can be used to corroborate theoretical predictions . among the latter ,
we find works about the location of the critical point in disordered spin-1/2 chains @xcite , analysis of the relation between the distribution of the wavefunction coefficients and the onset of localization @xcite , various efforts to identify the quasi - local integrals of motion in the mbl phase @xcite , and descriptions of the evolution of the entanglement entropy @xcite , few - body observables @xcite , and the loschmidt echo @xcite .
our goal in this work is to characterize the evolution of isolated disordered systems with interaction from very short to very long times .
since mbl is a dynamical transition , identifying general features of the dynamics of interacting systems is essential for the further developments of the field .
motivated by the results for noninteracting systems , our focus is on the decay of the survival probability and its relationship with the onset of multifractal states .
we consider a one - dimensional ( 1d ) disordered spin-1/2 system and analyze the evolution at different time scales of both the survival probability and the time - averaged survival probability . at short times
the decay is very fast and similar to that of clean systems .
afterwards , the decay slows down and shows an anomalous powerlaw behavior .
the exponent of this algebraic decay coincides with @xmath18 . at very long - times , the decay eventually saturates to @xmath23 .
we investigate the 1d isotropic heisenberg spin-1/2 system with two - body nearest - neighbor interaction , @xmath24 sites , and periodic boundary conditions . the hamiltonian is @xmath25 .
\label{ham}\ ] ] above @xmath26 , @xmath27 are spin operators , and @xmath28 sets the energy scale .
random static magnetic fields act on each site @xmath29 , the amplitudes @xmath30 being random numbers from a uniform distribution @xmath31 $ ] .
the total spin in the @xmath32-direction , @xmath33 , is conserved .
we work with the largest subspace , @xmath34 , of dimension @xmath35 .
localization in this symmetry sector guarantees localization in smaller sectors .
the dependence on @xmath36 of the level statistics and of the level of delocalization of the eigenstates of @xmath37 ( [ ham ] ) has been studied for at least a decade @xcite .
when @xmath38 , the system is analytically solvable with the bethe ansatz . if all the trivial symmetries of the hamiltonian are taken into account , one verifies that the level spacing distribution of neighboring levels is poisson .
in addition to the total spin in the @xmath32-direction , the other symmetries of the isotropic model at @xmath34 are : translational invariance , parity , spin reversal , and conservation of total spin . as @xmath36 increases from zero , the level spacing distribution eventually becomes wigner - dyson , indicating a transition to the chaotic regime .
the value of @xmath36 at which level repulsion becomes evident decreases as the system size increases . in parallel , the level of delocalization of the eigenstates in real space increases substantially .
the presence of disorder breaks the additional symmetries mentioned above and if the disorder is weak the states can spread out significantly . as @xmath36 further increases and becomes larger than the coupling strength , @xmath39 , the spreading of the eigenstates recedes and they become more localized in real space .
the critical point for the transition to the mbl phase has been identified as @xmath40 in @xcite and @xmath41 in @xcite . in any study of the structure of the eigenstates ,
the choice of basis is essential . since here
the goal is to investigate localization in real space , that is the level of confinement of the spin excitations in the lattice , the natural basis is that of the configuration space , which we refer to as the site - basis and is also known as the computational basis .
the site - basis vectors @xmath42 correspond to states where the spin on each site either points up or down along the @xmath32-axis , such as @xmath43 .
to study the dynamics of the disordered chain ( [ ham ] ) , we take as initial state a single site - basis vector , @xmath44 .
this is equivalent to a quench , where the initial hamiltonian is the ising part of the hamiltonian @xmath45 and the final hamiltonian is @xmath37 ( [ ham ] ) . to quantify how fast the initial state changes in time , we concentrate on the behavior of the survival probability , @xmath46 where @xmath47 are the eigenvalues of @xmath37 and @xmath48 is the overlap of the initial state with the eigenstates @xmath49 of @xmath37 .
@xmath50 measures the probability for finding the system still in @xmath51 at time @xmath52 .
the distribution in energy @xmath53 of the components @xmath54 of the initial state is often referred to as local density of states ( ldos ) .
if the envelope of this distribution is known , an analytical expression for @xmath50 can be obtained from the fourier transform , @xmath55 . for strong quenches ,
that is when the initial and final hamiltonians are very different , the envelope of @xmath56 is a gaussian with mean corresponding to the energy of the initial state , @xmath57 and width @xmath58 notice that the width depends only on the off - diagonal elements of the hamiltonian matrix written in the site - basis and is therefore independent of the diagonal disorder .
the gaussian shape of the ldos reflects the density of states , which for systems with two - body interaction is also gaussian @xcite . in the absence of disorder ,
the envelope @xmath56 is particularly well filled for initial states with energy @xmath59 near the center of the spectrum of @xmath37 @xcite .
its gaussian shape leads to the gaussian decay @xmath60 .
this behavior may persist until saturation or be followed by an exponential ( see refs .
@xcite and references therein ) . in fig .
[ fig : ldos ] we analyze the survival probability and the ldos in the presence of disorder .
the average of @xmath50 over different disorder realizations and different initial states is denoted by @xmath61 . for each system size and each realization , we select as initial states , only 10% of all the @xmath8 site - basis vectors .
they are the ones with energy @xmath59 closest to the middle of the spectrum of @xmath37 .
since the density of states is gaussian , the center of the spectrum contains the most delocalized states .
localization in this region assures localization in other parts of the spectrum . for each @xmath24 ,
the total number of data points for the average , including initial states and realizations , adds up to @xmath62 .
data points for @xmath63 from bottom to top ( a ) and ldos for a single realization for the bottom panel @xmath64 ( d ) , the middle @xmath65 ( c ) , and the top @xmath66 ( b ) ; @xmath67 . in ( a ) :
the dashed line indicates @xmath68 for @xmath64 and horizontal lines correspond to the saturation point , @xmath69 .
the envelopes ( solid lines ) of the distributions in panels ( b ) , ( c ) , and ( d ) are gaussians with center @xmath70 [ eq .
( [ eq : mean ] ) ] and width @xmath71 [ eq . ( [ eq : width ] ) ] . ] figure [ fig : ldos ] ( a ) displays @xmath61 for different values of @xmath36 .
the initial decay is very fast until @xmath72 . for small disorder ,
the initial evolution is purely gaussian . as @xmath36 increases , the interval of the gaussian decay shrinks until only the quadratic part persists , @xmath73 .
this is followed by a possible exponential behavior , the time interval being too short for certainty .
after the initial fast evolution , oscillations appear .
the time interval of these oscillations as well as their amplitudes increase with the disorder strength .
the oscillations eventually fade away and give place to a powerlaw decay with exponent @xmath74 .
the initial state finds new channels that give continuation to its evolution .
the couplings at higher order in perturbation theory become gradually effective .
the long - time powerlaw behavior reflects the onset of multifractal states @xcite .
our results indicate that multifractal many - body states can occur even at small @xmath36 .
as the disorder increases and the eigenstates become less extended , the powerlaw exponent naturally decreases . for @xmath67 ,
the decay after the oscillations is hardly noticeable for @xmath75 . at very long times
, the decay eventually saturates .
the saturation point is derived from the infinite time average of eq.([eq : fidelity ] ) , @xmath76 the value of this infinite time average naturally increases with the disorder strength .
figures [ fig : ldos ] ( b ) , ( c ) , and ( d ) display representative ldos for three values of @xmath36 .
the widths of the three distributions are equivalent , because according to eq .
( [ eq : width ] ) , @xmath71 does not depend on the disorder strength .
this explains the indistinguishable initial decay for all curves in fig .
[ fig : ldos ] ( a ) . at small @xmath36 [ fig .
[ fig : ldos ] ( d ) ] , the gaussian envelope of the distribution is still well filled , indicating a very delocalized initial state .
this is independent of the realization , provided @xmath59 be near the center of the spectrum . as the disorder increases , the multifractal structures of the eigenstates spread to larger scales and the coefficients @xmath77 fluctuate strongly . as a result ,
the ldos becomes more sparse [ figs .
[ fig : ldos ] ( b ) and ( c ) ] , justifying the oscillations and subsequent powerlaw decay in fig .
[ fig : ldos ] ( a ) .
the oscillations are due to the small number of states energetically accessible to the initial state in low order of perturbation theory ; a number that decreases as @xmath36 increases .
these oscillations are not random fluctuations that can be averaged out with enough realizations , as those at very long times .
they are connected with the approach to the mbl phase and the onset of quasi - integrals of motion @xcite . in fig . [
fig : sponecase ] , we analyze the survival probability for different system sizes and four values of @xmath36
. the strengths of the disorder are small in figs .
[ fig : sponecase ] ( a ) , ( b ) , while in figs .
[ fig : sponecase ] ( c ) , ( d ) , they coincide , within errors , with the critical point @xmath78 obtained in refs .
the fast evolution for @xmath79 is separated from the later powerlaw decay either by a small plateau ( a ) or by visible oscillations ( c ) , ( d ) .
the scope of the powerlaw behavior increases with system size and with disorder strength[compare the time where saturation takes place in ( a ) with the time in ( c ) , for example ] .
this suggests that for very large @xmath24 the algebraic decay may persist for @xmath80 .
data point for @xmath81 ( a ) , @xmath65 ( b ) , @xmath82 ( c ) , and @xmath66 ( b ) for @xmath83 from top to bottom .
dashed lines give @xmath84 , where @xmath85 ( a ) , @xmath86 ( b ) , @xmath87 ( c ) , and @xmath88 ( d ) . ] the dashed lines in fig .
[ fig : sponecase ] correspond to an algebraic decay described by the generalized dimension , @xmath89 . as shown in fig .
[ fig : supp01 ] , @xmath18 is extracted from the best linear fit to @xmath90 vs @xmath91 for @xmath92 .
the error bars are standard deviations over @xmath62 different values of @xmath93 for each @xmath24 . at small disorder ,
the error bars are small . as @xmath36 approaches @xmath78 , the dispersion of the values of @xmath93 and therefore the uncertainty in the value of @xmath18 increases .
at very large disorder , the error bars decrease again .
vs @xmath94 ( dark circle ) and @xmath95 vs @xmath94 ( light square ) for @xmath96 ( a ) , @xmath65 ( b ) , @xmath97 ( c ) , @xmath66 ( d ) , @xmath98 ( e ) , and @xmath99 ( f ) .
error bars are standard deviations over @xmath62 values of @xmath100 ( dark color ) or of @xmath101 ( light color).,scaledwidth=40.0% ] for small disorder , @xmath102 , the system is still close to the metallic phase and the decay is diffusive , @xmath103 . in this case
, the exponent of the numerical powerlaw decay agrees extremely well with @xmath18 when the system size is large [ see fig . [
fig : sponecase ] ( a ) ] . as @xmath36 increases
, @xmath18 decreases , but not as fast as the numerical exponent .
for @xmath65 [ fig .
[ fig : sponecase ] ( b ) ] , the agreement between the numerical curve and @xmath89 is not very good anymore . in the vicinity of the critical point ,
[ fig : sponecase ] ( c ) and ( d ) , oscillations are seen approximately in the same time interval of the algebraic decay of fig .
[ fig : sponecase ] ( a ) .
the generalized dimension is now @xmath104 and it agrees well with the rate of the damping of those oscillations , while the powerlaw decay appears now latter in time .
as @xmath24 increases , the amplitudes of the oscillations decrease and the slope of the subsequent powerlaw decay becomes more pronounced and closer to @xmath84 .
it is thus plausible to expect that for very large system sizes , @xmath18 might be able to capture the algebraic decay also for large disorder .
this expectation is further supported by the results below for the time - averaged survival probability .
in the analysis of the dynamics of noninteracting systems at the mobility edge @xcite , the commonly employed quantity is the time - averaged survival probability , which smoothes the fluctuations in @xmath61 .
it is defined as , @xmath105 to reduce also the fluctuations in the values of ipr@xmath106 , one often deals with the so - called typical inverse participation ratio , ipr@xmath107 .
the scaling analysis of ipr@xmath108 gives @xmath109 , as shown in fig .
[ fig : supp01 ] .
the error bars for ipr@xmath108 in that figure are , of course , smaller than those for the regular @xmath100 , since now we deal with the dispersions in the values of @xmath110 .
( a ) , @xmath111 ( b ) , @xmath97 ( c ) , and @xmath66 ( d ) for @xmath83 from top to bottom .
dashed lines gives @xmath112 , where @xmath113 ( a ) , @xmath114 ( b ) , @xmath115 ( c ) , and @xmath116 ( d ) . ] in fig .
[ fig : averagesp ] , we compare @xmath117 with @xmath118 . when the system is still close to the metallic phase , as in fig .
[ fig : averagesp ] ( a ) , the decay of @xmath117 is smooth all the way to saturation and in excellent agreement with @xmath119 , especially for @xmath67 . as the disorder increases , the powerlaw exponent decreases , but the short - time dynamics does not change much
this creates an abrupt contrast between the two time scales , resulting in a visible elbow [ see fig .
[ fig : averagesp ] ( c ) and ( d ) ] .
as @xmath36 increases , we also notice that the time interval for the agreement between the algebraic decay of @xmath117 and @xmath119 shortens and starts later in time ( compare the four panels ) . yet , for a fixed disorder , the agreement also improves with @xmath24 .
this indicates that for system sizes larger than available for exact diagonalization , @xmath109 should be able to describe the powerlaw decay for long times , even when the disorder is strong .
the exponent of the powerlaw decay of @xmath50 contains important information about the system : \(i ) because it coincides with the generalized dimension @xmath18 , it indicates the level of delocalization of the initial state . since @xmath120 ,
as suggested in @xcite and confirmed by us for our model , the powerlaw exponent also manifests the level of delocalization of the eigenstates .
\(ii ) it gives information about the correlations between the components @xmath121 , because the algebraic decay implies that @xcite , @xmath122 when the eigenstates , and consequently the initial state , are extended and thus similar to random vectors , as it happens in the chaotic domain ( @xmath102 ) for states close to the middle of the spectrum , the components @xmath121 are uncorrelated random numbers . in this case @xmath123 and the dynamics is diffusive ( @xmath103 ) , as obtained also in @xcite . as the disorder increases ,
the states become multifractal ; the components @xmath121 show large fluctuations and become gradually more correlated , so @xmath124 with @xmath125 , resulting in a subdiffusive dynamics . the limited spreading of the initial state quantified by @xmath18 reflects , as made explicit by eq .
( [ eq : correlations ] ) , the level of correlations between the components @xmath121 . figure [ fig : distribution ] ( a ) shows how @xmath126 and @xmath127 depend on the disorder strength . in parallel with the standard deviations in fig .
[ fig : supp01 ] , the error bars are larger for @xmath126 than for @xmath127 . we show only the latter to simplify the figure . within errors ,
the two generalized dimensions coincide . as @xmath36 increases , the number of states that contribute to the evolution of the initial state shrinks and the generalized dimensions decrease .
the decay is evident for @xmath128 and it becomes extremely slow afterwards .
we avoid an analysis of what happens for @xmath129 , because for the very small system sizes @xmath130 we actually see an approach to on - site localization .
even though it is not clear at this point how to identify the mbl critical point from fig .
[ fig : distribution ] ( a ) , a comparison with previous studies is instructive .
the values of @xmath36 for the mid - point between a metal and an insulator , that is @xmath131 ( @xmath132 ) , and for the inflection point of the fitting curve ( @xmath133 ) are not too far from the critical points found in @xcite .
in addition , the point of an almost halt in the decay of the values of the generalized dimensions , @xmath134 , is very close to the critical point @xmath135 obtained in ref .
@xcite for states that , as in our case , live close to the middle of the spectrum .
( dark circle ) and @xmath127 ( light square ) vs disorder strength ( a ) and the distribution of @xmath136 for @xmath66 and @xmath137 from highest to lowest peak ( b ) . in ( a ) :
solid lines are fitting curves .
only the error bars for @xmath127 are shown .
they are smaller than those for @xmath126 , in accordance with the standard deviations in fig .
[ fig : supp01 ] . ]
we expect @xmath126 and @xmath127 to get closer for scaling analysis performed with larger system sizes than the very few ones now available for exact diagonalization . as @xmath24 increases
, they should also better agree with the powerlaw exponent of @xmath50 .
these claims find support already in the results for @xmath138 and @xmath139 .
if the powerlaw behavior is indeed to be described by the generalized dimension , then the algebraic decays for different system sizes must coincide .
in figs . [
fig : sponecase ] and [ fig : averagesp ] , the slopes are visibly different for small @xmath24 s , but they get closer for @xmath138 and @xmath139 .
this suggests that the scaling analysis should become more accurate for sizes @xmath140
. figure [ fig : distribution ] ( b ) endorses the proximity of the results for @xmath138 and @xmath139 .
it shows the distribution of the inverse participation ratios .
ipr@xmath106 fluctuates with disorder realization and initial state .
however , the validity of eq .
( [ eq : d2forinitial ] ) presupposes that @xmath126 does not depend strongly on what is used on the left side of that equation , whether it is @xmath141 , ipr@xmath142 , or the most probable value of ipr@xmath106 .
this implies that the distribution of ipr@xmath106 normalized to its median @xmath143 must have a scale invariant shape @xcite . as seen in fig .
[ fig : distribution ] ( b ) , the distribution @xmath144 $ ] broadens considerably from @xmath145 to @xmath146 , but the shapes are similar for @xmath138 and @xmath139 . in noninteracting disordered systems described by the powerlaw random banded matrix , numerical evidence for the scale invariance of @xmath147 $ ] was achieved @xcite already for @xmath148 , in contrast with the @xmath149 needed here
. the existence of more correlations between the matrix elements of our system when compared to random matrices may justify such large difference . the number of nonrandom elements in the hamiltonian matrix of eq .
( [ ham ] ) is much larger than in noninteracting systems , such as those described by the tight - binding model or the powerlaw random banded matrix .
we studied the dynamics of an isolated disordered 1d heisenberg model as it approaches the mbl phase .
the analysis was based on the entire evolution of the survival probability @xmath50 , from @xmath150 to @xmath151 , for initial states corresponding to site - basis vectors .
@xmath50 is one of the simplest quantities that can reveal the multifractality of the eigenstates .
it also appears explicitly in the evolution of observables @xcite .
the dynamics of clean and disordered interacting systems is comparable at short times . for both
, the gaussian decay rate of @xmath50 coincides with the width of the ldos . in the presence of disorder , the ldos gets sparse , reflecting the reduced number of states participating in the dynamics and the multifractality of the eigenstates . as a result ,
the behavior of @xmath50 at long times becomes powerlaw .
the exponent of the powerlaw decay coincides with the generalized dimension @xmath15 .
this finding establishes a parallel with previous works about the dynamics of noninteracting systems at criticality and may help advance our understanding of transport properties in interacting systems .
it also implies that from @xmath50 , one can infer the level of delocalization of the initial states and eigenstates , as well as the correlations of their components .
this is advantageous , since numerical methods other than exact diagonalization are available for studying dynamics , which gives access to larger system sizes .
the dynamics can also be studied experimentally with quantum simulators .
this work was supported by the nsf grant no .
ejth acknowledges support from conacyt , mexico .
lfs thanks the itamp hospitality , where part of this work was done .
we thank j. a. mendez - bermudez and z. papi for useful suggestions . | the isolated one - dimensional heisenberg model with static random magnetic fields has become paradigmatic for the analysis of many - body localization . here , we study the dynamics of this system initially prepared in a highly - excited nonstationary state .
our focus is on the probability for finding the initial state later in time , the so - called survival probability .
two distinct behaviors are identified before equilibration . at short times
, the decay is very fast and equivalent to that of clean systems .
it subsequently slows down and develops a powerlaw behavior with an exponent that coincides with the multifractal dimension of the eigenstates . |
sub - velocity precision can now be achieved for m - dwarfs resulting from stabilized spectrographs ( such as harps , * ? ? ?
* ) and specialized spectral analysis techniques @xcite .
this increased sensitivity enables the detection of planets of a few - earth masses orbiting in the star s habitable zone . recently , statistical population analyses of the nasa / kepler mission @xcite , and ground - based doppler surveys @xcite suggests that every low - mass star has at least one ( or more ) planet with an orbital period less than 50 days .
very small planets in very short - period orbits have recently been discovered by the kepler mission ( e.g. koi-1843 , * ? ? ?
we recently started the cool tiny beats survey to charaterise the doppler variability over short time - scales and search for such small planets in short period orbits around nearby low - mass stars .
the program uses the high accuracy radial velocity planet searcher spectrograph ( harps ) installed at the 3.6 m eso telescope at la silla / chile and its northern counterpart at the telescopio nationale galileo / la palma .
[ cols= " < , < , < , < , < , < , < , < , < , < , < " , ]
at only 3.91 pc , kapteyn s star ( or gj 191 , hd 33793 ) is the closest halo star to the sun @xcite .
it is subluminous with respect to main sequence stars of the same spectral type and was spectroscopically classified as an m1.0 sub - dwarf by @xcite .
a radius of [email protected] r@xmath1 was directly measured by @xcite using interferometry .
this combined with an estimate of its total luminosity was then used to derive an effective temperature of 3570@xmath0156 k and its mass was estimated to be [email protected] m@xmath1 .
@xcite determined a metallicity of @xmath2 $ ] = -0.86 which coincides with several later estimates within 0.05 dex .
we independently obtained estimates of its metallicity and temperature by comparing its observed colors ( b , v , j , h , and k ) to synthetically generated ones from the phoenix library @xcite , obtaining compatible values of @xmath3}=-0.89 $ ] and @xmath4 k. only an upper limit has been measured for its projected rotation velocity ( @xmath5 , * ? ? ? * ) , and its x - ray luminosity has been measured to be comparable to other multi - planet host m - dwarfs such as gj 876 and gj 581 @xcite .
therefore , planets of a few earth - masses orbiting such an inactive m - star should be detectable using doppler spectroscopy @xcite . a precise age estimation of the star can not be obtained from models , as they change very little for m @xmath6 0.6 m@xmath1 in the range of ages between 0.4 and 15 gyr @xcite . the low metallicity and halo kinematics suggest an ancient origin , which is consistent with its low - activity and slow rotation .
the observations comprise both new and archival data from harps , hires and pfs spectrometers .
harps is a stabilized high - resolution spectrometer ( resolving power @xmath7 110 000 ) covering from 380 to 680 nm .
the harps data we use come from two programs ; the cool tiny beats survey and archival data from the first harps - gto survey . the cool tiny beats data were obtained in two 12 night runs in may 2013 ( 11 spectra ) and in dec 2013 ( 55 spectra ) .
the harps - gto data contain 30 spectra taken between 2003 and 2009 .
@xcite reported some possible signals on kapteyn s star , but no detection claim could be made at the time .
doppler measurements were obtained using the least - squares template - matching approach as implemented by the harps - terra software @xcite . for each spectrum
, we also use two measurements of the symmetry of the mean - line profile as provided by the harps - data reduction software .
these are the bisector span ( bis ) and the full - width - at - half - maximum ( fwhm ) of the cross - correlation function .
bis and fwhm are known to correlate with activity induced features that can cause spurious doppler signals .
for example , changes in the line - symmetry caused by co - rotating dark and magnetic spots induce changes in the symmetry of the lines that produce spurious doppler shifts ( e.g. , * ? ? ? * ) .
the variability in the ca ii h+k emission lines was also measured by harps - terra through the s - index .
the variability in the s - index is also known to correlate with magnetic activity of the star @xcite , and localized active regions such as spots .
any doppler signal with a period equivalent to variability detected in the bis , fwhm and s - index is likely to be spurious .
we include 30 doppler measurements spanning between 1999 and 2008 using the hires spectrometer at keck @xcite .
these measurements were obtained using the iodine cell technique as described in @xcite .
a new hires stellar template was generated by deconvolving a high signal - to - noise iodine - free spectrum of the star applying maximum likelihood deconvolution with boosting @xcite .
the star is never at an altitude greater than 26 deg from hawaii , which could explain the lower accuracy compared to the harps measurements .
the long baseline of hires puts strong constraints on the long - period variability and mitigates alias ambiguities in the harps data .
we also included 8 new doppler measurements obtained with pfs @xcite at magellan / las campanas observatory , also using the iodine cell technique . while the statistical contribution of pfs is small , the significance of the preferred solution increases thus providing further support to the signals .
all spectroscopic measurements used are given in table [ tab : data ] .
the initial signal detection is performed using log - likelihood periodograms @xcite and tempered markov chain samplings using the delayed - rejection adaptive - metropolis method ( or dram , * ? ? ?
* ) as implemented in @xcite .
these methods produce maps of the maximum likelihood and posterior densities as a function of the period being investigated . as in classic periodograms
, preferred periods will appear as peaks ( local probability maxima ) .
the significance is then assessed using the bayesian criteria described in @xcite .
we say that a doppler signal is detected if 1 ) its period is well - constrained , 2 ) its amplitude is statistically significantly different from zero , and 3 ) inclusion of the signal in the model increases the bayesian evidence by a factor of @xmath8 , which is more conservative than usual because important effects might be missing in the model , especially when combining data from different instruments .
the model of the observations is fully encoded in the definition of the likelihood function .
in addition to the usual keplerian parameters , our model includes three nuisance parameters for each instrument ( ins ) : a constant offset @xmath9 , an extra white - noise term @xmath10 and a moving average coefficient @xmath11 that quantifies the amount of correlation between consecutive measurements .
@xmath12 is bound between @xmath13 and @xmath14 ( anticorrelation ) and a value consistent with zero implies no significant correlation .
our model also assumes an exponential decay of the correlation that depends on a characteristic time - scale @xmath15 . adding it as a free parameter led to overparameterization , so a fixed value of @xmath16 days was set by default .
linear trends caused by the presence of planets with periods much longer than the time - baseline is parameterized as @xmath17 , where @xmath18 is also a free parameter of the doppler model and @xmath19 is an arbitrary reference epoch ( see table [ tab : parameters ] ) .
the log - likelihood periodograms use the same likelihood model but do not include correlations terms .
false alarm probability assessments using likelihood periodograms ( or fap , see * ? ? ?
* ) are also computed to double - check the significance of each detection .
the search for signals in the doppler time - series is summarized in figure [ fig : periodograms ] .
two very significant periods are detected at 121 and 48.6 days .
the bayesian evidence ratios are @xmath20 for the one - planet model against the no - planet model , and @xmath21 for the two - planet solution against the one - planet model .
these numbers are well in excess of the 10@xmath22 threshold indicating very confident detections .
the faps as derived from the log - likelihood periodograms are @xmath23 and @xmath24 respectively , which are also very small further supporting the detections ( acceptable detection thresholds using periodogram methods are typically between 1% and 0.1% ) .
the support of the signals by the three data sets is illustrated in the phase - folded curves in figure [ fig : phased ] .
the uneven sampling can produce biases in the significance estimates . to investigate this
, we also performed the same analysis using nightly averaged measurements .
the same two signals are detected well above our significance thresholds , thus providing further confidence in the detections .
as for other stars , further follow - up over the next few years is desirable to confirm that both low - amplitude signals remain coherent over time .
we investigated linear correlations with activity indices by including them in the likelihood model of the harps data . while we could still easily detect the planetary signals , these models had lower probabilities due to over - parameterization and are not included in the final solution
. we also verified whether if any indicator of activity showed variability in similar time - scales as the doppler data .
while the bis does not show any hint of temporal coherence , the fwhm and s - index do share similar periodogram structures with tentative peaks between 140 to 2000 days .
however , neither had faps that were lower than 5% , indicating little significance .
lrr model parameter & kapteyn b & kapteyn c + @xmath25 [ day ] & @xmath26 & @xmath27 + @xmath28 [ ] & @xmath29 & @xmath30 + @xmath31 & @xmath32 & @xmath33 + @xmath34 [ deg ] & @xmath35 & @xmath36 + @xmath37 [ deg ] & @xmath38 & @xmath39 + @xmath18 [ yr@xmath40 & @xmath41 + @xmath42 [ ] & @xmath43 + @xmath44 [ ] & @xmath45 + @xmath46 [ ] & @xmath47 + @xmath48 & @xmath49 + @xmath50 & @xmath51 + @xmath52 & @xmath53 + + + @xmath54 [ deg ] & @xmath55 & @xmath56 + @xmath57 [ m@xmath58 & 4.8@xmath59 & 7.0@xmath60 + @xmath61 [ au ] & 0.168@xmath62 & 0.311@xmath63 + s / s@xmath64 & 40% & 12% + hz - range [ au ] & @xmath70.126 - 0.236 + p@xmath65/p@xmath66 & @xmath67 + we also analysed v - band photometry obtained from the latest release of the asas program @xcite .
a standard deviation computed using the central 80% percentile was used to remove 4-@xmath68 outliers resulting in 503 epochs spanning from dec 2000 to sep 2007 .
a log - likelihood periodogram analysis identified two significant signals at 340 and 1100 days , both with amplitudes of @xmath7 5 mmag .
the signal at 1100 days is compatible with other reports of periodicities in activity indices of m - dwarfs @xcite , and the 340-day variability might be caused by seasonal systematic errors , or by a long rotation period .
the periodogram structure of the photometry resembles those of the fwhm and the s - index , further supporting potential low - level activity changes happening at timescales of several hundred days .
given that the activity and doppler signals appear uncorrelated , we conclude that the simplest interpretation of the doppler data is the presence of two planets ( see table [ tab : parameters ] ) . with a period of 48.6 days
, kapteyn b lies well within the liquid water habitable zone of the star @xcite . assuming rocky composition , its properties and possible climates should be similar to those discussed for gj 667ce @xcite .
kapteyn c receives @xmath7 10% of earth s irradiance , implying that the possibility of it being able to support liquid water on its surface is less probable .
we used optimal bls @xcite to search for possible transit events in the asas photometry .
no trasits were found when searching for a signal in the range of parameters compatible with the doppler solution .
further observations at the predicted transit windows are needed to put meaningful constraints on possible transit signals .
the age of kapteyn star can be inferred from its membership to the galactic halo and peculiar element abundances @xcite . the current hierarchical milky way formation scenario suggests that streams of halo stars were originated as tidal debris from satellite dwarf galaxies being engulfed by the early milky way ( * ? ? ? * and references therein ) .
this scenario is supported by the age estimations of the stars in the inner halo ( @xmath71012 gyr , * ? ? ? * ) , and globular clusters .
the work by eggen and collaborators ( see , * ? ? ?
* as review summary ) established the existence of such high - velocity , metal - poor moving groups in the solar neighborhood .
kapteyn s star is the prototype member of one of these groups , which has been recently investigated by @xcite among others .
stars in kapteyn s group share a retrograde velocity of rotation around the galactic centre of about @xmath69 k@xcite .
spectroscopic observation of 16 members has shown that the group likely originated from the same progenitor structure as the peculiar globular cluster @xmath34 centauri , but not from within the cluster itself @xcite .
the origin of kapteyn s star within a merging dwarf - galaxy sets its most likely age around the young halo s one ( @xmath70 gyr ) and should not be older than 13.7 gyrs , which is the current estimate of the age of the universe @xcite .
numerical integration of several possible orbits over 10@xmath22yr to 1 myr using _ mercury 6 _ @xcite shows very small changes in the orbital parameters . despite the the periods for the two planet candidates are compatible with a 5:2 period commensurability ( see table [ tab : parameters ] ) , the analysis of the resonant angles for the best fit solution did not support the presence of dynamical mean - motion resonances , which existence also depend on a number of other properties such total planet masses and mutual inclinations .
we found , however , that the difference in periastron angles @xmath71 librates around 180 deg ; which corresponds to a long - term stable configuration called apsidal locking .
this is sufficient to show that the proposed system is compatible with long - term physically - viable solutions .
it has been suggested that two - planet systems that underwent weak dissipation ( slow migration ) should always end - up in absidal locking @xcite , which has consequences on the likely planet - formation scenario .
a more thorough analysis to properly quantify the significance of the absidal locking and possible mean - motion ressonances requires a much more extended discussion ( e.g. , * ? ? ? * ) which is beyond the scope of this paper .
the detection of two super - earths here is consistent with the idea that low - metallicity stars are more prone to the formation of low - mass planets rather than gas giants @xcite .
this is further supported by a significant paucity of lowest - mass planets in doppler searches of metal - rich stars @xcite , and non - comfirmation of previous claims of gas giants around extremely metal - poor stars @xcite .
these observational findings are compatible with recent population synthesis experiments such as those in @xcite .
once the planets signals are included in the doppler model , the rv residuals variability are reduced to instrumental noise ( i.e. extra jitter term is compatible with reported stability of harps , and correlation coefficients compatible with 0 ) .
this indicates that the star is very doppler stable , possibly more stable than the instruments themselves .
this indeed would be expected because pulsation and convective motions are thought to be much smaller in inactive low - mass stars than in earlier types . at the likely age of the system
, most g and k dwarfs are evolving away from the main sequence into giants , which makes the doppler detection of small planets unfeasible due to increased activity levels ( e.g. , * ? ? ? * ) . as a result
, original architectures of the first planetary systems can only be explored by observing venerable low - mass stars which are still on the main - sequence such as kapteyn s star .
* acknowledgments . *
we thank r.p.nelson , j.chanam and r. baluev ( referee ) for constructive comments and discussions .
we acknowledge funding from : dfg / germany through crc-963 ( cjm ) ; dfg / germany 1664/9 - 1 ( ar ) ; alexander von humboldt foundation / germany ( jm ) ; erc - fp7/eu grant number 27347 ( mz ) , aya2011 - 30147-c03 - 01 by mineco / spain , feder funds / eu , and 2011 fqm 7363 of junta de andaluca / spain ( cr - l and pja ) ; jae - doc program ( cr - l ) ; fpi bes-2011 - 049647 mineco / spain ( zmb ) ; cata ( pb06 , conicyt)/chile ( jsj ) ; conicyt - pfcha / doctorado nacional / chile ( md);nsf / usa grants ast-0307493 and ast-0908870 ( ssv ) ; and nasa grant nnx13af60 g s02 ( rpb ) .
based on observations made with eso telescopes under programme i d 191.c-0505 and eso s science archive facility ( req .
number ganglada95087 ) .
we acknowledges the effort of the harps team program obtaining data within program 072.c-0488 .
some data were obtained at the w.m .
keck obs . made possible by the support of the w.m .
keck foundation and operated among caltech , univ .
of california and nasa .
the authors are most fortunate to conduct observations from the sacred mountain of mauna kea .
this study uses data obtained at magellan , operated by the carnegie inst .
, harvard univ .
, univ . of michigan ,
univ . of arizona , and the massachusetts inst . of technology .
g. , butler r. p. , 2012 ,
apjs , 200 , 15 g. , tuomi m. , gerlach e. , et al .
2013 , a&a , 556 , a126 r. v. , 2008 , mnras , 385 , 1279 r. v. , 2009 , mnras , 393 , 969 i. , chabrier g. , allard f. , hauschildt p. h. , 1997 , a&a , 327 , 1054 j. r. , jeffers s. v. , jones h. r. a. , 2011 , mnras , 412 , 1599 x. , delfosse x. , udry s. , et al .
2013 , a&a , 549 , a109 m. k. , basri g. , marcy g. w. , west a. a. , zhang j. , 2010 , aj , 139 , 504 l. a. , latham d. w. , johansen a. , et al .
2012 , nature , 486 , 375 r. p. , marcy g. w. , williams e. , et al .
1996 , pasp , 108 , 500 j. e. , 1999 , mnras , 304 , 793 j. d. , shectman s. a. , butler r. p. , et al .
2010 , proc .
spie , 7735 s. , sozzetti a. , bonomo a. s. , et al .
2013 , a&a , 554 , a29 c. d. , charbonneau d. , 2013 , apj , 767 , 95 o. j. , 1996 , aj , 112 , 1595 j. e. , 1997 , aj , 113 , 806 j. , santos n. c. , bonfils x. , et al .
2012 , a&a , 541 , a9 haario h. , laine m. , mira a. , saksman e. , 2006 , statistics and computing , 16 , 339 t .- o . , wende - von berg s. , dreizler s. , et al .
2013 , a&a , 553 , a6 j. s. , jones h. r. a. , tuomi m. , et al .
2013 , apj , 766 , 67 p. , weiss a. , 2011 , a&a , 533 , a59 m. i. , jenkins j. s. , 2014 , a&a , 562 , a129 r. j. , 2010 , a&a rev . , 18 , 567 r. k. , ramirez r. , kasting et al .
2013 , apj , 765 , 131 e. , innanen k. , dawson p. c. , et al .
2005 , a&a , 438 , 957 m. , pepe f. , queloz d. , bouchy f. , et al .
2003 , the messenger , 114 , 20 t. a. , rodrguez a. , 2011 , mnras , 415 , 2275 c. , alibert y. , benz w. , klahr h. , henning t. , 2012 , a&a , 541 , a97
m. , matousec v. , 2009 , dsp , 19 , 372 g. , niedzielski a. , wolszczan a. , adamw m. , maciejewski g. , 2013 , apj , 770 , 53 a. , 2014 , a&a , 561 , a138 a. , dreizler s. , 2013 , a&a , 555 , a58 2013 , arxiv:1303.5062 g. , 1997 , aca , 47 , 467 a. , shulyak d. , anglada - escud g. , et al .
2013 , a&a , 552 , a103 d. , kervella p. , forveille t. , queloz d. , 2003 , a&a , 397 , l5 m. , anglada - escude g. , jenkins j. s. , jones h. r. a. , 2014 , arxiv:1405.2016 m. , jones h. r. a. , barnes j. r. , et al .
2014 , mnras s. , santos n. c. , 2007 , ara&a , 45 , 397 f. , 2007 , a&a , 474 , 653 s. s. , allen s. l. , bigelow b. c. , et al .
1994 , spie conf .
series , 2198 , 362 l. m. , johns - krull c. m. , hawley s. l. , 2008 , apj , 677 , 593 v. m. , wallerstein g. , 2005 , mnras , 356 , 963 e. , freeman k. , williams m. , 2010 , aj , 139 , 636
.... spectroscopic measurements on kapteyn 's star ( anglada - escude+ , 2014 ) = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = description : time - series of spectroscopic measurements used in the paper .
median value and a perspective acceleration were subtracted to each rvs set ( ins .
1 is harps , 2 is hires , 3 is pfs ) .
measurements of the fwhm , bis of the cross - correlation profiles and measurements of the s - index are provided for harps data only .
uncertainty in the fwhm is 2.5 times the uncertainty in bis .
check 2012apjs .. 200 ... 15a , for more detailed definitions of the measurements and their uses . -------------------------------------------------------------------------------- bytes format units label explanations -------------------------------------------------------------------------------- 1- 13 f13.5 d jd barycentric julian date 15- 19 f5.2 m / s rvel radial velocity 21- 24 f4.2
m / s ervel uncertainy in rv 26 i1 --- ins instrument used 28- 33 f6.2 m / s bis bisector span of the ccf 35- 38 f4.2 m / s ebis uncertainty in bis 40- 46 f7.5 km / s fwhm fwhm of the ccf ( 1 ) 48- 53 f6.4 --- sidx caii h+k s - index in the mount wilson system 55- 60 f6.4 --- esidx uncertainty in s - index -------------------------------------------------------------------------------- = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = 2452985.74111 3.11 0.80 1 -4.64 0.86 3.21561 0.2966 0.0042 2452996.76321 3.00 0.27 1 -9.75 0.56 3.19206 0.2699 0.0033 2453337.80096 -2.65 0.89 1 -8.17 0.76 3.18036 0.1859 0.0036 2453668.83537 -2.57 0.52 1 -10.27 0.49 3.21216 0.2678 0.0026 2453670.73947 -3.76 0.59 1 -13.27 0.41 3.20937 0.2636 0.0022 2454141.57444 0.56 0.53 1 -13.62 0.37 3.20616 0.2968 0.0021 2454167.53053 -1.22 0.56 1 -14.86 0.38 3.20717 0.2616 0.0020 2454169.50521 -1.66 0.48 1 -13.34 0.34 3.19959 0.2578 0.0018 2454173.50080 -1.57 0.48 1 -13.76 0.36 3.19892 0.2540 0.0019 2454197.51207 2.47 0.54 1 -16.21 0.43 3.20409 0.2470 0.0022 2454229.46873 2.59 0.73 1 -13.00 0.56 3.20358 0.2926 0.0034 2454342.86258 2.61 0.59 1 -13.37 0.39 3.20687 0.2632 0.0020 2454345.87966 0.29 0.45 1 -12.34 0.33 3.20096 0.2549 0.0017 2454349.83417 0.35 0.45 1 -17.60 0.35 3.20115 0.2346 0.0017 2454424.73161 2.40 0.54 1 -14.00 0.37 3.19613 0.2305 0.0018 2454429.71897 4.68 0.56 1 -12.71 0.39 3.19832 0.2390 0.0020 2454449.76051 1.27 0.61 1 -13.88 0.45 3.18970 0.2318 0.0023 2454450.64700 1.67 0.57 1 -13.85 0.40 3.19968 0.2346 0.0020 2454452.70683 1.15 0.74 1 -17.21 0.53 3.19416 0.2540 0.0028 2454459.64359 0.16 0.56 1 -15.85 0.35 3.20016 0.2377 0.0017 2454462.70572 1.82 0.55 1 -10.81 0.35 3.20233 0.2457 0.0018 2454463.69854 2.61 0.49 1 -12.56 0.43 3.19982 0.2341 0.0022 2454464.68243 1.07 0.53 1 -12.38 0.42 3.20015 0.2452 0.0021 2454733.89725 -0.93 0.60 1 -15.14 0.36 3.20390 0.2536 0.0018 2454754.83883 -2.37 0.57 1 -13.84 0.41 3.20737 0.2538 0.0021 2454767.75805 0.08 0.58 1 -14.73 0.43 3.20601 0.2393 0.0021 2454879.55605 -2.33 0.59 1 -15.14 0.43 3.19513 0.4865 0.0032 2454879.69685 -1.36 0.85 1 -14.42 0.68 3.19309 0.2085 0.0034 2454880.52165 -0.01 1.56 1 -0.65 1.28 3.19636 0.2472 0.0053 2454880.67387 -2.96 0.76 1 -13.39 0.57 3.19398 0.2071 0.0028 2456417.46007 -0.10 0.60 1 -14.56 0.68 3.20851 0.2682 0.0044 2456417.46745 -0.54 0.68 1 -7.42 0.75 3.20894 0.2566 0.0047 2456418.46471 1.80 0.43 1 -16.75 0.55 3.20068 0.2761 0.0040 2456420.47541 0.63 0.39 1 -15.72 0.47 3.19810 0.3089 0.0037 2456421.46438 1.82 0.76 1 -18.81 0.95 3.20174 0.2652 0.0056 2456422.45767 2.00 0.72 1 -13.54 0.71 3.19709 0.2700 0.0049 2456423.46236 2.00 0.73 1 -11.24 0.89 3.19925 0.1930 0.0048 2456424.44602 1.03 0.93 1 -7.98 0.95 3.20863 0.4363 0.0069 2456426.46387 0.84 0.59 1 -13.18 0.68 3.19599 0.2540 0.0046 2456427.44860 3.42 0.75 1 -14.61 0.78 3.20405 0.2690 0.0050 2456428.46156 1.08 0.55 1 -10.70 0.68 3.20416 0.4196 0.0058 2456656.56057 0.24 0.75 1 -14.31 0.68 3.21574 0.2927 0.0042 2456656.59693 0.36 0.60 1 -11.10 0.51 3.20988 0.2697 0.0033 2456656.64552 0.26 0.83 1 -22.11 0.72 3.21746 0.2505 0.0041 2456656.69364 3.00 0.55 1 -12.33 0.55 3.20616 0.2597 0.0037 2456656.74624 1.87 0.58 1 -15.74 0.55 3.20590 0.2707 0.0037 2456657.55783 1.33 0.75 1 -16.59 0.62 3.21076 0.2827 0.0040 2456657.59109 2.32 0.53 1 -13.13 0.58 3.20375 0.2525 0.0037 2456657.64325 1.65 0.64 1 -11.93 0.58 3.21207 0.2984 0.0038 2456657.70141 2.59 0.57 1 -19.43 0.52 3.19999 0.2884 0.0040 2456657.75751 0.63 0.62 1 -19.62 0.50 3.21424 0.2675 0.0032 2456658.57385 2.80 0.58 1 -15.94 0.59 3.20647 0.2747 0.0038 2456658.60490 1.03 0.51 1 -13.50 0.57 3.21336 0.2671 0.0036 2456658.70151 1.26 0.65 1 -18.40 0.51 3.22847 0.2596 0.0029 2456658.70928 0.49 0.59 1 -12.00 0.50 3.23253 0.2737 0.0029 2456658.76592 2.35 0.43 1 -16.13 0.46 3.21731 0.2722 0.0031 2456658.84514 1.04 0.55 1 -14.10 0.49 3.19162 0.2733 0.0037 2456659.63210 1.86 0.53 1 -17.67 0.47 3.21368 0.2745 0.0030 2456659.69120 0.19 0.57 1 -11.60 0.48 3.20922 0.2819 0.0034 2456659.74851 0.31 0.45 1 -15.85 0.49 3.20449 0.2833 0.0036 2456659.79823 1.45 0.66 1 -10.39 0.54 3.19687 0.2690 0.0040 2456660.53935 0.99 0.74 1 -13.38 0.75 3.20692 0.2893 0.0046 2456660.59008 0.61 0.61 1 -16.00 0.59 3.20940 0.2862 0.0039 2456660.62526 1.24 0.52 1 -14.84 0.51 3.21176 0.2766 0.0033 2456660.68361 2.24 0.61 1 -15.18 0.55 3.20553 0.2735 0.0038 2456660.74382 1.46 0.63 1 -15.76 0.66 3.21245 0.2705 0.0042 2456660.79408 3.28 0.76 1 -16.45 0.83 3.20489 0.2632 0.0051 2456660.82165 3.14 0.98 1 -10.63 1.01 3.20620 0.3202 0.0065 2456661.53634 2.80 0.51 1 -18.11 0.54 3.21203 0.2905 0.0035 2456661.56643 2.21 0.55 1 -12.63 0.56 3.20931 0.2886 0.0037 2456661.59856 1.54 0.67 1 -13.82 0.54 3.20993 0.2781 0.0035 2456661.65691 0.66 0.52 1 -11.49 0.57 3.21339 0.2699 0.0038 2456661.71555 0.18 0.48 1 -15.70 0.56 3.21309 0.2501 0.0036 2456661.77193 0.66 0.81 1 -9.30 0.68 3.20710 0.2444 0.0042 2456665.54012 3.21 0.62 1 -15.37 0.55 3.21136 0.2941 0.0035 2456665.61482 0.85 0.41 1 -13.40 0.52 3.21657 0.3057 0.0035 2456665.63272 0.89 0.86 1 -14.43 0.66 3.21348 0.3073 0.0046 2456665.64914 1.06 0.59 1 -15.94 0.57 3.21032 0.2836 0.0040 2456665.66524 2.73 0.69 1 -15.27 0.66 3.21198 0.2955 0.0046 2456665.68169 0.33 0.65 1 -17.38 0.60 3.21303 0.2737 0.0040 2456665.69927 1.95 0.59 1 -15.54 0.61 3.21404 0.2957 0.0043 2456665.73292 1.00 0.71 1 -12.16 0.64 3.20612 0.2586 0.0042 2456665.74854 0.60 0.82 1 -17.69 0.75 3.20405 0.2949 0.0053 2456665.76395 1.46 0.83 1 -19.09 0.82 3.20640 0.2686 0.0054 2456666.53144 1.56 0.53 1 -22.85 0.69 3.21604 0.3345 0.0049 2456666.62771 1.60 0.59 1 -18.01 0.59 3.22144 0.2866 0.0039 2456666.67780 0.95 0.64 1 -14.63 0.63 3.21306 0.2783 0.0043 2456666.74492 1.06 0.62 1 -15.17 0.59 3.20733 0.2932 0.0045 2456666.76073 0.33 0.66 1 -15.40 0.66 3.21056 0.2877 0.0047 2456667.54294 2.00 0.73 1 -16.22 0.82 3.22648 0.3151 0.0048 2456667.59735 0.29 0.61 1 -14.08 0.62 3.21076 0.2695 0.0040 2456667.60350 1.62 0.56 1 -13.13 0.62 3.21454 0.3004 0.0042 2456667.67412 2.21 0.79 1 -16.97 0.72 3.21627 0.3009 0.0049 2456667.74602 1.04 0.57 1 -14.67 0.64 3.20707 0.3147 0.0048 2456667.79693 2.86 0.72 1 -9.98 0.78 3.21002 0.2869 0.0052 2451170.92507 5.00 1.59 2 0.00 0.00 0.00000 0.0000 0.0000 2451580.81122 -4.06 1.71 2 0.00 0.00 0.00000 0.0000 0.0000 2451899.97215 2.63 1.79 2 0.00 0.00 0.00000 0.0000 0.0000 2452189.14851 1.79 2.37 2 0.00 0.00 0.00000 0.0000 0.0000 2452235.90051 -0.81 1.98 2 0.00 0.00 0.00000 0.0000 0.0000 2452538.06980 7.73 1.88 2 0.00 0.00 0.00000 0.0000 0.0000 2452651.88827 -7.96 2.36 2 0.00 0.00 0.00000 0.0000 0.0000 2452711.72343 -5.73 1.85 2 0.00 0.00 0.00000 0.0000 0.0000 2452712.74594 0.38 1.55 2 0.00 0.00 0.00000 0.0000 0.0000 2452988.87863 0.00 1.92 2 0.00 0.00 0.00000 0.0000 0.0000 2453339.97662 -1.19 1.85 2 0.00 0.00 0.00000 0.0000 0.0000 2453368.91795 0.24 1.96 2 0.00 0.00 0.00000 0.0000 0.0000 2453425.72302 -6.66 1.38 2 0.00 0.00 0.00000 0.0000 0.0000 2453425.72927 -7.53 1.88 2 0.00 0.00 0.00000 0.0000 0.0000 2453723.91920 1.93 1.74 2 0.00 0.00 0.00000 0.0000 0.0000 2453776.76985 -1.62 1.52 2 0.00 0.00 0.00000 0.0000 0.0000 2453776.77608 -2.64 1.50 2 0.00 0.00 0.00000 0.0000 0.0000 2453982.10746 4.57 1.30 2 0.00 0.00 0.00000 0.0000 0.0000 2454085.98215 3.44 2.05 2 0.00 0.00 0.00000 0.0000 0.0000 2454085.98875 2.72 2.06 2 0.00 0.00 0.00000 0.0000 0.0000 2454138.73882 -1.22 1.57 2 0.00
0.00 0.00000 0.0000 0.0000 2454398.04694 -1.97 1.66 2 0.00 0.00 0.00000 0.0000 0.0000 2454430.03169 0.71 1.99 2 0.00 0.00 0.00000 0.0000 0.0000 2454454.98751 -3.58 1.56 2 0.00 0.00 0.00000 0.0000 0.0000 2454455.85705 -1.14 1.31 2 0.00 0.00 0.00000 0.0000 0.0000 2454460.82264 -2.21 1.97 2 0.00 0.00 0.00000 0.0000 0.0000 2454718.13512 1.39 1.52 2 0.00 0.00 0.00000 0.0000 0.0000 2454719.13273 1.46 1.28 2 0.00 0.00 0.00000 0.0000 0.0000 2454723.12422 0.03 1.45 2 0.00 0.00 0.00000 0.0000 0.0000 2454724.13185 -0.17 1.57 2 0.00 0.00 0.00000 0.0000 0.0000 2454725.14330 1.29 1.68
2 0.00 0.00 0.00000 0.0000 0.0000 2455490.97157 -5.19 1.53 2 0.00 0.00 0.00000 0.0000 0.0000 2456695.57403 -0.31 0.64 3 0.00 0.00 0.00000 0.0000 0.0000 2456697.60659 0.21 0.54 3 0.00 0.00 0.00000 0.0000 0.0000 2456700.62164 0.00 0.53 3 0.00 0.00 0.00000 0.0000 0.0000 2456729.56735 1.40 0.54 3 0.00 0.00 0.00000 0.0000 0.0000 2456730.55791 -0.06 0.57 3 0.00 0.00 0.00000 0.0000 0.0000 2456731.52289 -1.26 0.84 3 0.00 0.00 0.00000 0.0000 0.0000 2456732.54351 -0.59 0.77 3 0.00 0.00 0.00000 0.0000 0.0000
2456735.53824 0.06 0.65 3 0.00 0.00 0.00000 0.0000 0.0000 .... | exoplanets of a few earth masses can be now detected around nearby low - mass stars using doppler spectroscopy . in this paper
, we investigate the radial velocity variations of kapteyn s star , which is both a sub - dwarf m - star and the nearest halo object to the sun .
the observations comprise archival and new harps , hires and pfs doppler measurements .
two doppler signals are detected at periods of 48 and 120 days using likelihood periodograms and a bayesian analysis of the data . using the same techniques ,
the activity indicies and archival asas-3 photometry show evidence for low - level activity periodicities of the order of several hundred days . however , there are no significant correlations with the radial velocity variations on the same time - scales .
the inclusion of planetary keplerian signals in the model results in levels of correlated and excess white noise that are remarkably low compared to younger g , k and m dwarfs . we conclude that kapteyn s star is most probably orbited by two super - earth mass planets , one of which is orbiting in its circumstellar habitable zone , becoming the oldest potentially habitable planet known to date .
the presence and long - term survival of a planetary system seems a remarkable feat given the peculiar origin and kinematic history of kapteyn s star .
the detection of super - earth mass planets around halo stars provides important insights into planet - formation processes in the early days of the milky way .
[ firstpage ] techniques : radial velocities stars : individual : kapteyn s star , planetary systems |
during the last two , three decades , the field of crowd dynamics has emerged as the _ natural _
sciences reaction to questions arising from _ social _ sciences , population biology and urban planning .
see e.g. @xcite for an example of a problem addressed in psychology , or @xcite for an illustration of the civil engineering aspects .
the roots and philosophy of crowd dynamics are very much in the spirit of statistical mechanics , molecular dynamics , interacting particle systems methods and the theory of granular matter , as such treating individual humans nearly as non - living material ( cf .
e.g. the nice overview @xcite and references cited therein ) .
a justification for this approach lies in the fact that the individuals personal will is more or less averaged out if one looks at the crowd as a whole . from this perspective
, it can be considered as ( stochastic ) noise , superimposed on some ` clean ' ( deterministic ) dynamics .
+ to illustrate the thin borderlines between several fields of study , the reader is referred to e.g. @xcite for the dynamics of non - living particles , @xcite for studies of tumbling or self - propelled living particles ( like bacteria ) , or @xcite for crowd dynamics .
although these fields all focus their own specific real - world scenario , their way of thinking and posed questions are very much alike . + + however , an evident and important difference between people and molecules or grains ( apart from people s own opinions , irritations etc . )
is the fact that people clearly have front and back sides .
our degree of perceiving our surroundings highly depends on the direction of looking .
we mainly base our walking behaviour on what we see , and clearly what happens in front of us thus has more influence than what happens behind us ( this statement is also supported e.g. by @xcite ) .
a modification or extension of physics - inspired models is needed to incorporate this kind of anisotropy in the interactions between individuals .
this paper investigates the effect of anisotropy on the global behaviour of a group of pedestrians .
+ + our focus is on the simulation of a scenario where pedestrians move in a long corridor .
we might relate this situation to evacuation of people from a building ( cf .
it is sane to assume that these evacuees have an intrinsic _ drive _ to move towards the exit ( i.e. one side of the corridor ) , and moreover that there view is focused in the same direction .
investigating the effect of anisotropy on the large - scale behaviour of the crowd therefore relates to assessing the escape process .
+ + in section [ sec : level1 ] of this paper the model is presented and explained .
section [ sec : sim setup and results ] is the main part of the paper .
it describes the exact scenario of our simulations and the definitions of the quantities we use for assessing the results ( polarization index , projected density , morisita index ) .
moreover , in this section the simulation results are presented and discussed .
conclusions and an outlook on possible future work are given in section [ sec : conclusion ] .
we represent pedestrians by point particles having masses @xmath0 .
they are located in a long corridor of length @xmath1 and width @xmath2 . here
, the word ` long ' refers to the fact that at the time scales we focus on , the pedestrians are not able to reach the end of the corridor .
interactions between pedestrians are short - ranged .
we therefore suppose that the correlation length in the system is less than or equal to a certain @xmath3 and we can subdivide the corridor in an array of rectangles ( width @xmath2 and length @xmath4 ) , which are all duplicates of each other .
we thus have a scenario with periodic boundary conditions .
our domain of interest is therefore a rectangular box @xmath5 \times [ -\frac{b}{2 } , \frac{b}{2 } ] , \ ] ] with periodic boundary conditions in one direction and impermeable walls in the other direction .
the corridor contains @xmath6 pedestrians .
for all @xmath7 @xmath8 and @xmath9 , the vector @xmath10 represents the position of the @xmath7-th pedestrian at time @xmath11 .
we denote its velocity by @xmath12 .
+ + we assume that the governing equation of motion is @xmath13 the equation describes the motion of the @xmath7-th individual , which has mass @xmath0 and which moves with velocity @xmath12
. however , he / she _ tries _ to move according to its desired velocity @xmath14 . here
, @xmath15 is the characteristic relaxation time related to attaining the desired velocity .
its actual velocity is moreover perturbed by two ` forces ' .
the word ` force ' is used since ( [ newton_overdamped ] ) can be regarded as an overdamped limit of a newton - like equation ( cf .
@xcite for this newton - like way of modelling ) .
+ one could argue whether the social force is the right concept to use to drive the pedestrians , or maybe ideas like social pressure ( as in a darcy - like law ) or cognitive - based heuristics ( see e.g. @xcite ) are more appropriate . here
we avoid any polemic by deciding to choose a framework based on social forces and leave for later any further developments of other possible approaches .
+ + there is a physical force @xmath16 that acts on the individual to describe the effect of the non - living environment ( geometry ) . in this paper
we only take into account the influence of walls on pedestrians , that is : @xmath16 = @xmath17 .
furthermore , pedestrian @xmath7 experiences a so - called social force @xmath18 due to the presence of other individuals , which influences the motion of this particular pedestrian @xmath7 .
+ + individuals are influenced by the walls as soon as they come too close , i.e. within a distance @xmath19 .
we model these impermeable walls by means of a strong repulsive force @xmath17 acting on pedestrian @xmath7 : @xmath20 here , @xmath21 is the unit normal pointing from the corresponding wall into the corridor , @xmath22 is the strength of the repulsive force and @xmath23 is the distance to the wall for pedestrian @xmath7 .
the word ` strong ' here implies that this force is not just a contact force , but has a longer range .
typically , this makes individuals avoid walls before touching them . for small
@xmath23 is the same as in the interactions between individuals ( cf .
( [ fsoc])([u_attr_rep ] ) ) , be it with different parameters . ] + + furthermore , very much in the spirit of @xcite , we specify the social force by @xmath24 where : * @xmath25 is the collection of the position vectors of all individuals which are within a distance @xmath26 to pedestrian @xmath7 .
in other words , pedestrians interact only when they are close enough to each other ; * we assume that the interaction potential @xmath27 depends only on the relative position of the two pedestrians @xmath7 and @xmath28 and not on their relative velocity .
( 0,0)(0,1)node[anchor = north east]@xmath29 ; ( 0,0)(1,0)node[anchor = north east]@xmath30 ; ( 0,0)(1,2)node[anchor = north east]@xmath31 ; ( 0,0)(3,4)node[anchor = north east ] ; ( 1,2)(3,4 ) node[midway , sloped , above]@xmath32 node[anchor = north west]@xmath33 ; ( 1,2)(3.5,2 ) node[midway , sloped , below]@xmath14 ; ( 2.0 , 2 ) arc ( 0:45:1.0 ) ; at ( 2.25,2.5 ) @xmath34 ; specifically , @xmath27 takes the form , @xmath35 here , @xmath36 $ ] is an _ anisotropy parameter _
, for which @xmath37 means that the potential is isotropic , and @xmath38 means that the anisotropy effects are maximal .
the angle of perception @xmath34 is the angle under which an individual positioned in @xmath31 , while moving in the direction of @xmath14 , perceives location @xmath33 , see figure [ figure geometry ] .
the precise type of interaction is hidden in the structure of the function @xmath39 . in particular , we distinguish between two types of interactions , namely * _ only repulsive interaction _ ( for simplicity denoted by ) .
pedestrians repel each other when their separation distance is smaller than @xmath40 ( called radius of repulsion in ) and do not interact at larger distances ; * _ both repulsive and attractive interaction _ ( denoted by ) . for this type of interaction , individuals repel each other when their separation distance is smaller than @xmath41 ( called repulsive radius in ) .
however , when they are separated by a distance between @xmath41 and @xmath42 ( called attractive radius in ar case ) , they are attracted to one another .
they do not interact outside these regions .
for these two cases @xmath43 , respectively @xmath44 .
the second case is an extension of the first one .
the we might regard as a population of individualistic people that simply try to avoid each other . in
the there is also some _ social cohesion _ , as they try to keep the group together . in the following ,
we describe the precise structure of our potentials .
in the repulsive case we take @xmath45 see figure [ prim_r ] for an example of an interaction potential of the above form . ) . ]
if we want to involve a both attractive and repulsive way of interactions , then we take @xmath46 where @xmath47 in the above , the constants @xmath48 and @xmath49 are such that @xmath39 is a continuous function .
figure [ prim_ar ] shows an example of this kind of interaction potentials . )
([u attr - rep parabola ] ) . ]
for all simulations , the number of pedestrians @xmath50 is an integer multiple of 10 .
initially , we place these individuals on a lattice of @xmath51 rows of 10 pedestrians each .
the distance between two succeeding rows is always @xmath52 .
note that there is an @xmath50 here , not @xmath53 : the pedestrians are thus distributed initially in a domain about one tenth the length of the total corridor ( see figure [ pos_0 ] ) .
the distance between two pedestrians in the same row is @xmath54 .
let @xmath55 denote the position of pedestrian @xmath56 , where @xmath57 and @xmath58 .
then the coordinates are @xmath59 and @xmath60 .
we use these initial conditions for _ all _ simulations .
+ + s. this initial configuration is used in all simulations .
the markers that indicate the individuals positions are smaller than in figures [ pos_r_00][pos_ar_10 ] , just to avoid overlap here . ]
the simulation time corresponds to a real life time @xmath61 s. this time is large enough to witness a stable profile for all simulations .
we give in table [ modpar ] a summary of the parameters and their values used in both the ` repulsive ' and ` attractive and repulsive ' interaction potentials .
m a _ corridor_. in our setting , the word _ corridor
_ refers to the ratio between length and width , rather than to the absolute values of @xmath2 and @xmath4 .
note that this ratio of the ` real ' corridor is in fact even bigger than @xmath62 , namely @xmath63 , cf .
the explanation in the beginning of section [ sec : level1 ] . ] + .model parameters , symbols , units and reference values .
we do simulations in two cases .
the : pedestrians interact in a both attractive and repulsive way and the : pedestrians only interact in a repulsive way .
nb : @xmath64 and @xmath65 relate to the morisita index and are specified in section [ sect : def morisita ] . +
[ cols="<,^ , > " , ] to investigate the influence of the initial conditions , we also calculate the morisita index corresponding to random initial conditions .
more details are given in table [ morisita random ic ] .
both in the and in the the mean morisita index over 10 random runs is significantly lower than the morisita index for regular initial conditions .
the difference is even much larger than the standard deviation .
+ however , there is consistency , in the sense that in both cases ( r and ar ) the values are lower .
the behaviour of a crowd of pedestrians inside a corridor , in which the individuals interact via an anisotropic way , can be distinguished clearly from the case in which pedestrians interact in a completely isotropic way .
in particular , we observe the following differences compared to the isotropic case : 1 .
the polarization index increases with increasing anisotropy ( i.e. increasing @xmath66 ) ; 2 .
the projected density along the @xmath30 axis shows a symmetric profile around the center of mass for the isotropic case .
however , increasing anisotropy implies loss of symmetry ; 3 .
the morisita index , as a measure of clustering , depends clearly on the anisotropy .
it increases ( with increasing anisotropy ) in the and , roughly speaking , decreases in the . 4 . in case of repulsive interactions ,
the crowd tends to fill the whole corridor .
if attraction is included , the group stays compact .
increasing @xmath66 however seems to diminish this kind of _ social cohesion _ as individuals do not look behind . as a result of our study , many new questions arose .
future research should be concentrated on the following three directions : * most obviously : what is the effect of a further increase of the number of pedestrians ?
do the observed relations between the measured quantities and the number of pedestrians still hold ? how do the observed limiting values of the polarization for large @xmath11 depend on @xmath50 ? what about the limit @xmath67 in this case ?
+ can we gain more insight in the strange issues of the ( cf .
figure [ niar ] : intersecting curves for @xmath68 and @xmath69 ) by further measurements of the morisita index for increasing @xmath50 ?
e.g. can we extrapolate information for @xmath70 back into the interval @xmath71 $ ] ?
some preliminary comments in this direction are given in [ appendix ] .
+ if @xmath50 increases , the natural thing to do is to consider the discrete - to - continuum limit ( i.e. construct educated procedures to derive mean - field limit equations ) .
does such limit exist , can we derive it , and can we compare the effect of anisotropy in the limit to the observations of the current work ? * how much does the large time behaviour of the crowd depend on the initial conditions ?
the initial distribution of pedestrians in this paper is not a realistic situation .
people starting to enter a corridor are in real life never distributed in a crystalline structure . however , for an escape situation ( for example in the case of fire ) it seems reasonable to assume that a group of people starts , being clustered , at one side of a corridor .
therefore , as an extension of this research , we propose to use as initial distribution a more realistic configuration in which people are placed at one side of the corridor , with their positions ( slightly ) perturbed from the grid points .
averaging over a large collection of such perturbed initial distributions , will lead to effective results . are these averaged results comparable to the ones presented in this paper ?
in other words : is averaging the results basically the same as removing the fluctuations from the initial conditions ?
+ in the paper we have included some preliminary results ( in the ends of sections [ sect : results polarization ] and [ sect : results morisita ] ) in this direction .
there we took the other extreme : random initial conditions over the whole corridor . *
what happens if we try to make our model more realistic : e.g. change the shape of the domain @xmath72 , or allow variation in the direction and magnitude of individuals desired velocity ? including more sophisticated active parts in the boundary ( doors ) or impermeable objects within the domain , automatically leads to questions about the efficiency of the flow ( such issues are also addressed e.g. in @xcite ) .
which geometry leads to the fastest evacuation ?
first steps in this direction have been made in @xcite .
the issues addressed in this paper show that anisotropy related to perception has nontrivial effects on the global dynamics of a crowd . certainly , these effects can not be neglected .
more work , both numerically and analytically , is needed to extend and formalize our results .
the authors thank c. storm , f. toschi , f. van de ven , h. wyss ( all with tu eindhoven ) and r. fetecu ( simon fraser univ .
canada ) for a series of fruitful discussions on the dynamics of self - propelled particles with anisotropic motion potentials .
they are also grateful to j. lega ( univ . of arizona usa ) for sharing her thoughts .
moreover , they thank the anonymous referees for their comments and suggestions for improvements .
je kindly acknowledges the financial support of the netherlands organisation for scientific research ( nwo ) , graduate programme 2010 .
let us start by saying that simulation of large numbers of individuals is beyond the scope of this paper .
our current implementation is inadequate for simulating system sizes one is used to in molecular dynamics . in our perspective this paper aims primarily at getting insight about what features to expect .
a second stage ( and follow - up paper ) is to optimize the implementation and increase the system size .
+ + looking ahead , we provide here some preliminary results for @xmath73 in the . in figures
[ p1000 ] and [ i 1000 ] we show the time - averaged polarization and the morisita index , respectively , as a function of @xmath50 . compared to figures [ npar ] and [ niar ] , the graphs have been continued by incorporation of the values at @xmath73 . note the logarithmic scale of the horizontal axes . + as a function of the number of pedestrians @xmath50 .
results in the , for several different values of @xmath66 at time @xmath74s .
this is an extension of figure [ niar ] including the value for @xmath73 with logarithmic scaling on the horizontal axis . ] as a function of the number of pedestrians @xmath50 .
results in the , for several different values of @xmath66 at time @xmath74s .
this is an extension of figure [ niar ] including the value for @xmath73 with logarithmic scaling on the horizontal axis . ]
note moreover that we have extended figures [ npar ] and [ niar ] by only one data point each .
the linear interpolation between the values at @xmath75 and @xmath73 is therefore probably not very meaningful .
what we are interested in , is the general trend .
+ + for the polarization , the ordering as a function of @xmath66 remains as we observed it .
what requires more investigation is the increasing trend : that is , increasing @xmath76 for increasing @xmath50 . in figure [ npar ]
the graphs seem to stop growing as @xmath50 goes towards @xmath77 .
an issue here might be that the time interval of 100 s is simply too short for larger @xmath50 . + + in the morisita plots we see a downward trend .
we see that the curve for @xmath68 remains beneath the other two , also for @xmath73 .
we can thus regard the absence of ordering of the curves for @xmath50 smaller than @xmath78 as an exception . +
a remark needs to be made about the fact that , at @xmath73 , the morisita index for @xmath79 is smaller than for @xmath69 .
the values are @xmath80 for @xmath68 , @xmath81 for @xmath69 and @xmath82 for @xmath79 .
the ordering of the curves therefore seems to be lost here also , be it that the difference is small compared to the magnitude of @xmath83 .
99 u. ascher and l. r. petzold , computer methods for ordinary differential equations and differential - algebraic equations , philadephia , siam , 1988 . | we consider a microscopic model ( a system of self - propelled particles ) to study the behaviour of a large group of pedestrians walking in a corridor .
our point of interest is the effect of anisotropic interactions on the global behaviour of the crowd .
the anisotropy we have in mind reflects the fact that people do not perceive ( i.e. see , hear , feel or smell ) their environment equally well in all directions .
the dynamics of the individuals in our model follow from a system of newton - like equations in the overdamped limit .
the instantaneous velocity is modelled in such a way that it accounts for the angle under which an individual perceives another individual . + we investigate the effects of this perception anisotropy by means of simulations , very much in the spirit of molecular dynamics .
we define a number of characteristic quantifiers ( including the polarization index and morisita index ) that serve as measures for e.g. organization and clustering , and we use these indices to investigate the influence of anisotropy on the global behaviour of the crowd .
the goal of the paper is to investigate the potentiality of this model ; extensive statistical analysis of simulation data , or reproducing any specific real - life situation are beyond its scope .
_ keywords _ : traffic and crowd dynamics , interacting agent models , self - propelled particles , pattern formation ( theory ) |
we study numerically the fourier transformed gp equation @xmath8\frac{\partial}{\partial t}\tilde{\phi}(\vec{k},t)=[k^2-\mu(t)]\tilde{\phi}(\vec{k},t)+\frac{g}{l^6}\sum_{\vec{k}_1,\vec{k}_2}\tilde{\phi}(\vec{k}_1,t)\tilde{\phi}^\ast(\vec{k}_2,t)\tilde{\phi}(\vec{k}-\vec{k}_1+\vec{k}_2,t)+\frac{1}{l^3}\sum_{\vec{k}_1}\tilde{v}(\vec{k}_1,t)\tilde{\phi}(\vec{k}-\vec{k}_1,t).\label{eq - fourier - gp}\ ] ] here @xmath9 is the fourier transformation of the macroscopic wave function @xmath10 $ ] , @xmath11 is the coupling constant , @xmath12 the system size .
the dissipation is given by @xmath13 which works only in the scales smaller than the healing length @xmath14 to remove compressible short - wavelength excitations .
the chemical potential @xmath15 depends on time to conserve the total number of particles @xmath16 .
the fourier transformation of the moving random potential @xmath17 is @xmath18 , which moves so that it may satisfy the two - point correlation @xmath19,\label{eq - random - potential}\ ] ] where @xmath20 , @xmath21 and @xmath22 are the characteristic strength and scales of space and time , respectively .
thus @xmath18 injectes the energy to the fluid at the scale @xmath21 to set the energy - containing range .
we numerically solve eq .
( [ eq - fourier - gp ] ) in a periodic box with spatial resolution containing @xmath23 grid points .
we consider the case of @xmath24 . as numerical parameters
, we use a spatial resolution @xmath25 and @xmath26 , where the length scale is normalized by the healing length @xmath27 .
the parameters of the moving random potential are @xmath28 , @xmath29 and @xmath30 .
the numerical integral is made by the pseudo - spectral method in space and the runge - kutta - verner method in time with a time resolution @xmath31 .
we start from an initial configuration in which both the density and the phase of the wave function are uniform .
the system gets to steady turbulence after @xmath32 when both the total energy @xmath3 and the incompressible kinetic energy @xmath33 , given by are almost constant in time . here
@xmath35 denotes the incompressible part @xmath36 of the vector fields . in the steady state
we can obtain the dissipation rate @xmath5 from the time derivative of the incompressible kinetic energy @xmath37 after switching off the moving random potential . in this steady turbulence
, we can clearly define the energy - containing range @xmath38 in which the system takes the energy from the moving random potential , the energy - dissipative range @xmath39 in which the energy are dissipated by @xmath40 , and the inertial range @xmath41 between these two ranges .
figure [ fig - steady ] ( a ) shows the energy spectrum of the incompressible kinetic energy @xmath42 defined as @xmath43 . in the inertial range ,
the numerically obtained spectrum is quantitatively consistent with the kolmogorov law using the above obtained @xmath5 .
dependence of the energy spectrum @xmath42 ( a ) and the energy flux @xmath44 ( b ) of the incompressible kinetic energy on the wave number @xmath4 .
solid lines refer to the kolmogorov law @xmath45 in ( a ) and the energy dissipation rate @xmath5 in ( b ) , respectively.,title="fig : " ] + ( a ) dependence of the energy spectrum @xmath42 ( a ) and the energy flux @xmath44 ( b ) of the incompressible kinetic energy on the wave number @xmath4 .
solid lines refer to the kolmogorov law @xmath45 in ( a ) and the energy dissipation rate @xmath5 in ( b ) , respectively.,title="fig : " ] + ( b ) figure [ fig - steady ] ( b ) shows the flux @xmath46 of the incompressible kinetic energy which is given by the energy budget equation of eq .
( [ eq - fourier - gp ] ) .
the energy flux @xmath46 is about independent of @xmath4 in the inertial range and close to the energy dissipation rate @xmath5 .
this result supports strongly the scenario of qt played by quantized vortices ; the energy - containing range , the inertial range and the energy - dissipating range are , respectively , made by their nucleation , their richardson cascade and their decay to compressible excitations , as believed in eddies of ct .
9 u. frisch , _ turbulence _ , ( cambridge university press , cambridge , 1995 ) .
m. kobayashi and m. tsubota , phys .
lett . * 94 * , 065302 ( 2005 ) .
j. maurer and p. tabeling , europhys .
lett . * 43 * , 2934 ( 1998 ) .
s. r. stalp , l. skrbek and r. j. donnelly , phys .
lett . * 82 * , 48314834 ( 1999 ) . c. nore , m. abid and m. e. brachet , phys .
lett . * 78 * , 38963899 ( 1997 ) ; _ phys .
fluids . _ ,
* 9 * , 26442669 ( 1997 ) .
t. araki , m. tsubota and s. k. nemirovskii , phys .
* 89 * , 145301 ( 2002 ) . | quantum turbulence is numerically studied by solving the gross - pitaevskii equation . introducing both the energy dissipation at small scales and the energy injection at large scales
, we succeed in obtaining the steady turbulence made by the balance of the injection and the dissipation of the energy .
the inertial range takes the kolmogorov spectrum and the energy flux is consistent with the energy dissipation rate at small scales .
these results confirm the properties of the inertial range of quantum turbulence , proposing a prototype of turbulence much simpler than conventional classical turbulence .
quantum turbulence ( qt ) consisting of quantized vortices has become a great subject not only for investigating the issues of superfluidity in the proper field of low temperature physics but also for studying a prototype much simpler than conventional classical turbulence ( ct ) .
the steady state of fully developed ct of an incompressible fluid takes the characteristic statistical law for the energy spectrum @xcite .
the energy , injected into the fluid at some large scales in the energy - containing range , is transferred in the inertial range to smaller scales without been dissipated . in the inertial range the energy spectrum takes the kolmogorov law @xmath0 here the energy spectrum @xmath1 is defined as @xmath2 , where @xmath3 is the kinetic energy per unit mass and @xmath4 is the wave number from the fourier transformation of the velocity field . the energy transferred to the energy - dissipative range
is dissipated by the viscosity with the dissipation rate @xmath5 of eq .
( [ eq - kolmogorov ] ) , which is equal to the energy flux @xmath6 in the inertial range .
the kolmogorov constant @xmath7 is a dimensionless parameter of order unity .
the inertial range is believed to be sustained by the self - similar richardson cascade process , in which large eddies are broken up to smaller eddies through reconnections .
however , the richardson cascade has never been confirmed clearly in ct , because it is impossible to definitely identify each eddy .
in contrast , quantized vortices are definite and stable topological defects with the quantized circulation , consequently qt may give us a prototype of turbulence much simpler than ct @xcite . some experimental @xcite or theoretical works
@xcite studied the kolmogorov law for qt , but it was very difficult to control the energy dissipation rate @xmath5 .
recently we made the numerical simulation of the gross - pitaevskii ( gp ) equation which describes a quantum fluid at zero temperature @xcite .
here we introduced a dissipative term that works only in the smaller scales than the healing length , in order to remove compressible short - wavelength excitations created through vortex reconnections which may hinder the cascade process .
we thus obtained the kolmogorov spectrum for decaying turbulence starting from the random configuration of the velocity field . in this work
, we also inject energy into the fluid at large scales by moving the random potential , thus obtaining steady turbulence sustained by the balance of the injection and dissipation of the energy .
the energy spectrum of the incompressible kinetic energy still takes the kolmogorov law , and the energy flux @xmath6 is almost independent of the scales in the inertial range and consistent with the energy dissipation rate in the energy - dissipative range .
these results make clear the properties of the inertial range of qt . |
among spiral galaxies , one of the most extensively observed object is undoubtedly the system of m51 ( ngc 5194 ) and its companion ngc 5195 . the first galaxy in which spiral structure was observed since 1845 , this system has been studied in every e.m .
wavelengths with ground based telescopes , balloon borne and space instrumentation .
several studies in morphology , spectroscopy , photometry , and polarimetry has been made in order to define the dynamic , the chemical constitution in stars , dust and gas , and the origin and evolution of the spectacular spiral pattern .
the system was first suspected to be gravitationally interacting by zwicky ( 1953 ) , but the confirm was due to the pioneering works of toomre & toomre ( 1972 ) .
the numerical simulations of these authors , starting from an undifferentiated disk , reproduced the formation of outer arms , tails and bridge structures , demonstrating a gravitational interaction with the companion .
the system has been investigated in all his components , neutral gas ( hi ) , ionized gas ( hii ) , dust , and stars , and during the last years the studies went on also using balloon borne instrumentation to investigate the far uv , and hubble space telescope to investigate with the highest resolution the fine central spiral structure ( panagia et al .
1995 ) . as a result of this large effort the most part of the characteristics of the system have been explained , but even if we have a general convincing physical model on the origin and dynamic of this structure , more accurate simulations and observations disclosed , as usual , some discrepancies .
in fact , the self - gravitating numerical simulations of toomre ( 1978 ) failed to reproduce the inner regions of this system , as well as the efforts by byrd & howard ( 1989 ) , that obtained a model in which they correctly derived the origin of the two large outer material arms from a gravitational encounter in progress since 70 millions of years , but they were not able to reproduce the inner density wave structure .
observations on inner structure from zaritsky et al .
( 1993 ) , using near ir ( k band , 2.2@xmath0 ) photometry in order to overcome the dust and gas reddening and absorption , indicate that the data are consistent with a coherent presence of the two arms winding through about three revolutions toward the galactic centre . according to the authors these results
are not consistent with the current theories of the spiral structures . on the other hand , using balloon borne detectors , to detect the far uv ( @xmath12000 ) , petit et al .
( 1996 ) , found different distributions between the evolved , hot , intermediate massive stars ( 2@xmath25 m@xmath3 ) , and the very massive stars coexisting with the ionized gas ( hii regions ) .
this scenario is compatible with the wave density triggered star formation , and then it s in agreement with the theory of the unperturbed spiral structure . [ h ] the hypothesis that density waves and material arms due to tidal interaction can coexist in the same system is already present in a paper of sundelius ( 1989 ) , in which , with a self - gravitating numerical simulation , in the late evolution of the interacting system appear both the density waves and the material arms .
the idea that in the m51 system the outer structure coming from the gravitational interaction , and the inner structure can be distinct is present also in elmgreen ( 1989 ) , and in howard & byrd ( 1990 ) .
the hypothesis of elmgreen is that the inner density waves were triggered from the tidal waves due to the encounter with the companion .
howard & byrd on the contrary , with the results of a numerical self - gravitating simulation suggested that the large outer hi regions detected in the radio from rots et al .
( 1990 ) can be originated by a precedent encounter of m51 with the same companion .
similar conclusions are in rix & rieke ( 1993 ) which affirm that some morphological characteristics revealed with photometric observations in the near ir ( k band ) , can be due to the interference between the inner spiral structure and the material tidal arms stimulated by the encounter . in the paper , the authors suppose that the inner structure was pre - existing to the encounter . having in mind these dilemma
, our team attempts to investigate the main arms structures , in order to detect features and/or anomalies in the trend of the density waves which could help in finding some answers to the questions described .
the results presented in this paper have been carried out from b , v , and i broad bands observations of the m51 system . how we cited in the introduction ,
several authors used , as star formation tracer , near ir photometric bands , to avoid extinction and reddening due to dust and gas , and to detect a single star population . ;
b ) @xmath4 ; c ) @xmath5 ; d ) @xmath6 .
the angles are measured starting from the west point counteclockwise ] ; f ) @xmath7 ; g ) @xmath8 ; h ) @xmath9 .
the ascissa are in arcsec from the center of the galaxy . ] ; b ) @xmath10 ; c ) @xmath11 ; d ) @xmath12 .
the ascissa are in arcsec from the center of the galaxy . ]
since our target was the investigation of the density of the supermassive , just born , short living stars connected with the higher density regions of the galactic materials , across the spiral arms , a color index tracing the temperatures seemed us more fitting the goal .
the images were obtained with a ccd thomson ( 384x576)x23@xmath0 uv coated at the prime focus of the 60x90x180 cm schmidt telescope of the campo imperatore observatory , since 1990/92 , as part of a more general observational program about the radial luminosities and color gradients in face - on nearby spiral galaxies ( pompei & natali 1997 ) .
the m51 system was included in the sample , in order to detect differences with respect to the non - interacting galaxies . due to the scale factor of the telescope ( 2.64 arcsec / pixel ) ,
the ccd single frame ( 17x25 arcmin ) , detected both m51 and companion , including all parts of the outer arms as well as the sky background . in the cited paper
the analysis of the images we carried out , was the traditional one used in the studies of radial luminosities and color gradients of galaxies .
the * @xmath0(r ) * derived from the medians of the luminosities * @xmath0(r,@xmath13 ) * , were computed over * all the azimuthal angles @xmath13*. the median , as well known , enhances the disk , fading out the characteristics of the spiral arms .
the results about the m51 system did nt point out any specific feature with respect to the non - interacting galaxies .
only the radial color trends of this galaxy showed more noise with respect to the signal . in the new measurements , on the contrary , the target being to enhance the spiral arms special features , we tried to use as tracer of the star formation loci , the ( b - i ) vs r , averaged for * small * @xmath14 * azimuthal angles * and computed for different * @xmath13*. the preprocessing of the frames i.e. flatting - field , residual gradient subtraction , cosmic rays removing , has been the same as in the previous study .
the specific use of the ( b - i ) color index as tracer , has been suggested from the results of a study about the characteristic of this index , which , close correlated to the temperature of the object like the more used ( b - v ) , presents minor relative errors with respect to the ( b - v ) itself ( natali el al .
regarding our target , this tool has been an important factor to reveal features otherwise hidden in the noise . for each photometric band
we computed the surface mean ( not the median ) luminosity * @xmath0(r ) * covering only a * sector * of the galaxy with an angle @xmath15 ; fig.1 . in fig.2 and fig.3
we report four radial color indices for different azimuthal angles selected in the east part , and in the west part respectively of the galaxy .
in addition , to have a check , we derived the same sector measurements from ngc 1232 and ngc 3184 , two face - on , not - interacting , spiral galaxies .
relative images and radial color indices are in fig.4 , fig.5 , fig.6 and fig.7 .
as discussed in the previous section , we will consider the color indices as density waves tracer for the galactic _
materia_. the most evident feature striking in the curves of fig.2 and fig.3 , are the different structures of the color indices and then of the density waves in the east part of the galaxy , close to the probable orbit of the companion , with respect to the west part . in the east part ,
the color indices show in almost all the analyzed azimuthal angles a characteristic saw - toothed structure .
in fact , the trends of the density wave going toward the inner regions rise slownly , ( smaller b - i means younger stars , and then higher star formation rate , and then higher material density ) , dropping after the maxima very quickly .
this feature is more evident for the @xmath16 azimuthal angle ( fig.1d ) . for this angle
the saw - toothed shape of the density wave is clearly present as far as the inner spiral arm .
moreover , the saw - toothed shape is present also in the outer arm where the density wave , and the sfr , is more faint with respect to the central one .
the same characteristics of the radial color index are present also with different angle averages i.e. for * @xmath17 * , * @xmath18 * , and * @xmath19*. this characteristic is not evident in the diaphram west - east photoelectric scans of m51 reported by burkhead ( 1978 ) . about the color index variations , they are stronger in the east part , ( @xmath20 ) , with respect to the west part , ( @xmath21 ) . ; b ) @xmath22 ; c ) @xmath6 ; d ) @xmath23 .
the ascissa are in arcsec from the center of the galaxy . ]
a difference between the east and west part , about the sfr intensity in m51 , was detected with far uv , balloon borne instrumentation early in the 1990 by bohlin et et al .
( 1990 ) . in the west part ,
when it is possible to extract the spiral arms signal from the noise , ( @xmath9 , or @xmath24 ) the radial color index do not show different rise and fall trends ( fig.3 ) .
we detected a similar behaviour in the _ normal _ spiral galaxies ngc 1232 and ngc 3184 ( fig.4 and fig.6 ) , where the radial color indices present small symmetrical variations in correspondence of the relative spiral arms ( fig.5 and fig.7 ) .
these results suggest that the saw - toothed morphology we detect in the east part of m51 can be due to the tidal forces triggered by the encounter with the companion .
the asymmetrical shape of the density wave is consistent with a scenario of a progressive wave propagating in a medium in which the phase velocity is modulated by a pre - existing quasi - steady density wave .
in fact , crossing the higher density regions , the front of the progressive wave will decrease his phase velocity generating the observed asymmetries .
as reported in the introduction the m51 system presents some characteristics which remain unexplained .
several authors derived from their measurements and/or numerical simulations , the suspect that the companion is responsible only in part of the observed scenario . in many papers
it appears the hypothesis that a quasi - steady inner density wave is coexisting with a material tidally induced spiral pattern .
the results presented in this paper show a peculiar density shape in the east part of m51 clearly connected with the gravitational perturbation of the companion , not revealed in the west part of the galaxy or in similar not - perturbed spiral galaxies .
the saw - toothed shape of the material arm density in the east part of m51 , is similar to that of a progressive wave propagating in a medium with a preexisting oscillating density .
the variations of the medium density in fact , will produce an oscillating group velocity , generating then , the observed shapes in the progressive wave . in this case it should have to assume the whirlpool was a spiral galaxy before the encounter .
we think that our conclusions are not definitive and even if , it is very simple and exciting , it raises new striking questions .
in fact , if the galaxy was a spiral before the encounter , the numerical simulations of the system must start from a pre - existing spiral pattern and not from an undifferentiated disk . and then , can the spiral survive to the disrupting tidal forces ?
if so , which are the new impact parameters as the time - scale of the encounter , the minimum distance of the barycentres , the orbits and the initial velocities and masses of the two objects ? what about on other m51-type spirals ?
do they present the same kind of density structure in their arms ?
if so , the progressive wave hypothesis could be strengthened , but only further observational data can solve the new question .
rots a.h.,bosma a. , van der hults j.m.,athanassoula e. and crane p.c.,1990 , `` dynamical status of m51 '' in dynamics and interactions of galaxies .
proceedings of a conference held in heidelberg . r.wichen ed . | using the b - i color index in the radial photometry of m51 , applied to defined azimuthal sectors , we reveal asymmetries in the shape of the density pattern of the galactic arms . the anomalous shape is present only in the side close to the companion and not in the far side .
two more non - interacting check - spirals do not show similar behaviour .
the anomalous sawtooth - like shape is similar to that produced by a progressive wave propagating in a preexisting density pattern . in this hypothesis
we have to conclude that the galaxy m51 could be a spiral galaxy formerly , i.e. before the encounter with the companion . |
the colour strings approach @xcite is widely applied for the description of the soft part of the hadronic and nuclear interactions at high energies . in the frame work of this approach
the string fusion model was suggested in papers @xcite .
later it was developed @xcite-@xcite and applied for the description of the long - range multiplicity and @xmath0 correlations in relativistic nuclear collisions @xcite-@xcite . in the paper
@xcite we have formulated some simple cellular analog of the string fusion model , which enables explicit analytical calculations of the correlation functions in some asymptotic cases and can simplify calculations in the case of real nuclear collisions .
in that paper we have checked up the assumptions of the cellular approach and the validity of a suggested gauss approximation in the simplest ( no fusion ) case when the explicit solution of the model can be found . in the present paper in the framework of proposed cellular approach
we calculate @xmath0@xmath1 and @xmath1@xmath1 correlation functions and correlation coefficients in two cases : with local and with global string fusion .
the calculations are done both numerically and in some asymptotic cases analytically using the gauss approximation .
the results of both calculations are in a good agreement , which proofs the validity of proposed gauss approximation . at large string density @xmath2 the connection between the @xmath0@xmath1 and @xmath1@xmath1 correlation coefficients
are found in both cases : with local or global string fusion . at that for the correlation coefficient the scaling takes place .
it depends only on one combination @xmath3 of the variables @xmath4 and @xmath5 ( @xmath5 is the mean multiplicity emitting by a single string ) .
the paper organized as follows . in the next section we recall the formulation of the cellular approach in the case with a _
local _ string fusion . in the section 3
we develop gauss approximation at large string density @xmath4 in the _ local _ fusion case . in the section 4
the @xmath0@xmath1 and @xmath1@xmath1 correlation coefficients are calculated at large @xmath4 in the case with _ local _ string fusion and connection between these two coefficients is found . in the section 5
the @xmath0@xmath1 and @xmath1@xmath1 correlation functions and correlation coefficients are calculated at large @xmath4 for `` homogeneous '' situation ( a constant mean density of strings ) in the case with _
local _ fusion .
it s shown that in this situation the @xmath0@xmath1 and @xmath1@xmath1 correlation coefficients become equal and the @xmath3-scaling takes place .
the obtained results are compared with the results of numerical calculations using formulas of section 2 . in the section 6 we recall the formulation of the cellular approach in the case with a _ global _ string fusion .
the exact closed formulas for the @xmath0@xmath1 and @xmath1@xmath1 correlation functions in this case are obtained . in the section 7
we develop gauss approximation at large string density @xmath4 for the _ global _ fusion case .
the @xmath0@xmath1 and @xmath1@xmath1 correlation functions and coefficients for this case are calculated and compared with the results of numerical calculations using formulas of section 6 .
the behavior of the correlations at small string density is studied in appendix a.
let us recall the formulas obtained for this case in @xcite .
we consider the collision of nuclei in two stage scenario when at first stage the colour strings are formed , and at the second stage these strings , or some other ( higher colour ) strings formed due to fusion of primary strings , are decaying , emitting observed particles . in principle
, one can consider two types of fusion .
the case with a local fusion corresponds to the model , where colour fields are summing up only locally and the global fusion case corresponds to the model , where colour fields are summing up globally - all over the cluster area - into one average colour field , the last case corresponds to the summing of the sources colour charges .
( in section 5 of @xcite we have referred to these cases as a ) and b ) correspondingly . ) in the transverse plane depending on the impact parameter @xmath6 we have some interaction area @xmath7 .
let us split this area on the cells of order of the transverse string size .
then we ll have @xmath8 cells , where @xmath9 is the transverse area of the string and @xmath10 is the string radius . in the case with a local fusion
the assumption of the model is that if the number of strings belonging to the @xmath11-th cell is @xmath12 , then they form higher colour string , which emits in average @xmath13 particles with mean @xmath14 equal to @xmath15 , compared with @xmath16 particles with @xmath17 emitting by a single string .
if we denote by @xmath18 and @xmath19 - the number and the average number of particles emitted by the higher string from @xmath11-th cell in a given rapidity interval , then _
i=_0 [ fus1 ] from event to event the number of strings @xmath12 in @xmath11-th cell will fluctuate around some average value - @xmath20 .
clear that in the case of real nuclear collisions these average values @xmath20 will be different for different cells .
they will depend on the position ( @xmath21 ) of the @xmath11-th cell in the interaction area ( @xmath21 is two dimensional vector in transverse plane ) . to get the physical answer we have to sum the contributions from different cells , which corresponds to integration over @xmath21 in transverse plane .
the average local density of primary strings @xmath20 in the point @xmath21 of transverse plane is uniquely determined by the distributions of nuclear densities and the value of the impact parameter - @xmath6 .
they can be calculated , for example , in glauber approximation .
we ll do this later in a separate paper . in present paper
we consider that all @xmath20 are already fixed from these considerations at given value of the impact parameter - @xmath6 .
let us introduce some quantities , which will play important role in our consideration : @xmath22 r=_i=1^m , = _ i=1^m [ not ] then clear that @xmath23 is the number of strings in the given event and @xmath24 is the mean number of strings for this type of events ( at the fixed impact parameter @xmath6 ) . to go to long - range rapidity correlations we have to consider two rapidity windows @xmath25 ( forward ) and @xmath26 ( backward ) .
each event corresponds to a certain configuration @xmath27 of strings and certain numbers of charged particles @xmath28 emitted by these strings in the forward rapidity window .
then the total number of particles produced in the forward rapidity window will be equal to @xmath29 : n_f=_i=1^m n_i [ nf ] the probability to detect @xmath29 particles in the forward rapidity window for a given configuration @xmath27 of strings is equal to p^_\{^_1, ... ,^_m } ( ) = _
\{n^_1, ... ,n^_m } _ , _ i n_i _ i=1^m p^__i ( n_i ) [ pnf ] where @xmath30 is the probability of the emission of @xmath18 particles by the string @xmath12 in the forward rapidity window . by our assumption ( [ fus1 ] ) _ i_n_i=0^ n_i p^__i ( n_i ) = _ 0 [ fusion ] if we denote else by @xmath31 the probability of realization of the string configuration @xmath27 in the given event , then the average value of some quantity @xmath32 under condition of the production of @xmath29 particles in the forward window will be equal to = [ avo ] one has to omit in this @xmath33-fold sums one term , when all @xmath34 , which corresponds to the absence of inelastic interaction between the nucleons of the colliding nuclei ( see details in appendix a ) .
if the @xmath32 in the number of particles produced in the backward rapidity window @xmath35 in the given event , then we have to use for @xmath36 correlations : = _ 0 _ i=1^m = _ 0 r [ onn ] if the @xmath32 in the mean squared transverse momentum of particles produced in the backward rapidity window @xmath37 in the given event , then we have to use for @xmath38 correlations : = _
i=1^m p^2 = p^2 = p^2 [ optn ] later we ll assume that numbers of primary strings in each cell @xmath12 fluctuate independently around some average quantities @xmath20 uniquely determined by the distributions of nuclear densities and the value of the impact parameter - @xmath6 ( see above ) , then w ( ) = _ i=1^m w ( _ i ) , _
i w ( _ i ) = [ wceta ] for clearness we ll sometimes address to a simple `` homogeneous '' case , when all @xmath39 ( but not the @xmath12 , which fluctuate from event to event ! ) is equal each other in the interaction area @xmath40 ( a constant mean density of strings ) .
the parameter @xmath4 coincides in this case with the parameter @xmath4 used in the papers @xcite and has the meaning of the mean number of strings per area of one string ( @xmath41 ) . in general case
the parameters @xmath39 have the same meaning , but with mean string density depending on the point @xmath42 in the transverse interaction plane + ( @xmath43 ) . as it was shown in @xcite
if we assume else the poissonian form of @xmath30 ( @xmath44 is the poisson distribution with @xmath45 ) : p^__i ( n_i ) = ^__0(n_i ) e^-_0 [ poisson ] then we find the poissonian distribution for p^_\{^_1, ...
,^_m } ( ) = ^__0 ( ) [ poisso ] with @xmath46 and @xmath47 , which we can replace at large @xmath48 by gauss distribution : p^_\{^_1, ... ,^_m } ( ) = e^- [ pgausa ] it was also shown in the section 6 of @xcite that if we assume the binomial form of @xmath30 then we find the binomial distribution for @xmath49 with @xmath46 and @xmath50 , where @xmath51 corresponds to the poisson limit and @xmath52 corresponds to the case when each higher string emits fixed number of particles - @xmath53 in each event .
moreover appealing to the central limit theorem of the probability theory one can state that at large @xmath33 we ll have formula ( [ pgausa ] ) for any type of @xmath30 .
now we ll evaluate @xmath54 at the condition that all @xmath55 . at this condition
we can use gauss approximation for each @xmath56 w ( _ i ) = e^- [ wgausa ] with @xmath57 , where again @xmath58 corresponds to the poisson limit and @xmath59 corresponds to the case with a fixed number of strings @xmath60 ( see the section 6 of @xcite ) . as in @xcite
we have then = _ 0 [ bf1a ] and = p^2 [ pbf1a ] recall that @xmath61 and @xmath62 . here
( _ i,)=_i=1^m + [ phia ] further we take out factors before exponent at the point , where @xmath63 is minimal , after that the rest integrals in the numerator and in the denominator are being reduced , and we find = _ 0 r^ * [ bf2a ] and = p^2 [ pbf2a ] the @xmath64 and @xmath65 are the values of @xmath23 and @xmath66 in the point @xmath67 , where @xmath68 is minimal : = 0 [ dphia ] this leads to the system of equations : -1= ( -1 ) [ etistara ] recall that @xmath69 and = [ kappa ] for the meaning of @xmath70 see the end of the previous section and the section 6 of @xcite , for both poissonian distributions @xmath71 and @xmath70 is the relative width of the @xmath72 and @xmath73 distributions in other cases . the ( [ etistara ] ) defines @xmath74 as function of @xmath75 .
introducing short notations z_i= , f== , a_i= [ nota ] we can rewrite ( [ etistara ] ) as z_i^3-z_i^=a_i ( f^2 -1 ) [ zia ] where @xmath76 , @xmath77 and @xmath78
. the ( [ zia ] ) defines @xmath79 as function of @xmath80 : @xmath81 .
then we can calculate @xmath82 and @xmath83 as a function of @xmath75 using ( [ bf2a ] ) and ( [ pbf2a ] ) .
the correlation coefficients are defined in the same way as in the section 4 of @xcite : b [ ba ] and [ betaa ] or for `` relative '' quantities [ bba ] and [ bbetaa ] ( note the same definition of @xmath84@xmath1 correlation coefficient @xmath85 in @xcite ( see formula ( 44 ) in @xcite ) , see also remark in appendix b ) . in short notation using ( [ bf2a ] ) and ( [ pbf2a ] ) we have : = [ bb1a ] and = [ bbeta1a ] we ca nt solve the equations ( [ zia ] ) for to find @xmath81 explicitly , but to calculate the correlation coefficients we need to know only @xmath86 , which can be done explicitly .
we see that at @xmath87 ( [ zia ] ) has the obvious solution : f=1 , z_i=1 , ^*_i= , r^*= , n^*=[solzia ] we need to calculate @xmath88 only at @xmath87 . differentiating ( [ zia ] ) on @xmath80 and using then again ( [ solzia ] ) we find z_i(1)=a_i [ solzi1a ] with @xmath89
. then = = _
i=1^m z_i(1 ) = [ bb2a ] and = - = - [ bbeta2a ] using = 2_i=1^m z_i(1)= [ dndfa ] we have = ( -1 ) = ( -1 ) [ bbeta3a ] we see the connection between @xmath84@xmath1 and @xmath1@xmath1 correlation coefficients .
note that due to obvious inequality : ( _ i=1^m ) ^2 m _ i=1^m [ ineqv ] we have @xmath90 and hence always @xmath91 .
clear that at equal @xmath92 we have @xmath93 , @xmath94 , @xmath95 and = = [ bbb3a ] from ( [ bb2a ] ) we see that @xmath1@xmath1 correlation coefficient is always positive .
can @xmath84@xmath1 correlation coefficient be negative ?
let us consider _ nonhomogeneous _ situation when @xmath96 at @xmath97 , @xmath98 at @xmath99 , @xmath100 and @xmath101 .
then @xmath102 , @xmath103 and we have ( 2 - 1 ) [ bbeta4a ] we see that at @xmath104 we can have @xmath105 .
let us consider for clearness _ homogeneous _ case , when all @xmath39 is equal each other in the interaction area @xmath40 ( a constant mean density of strings ) .
in this case we can explicitly calculate at large string density @xmath4 not only the @xmath0@xmath1 and @xmath1@xmath1 correlation coefficients , but also the corresponding correlation functions for the version with a local string fusion .
we have seen in the end of previous section that in this case the coefficients for @xmath1@xmath1 and @xmath84@xmath1 correlations defined as ( [ bba ] ) and ( [ bbetaa ] ) are coincide : = = = [ bb3a ] where @xmath107 . in this homogenous case
@xmath92 we can also calculate the correlation functions @xmath38 and @xmath36 at any @xmath75 .
due to symmetry , the system of equations ( [ zia ] ) have symmetrical solution @xmath108 and can be reduced to one equation z^3-z = a ( -1 ) [ za ] because @xmath109 , @xmath110 and @xmath111 with f== , a= [ not1a ] the ( [ za ] ) defines the function @xmath112 and then using ( [ bf2a ] ) and ( [ pbf2a ] ) we can calculate = _ 0 r^*= _ 0 m z(f)= z(f ) [ bf3a ] and = p^2 = p^2 z(f)= z(f ) [ pbf3a ] so we have at any @xmath113 : = = z(f ) [ pna ] from ( [ bb3a ] ) and ( [ za ] ) we see that in this homogenous case at large string density @xmath4 there is an remarkable _ scaling_. the @xmath84@xmath1 and @xmath1@xmath1 correlation coefficients and correlation functions depend only on one combination @xmath107 of parameters .
we present the correlation function @xmath114 ( [ pna ] ) in figs .
[ zf1],[zf2 ] and the correlation coefficient @xmath115 ( [ bb3a ] ) as function of @xmath4 in figs .
[ b1]-[b4 ] ( the solid lines ) .
we present also in figs .
[ b1]-[b4 ] the results of our direct numerical mc calculations of the @xmath84@xmath1 and @xmath1@xmath1 correlation coefficients in the local fusion case based on formulas ( [ avo]-[optn ] ) ( empty and filled points correspondingly ) .
we see that in the case with local fusion at small string density we have large @xmath1@xmath1 correlations ( the same as in the case without string fusion @xcite ) and no @xmath0@xmath1 correlations .
( the analysis at very small values of @xmath116 see in appendix a. ) at large string density in the homogeneous case the @xmath0@xmath1 and @xmath1@xmath1 correlation coefficients become equal and the @xmath3-scaling takes place .
we see also that in this limit our gauss asymptotic is in a good agreement with results of the numerical calculations and @xmath33-independence takes place .
in this case at first stage we also have @xmath8 cells ( like in the case with local fusion ) with @xmath12 , @xmath117 fluctuated around @xmath39 . then ( unlike the local fusion case ) we have to find average @xmath118 for given event , and then to generate particles from one cluster with average multiplicity equal to @xmath119 @xmath120 @xmath121 @xmath122 .
the general formulae for this case was obtained in the section 5 of @xcite : = [ avocl ] where @xmath123 with @xmath8 .
the assumption @xmath124 is essential , as only in this situation we can consider that the transverse area of the cluster @xmath125 is equal to all interaction area @xmath7 ( @xmath6-impact parameter ) . the @xmath126 is the rates of the backward production from configuration @xmath127 .
we have to use for @xmath36 correlations : = _ c = _ 0 m [ onncl ] and for @xmath38 correlations : = p^2 = p^2 [ optncl ] we see that the difference with the case of local fusion consists in replace @xmath128 . as a consequence calculations in the case of global string fusion
are much more simple , as we can reduce all sums @xmath54 to one sum @xmath129 , as in the no fusion case ( see section 3 in @xcite ) .
so in the global fusion case we can write simple formulas : = [ bfcl ] and = [ pbfcl ] where @xmath130 is given by the formula w(n ) = _ \{^_1,
... ,^_m } _ n , _
i w ( _ i ) [ wn ] we see that in the case of global fusion ( one cluster at large @xmath39 with area @xmath125 being equal to all interaction area @xmath7 ) @xmath1@xmath1 and @xmath84@xmath1 correlations are connected = = [ pnb ] note that unlike the local fusion case in this case we find this result without any assumptions on the properties of @xmath131 and @xmath73 distributions and for arbitrary ( even nonequal ) @xmath39 .
clear that in this case the results can depend only on mean number of strings @xmath24 and on combination @xmath132 : @xmath133 below we ll calculate numerically the correlation functions on formulas ( [ bfcl ] ) and ( [ pbfcl ] ) , but at first we would like to find explicit formulas for global fusion case in gauss approximation .
we ll see that results really depend only on one combination of the variables ( [ mum ] ) , namely on @xmath134 and the @xmath135 takes place as in the case with local fusion .
acting as in the no fusion case in section 4 of @xcite ( see also calculations in the gauss approximation in the section 3 of the present paper ) we find = _ m = _ 0 [ bf2cl ] or keeping in mind ( [ pnb ] ) = = [ bf3cl ] the @xmath64 is the value of @xmath23 at which the function ( n,)=+ [ phicl ] gains its minimum . in short notations z , f== , a= = [ notcl ] we find the equation z^3-z = a ( -1 ) [ zcl ] which defines the function @xmath112 and then using ( [ bf3cl ] ) we can calculate correlation functions = = z()=z(f ) [ bf4cl ] and correlation coefficients for the global fusion case = = z(1)= = = [ b4cl ] we see again that in gauss approximation there is the same remarkable @xmath135 .
the @xmath84@xmath1 and @xmath1@xmath1 correlations depend only on one combination of parameters : @xmath136
. note that unlike the local fusion case in this case we find this result for arbitrary ( even not equal ) @xmath39 . in the homogeneous situation
all @xmath40 and we have = m , a= [ not1cl ] and = = = [ b5cl ] we see that in the @xmath137 situation in gauss approximation the results with @xmath138 and @xmath139 fusion coincide ( as we have expected in section 5 of @xcite ) .
we have the same equations ( [ za ] ) and ( [ zcl ] ) with the same value of parameter @xmath140 ( [ not1a ] ) and ( [ not1cl ] ) .
note that in the no fusion case we have had _ very different _
equation for @xmath114 ( see section 4 of @xcite ) . unlike the local fusion case in the global fusion case we can control the validity of gauss approximation making calculations on exact formulas ( [ bfcl ] ) and ( [ pbfcl ] ) at different values of @xmath33 . along with
the correlation function @xmath114 ( [ pna ] ) , ( [ bf4cl ] ) in figs .
[ zf1],[zf2 ] and the correlation coefficient @xmath115 ( [ bb3a]),([b4cl ] ) in figs .
[ b1]-[b4 ] , calculated on our scaling formulas , which in gauss approximation are the same for local and global fusion ( the solid lines ) , we present at the same pictures the results of exact calculations in the global fusion case on formulas ( [ bfcl ] ) and ( [ pbfcl ] ) at different values of @xmath33 ( the dotted and dashed lines ) .
we present also in figs .
[ b1]-[b4 ] the results of our direct numerical mc calculations of the @xmath84@xmath1 and @xmath1@xmath1 correlation coefficients in the global fusion case based on formulas ( [ avocl]-[optncl ] ) ( half - filled squares ) .
we see that in this case the gauss approximation works very well and the @xmath3-scaling is not an artifact of this approximation .
more over along with the @xmath3-scaling at large @xmath4 we have also @xmath33-independence for correlation coefficients @xmath85 and @xmath141 ( see figs . [ b1]-[b4 ] ) starting very early ( from @xmath142 ) .
in conclusion let us compare the results obtained in the present paper in the case with string fusion with results obtained in @xcite in the case without string fusion .
we have obtained in the paper @xcite in the case * without * string fusion : 1 . for @xmath1@xmath1 correlations : @xmath143 with @xmath144 2 . for @xmath84@xmath1 correlations : @xmath145 in the present paper in the case with the * global * string fusion and in the * local * fusion case for * homogeneous * situation ( @xmath40 ) we find at large @xmath4 : 1 . for @xmath1@xmath1 correlations : @xmath143 with @xmath146 2 . for @xmath84@xmath1 correlations : @xmath147 with the same @xmath140 we see that with fusion the @xmath1@xmath1 correlations became weaker , but now as compensation we have the @xmath84@xmath1 correlations of the same strength .
we see also @xmath3-scaling in this case . for *
nonhomogeneous * situation ( different @xmath39 ) in the case with * local * string fusion we have find at large @xmath39 : 1 . for @xmath1@xmath1 correlations
: @xmath143 with @xmath148 2 . for @xmath84@xmath1 correlations : @xmath149 and hence @xmath91 @xmath150 as we have demonstrated above ( see ( [ bbeta2a ] ) and ( [ bbeta3a ] ) ) , this leads to @xmath85 smaller than @xmath141 : @xmath91 .
it s possible situation ( [ bbeta4a ] ) , in which @xmath151 . at * small * string density as it s shown in appendix a
the two types of limit at @xmath152 can be studied . 1 .
if one keeps @xmath153 , then we have @xmath154 ( because the configurations with @xmath155 , are not considered as events ) and we have nor @xmath0@xmath1 nor @xmath1@xmath1 correlation .
if one keeps @xmath156 , then @xmath157 and @xmath158 , hence the strings will be far separated in transverse plane and we ll have the same results as in the case without string fusion @xcite .
note that the results obtained in our cellular approach are in a good agreement with the results obtained in the framework of the real string fusion model taking into account detail geometry of strings overlapping @xcite .
the authors would like to thank m.a .
braun and g.a .
feofilov for numerous valuable encouraging discussions .
the work has been partially supported by the russian foundation for fundamental research under grant no .
01 - 02 - 17137-a .
in this appendix we calculate the correlation functions and the correlation coefficients at small string density in the case with local string fusion .
( in the case with global fusion this limit has no physical sense ( see discussion in the section 5 of @xcite ) ) . for clearness
we consider `` homogeneous '' case , when all @xmath39 is equal each other in the interaction area @xmath40 .
then in each cell @xmath11 ( @xmath117 ) the @xmath12 fluctuate around this mean value according to poisson law ( @xmath44 is the poisson distribution with @xmath45 ) : w(_i)=_^(_i)e^- [ ap : w ] we ll assume also the poissonian form of @xmath30 : p^__i ( n_i ) = ^__0(n_i ) e^-_0 [ ap : poisson ] then we have the poissonian distribution for p^_\{^_1, ... ,^_m } ( ) = ^__0 ( ) [ ap : poisso ] with @xmath159 .
then we find from ( [ avo ] ) for @xmath1@xmath1 correlations : n_b^_n_f^^= [ ap : nn ] and for @xmath84@xmath1 correlations : p_tb^2 _ n_f^^= [ ap : ptn ] recall that @xmath160 the @xmath23 is the number of strings in the given event .
one has to omit in all @xmath33-fold sums in ( [ ap : nn ] ) and ( [ ap : ptn ] ) one term , when all @xmath34 , which corresponds to the absence of inelastic interactions between the nucleons of the colliding nuclei ( we denote this fact by @xmath161 ) . the probability @xmath162 to detect @xmath29 particles in the forward rapidity window , which enters denominators of ( [ ap : nn ] ) and ( [ ap : ptn ] ) , is p(n_f)=c_\{_1^, ... ,_m^}^ ( _ i w(_i ) ) _ _ 0^r^(n_f^ ) [ ap : pnf ] where from normalization condition we have c== [ ap : c ] clear that this factor @xmath163 is canceling in the numerator and in the denominator of ( [ ap : nn ] ) and ( [ ap : ptn ] ) , but if we calculate the mean number of strings @xmath24 , we find = c_\{_1^, ... ,_m^}^ ( _ i w(_i ) ) n = c_\{_1^, ... ,_m^}^ ( _ i w(_i ) ) ( _ i _ i ) = [ ap : avn ] and for the @xmath164 at small @xmath165 we have = _
n^_f n^_f p(n^_f)= c^_0f_\{_1^, ... ,_m^}^ ( _ i w(_i ) ) r = ^_0f= ^_0f[ap : avf ] because for any @xmath166 we have @xmath167 at @xmath152 , as the main contribution comes from the term @xmath168 .
there are two possibilities when @xmath152 : 1 .
@xmath153 and then @xmath169 and @xmath170 ( see ( [ ap : avn ] ) ) 2 .
@xmath156 and then @xmath157 ( ( see ( [ ap : avn ] ) ) and @xmath158 we ll investigate both these possibilities . the first one means that in the limit we have @xmath171 ( because the configurations with @xmath155 , are not considered as events ) .
clear that in this situation we ll have nor @xmath0@xmath1 nor @xmath1@xmath1 correlation , as we ll have no fluctuations in the number of strings ( see discussion in the end of the section 4 of @xcite ) .
detail calculations are presented below . in the second case with @xmath156
we ll have the fluctuations in the number of strings @xmath23 , but in the limit @xmath152 we ll have @xmath158 , the strings will be far separated in transverse plane and the strings fusion will plays no role .
so in this case we ll have the same results as in the no fusion case , considered in @xcite : large @xmath1@xmath1 correlation with a correlation coefficient equal to @xmath172 and no @xmath0@xmath1 correlation .
( see calculations below . )
* detail calculations . *
let us evaluate the @xmath33-fold sums in ( [ ap : nn ] ) and ( [ ap : ptn ] ) at @xmath152 keeping all terms of order @xmath173 , @xmath174 with any @xmath175 and omitting all terms of order @xmath176 and higher .
the terms of order @xmath173 originates from the summands in ( [ ap : nn ] ) and ( [ ap : ptn ] ) with @xmath177 and other @xmath178 .
the terms of order @xmath174 originates from the summands in ( [ ap : nn ] ) and ( [ ap : ptn ] ) with @xmath179 , @xmath180 and other @xmath178 . keeping this into mind
we find for @xmath162 : p(n_f)=c(g_0 + g_1 ) [ ap : pnf1 ] and for @xmath1@xmath1 correlations : n_b^_n_f^^=^_0b [ ap : nn1 ] for @xmath84@xmath1 correlations we have : p_tb^2 _ n_f^^=p^2 = p^2 ( 1 + ) [ ap : ptn1 ] here g_0=p_0=_k=1^m c_m^k ^k ^__0 k(n^_f ) [ ap : coeff1 ] @xmath181 @xmath182 @xmath183 @xmath184 where @xmath185 and @xmath186 . note that at @xmath187 and @xmath157 we have for @xmath162 : p(n_f)=c e^-m ( _ 0 + e_^-^_0 _ 1 ) [ ap : pnf2 ] and for @xmath1@xmath1 correlations : n_b^_n_f^^=^_0b [ ap : nn2 ] for @xmath84@xmath1 correlations we have : p_tb^2 _ n_f^^=p^2 = p^2 ( 1 + e_^-^_0 ) [ ap : ptn2 ] here _ 0=_0=_k=1^ k_^n^_f [ ap : coeff2 ] @xmath188 @xmath189 @xmath190 @xmath191 where @xmath192 .
for the control of calculations we have used also the following explicit formulas at @xmath193 : _ 0(0)=_0(0)=e^d-1 [ ap : coeff3 ] @xmath194
@xmath195 @xmath196 @xmath197 @xmath198 @xmath199 @xmath200 @xmath201 @xmath202 the results of the calculations using the asymptotic formulas ( [ ap : nn1][ap : coeff1 ] ) in the first case ( @xmath153 , @xmath152 , @xmath169 and @xmath170 ) are shown in the figs .
[ sm1]-[sm2 ] ( lines ) together with results of direct mc numerical calculations using formulas ( [ ap : nn ] ) and ( [ ap : ptn])(points ) .
as we have expected in this case both @xmath1@xmath1 and @xmath0@xmath1 correlation coefficients go to zero when @xmath203 , i.e. at @xmath204
. remember that we have @xmath33-independence for the correlation coefficients at large @xmath4 .
now we see that in this limit it disappears at @xmath205 .
this is also the reason for nonlinear dependence of the correlation coefficients on @xmath4 in this region which one can see in figs .
[ sm1]-[sm2 ] for @xmath206 .
we see also that using asymptotic formulas ( [ ap : nn1][ap : coeff1 ] ) we can calculate @xmath1@xmath1 correlation coefficients in wider region of small @xmath4 , than @xmath0@xmath1 correlation coefficients , because we have the contributions of order @xmath173 and @xmath174 for @xmath1@xmath1 correlations and only first non - trivial contribution of order @xmath174 for @xmath0@xmath1 correlations .
note the very good agreement between results of the calculations on the asymptotic formulas ( [ ap : nn1][ap : coeff1 ] ) ( lines in the figs .
[ sm1]-[sm2 ] ) and the results of direct mc numerical calculations on formulas ( [ ap : nn ] ) and ( [ ap : ptn ] ) ( points in the figs .
[ sm1]-[sm2 ] ) . in the second case when @xmath152 we keep @xmath156 and then due to ( [ ap : avn ] ) @xmath157 so @xmath158 .
we can use formulas ( [ ap : nn2][ap : coeff2 ] ) with @xmath207 , then in the limit @xmath152 we find for @xmath1@xmath1 correlations : n_b^_n_f^^= ^_0b = ^_0b [ ap : nn3 ] here we multiply both the numerator and the denominator by @xmath208 .
recall that @xmath44 is the poisson distribution with @xmath45 , then we see that formula ( [ ap : nn3 ] ) coincides with the formula for @xmath1@xmath1 correlations obtained in the paper @xcite in the case _ without _ string fusion ( see formula ( 23 ) in @xcite ) .
so we ll have the same result for the @xmath1@xmath1 correlation coefficient @xmath172 as in the no fusion case ( see figs .
[ sm1]-[sm2 ] ) . for @xmath0@xmath1 correlation coefficients in this limit
we find from ( [ ap : ptn2 ] ) and ( [ ap : coeff2 ] ) with @xmath207 : p_tb^2_n_f^^= p^2 ( 1 + o ( ) ) [ ap : ptn3 ] so we have no @xmath0@xmath1 correlation as in the case without string fusion @xcite .
it s possible instead of ( [ bba ] ) and ( [ bbetaa ] ) to use the following definitions for the correlation coefficients : [ a1 ] and [ a2 ] where = [ a3 ] and = [ a4 ] clear that = _ ( ) _ p ( ) = [ a5 ] and the same for @xmath211 . | the long - range @xmath0 and multiplicity(@xmath1 ) correlations in high - energy nuclear collisions are studied in the framework of a simple cellular analog of the string fusion model .
two cases with local and global string fusion is considered .
the @xmath0@xmath1 and @xmath1@xmath1 correlation functions and correlation coefficients are calculated analytically in some asymptotic cases using suggested gauss approximation .
it s shown that at large string density the @xmath0@xmath1 and @xmath1@xmath1 correlation coefficients are connected and the scaling takes place .
the behavior of the correlations at small string density is also studied .
the asymptotic results are compared with results of the numerical calculations in the framework of proposed cellular approach .
-2.5 cm # 1 .
|^_#1 # 1 # 1 # 1#2 ^_#2 # 1 # 1#2#3#4#1 * # 2 * , # 4 ( # 3 ) |
supernova remnants ( snrs ) with a center - filled or composite morphology often host young pulsars whose strong relativistic winds can lead to the formation of pulsar wind nebulae ( pwne ) . in recent years
the high spatial resolution of _ chandra _ has made it possible to detect the pwne , to pinpoint the locations of the compact objects , and to constrain their nature . a new , yet unproven , case
in which a putative pulsar is suspected to power a pwn is g18.951.1 . for this composite snr , which is considered to be located in the sagittarius arm at a distance of @xmath22kpc @xcite ,
an x - ray luminosity of @xmath17erg s@xmath7 ( 0.110kev ) has been reported by @xcite .
g18.951.1 was discovered as a shell - like non - thermal extended radio source @xcite .
further observations carried out at ghz frequencies by @xcite , @xcite , and @xcite detected an elongated synchrotron nebula in the center of g18.951.1 and a region of enhanced radio emission at its tip , much like the central pwn in w44 ( e.g. , * ? ? ?
another region of enhanced radio emission is located along the western shell of g18.951.1 .
the age of the remnant and its radio diameter are considered to be ( 4.4@xmath18yr @xcite and @xmath233@xmath19 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) which is equivalent to @xmath218@xmath8pc , where @xmath8 is the distance in units of 2kpc . in a previous x - ray study carried out with _ rosat _ and _ asca _
, @xcite mapped the diffuse x - ray emission of g18.951.1 . a reproduction of the _ asca _ image in the 0.54kev energy band is shown in fig .
[ f1]a together with radio contours at 10.55ghz @xcite .
the radio emission consists of a central elongated radio - bright nebula ( @xmath20 erg s@xmath7 , * ? ? ?
* ) running from the nw to the se and another region of enhanced radio emission located at the western shell of g18.951.1 .
@xcite also detected a localized region of hard x - ray emission in the 410kev energy band whose peak emission is centered at ra=18@xmath2129@xmath22033 , dec=@xmath2352589 ( see also their fig . 2 for its location ) .
this hard x - ray emission is slightly offset from the region of enhanced radio synchrotron emission , similar to g326.31.8 @xcite .
harrus et al . found that the x - ray emission from g18.951.1 is predominantly thermal , heavily absorbed with a column density of about @xmath3@xmath4 , and can be best described by a non - equilibrium ionization ( nei ) model with a temperature of @xmath20.9kev and an ionization timescale of @xmath24@xmath25s@xmath7 . in a reanalysis of the _ asca _ data of the localized hard x - ray emission
, we confirmed the existence of a non - thermal component , which seems to dominate the spectrum above 2kev .
although _ asca _ s spatial resolution is insufficient to detect a point source within this region , it is possible that the hard x - ray emission indicates the location of an undetected pulsar .
we have learned from _
chandra _
observations that many young pulsars , like those in 3c 58 and g21.5 - 0.9 as well as currently undetected pulsars , e.g. , the one in g327.1 - 1.1 , are embedded in extended non - thermal emission from the pwne . in the case of g18.951.1
the extended emission is most likely also contaminated by diffuse thermal emission from the snr , but any non - thermal emission from a possible point source remained unresolved by _
asca_. with _
chandra _ we now have the opportunity to detect a point source in this extended emission .
therefore , we conducted a follow - up _ chandra _ observation of g18.95 - 1.1 in order to search for an x - ray point source embedded in the extended emission , to determine its nature , and to characterize the spatial and spectral distribution of the hard emission to test if it is consistent with an x - ray synchrotron nebula .
a 45ks _ chandra _ observation was executed on august 3rd , 2009 ( obsid 10098 ) using acis - i in vfaint mode with the aimpoint of the acis - i array centered onto the region in which the 410kev emission from _ asca _ overlapped the radio synchrotron emission ( see fig .
data reduction was carried out with ciao4.2 and caldb version 4.2.2 , following standard procedures , except for the charge transfer inefficiency ( cti ) correction .
as during the last 11ks of the observation the focal plane temperature was warmer than -118.7c ( increasing from -119.0c to -118.4c ) , we applied a time - dependent cti correction following @xcite .
the effects of this correction are marginal in the 0.358.0kev energy band as it adjusted the gain on average by @xmath20.5% to account for the increased cti at higher temperatures .
10.55ghz radio continuum contours @xcite overlaid onto a _
_ counts image ( bin=8 , 0.352.6kev , point source - subtracted ) .
the inlay shows the unbinned counts image of the point source in the full band and a saotrace simulation of the psf ( 90% encircled energy ) .
both images are displayed on a square root scale and plotted between 0 and 12 countspix@xmath7 .
spectral extraction regions are shown as black ellipses.,width=283,height=279 ]
in fig . [ f1]b we show the high - resolution _ chandra _ x - ray composite image of the inner part of g18.951.1 .
it reveals a region of hard x - ray emission at ra=18@xmath2129@xmath22131 , dec=@xmath2351134 ( @xmath26=@xmath2758584 , @xmath28=@xmath2959591 ) , which is shown in greater detail in fig .
this hard emission seems to originate from a point - like source and its immediate vicinity ( its extent corresponds to region e1 highlighted in fig .
the point - like source is located at the northern tip of a trail of softer diffuse x - ray emission extending towards the se and pointing back towards the center of the remnant .
the extended emission is well aligned with the radio synchrotron nebula , but does not appear to cover the full extent of the latter ( see red contours in fig .
[ f2 ] ) . comparing the radial profile of this source with
the one from an saotrace simulation , confirmed that this source is indeed a point source .
this object , cxouj182913.1 - 125113 hereafter , is the only x - ray point source detected within a circle of 14 radius centered on the position of the source .
cxouj182913.1 - 125113 has no known radio , 2mass , optical , or @xmath30-ray counterpart .
the brightest point source in the field ( the bright white object near the southern edge of the fov in fig . [ f1]b ) was classified by @xcite as a late - type star and appears not to be associated with g18.951.1 .
there are other sources in the field which could be pulsar candidates , notably the pair of sources near ra=18@xmath2128@xmath22360 , dec=@xmath2351008 ( @xmath26=@xmath2754574 , @xmath28=@xmath2951520 ) . however , these sources have soft x - ray spectra as indicated by their colors in fig .
[ f1]b , are @xmath31 away from the diffuse nebula , and can be associcated with foreground stars in the usno - b1.0 catalog ( monet et al .
the other point - like objects in the field which appear on or outside the edge of the fov are most likely artifacts of the smoothing in regions with low and/or variable exposure .
in addition , faint enhancements in the diffuse emission or in the background might appear to be point - like after smoothing .
although the searches for pulsations in the radio by @xcite proved unsuccessful ( their closest search position is @xmath2@xmath32 away from cxouj182913.1 - 125113 ) , the coincidence of the x - ray and radio synchrotron emission as well as the cometary - shaped morphology of the diffuse x - ray emission , with a hard x - ray point source at the tip , seems to point at a pulsar and its wind nebula . to test
if the point source and the extended emission are consistent with this interpretation , we extracted spectra for the putative pulsar and from three extended regions to the south of this source , using the ciao tool specextract . for the putative pulsar we extracted all counts within a radius of @xmath33 , while the three elliptical extraction regions shown in fig .
[ f2 ] were used for the extended emission from which only the point source was excluded .
two additional spectra were extracted which cover different fractions of the extended x - ray emission .
they contain all counts in the e2@xmath34e1 and e3@xmath34e2 regions .
this approach was chosen because we wanted to see how the spectra and the x - ray luminosity change as a function of increasing distance from the pulsar candidate .
background for the pulsar candidate was taken from an annular region centered on the position of this source and from regions close to the acis chip edges for the extended x - ray emission .
all spectra were grouped with dmgroup until at least a signal - to - noise ratio of 3 ( 5 ) was achieved for the pulsar candidate ( extended sources ) .
top : best - fitted powerlaw model for the point source ( ps ) .
bottom : spectra extracted from regions e1 to e3 can be best fitted by a combination of a @xmath35 model and a powerlaw.,width=283,height=453 ] ccccccccccc e1 & 0.81/44 & 9.21e-01 & 4.74e-04 & 2.13e-05 & 1.97@xmath36 & 0.48 ( frozen ) & 1.4@xmath37 & 3.72@xmath38 & 0.60@xmath39 + e2 & 0.71/47 & 9.11e-01 & 4.53e-04 & 1.62e-04 & 1.56@xmath40 & 0.48 ( frozen ) & 1.9@xmath41 & 9.36@xmath42 & 1.76@xmath43 + e3 & 0.91/35 & 6.22e-01 & 7.31e-03 & 2.45e-04 & 1.33@xmath44 & 0.48@xmath45 & 1.8@xmath46 & 54.4@xmath47 & 4.40@xmath48 + + e2e1 & 0.82/47 & 8.04e-01 & 1.20e-03 & 7.23e-05 & 1.54@xmath49 & 0.48 ( frozen ) & 1.7@xmath50 & 9.68@xmath51 & 1.28@xmath52 & + e3e2 & 1.59/22 & 3.90e-02 & 5.35e-03 & 1.31e-04 & 1.30@xmath53 & 0.46@xmath54 & 1.9@xmath55 & 37.5@xmath56 & 2.12@xmath57 & + [ t1 ] the spectrum labeled ps in fig . [ f4 ] shows the x - ray emission from the point source .
it contains 168 net counts and was fitted in xspec ( v12.6.0 ) with an absorbed powerlaw model and alternatively with an absorbed thermal plasma model . to simulate the absorption the @xmath58 model @xcite was used while the thermal emission was modeled with the @xmath59 model @xcite .
the @xmath1 was allowed to vary , but was initially set to the galactic value of @xmath60@xmath4 @xcite .
the best - fitted model ( @xmath61 ) is the absorbed powerlaw with a photon index of @xmath62 and an @xmath63@xmath4 ( all uncertainties are given for a 90% confidence level ) .
the predicted @xmath1 is in good agreement with previous measurements from @xcite and @xcite .
although the thermal plasma model provides an equally good fit , the predicted @xmath64 of 16kev is unreasonably high and therefore makes a thermal nature of the source very unlikely .
the unabsorbed x - ray luminosities predicted by the powerlaw model in the 0.58.0kev and 2.08.0kev energy bands amount to @xmath65@xmath6 erg s@xmath7 and @xmath66@xmath6 erg s@xmath7 , respectively .
the powerlaw index of 1.6 is consistent with the pulsar hypothesis and the derived luminosity is towards the low end of the distribution of pulsar luminosities .
this may not be surprising since the brighter objects have already been discovered . under the simplistic assumption that the stellar remnant traverses the distance from the geometric center of the snr ( to be assumed at ra=18@xmath2129@xmath22239 , dec=@xmath2359051 or equivalently @xmath26=@xmath2753136 , @xmath28=@xmath6705577 ) to its current location ( 4.3pc @xmath6 ) at a constant speed within 4400 to 6100 years
, transverse velocities between 700kms@xmath7 and 960kms@xmath7 are estimated .
although these values are significantly higher than the typical ones of about 500kms@xmath7 @xcite , they are not unusual for pulsars with periods significantly larger than several milliseconds @xcite . a pulsar would be consistent with the derived spectral index from the powerlaw fit and the displacement between the radio synchrotron nebula and the x - ray point source .
the spectra of the purely diffuse x - ray emission extracted from regions e1 , e2 , and e3 are shown in the lower panel of fig .
[ f4 ] together with the best - fitted model ( @xmath68 ) . fitting
the spectra with a single vnei model @xcite resulted in @xmath69 values @xmath702 , while a vnei model with an additional component ( e.g. , @xmath71 ) produced ionization timescales of @xmath72s@xmath25 which indicate that the plasma is in collisional ionization equilibrium .
the spectrum labeled e3 covers the bulk of the extended x - ray emission near the center of g18.951.1 .
it has 8975 net counts and provides evidence of emission around 1.35kev .
the model predicts a thermal plasma temperature of 0.48kev and a photon index of @xmath73=1.8 .
the best - fit results are shown in table [ t1 ] and are given for a 90% confidence level . in case of e3 , we freed the magnesium abundance to remove significant residuals around this line , yielding a relative mg abundance of 1.4@xmath74 .
this ratio is consistent with the galactic value and agrees well with that reported by @xcite .
the x - ray luminosities for e3 provide evidence that the gas is predominantly thermal , as 87% of @xmath12 is emitted by the thermal plasma component in the 0.58kev energy band .
the thermal emission could originate from the snr s shell seen along the line - of - sight or from the pwn if it has been disturbed by the reverse shock so that thermal emission and non - thermal emission are intertwined . for e3 , as well as for e1 and e2 ,
the fitted hi column density is consistent with the average galactic value of @xmath75@xmath4 .
a column that high seems to imply a distance much larger than 2kpc .
however , as @xcite and @xcite argue , the alternative distance of 15kpc suggested by @xcite appears unlikely , as the derived radio luminosity at this distance would be twice that of the crab nebula , and the swept up mass of about 3000@xmath76 and the explosion energy of @xmath77erg are unreasonably high . even if we determine a distance based on the best - fit , assuming @xmath78@xmath4 and an hi volume density of @xmath79@xmath25 @xcite , yielding a distance of about 11kpc , the explosion energy of @xmath80erg as well as the swept up mass of @xmath81@xmath82m@xmath83 are still unreasonably high .
we therefore adopt a distance of 2kpc for g18.951.1 and quote our results as a function of distance .
the line emission seen in e3 remains undetected in the spectra for regions e2 and e1 ( 1685 and 435 net counts , respectively ) , but the bumps at about 1.87kev indicate the presence of emission and therefore a partly thermal origin of the gas .
the photon index does not steepen significantly from e1 to e2 and the emission remains soft . for e1 73% and for e2 51% of the x - ray luminosity
is produced by the thermal plasma component .
fitting the spectra e1 , e2 , and e3 with only an absorbed powerlaw generally produces significantly worse fits ( @xmath84 ) with photon indices of @xmath73=1.9 ( e1 ) , 2.2 , ( e2 ) , and 3.9 ( e3 ) , respectively .
moreover , the potential emission features make a pure non - thermal origin of the x - ray emission appear very unlikely . because the pulsar and the pwn candidate are embedded within g18.951.1 , the non - thermal emission from these objects is expected to be highly contaminated by the thermal emission from the snr .
we therefore adopt the temperature obtained from spectrum e3 to be representative for the inner part of g18.951.1 and freeze the @xmath64 for e1 , e2 , and @xmath85 , respectively ( see table [ t1 ] ) .
this approach seemed to be justified , because spectra extracted from diffuse emission north of the pulsar candidate also suggested the @xmath64 to range from about 0.40 to 0.55kev with relatively large uncertainties of 0.15kev and values which are consistent with those listed in table [ t1 ] .
for the extended x - ray emission to be consistent with a pwn , several criteria have to be met ( see e.g. * ? ? ?
* ; * ? ? ?
first , the x - ray surface brightness should decrease with increasing distance from the point source .
second , the radio spectral index @xmath86 should be relatively flat ( @xmath87 , @xmath88 ) .
third , the photon index @xmath73 should be steeper at x - ray energies than in the radio .
as a consequence , the extent of the x - ray emission can be smaller than the extent of the radio emission , as the synchrotron lifetime of the electrons in the x - ray regime can be shorter than those at radio wavelengths . to test these criteria and to model how much the spectra change with increasing distance from the pulsar
, we discuss spectra extracted from regions e1 , @xmath85 , and @xmath89 and focus on the hard , non - thermal x - ray emission emitted above 2kev , which we solely attribute to the putative pulsar and the pwn . as can be seen from table [ t1 ] , the fraction of the x - ray luminosity above 2kev decreases from 16% ( e1 ) , to 13% ( @xmath85 ) , down to 6% ( @xmath89 ) . if we take the hard x - ray luminosities and divide by the area of the extraction regions , we get surface brightnesses which decrease from @xmath90@xmath6ergs@xmath7arcsec@xmath91 for e1 , to @xmath92@xmath6ergs@xmath7arcsec@xmath91 for @xmath85 , down to @xmath93@xmath6ergs@xmath7 arcsec@xmath91 for @xmath89 .
hence , the first criterion is fulfilled . for the inner region of g18.951.1 ,
flat radio spectral indices of @xmath94 @xcite and @xmath95 @xcite have been reported . as the radio emission
is also polarized in this region , @xcite report an integrated polarization percentage of about 2.5% at 4.75ghz for the whole remnant , the second criterion is met , too .
the average photon index derived from the x - ray spectral fits is @xmath96 , significantly steeper than the radio index , thereby satisfying the third criterion .
there is no significant steepening towards the edge of the nebula ( see column 8 , table [ t1 ] ) .
we also fitted the spectra with the fixed to the value of e3 , but the uncertainties were still too large to say with any confidence that there is a clear trend in photon index .
furthermore , the major half - axis diameter of the x - ray emission in region e3 is about @xmath97 ( @xmath814pc@xmath8 ) , which is much less than the extent of the radio synchrotron contours at 10.55ghz in fig .
[ f2 ] indicate .
we conclude , that the extended hard x - ray emission originates from a pwn with a photon index between @xmath10 and an unabsorbed luminosity of @xmath98@xmath6 erg s@xmath7 in the 28kev energy band .
we can now begin to estimate some of the fundamental properties of the putative pulsar in g18.95 - 1.1 . using the relationship between the spin - down energy loss rate ( @xmath99 ) and the non - thermal x - ray luminosity of the pulsar and the pwn @xcite of @xmath100@xmath6 erg s@xmath7
, we estimate @xmath99 to be @xmath101 erg s@xmath7 .
following @xcite , adopting an average age of g18.951.1 of @xmath102yr ( see * ? ? ?
* ) , a canonical moment of inertia of the neutron star of @xmath103g@xmath104 , and assuming further a constant braking index of 3 , that spin down happens via magnetic dipole radiation , and that the current value of the pulsar s spin period is much larger than its initial one , the current spin period ( @xmath15 ) , the period derivative ( @xmath16 ) , and the surface magnetic field of the pulsar is estimated to be @xmath105s , @xmath106ss@xmath7 , and @xmath107 g , respectively .
these parameters can only serve as rough estimates as they are derived from the relation between @xmath12 and @xmath99 which can vary by about one order of magnitude ( see e.g. , * ? ? ?
it should also be kept in mind that @xmath108 gives the field strength at the magnetic equator and that the field strength at the magnetic poles can be a factor of 2 higher .
if the inferred values for the period , period derivative , and magnetic field are correct , this pulsar candidate occupies a highly interesting and sparsely populated place in the @xmath109 diagram .
it would be located close to the line of the quantum critical field of @xmath110 g ( see e.g. , * ? ? ?
* ) , which divides normal radio pulsars and the more exotic radio - quiet objects , such as the anomalous x - ray pulsars ( axp ) and soft gamma - ray repeaters ( sgrs ) .
we report the detection of a faint x - ray point source , cxouj182913.1 - 125113 , in the inner part of the composite snr g18.951.1 and a trail of diffuse emission pointing towards the center of the snr .
these sources are considered to be the pulsar and its wind nebula .
the best fit to the spectrum of the pulsar candidate is an absorbed powerlaw model ( @xmath73=1.6 , = @xmath3@xmath4 ) with an unabsorbed luminosity in the 0.58kev energy band of @xmath12@xmath111@xmath112@xmath6 erg s@xmath7 .
the x - ray luminosity of the pwn ( @xmath10 , = @xmath113@xmath4 ) amounts to @xmath12@xmath111@xmath114@xmath6 erg s@xmath7 in the 0.58kev energy band .
it appears as if the spectrum extracted from the region which contains the pwn shows thermal emission from the snr along the line - of - sight or the pwn has been disturbed by the reverse shock so that thermal and non - thermal emission is mixed .
although pulsations have not been reported for the pulsar , we estimate the spin period and the period derivative to be @xmath15=@xmath115s and @xmath16=@xmath116ss@xmath7 , assuming the dipole approximation .
compared to other rotation - powered pulsars , a high magnetic field of @xmath107 g is implied by its location in the @xmath15@xmath16 diagram .
this value is close to that of the quantum critical field and makes this source a highly interesting object for follow - up observations .
16 borkowski , k. j. , lyerly , w. j. , & reynolds , s. p. 2001
, , 548 , 820 dickey , j. m. , & lockman , f. j. 1990 , , 28 , 215 frst , e. , reich , w. , reich , p. , sofue , y. , & handa , t. 1985 , , 314 , 720 frst , e. , hummel , e. , reich , w. , sofue , y. , sieber , w. , reif , k. , & dettmar , r .- j .
1989 , , 209 , 361 frst , e. , reich , w. , & aschenbach , b. 1997 , , 319 , 655 gaensler , b. m. , & slane , p. o. 2006 , , 44 , 17 harrus , i. m. , slane , p. o. , hughes , j. p. , & plucinsky , p. p. 2004
, , 603 , 152 kondratiev , v. i. , mclaughlin , m. a. , lorimer , d. r. , burgay , m. , possenti , a. , turolla , r. , popov , s. b. , & zane , s. 2009 , , 702 , 692 monet , d. g. , et al .
2003 , , 125 , 984 nakanishi , h. , & sofue , y. 2003 , , 55 , 191 odegard , n. 1986 , , 92 , 1372 patnaik , a. r. , velusamy , t. , & venugopal , v. r. 1988 , , 332 , 136 petre , r. , kuntz , k. d. , & shelton , r. l. 2002 , , 579 , 404 plucinsky , p. p. , dickel , j. r. , slane , p. o. , edgar , r. j. , gaetz , t. j. , & smith , r. k. 2002 , aps meeting abstracts , 17037 possenti , a. , cerutti , r. , colpi , m. , & mereghetti , s. 2002 , , 387 , 993 posson - brown , j. , grant , c. , allen , g. , plucinsky , p. p. , & edgar , r. j. 2010 , http://cxc.harvard.edu/twiki/bin/view/aciscal/ + tcti_howto reich , w. , frst , e. , haslam , c. g. t. , steffen , p. , & reif , k. 1984 , , 58 , 197 safi - harb , s. 2004 , boletin de la asociacion argentina de astronomia la plata argentina , 47 , 277 seward , f. d. , & wang , z .- r .
1988 , , 332 , 199 smith , r. k. , brickhouse , n. s. , liedahl , d. a. , & raymond , j. c. 2001 , , 556 , l91 toscano , m. , sandhu , j. s. , bailes , m. , manchester , r. n. , britton , m. c. , kulkarni , s. r. , anderson , s. b. , & stappers , b. w. 1999 , , 307 , 925 wilms , j. , allen , a. , & mccray , r. 2000 , , 542 , 914 | using the _ chandra _ x - ray observatory , we have pinpointed the location of a faint x - ray point source ( cxouj182913.1 - 125113 ) and an associated diffuse nebula in the composite supernova remnant g18.95 - 1.1 .
these objects appear to be the long - sought pulsar and its wind nebula .
the x - ray spectrum of the point source is best described by an absorbed powerlaw model with @xmath0 and an @xmath1 of @xmath2@xmath3@xmath4 .
this model predicts a relatively low unabsorbed x - ray luminosity of about @xmath5@xmath6erg s@xmath7 , where @xmath8 is the distance in units of 2kpc .
the best - fitted model of the diffuse nebula is a combination of thermal ( @xmath9kev ) and non - thermal ( @xmath10 ) emission .
the unabsorbed x - ray luminosity of @xmath11@xmath6erg s@xmath7 in the 0.58kev energy band seems to be largely dominated by the thermal component from the snr , providing 87% of @xmath12 in this band .
no radio or x - ray pulsations have been reported for cxouj182913.1 - 125113 .
if we assume an age of @xmath25300yr for g18.95 - 1.1 and use the x - ray luminosity for the pulsar and the wind nebula together with the relationship between spin - down luminosity ( via magnetic dipole radiation ) and period , we estimate the pulsar s period to be @xmath13s . compared to other rotation - powered pulsars , a magnetic field of @xmath14 g
is implied by its location in the @xmath15@xmath16 diagram , a value which is close to that of the quantum critical field . |
as various technologies for submillimeter ( submm ) wave observations have advanced , the prospects for very long baseline interferometry ( vlbi ) at submm wavelengths have become a reality .
operations at shorter wavelengths improve proportionately the spatial resolution as compared with current centimeter and millimeter vlbi observations ; we anticipate that it will be possible to resolve astronomical sources in greater details , by a factor of 10 .
the demand for better angular resolution is quite strong , especially for the direct imaging of nearby supermassive black holes ( smbhs ) ; sagittarius a@xmath7 ( sgr a@xmath7 ) , which is located at the center of our galaxy and therefore the nearest smbh , has been imaged at various wavelengths using the vlbi technique .
the observed size of sgr a@xmath7 is obviously affected by the interstellar scattering at 3 mm or longer wavelengths , following the @xmath8 scattering law ( e.g. , * ? ? ?
* ; * ? ? ?
this effect , however , lessens at shorter wavelengths , and the size is observed to deviate from the @xmath8 scattering law at 1.3 mm @xcite .
this has strongly motivated submm - vlbi observations toward nearby smbhs to resolve and image emission from their vicinities .
upper limits of the intrinsic sizes of smbhs have been measured so far for sgr a@xmath7 @xcite and the nucleus of m87 @xcite ; upper limits to the sizes of both sources are about @xmath9arcsec .
these results indicate that much longer baselines and/or higher frequencies are needed to resolve and image the smbhs .
we , therefore , started to look for a new site to perform submm - vlbi observations with substantially longer baselines than before .
for the site selection , we set criteria as follows : 1 . annual precipitable water vapor ( pwv ) of less than 3 mm for low submm opacity .
longest possible baselines with existing submm telescopes , for obtaining the highest angular resolution ( this also means that we do not consider sites that already have submm telescopes ) .
3 . overlapping sky coverage with the atacama large millimeter / submillimeter array ( alma ) to achieve the highest possible sensitivity .
accessibility to the site for maintenance and operations .
we checked the satellite - based pwv data measured by the moderate resolution imaging spectroradiometer ( modis ) on the nasa aqua and terra satellites for potential sites with respect to the locations of the available submm telescopes .
we found three potential broad regions of interest ; western china and tibet , the highest mountains of southern alaska , and the high arctic polar desert , including northern canada and greenland .
the western china and tibet regions do not have common sky with alma , so they do not meet the criterion ( 3 ) .
the tallest peaks in alaska ( e.g. , denali , or former official name mount mckinley ) are protected or otherwise inaccessible , so they do not meet the criterion ( 4 ) .
the summit of the greenland ice sheet , on the other hand , has low pwv conditions throughout the year , has a common sky coverage with alma , will create the longest baseline length of about 9000 km for the submm - vlbi , and already has a research facility , summit camp , which is operated by ch2 m hill .
polar services ( cps ) for the u.s . national science foundation ( nsf ) .
the summit camp , therefore , meets all four criteria .
the greenland summit camp is located at @xmath10 n latitude and @xmath11 w longitude , at an altitude of 3200 m. the temperature is very low , with winter temperatures between @xmath12c and @xmath13c , and summer temperatures between @xmath14c and @xmath12c @xcite . due to the combination of the high altitude and the low temperature , very low opacity is expected .
furthermore , the nsf is currently funding the integrated characterization of energy , clouds , atmospheric state , and precipitation ( icecaps ) project at the summit camp , which is using active and passive ground - based remote sensors , including two radiometers that observe 16 frequencies from 22.2 ghz to 150.0 ghz , to provide the first complete description of cloud properties above this site @xcite .
they determine the annual cycle of pwv using the radiometer observations at the central frequencies of 23.8 , 31.4 , 90 , and 150 ghz as well as from radiosondes that are launched twice daily by icecaps technicians .
we , therefore , decided to put a 225 ghz tipping radiometer at this site to measure the atmospheric opacity conditions for possible submm vlbi operations ( see the next section for the reason to choose the measurement frequency ) .
our group is in the process of deploying a 12 m diameter submm telescope to greenland .
overall explanations of this project , the greenland telescope ( glt ) project , are described in @xcite for submm - vlbi science and technical details , and in @xcite for single - dish science cases .
in addition , more detailed information about antennas , receivers , and software for the glt project are in @xcite , @xcite , and @xcite , respectively .
for the opacity measurement at the summit of the greenland ice sheet , we procured a 225 ghz tipping radiometer , rpg-225 radiometer , from radiometer physics gmbh ( rpg ) .
this instrument has the ability to operate in very cold environments , and indeed some units are operating at arctic and antarctic areas ( two radiometers operating as part of icecaps mentioned above are also from rpg ) .
the reason for the choice of this operating frequency is that there are many site survey results from all over the world , including the current submm telescope sites , such as the summit of mauna kea , the alma ( chajnantor and pampa la bola ) site , and south pole .
the radiometer has an uncooled double side band heterodyne receiver with a bandwidth of 1 ghz . a tipping paraboloid mirror , which can rotate @xmath15 with its half power beam width of @xmath16 , is installed in front of the feed horn .
a gortex window covers @xmath17 from zenith to allow sampling of the sky signal , and a black body target , whose temperature is monitored by a thermometer , is located at the bottom ( i.e. , @xmath18 from zenith ) .
a 140 w blower is located below the window and provides heating to prevent ice formation and accumulation on the window .
we obtained the radiometer in the autumn of 2010 , deployed it on the roof of our institute in taipei , and conducted functional and gain stability tests .
we then moved the radiometer to the summit of mauna kea , hawaii , in the end of 2010 , to check the consistency with the 225 ghz tipping radiometer at the caltech submillimeter observatory ( cso ) .
we put our radiometer near the cso with the same tipping direction , measured the opacity for about 2 weeks , and confirmed that the results were consistent with each other ( linear regression coefficient = 1.04 ) . after this , we moved the radiometer to the polar environment atmospheric research laboratory ( pearl ) , located on a ridge at an altitude of 610 m at @xmath19 n and @xmath20 w , 15 km away from the eureka weather station on ellesmere island , canada .
we measured the atmospheric opacity for 3 months between late winter to early spring , and the results are reported in @xcite and @xcite . after this measurement , we moved the radiometer to the greenland summit camp .
the radiometer was installed on the roof of the mobile science facility ( msf ; * ? ? ?
* ) , which was built by cps in support of the icecaps project .
the roof deck is at about 3 m above the snow surface ( fig .
[ fig - rpg ] ) .
the measurement was started from august 17th , 2011 , and data were collected until february 12th , 2015 .
we present here the data for these @xmath21 years .
to measure the atmospheric opacity , we adopt the tipping method ; we observe five angles from horizon ( @xmath22 = zenith , @xmath23 , @xmath24 , @xmath25 , and @xmath26 , which corresponds to @xmath27 of 1.0 , 1.5 , 2.0 , 2.5 , and 3.0 ) with 4 second integration at each angle .
we scan the mirror from south to north , namely observe the five angles in both the southern and the northern sky ( for the measurements at greenland ) . the black body target in the bottom of the radiometer
is observed before and after the scan for the gain calibration .
the total duration of a tipping measurement is 75 seconds , and each tipping is performed every 10 minutes . between the tipping measurements ,
the mirror is pointed toward the zenith and the sky data are recorded every 1 second .
the output voltages , together with the mirror position , the black body target temperature , and other monitoring data , are recorded every 1 second into the hard disk of the host computer in binary format .
the raw binary data are downloaded from greenland to taiwan regularly via internet . ]
the tipping data have been reduced with the typical data reduction method for tipping measurements ( e.g. , * ? ? ?
since the msf is a mobile facility on the snow , it can shake due to wind or human activities inside the facility , so that the radiometer tipping angle may be affected .
in addition , the leveling accuracy for the radiometer is limited , so that the radiometer may also have a small constant tilt .
we assume that the opacity is the same between the southern and northern skies , and the small difference in opacities derived from the southern and northern sky tippings is considered as the result of the tilt of msf and/or the radiometer .
based on this assumption , we calculate the tilt angle and correct for it when deriving the opacity . in case
the difference of the opacities between the southern and northern skies is large ( i.e. , when the tilt angle is calculated to be larger than two degrees ; this value is also to allow some tolerance for the different opacities between the northern and southern sky ) , we judge this is due to real opacity differences , and we flag the data .
[ fig - time ] displays the measured time variation of the 225 ghz opacity for the 3.5-year period at the greenland summit camp .
it is clear that there is a seasonal variation ; opacity in winter is low , but high in summer . fig .
[ fig - month ] shows the monthly quartile variations ( solid , dashed , and dotted lines are 25% , 50% , and 75% quartiles , respectively ) .
december and march tend to have the best opacity conditions , and july tends to have the worst .
note that the first and the last months of this measurement ( i.e. , aug .
2011 and feb .
2015 ) have fewer data than the other months , so the statistical significance is lower . in both diagrams , there is no significant annual difference .
we also made cumulative distribution plots and histograms of the measured 225 ghz opacity while separating the seasons into winter and summer ( fig .
[ fig - cdp ] ) . here
we define winter as between the beginning of november and the end of april ( solid line plots ) , and summer as may through october ( dashed line plots ) .
the quartiles for each season are 0.046 , 0.060 , and 0.080 for 25% , 50% , and 75% , respectively , in winter , and 0.089 , 0.118 , and 0.159 for summer ( see also the crosses and the values in the figure ) .
it is obvious that the opacity in winter is about half of that in summer at all the quartiles .
the histograms show that the opacities of 0.04 and 0.10 are the opacities that occur most often in winter and summer , respectively .
we then compared the opacity quartiles using our 3.5-year period statistics with those at the alma and the south pole sites , which are well - established sites for submm observations .
the 225 ghz opacity data for the alma site have been obtained from @xcite and @xcite , whose measurements have been made between april 1995 and april 2006 ( @xmath28 years ) , and that for the south pole site from @xcite measured between january and december 1992 ( 1 year ) . since both the alma and the south pole sites are located in the southern hemisphere , we define winter as between the beginning of may through the end of october , and summer as november through april .
the calculated quartiles for these three sites are listed in table [ tab - quartile ] .
ccccc gl summit camp & winter & 0.046 & 0.060 & 0.080 + & summer & 0.089 & 0.118 & 0.159 + alma & winter & 0.035 & 0.050 & 0.080 + & summer & 0.071 & 0.131 & 0.261 + south
pole & winter & 0.041 & 0.048 & 0.057 + & summer & 0.050 & 0.062 & 0.076 for winter , the alma site is the best at the 25% quartile , south pole is next , and greenland summit camp is the worst , but only by @xmath29 ( @xmath30 ) difference in opacity between each site . at the 50% quartile ,
the alma and south pole sites are almost the same , and the opacity is about 0.01 ( about 25% ) worse at greenland summit camp . at the 75% quartile , south pole is the best ( 40% better ) and the opacity is the same between the alma and the greenland sites . for summer ,
south pole is the best for the 25% quartile , alma is the next , and greenland is the worst with @xmath31 difference in opacity between each site . at the 50% and 75% quartiles , south pole is again the best , but greenland is next and the alma site is the worst .
the south pole site is about twice better in opacity as compared with the greenland site , and the alma site is significantly worse than these two sites .
we note here that there are many studies that compare the opacity conditions at the summit of mauna kea , which is also a well - established submm site , with those at the alma and south pole sites ; opacity quartiles of mauna kea are @xmath32 higher than those of the alma site @xcite . in summary , the south pole site has little seasonal differences in opacity over the annual cycle , with a factor of two difference between seasons . on the other hand ,
the alma site has large variations between winter and summer , with significantly worse conditions in summer , known as the bolivian winter around february and march .
the opacity conditions at the greenland summit camp are roughly intermediate between the south pole and the alma sites . the top row of fig .
[ fig - diurnal ] shows the diurnal opacity variations in winter ( left column ) and summer ( right column ) at the greenland summit camp .
solid , dashed , and dotted lines are 25% , 50% , and 75% of the hourly quartiles .
it is obvious that there is no diurnal variation in both seasons .
this can be easily explained by the polar conditions ; only nighttime in winter and daytime in summer .
it is also clear that the opacity in winter is half of that in summer , as mentioned in the cumulative distribution plot ( fig .
[ fig - cdp ] ) above .
the middle and bottom rows of fig .
[ fig - diurnal ] show the diurnal variation at the alma and the south pole sites , respectively .
it is obvious that there is a clear diurnal variation in the alma data , which is naturally explained by the mid - latitude conditions , leading to a strong diurnal cycle compared to polar regions , while the south pole data are very similar to the greenland summit camp ( i.e. , no diurnal variation ) . for winter
, the south pole site is always the best at all quartiles .
opacities at the 25% and 50% quartiles at the alma site are almost always statistically better than those at the greenland summit camp , but for the 75% quartile , the daytime opacity is statistically better at the greenland summit camp than at the alma site due to the diurnal variation .
cccccccccccccc gl summit camp & 17.7 & 6.4 & 2.0 & 1.2 & 0.8 & & 6.0 & 2.4 & 1.2 & 0.4 & & 1.2 & 0.4 + alma & 30.6 & 10.7 & 1.5 & 0.2 & 0.0 & & 17.6 & 4.7 & 0.0 & 0.0 & & 4.1 & 0.3 + south pole & 32.6 & 13.6 & 5.4 & 2.7 & 2.7 & & 8.1 & 2.7 & 0.0 & 0.0 & & 0.0 & 0.0 for summer , the south pole site is again always the best at all quartiles . for opacities at the 25% and 50% quartiles ,
the greenland summit camp is always better between @xmath33 and @xmath34 hour than the alma site ( i.e. , daytime at the alma site ) due to no diurnal variation . for opacities at the 75% quartile ,
the greenland summit camp is always better than the alma site .
we then calculated time durations of opacity conditions continuously lower than certain values .
we focused on opacities lower than 0.05 , 0.04 , and 0.03 , which are excellent opacity conditions that only occur in @xmath35 of the winter season at the greenland summit camp .
the resultant histograms for the greenland summit camp are presented in fig .
[ fig - duration - hist](a ) . since the tipping measurements are done every 10 minutes as mentioned in sect .
[ sect - meas ] , the lower limit of the measurements are located at 0.17 hour .
the counts ( vertical axis ) are normalized with the number of annual data points , assuming one data point takes 10 minutes ( sect .
[ sect - meas ] ) .
we intentionally cut the count limit to 70 in the plot for presentation purpose .
for opacity less than 0.05 or 0.04 at the greenland summit camp , there were several occasions for which more than 100 hours were continuously showing opacities lower than those values , and there are many occasions that continued for more than 10 hours or several hours .
the counts for durations between 1 hour and 20 hours are almost the same , and that for durations between 20 hours to several tens of hours are roughly half of the shorter duration . for the opacity less than 0.03 , there were several occasions when the opacity was continuously low for more than 10 hours of time , and more occasions exist for durations of more than an hour .
this is obviously due to the polar conditions ( no diurnal variation ) as mentioned above , and it is very difficult to achieve similar values at other submillimeter sites that are not at polar regions , such as in the hawaii ( mauna kea ) or northern chile ( alma ) site . indeed , we also calculated the time duration of opacity conditions for the alma site under the same opacity conditions as the greenland summit camp ( fig .
[ fig - duration - hist]b ) , and the difference is obvious . at the alma site , there are many occasions for the opacity less than 0.05 , 0.04 , and even 0.03 continues for up to 20 hours , but the occasions for the duration longer than 20 hours are significantly smaller , about one - fifth , than the shorter durations . the duration longer than 100 hours has only been recorded for opacity lower than 0.05 , and never happened for opacities lower than 0.04 or 0.03 in the 11 yr long data .
for the south pole site ( fig .
[ fig - duration - hist]c ) , the duration is obviously shorter than that of greenland summit camp .
there are many occasions for the opacity less than 0.05 , which continues up to 10 hours , but there are significantly fewer occasions for the duration longer than 10 hours .
similar to the alma site , the duration longer than 100 hours has only been recorded in opacity lower than 0.05 , and never happened for opacity lower than 0.04 or 0.03 . for the opacity lower than 0.03 ,
the duration is only up to about 10 hours , and has never been longer .
although the south pole data are taken only for a year and the opacity statistics are better than for the greenland summit camp ( sect .
[ sect - res - cdh ] , [ sect - res - diurnal ] ) , good opacity duration time is shorter .
table [ tab - duration ] shows that low opacity conditions continue for more than 24 , 50 , 100 , 150 , and 200 hours at the greenland summit camp , alma , and the south pole .
the alma and south pole sites have better atmospheric opacities over time durations longer than 24 hours and 50 hours at opacity conditions lower than 0.05 and 0.04 , respectively .
but for time durations longer than 100 hours at opacity conditions lower than 0.04 , the greenland summit camp clearly has lower atmospheric opacities .
we also made a table to show the statistics of low opacity conditions continues for more than 24 , 50 , 100 , 150 , and 200 hours in unit of counts per year ( table [ tab - duration ] ) .
the aforementioned statements are quantitatively shown in this table ; the alma and south pole sites have better statistics at time duration longer than 24 hours and 50 hours at opacity conditions lower than 0.05 and 0.04 .
but for the time duration longer than 100 hours at opacity conditions lower than 0.04 , the greenland summit camp clearly has better statistics .
these results are also clearly seen in fig .
[ fig - duration - cumul ] , which compares the cumulative distributions of the time durations for the three sites discussed above . figs .
[ fig - duration - cumul](a ) and ( b ) show the cumulative distributions of the time durations of opacities less than 0.05 and 0.04 , respectively , and it is clear that the greenland summit camp ( solid line ) has a long tail toward the long duration of more than a hundred hours .
the alma site always exhibits higher cumulative distributions than that of the greenland summit camp , and reaches 100% around several tens of hours .
the south pole site shows the steepest cumulative distribution , and reaches 100% around a few tens of hours , much shorter than the other two sites .
the cumulative distributions of the time durations of opacity less than 0.03 ( fig .
[ fig - duration - cumul]c ) display very similar distribution between the greenland and the alma site , but the greenland summit camp shows a long tail up to several tens of hours .
again , the south pole site is much shorter , only up to ten hours . in summary , for the low opacity duration
, the greenland summit camp is the best site to have continuous low opacity conditions .
since the variation of opacity is directly related to the variation of sky temperature ( background ) , these long stable opacity conditions will be a significant advantage for astronomical observations , which need unusually stable sky background , such as thz observations or wide - field submillimeter continuum observations .
using our 225 ghz opacity data and the radiative transfer program `` _ am _ '' @xcite , we estimated the atmospheric transmission spectra between 0 ghz and 1600 ghz in both winter and summer at the greenland summit camp . first , from the nasa modern - era retrospective analysis for research and applications , version 2 ( * ? ? ?
* ; * ? ? ?
* merra-2 ) reanalysis data interpolated to the location of the greenland summit camp , we computed various percentiles of the temperature , h@xmath36o mixing ratio , and o@xmath37 mixing ratio at each merra pressure level , for the winter and summer months covering the same period as the radiometer measurements .
am _ model files were then constructed such that each percentile model used the corresponding percentile temperature and h@xmath36o profile , whereas all models used the 50% quartile o@xmath37 profile .
here we assumed that temperature and water vapor should be highly correlated , such that it is physically meaningful to associate a percentile profile with the corresponding percentile statistics on each level . on the other hand
, o@xmath37 is relatively uncorrelated with either temperature or water vapor , so that it makes sense to simply use the median profile . for each set of merra-2 percentile profiles
, we found a scaling factor on the tropospheric part of the h@xmath36o profile , which reproduced the corresponding 225 ghz opacity percentiles from our measurements .
the scale factor on the median merra h@xmath36o profile to match the median 225 ghz opacity was 1.09 in winter and 1.10 in summer , indicating a dry bias of approximately 10% for merra-2 relative to our measurements . from a radiative transfer model using the percentile profiles scaled to our measurements
, we then estimated the corresponding percentile atmospheric transmission spectra .
the estimated atmospheric transmission spectra for the 2% , 10% , 25% , and 50% opacity conditions in winter season are plotted in fig .
[ fig - est](a ) .
note that these opacity conditions correspond to the 225 ghz opacity of 0.031 , 0.037 , 0.046 , and 0.060 , respectively .
these spectra suggest that it is possible to observe astronomical sources with little atmospheric attenuation ( opacities @xmath38 ) most of the winter time at the greenland summit camp for frequencies lower than 450 ghz , and half of the time for frequencies between 450 ghz and 1000 ghz with opacities @xmath5 . for the thz windows ( 1035 ghz , 1350 ghz , and 1500 ghz )
, 10% of the winter time will have an atmospheric transmission of more than 10% .
the estimated atmospheric transmission spectra for the same opacity percentiles as above , but in summer season are plotted in fig .
[ fig - est](b ) .
note that these opacity conditions correspond to the 225 ghz opacity of 0.052 , 0.068 , 0.088 , and 0.118 , respectively .
these spectra suggest that it is possible to observe astronomical sources with little atmospheric attenuation ( opacities @xmath38 ) most of the summer time for frequencies below the 380 ghz water vapor line , and more than half of the time for the 450 ghz atmospheric window with opacities @xmath39 . for the windows between 450 ghz and 1000 ghz
, 25% of the summer time will have an atmospheric transmission of more than 10% .
the thz windows are totally opaque in summer .
we also calculated the relationships between 225 ghz opacity and other submillimeter atmospheric window opacities using the above estimates .
this is useful for estimating the opacities at higher frequencies from the observations of the opacity at 225 ghz .
relations between two opacities turned out to be all linear , and therefore the linear coefficients and offsets have been calculated .
the calculated values are in table [ tab - corr ] .
opacities at the 492 ghz , 675 ghz , and 875 ghz windows are about 25 30 times larger than the 225 ghz opacity , and those at the thz windows are about @xmath40 times larger .
rclc 345 ghz & 3.80 & @xmath41 & @xmath42 + 410 ghz & 7.34 & @xmath43 & @xmath44 + 492 ghz & 26.4 & @xmath45 & @xmath46 + 675 ghz & 27.7 & @xmath47 & @xmath48 + 875 ghz & 31.0 & @xmath49 & @xmath50 + 937 ghz & 55.5 & @xmath51 & @xmath52 + 1035 ghz & 140 & @xmath53 & @xmath54 + 1350 ghz & 129 & @xmath55 & @xmath46 + 1500 ghz & 130 & @xmath56 & @xmath57 [ tab - corr ] these values can be compared with the measurement results at the alma site @xcite : for the 345 ghz and 410 ghz windows , there are only a few % differences between the greenland summit camp and the alma site , but for higher frequencies , the greenland summit camp is about @xmath58% worse than the alma site for the 492 ghz , 1035 ghz , and 1350 ghz windows , and about 25% worse for the 675 ghz , 875 ghz , 937 ghz , and 1500 ghz windows ( table [ tab - corr ] ) .
this is due to the altitude difference .
the altitude of the greenland summit camp is only 3200 m , much lower than that of the alma site at 5000 m , and this difference increases the opacity in the pressure - broadened wings of saturated h@xmath36o lines at high frequencies . in addition , the lower altitude introduces larger dry air continuum absorption at the greenland summit camp ; the dry air continuum absorption increases at higher frequencies , up to the middle of the n@xmath59-n@xmath59 collision - induced absorption band near 3 thz @xcite .
the `` _ am _ '' program also estimates the pwvs together with the transmission spectra , and we present those in table [ tab - pwv ] . using these values , it is possible to derive the relation between the 225 ghz opacity and pwv at the greenland summit camp , which turned out to be @xmath60 } + 0.022,\ ] ] where @xmath61 is the 225 ghz opacity . with this equation , together with the relationship between the 225 ghz opacity and that of other frequencies ( table [ tab - corr ] ) , it is also possible to compare various atmospheric window opacities and pwv .
rcccc 2% & 0.0311 & 0.186 & 0.0516 & 0.623 + 10% & 0.0375 & 0.321 & 0.0679 & 0.965 + 25% & 0.0464 & 0.506 & 0.0885 & 1.396 + 50% & 0.0602 & 0.790 & 0.1178 & 2.054 + 75% & 0.0801 & 1.205 & 0.1586 & 2.942 [ tab - pwv ] the 3-year ( 2008 2010 ) pwv statistics and the atmospheric transmission spectra of the summit of greenland have also been derived by @xcite using the water vapor product from the infrared atmospheric sounding interferometer ( iasi ) instrument on the meteorological operation ( metop)-a satellite .
they found a median annual pwv of 0.94 mm over the summit of greenland .
our radiometer - derived median pwv for the 3-year subset of our data from october 2011 to september 2014 is 1.28 mm . to connect these two periods
, we note that the merra-2 median pwv for january 2008 december 2010 is 1.27 mm , whereas for october 2011
september 2014 it is 1.16 mm .
if we assume the merra-2 dry bias of approximately 10% is consistent between these periods , then the implied median pwv for 2008 2010 is 1.40 mm .
this would suggest that the pwv for summit in @xcite is systematically lower than the actual pwv by approximately 33% .
there are two reasons why a single satellite data set could produce significantly biased pwv statistics .
first , although iasi is able to retrieve surface temperature , the water vapor retrieval accuracy in the lower troposphere necessarily suffers from a lack of thermal contrast with the surface @xcite .
the second reason is temporal sampling bias .
typically satellite sounders are in sun - synchronous orbits that pass over a given point on the earth s surface at the same pair of local times each day ; these are 9:30 and 21:30 in the case of iasi @xcite . in comparison
, a reanalysis such as merra-2 can be expected to produce more realistic pwv statistics because it assimilates multiple satellite , surface , and upper air measurements using a model with realistic dynamics .
we present the 3.5-yr monitoring of the 225 ghz opacity at the summit of the greenland ice sheet using a tipping radiometer .
opacity variations clearly show a seasonal variation between winter ( @xmath62 nighttime ) and summer ( @xmath62 daytime ) , but no diurnal variation , with the opacity in winter being about half of that in summer .
this is similar to the opacity variations at the south pole site due to the polar conditions , but the absolute opacity value is about 10%-30% higher than that of the south pole site in winter , and about double in summer . in contrast , the alma site shows clear seasonal and diurnal variations due to mid - latitude conditions ; the opacity at the alma site is up to 20% better than at the greenland site in winter , but up to 40% worse in summer .
the estimated atmospheric transmission spectra suggest that most of the winter time is useable for astronomical observations at frequencies lower than 450 ghz , half of the time is useable for frequencies between 450 ghz and 1000 ghz , and 10% of the time is useable for the thz atmospheric windows .
most of summer time is useable at frequencies lower than 380 ghz , and half of the summer time is useful at the 450 ghz atmospheric window .
the linear correlations between 225 ghz and submillimeter atmospheric window opacities are derived , and opacities at the 492 ghz , 675 ghz , and 875 ghz windows are about 25 30 times larger , while those at thz windows are about @xmath40 times larger than the opacity at 225 ghz .
these opacities are up to 25% higher than those at the alma site , which is due to the altitude difference between the greenland summit camp ( 3200 m ) and the alma site ( 5000 m ) .
the biggest advantage of the opacity conditions at the greenland summit camp is the long time durations of low opacity . at the greenland summit camp ,
opacities lower than 0.04 or 0.05 can continue for more than a hundred hours occasionally , and opacities lower than 0.03 can continue for more than several tens of hours in some cases . in case of opacity lower than 0.04 , the greenland summit camp is the only site that can continue this condition for more than one hundred hours , indicating that the greenland site is suitable for observations that need stable opacity conditions for a long time .
we would like to thank cps for providing liquid nitrogen yearly , which was used for the calibration of the radiometer s detector , and to the icecaps technicians who assisted us in a timely fashion when needed .
we also would like to thank simon j. e. radford for providing us with the opacity data of the alma and the south pole sites .
the merra-2 data used in this study have been provided by the global modeling and assimilation office ( gmao ) at nasa goddard space flight center .
we are also grateful to the anonymous referee for helpful comments .
sm is supported by the national science council ( nsc ) and the ministry of science and technology ( most ) of taiwan , nsc 100 - 2112-m-001 - 006-my3 and most 103 - 2112-m-001 - 032-my3 .
99 asada , k. , martin - cocher , p. l. , chen , c .- p . ,
2012 , , 8444 , 84441j bower , g. c. , goss , w. m. , falcke , h. , backer , d. c. , & lithwick , y. 2006 , , 648 , l127 chamberlin , r. a. & bally , j. 1994 , , 33 , 1095 chamberlin , r. a. & bally , j. 1995 , int .
j. ir mm waves , 16 , 907 doeleman , s. s. , weintroub , j. , rogers , a. e. e. , et al .
2008 , , 455 , 78 doeleman , s. s. , fish , v. l. , schenck , d. e. , et al .
2012 , science , 338 , 355 grimes , p. k. , asada , k. , blundell , r. , et al .
2014 , , 9153 , 91531v hilton , f. , armante , r. , august , t. , et al .
2012 , bull .
american met .
soc . , 93 , 347 hirashita , h. , koch , p. m. , matsushita , s. , et al .
2016 , , 68 , 1 inoue , m. , algaba - marcos , j. c. , asada , k. , et al .
2014 , radio science , 49 , 564 laursen , e. v. 2010 , dmi technical report no.10 - 09 martin - cocher , p. l. , asada , k. , matsushita , s. , chen , m .-
, ho , p. t. p. , & chen , c .- p .
2014 , , 9147 , 91473n matsuo , h. , sakamoto , a. , & matsushita , s. 1998 , , 50 , 359 matsushita , s. , chen , m .-
, martin - cocher , p. , et al . , 2013 , iau symp . , 288 , 204 matsushita , s. , matsuo , h. , pardo , j. r. , & radford , s. j. e. 1999 , , 51 , 603 matsushita , s. , matsuo , h. , sakamoto , a. , & pardo , j. r. 2000 , , 4015 , 378 molod , a. , takacs , l. , suarez , m. , & bacmeister , j. 2015 , geoscie .
model dev . ,
8 , 1339 paine , s. 2014 , sma technical memo , 152 pardo , j. r. , cernicharo , j. , & serabyn , e. 2001a , ieee trans . on antennas and propagation , 49 , 1683 pardo , j. r. , serabyn , e. , & cernicharo , j. 2001b , , 68 , 419 patel , n. a. , nishioka , h. , & huang c .- w .
l. 2016 , , 9913 , 991308 radford , s. j. e. 2011 , rev .
astronoma y astrofsica ( serie de conferencias ) , 41 , 87 radford , s. j. e. & chamberlin , r. a. 2000 , alma memo , 334 radford , s. j. e. & peterson , j. b. 2016 , , 128 , 965 , 075001 raffin p. , ho , p. t. p. , asada , k. , et al .
2016 , , 9906 , 99060u rienecker , m. m. , suarez , m. j. , gelaro , r. , et al .
2011 , j. climate , 24 , 3624 shen , z .- q . , lo , k. y. , liang , m .- c .
, ho , p. t. p. , & zhao , j .- h .
2005 , , 438 , 62 shupe , m. d. , turner , d. d. , walden , v. p. , et al .
2013 , bull .
american met .
soc . , 94 , 169 tremblin , p. , schneider , n. , minier , v. , durand , g. al . , & urban , j. 2012 , , 548 , a65 wulfmeyer , v. , hardesty , r. m. , turner , d. d. , et al .
2015 , reviews of geophysics , 53 , 819 | we present the 3.5-yr monitoring results of 225 ghz opacity at the summit of the greenland ice sheet ( greenland summit camp ) at an altitude of 3200 m using a tipping radiometer .
we chose this site as our submillimeter telescope ( greenland telescope ; glt ) site , because conditions are expected to have low submillimeter opacity and because its location offers favorable baselines to existing submillimeter telescopes for global - scale very long baseline interferometry ( vlbi ) .
the site shows a clear seasonal variation with the average opacity lower by a factor of two during winter .
the 25% , 50% , and 75% quartiles of the 225 ghz opacity during the winter months of november through april are 0.046 , 0.060 , and 0.080 , respectively . for the winter quartiles of 25% and 50%
, the greenland site is about @xmath0 worse than the atacama large millimeter / submillimeter array ( alma ) or the south pole sites .
estimated atmospheric transmission spectra in winter season are similar to the alma site at lower frequencies ( @xmath1 ghz ) , which are transparent enough to perform astronomical observations almost all of the winter time with opacities @xmath2 , but @xmath3 higher opacities at higher frequencies ( @xmath4 ghz ) than those at the alma site .
this is due to the lower altitude of the greenland site and the resulting higher line wing opacity from pressure - broadened saturated water lines in addition to higher dry air continuum absorption at higher frequencies .
nevertheless , half of the winter time at the greenland summit camp can be used for astronomical observations at frequencies between 450 ghz and 1000 ghz with opacities @xmath5 , and 10% of the time show @xmath6 transmittance in the thz ( 1035 ghz , 1350 ghz , and 1500 ghz ) windows .
summer season is good for observations at frequencies lower than 380 ghz .
one major advantage of the greenland summit camp site in winter is that there is no diurnal variation due to the polar night condition , and therefore the durations of low - opacity conditions are significantly longer than at the alma site .
opacities lower than 0.05 or 0.04 can continue for more than 100 hours .
such long stable opacity conditions do not occur as often even at the south pole ; it happens only for the opacity lower than 0.05 .
since the opacity variation is directly related to the sky temperature ( background ) variation , the greenland summit camp is suitable for astronomical observations that need unusually stable sky background . |
the x - ray burst source and soft x - ray transient aquila x1 displays x - ray variability on all time scales .
usually the source is in its quiescent state with a very low x - ray luminosity , typically @xmath1 erg s@xmath2 ( verbunt et al .
1994 ) . within intervals of months
to years it shows outbursts characterized by a gradual increase in flux to a level that sometimes is comparable to the crab , but at other times is two orders of magnitude fainter .
the periodicity of these long - term outburts is unstable with recurrent periods of @xmath0 125 and 309 days ( priedhorsky & terrell 1984 ; kitamoto et al . 1993 ) .
the rise time is typically 10 days , the source remains at a flux maximum for 510 days and then decays back to quiescence in typically 3050 days ( campana et al .
1998 ; kitamoto et al .
the spectrum is soft and no regular pulsations have been detected ( cui et al .
1998 ; zhang , yu & zhang 1998 ) . during outbursts
, aquila x1 displays type - i x - ray bursts ( zhang et al .
1998 ) that can be explained in terms of runaway thermonuclear burning of matter on the surface of the neutron star . as in many other low - mass x - ray binaries ( lmxb ) ,
khz quasi - periodic oscillations ( qpo ) have been detected in the persistent flux of aquila x1 ( zhang et al .
1998 ; cui et al .
these khz qpo occurred in the frequency range 740830 hz , had a q value ( = @xmath3 ) of over 100 and an average fractional @xmath4 amplitude of 7% .
khz qpo may be caused by orbital motion of gas around the neutron star very close to its surface ( see van der klis 1999 for a recent review ) .
recent deep optical and infrared images and low resolution spectroscopy of this system suggest that the mass - donating companion is a v=21.6 k7v star , located at an estimated distance of 2.5 kpc ( callanan et al .
1999 ; chevalier et al . 1999 ) , and not the v=19.2 k1iv star , 0.48@xmath5 west , previously assumed to be the candidate ( chevalier & ilovaisky 1991 ; shahbaz , casares & charles 1997 ) . at radio wavelengths
aquila x-1 is unusual in being one of the only neutron star binaries to exhibit radio emission ( hjellming & han 1995 ) .
the late - type optical counterpart and the episodic outbursts define aquila x1 as a soft x - ray transient ( sxt ) , whereas the existence of type i bursts denotes a low magnetic field neutron star companion ( as opposed to black hole ) . prior to exosat ,
low mass x - ray binaries ( lmxb ) were classified on the basis of their x - ray luminosity ( schulz , hasinger & trmper 1989 ) into two classes : high and low luminosity systems .
the low luminosity systems included the so - called x - ray bursters . thanks to its wide orbit and large collecting area ,
the exosat satellite allowed long uninterrupted observations and detailed x - ray timing analysis .
taking into account the rapid aperiodic variability it turned out that low - mass x - ray binaries could be divided into two different subclasses , z and atoll sources , defined by the patterns that these sources display in x - ray color - color diagrams and the properties of the rapid x - ray variability ( hasinger & van der klis 1989 , hereafter hk89 ) .
the classification of aquila x1 in this scheme is not certain .
_ rxte _ data has suggested that aquila x1 is an atoll source ( cui et al .
1998 ) , though some properties appear anomalous . in this paper
we investigate the correlated x - ray timing and spectral variations of aquila x1 and present evidence for its classification as an atoll source .
the data used in this work were retrieved from the public _
rxte _ archive and correspond to two different sets of observations .
the first one took place between 1997 february 16 march 10 and consists of 12 separate observations ( one every two days , roughly ) , with a total usable time of about 96 ks ( see zhang et al .
these data correspond to the decay phase of an outburst .
the observations during the second set , 1997 august 11
september 10 , were conducted at a typical rate of once or twice per day and produced approximately 168 ks of usable data .
these observations began halfway through the rising phase of another outburst and finished half way through the decay phase .
five type i x - ray bursts ( two of them were already reported in zhang et al .
1998 ) as well as khz qpo were present in the data .
the bursts were excluded from our analysis .
another four snapshots taken during the rising phase of the 1998 february - march outburst ( cui et al .
1998 ) were also analysed in order to compare our results with those of other authors .
_ rxte _ carries two pointed instruments , the proportional counter array ( pca ) developed to cover the lower part of the energy range ( 260 kev ) , and the high energy x - ray timing experiment ( hexte ) sensitive to x - rays between 15 and 250 kev .
these instruments are equipped with collimators yielding a fwhm of one degree .
in addition , rxte carries an all - sky monitor ( asm ) that scans about 80% of the sky every orbit . in this work we analyzed data from the pca only since its large collecting area ( @xmath0 6500 @xmath6 ) , makes it the most appropiate instrument for timing studies .
background subtracted light curves corresponding to the energy ranges 2.03.5 kev , 3.56.0 kev , 6.09.7 kev and 9.716.0 kev were used to define the soft and hard colors as sc = 3.56.0/2.03.5 and hc = 9.716.0/6.09.7 , respectively . in a few of the observations one or two of the five detectors of the pca were switched off ; we only used the three detectors which were always on to calculate these count rates , and normalized the count rates to 5 detectors .
the color - color diagram of aquila x1 is shown in figure [ cd ] .
the data points in this color - color diagram fall into several distinct groups , the hard color being the defining quantity .
the statistical scatter in the uppermost ( open squeres ) and middle ( open circles ) groups in figure 1 is considerable .
these groups of points come from observations at the very end ( after march 5th ) of the decay of the 1997 march outburst .
the lower branch corresponds to observations from the rest of the 1997 february march run and from the entire 1997 august
september run .
the points represented with filled squares correspond to the 1998 march 2 observation .
the mean pca intensity of each group is given in table [ fitres ] .
at the rising part of the 1997 august - september outburst the source moved smoothly from left to right along the lower elongated branch , and moved back from right to left during the decay phase . during the 1997 february - march outburst the source also moved towards lower soft colors as the intensity decreased . in order to measure the source luminosity in the different states spanned in the color - color diagram we extracted energy spectra and fitted an absorbed blackbody plus power - law model to the data .
an iron line at 6.4 kev was added if necessary .
the 210 kev x - ray luminosity , assuming a distance of 2.5 kpc , varies from 2.0 @xmath7 10@xmath8 erg s@xmath2 in the uppermost group to 3.2 @xmath7 10@xmath9 erg s@xmath2 in the middle one .
the elongated lower group of points presents the highest count rate . here the source luminosity increases from left to right , from 2.4 @xmath7 10@xmath10 erg s@xmath2 to 3.9 @xmath7 10@xmath10 erg s@xmath2 .
these structures in the color - color diagram and associated count rate differences are similar to those of typical atoll sources .
the upper and middle group of points can be associated with the so - called island state , whereas the more elongated lower part would then represent the banana state . only by looking at the position occupied by the points of the 1998 observation it is not posssible to tell whether they define an island or a banana state and an analysis in terms of the aperiodic variability
is needed . in any case
they seem to indicate that the transition between the two spectral states is smooth .
the gap between the branches is probably observational : no data were obtained for two days between the banana and the island and for three days between the island and the extreme island states during the 1997 observations . in order to investigate the variability of aquila x1 as a function of the position in the color - color diagram we divided the color - color plane into several regions as shown in figure [ cd ] .
the two island branches define the first two groups .
the banana branch was divided so that each region contains approximately the same amount of data .
we then approximated the shape of the banana branch with a spline , and used the parameter @xmath11 to measure positions along this spline ( mndez et al .
we set @xmath11 to 2 at ( sc , hc)= ( 1.732,0.338 ) and to 3 at ( sc , hc)=(2.045,0.320 ) .
the intermediate positions are obtained through spline interpolation between the two defining values .
this could be done for the banana branch only since the discontinuity between the island and banana branches does not allow to define a unique path .
we arbitrarily assigned the values 1.1 and 1.5 to the two island states . in this way
each one of the regions in which we had divided the color - color diagram is then characterized by a value of the parameter @xmath11 . in order to study the source variability at low frequencies ( @xmath12 512 hz ) we divided the 260 kev pca light curve ( no energy selection was done ) of each observation into 256 s segments and calculated the fourier power spectrum of each segment up to a nyquist frequency of 1024 hz .
the high frequency end ( 9501024 hz ) of the power spectra was used to determine the underlying poisson noise , which was subtracted before performing the spectral fitting .
the power spectra were normalized to fractional @xmath4 squared per hertz ( van der klis 1995 ) . in a few observations one or two of the five pca units
were switch off .
we have used only those observations for which the five detector units were switched on .
this implied a loss of @xmath0 11% of the data . to avoid contributions to the power from the khz qpos we restricted the spectral fitting to the frequency interval 1/256512 hz .
the 256-s power spectra were then grouped according to the position of the source in the color - color diagram : for each region in the banana and each island state one mean power spectrum was obtained .
the power spectra were fitted using two broad noise components called the very - low frequency noise ( vlfn ) and the high - frequency noise ( hfn ) .
these two components are mathematically represented by a power - law and a power - law times an exponential cut - off , respectively ( hk89 ) .
the vlfn accounts for the low - frequency part of the spectrum , whereas the hfn dominates at higher frequencies .
figure [ pds ] shows six power spectra corresponding to different positions in the color - color diagram ; all the island states and the lower , middle and upper banana states are shown .
while the vlfn plus hfn model can satisfactorily describe the banana - state power spectra , it does not provide good fits for the faintest ( @xmath13 ) of the two island - state power spectra .
ford & van der klis ( 1998 ) used a broken power - law plus one low - frequency ( 1050 hz ) lorentzian , possibly representing lense - thirring precession ( stella & vietri 1998 ) , to fit the island - state power spectra of the atoll source 4u 172834 . in aquila x1
we find that a simple broken power - law gives good fits ( @xmath14 @xmath12 1.2 ) to the power spectra in both island states .
although described by different mathematical functions , the cut - off power - law of the banana state and the broken power - law of the island state are likely to represent the same type of noise component , namely hfn .
to be consistent , we have used the broken power - law model to fit the power spectra of the island states .
figure [ rms ] shows the variation of the amplitude ( as fractional @xmath4 ) of the vlfn ( circles ) , hfn ( squares ) and broken power - law ( triangles ) components .
the latter refers to the island states . unlike the hfn component
, the amplitude of the vlfn component does not show any clear trend , but remains at @xmath0 6 - 7% @xmath4 , irrespective of the position that the source occupies in the banana .
the maximum strength of the hfn is found during the island phases ( @xmath15 34% and 18% for the upper and lower island sates , respectively ) .
this is in contrast to the results of cui et al .
( 1998 ) who did not detect such hfn component in the first observations of the 1998 february - march outburst ( mjd 5087577 ) , despite the inference from the color diagrams that these data were in an island state .
these authors discuss the absence of an hfn in the island state as an unusual phenomenon .
we have reanalysed the first four observations in the cui et al .
( 1998 ) paper and found that there is , in fact , an hfn component with an @xmath4 of [email protected]% ( fig [ pds ] , table [ fitres ] ) in the first two observations .
as expected in the island , there is no vlfn ( 1.6% upper limit , 95% confidence level ) .
the mean hard and soft colors of the corresponding points are hc=1.84 , sc=0.34 and hc=1.86 , sc=0.31 for the first and second observation , respectively .
that is , in our color - color diagram they lie in between the lower island and the banana states .
this distribution of the island points in the color - color diagram suggests a continuous transition between the island and banana groups .
the other two observations produced points in the banana only . for the khz qpo analysis we produced power spectra using 64 s data segments and a nyquist frequency of 2048 hz .
khz qpo were only observed in a specific range in the color - color diagram , namely the lower banana near @xmath17 ( the filled circles in fig .
the 260 kev fractional amplitudes of the khz qpo ranged from @xmath0 4.5% to @xmath0 11.7% .
we did not detect qpo in the upper part of the banana , with a 95% confidence upper limit of 1.4% @xmath4 nor in the island with an upper limit of @xmath0 8% .
figure [ figrate ] shows the dependence of the frequency of the qpo as a function of the 216 kev pca count rate .
the two parallel groups at the lower left of the plot ( open and grey circles ) correspond to the 1997 february - march observations ( zhang et al .
1998a ) , while the rest come from the 1997 august - september set of observations .
mndez et al .
( 1999 ) found that there is a much better correlation between the frequency of the khz qpo in 4u 160852 and the position of the source on the color - color diagram than between frequency and flux .
similarly , the multi - valued dependence of @xmath18 on x - ray flux in aquila x1 is reduced to a single relation when @xmath18 is plotted against the @xmath11 parameter , which in our case is just a measure of the soft color ( fig .
[ fig_hc ] ) .
the hard color is not as sensitive to changes in the khz qpo frequency in aquila x1 as it is in 4u 160852 ( kaaret et al .
1998 ; mndez et al .
we have measured the color and timing properties of the lmxb aquila x1 , and found a behavior similar to that of other low - luminosity lmxbs .
the color - color diagram shows the classical atoll shape with the banana and island states .
the island state is not defined by a single group of points but it is split into a number of different groups as in other atoll sources , e.g. 4u 163653 ( prins & van der klis 1997 ) and 4u 160852 ( yoshida et al . 1993 ) . the lowest count rate and fluxes are detected in the island state with the hardest color and increase as the hard color decreases . based on the similarity of the color - color diagram of aquila x1 to that of other atoll sources it seems likely that aquila x1 can be placed in the group of atoll sources . however , the information provided by the color - color diagram is , by itself , not always sufficient to classify the source state .
it is not always possible to make a distinction between z and atoll sources or between the two spectral states ( island / banana ) within the atoll class ( hk89 ) .
moreover , in some atoll sources the island and banana branches have been seen to shift in the color - color diagram , both in soft and hard colors ( prins & van der klis 1997 ) .
the information obtained from the analysis of the noise components in the power spectra provides the key for an unambiguous classification .
the preliminary classification of aquila x1 as an atoll source on the basis of the color - color diagram is confirmed by the fast - timing analysis .
the power spectra of atoll sources ( hk89 ) are characterized by two broad noise components called the very - low frequency noise and the high - frequency noise .
the relative strengths of these two components vary in anticorrelation with each other and with the inferred mass accretion rate , @xmath19 , as measured by @xmath11 in our analysis : the vlfn component appears at the highest inferred @xmath19 , whereas that of the hfn decreases as @xmath19 increases .
these two components are present in the power spectra of aquila x1 ( see fig .
[ pds ] and table [ fitres ] ) .
as is commonly seen in atoll sources the vlfn is most prominent in the banana state , while the hfn dominates the island state .
the fractional amplitude of the hfn component decreases as the system moves from the island state to the lower banana and from here to the upper banana .
the vlfn component is practically undetectable ( @xmath4 @xmath12 2% ) in the island state , and although present in the banana state its amplitude does not change as expected ( see below ) .
another similarity between the power spectra of aquila x1 in the banana state and those of atoll sources is the presence of wiggles in the vlfn ( hk89 ) .
the increase of the break frequency as the x - ray count rate ( or flux ) increases , and the fact that the largest fractional amplitude and the hardest spectrum are observed in the two island states , represent further evidence in favor of an atoll classification for aquila x1 ( e.g. 4u 0614 + 09 , mndez et al .
it is worth noting that all these characteristics are also seen in black hole candidates during the low state , emphasizing the observational similarities between atoll sources and black holes systems ( van der klis 1994 ) .
one of the open questions in atoll sources is whether the transition from the island to the banana states occurs continuously or abruptly , that is , with the source jumping from one state to the other .
the trends of the spectral parameters just described seem to indicate that such transition is continuous .
this idea would be supported by the 1998 march data ( cui et al .
the upper limit to the @xmath4 of the vlfn component in the 1998 march observation ( @xmath20% , 95% confidence ) is lower than the value in the banana ( 57% ) , and the @xmath4 of the hfn is larger ( 6% compared to 3.5% ) .
the break frequency @xmath21 @xmath22 43 hz follows the same trend as the other two island states , namely , increases as the source moves into the banana .
an interesting difference between the island state of 1998 march and the other two island states is the source count rate . while the two 1997 march islands show the lowest pca intensity , the 1998 march one is comparable to that detected in the lower banana state .
it is worth noting that the 1997 island data were collected during the _ decay _ of an x - ray outburst , whereas that of 1998 correspond to the _ rise _ of the outburst .
a hysteresis effect may be present .
our observed vlfn amplitude does not change significantly in the banana state , contrary to what is observed in other atoll sources .
also , the power - law index of the vlfn becomes less steep as the source moves up in the banana branch , while usually in other atoll sources a slight increase , if any change , occurs .
this unusual behavior of the vlfn component may be related to nuclear burning on the surface of the neutron star . in periodic x - ray burters , like aquila x1
, the entire surface of the star is rapidly ( @xmath12 10 s ) burned by a fast propagating thermonuclear instability .
this can only happen at low mass accretion rates , where the envelope is convectively combustible most of the time .
the convective burning makes it difficult for slower combustion to occur , thus suppressing the vlfn ( bildsten 1993 ) .
yu et al . ( 1999 ) reported another unusual aspect of the vlfn in aquila x1 , namely , its disappearance after a type i burst , in which the 210 kev flux decreased by about 10% and the khz qpo frequency fell abruptly by @xmath0 37 hz . aquila x1 is unusual among the group of low - mass x - ray binaries with khz qpo in that only one khz qpo has been so far observed .
all other sources ( z and atoll ) have at least sometimes shown two simultaneous khz qpos .
nevertheless , the relationship between the frequency of the khz qpo and the x - ray flux in aquila x-1 is very similar to that seen in the atoll sources 4u 160852 ( mndez et al .
1999 ) and 4u 0614 + 091 ( ford et al . 1997 ) : on time scales longer than @xmath0 1 day there is no correlation between the khz frequency and the x - ray intensity ( or flux ) , whereas on short time scales ( @xmath0 hours ) , these two quantities correlate remarkably well ( zhang et al .
we also find that similarly to other sources , on time scales longer than a day , the qpo frequency correlates much better to color ( or @xmath11 ) than to count rate .
we have used _ rxte _ data to carry out a timing analysis with the aim of solving the issue of the classification of aquila x1 in the z / atoll scheme .
aquila x1 traces a pattern in the color - color diagram that is consistent with having an island and a banana branch .
the pca intensity correlates with the position of the source in the color color diagram , decreasing in the sense upper banana @xmath23 lower banana @xmath23 island state .
the power density spectra of the banana states can be described in terms of a power - law component , the so - called very low - frequency noise and a cut - off power - law component or high - frequency noise . in the island state
the high frequency noise component is best described in terms of a broken power - law , whereas the very low frequency noise is practically undetectable .
the characteristics of the power spectra also change in correlation with the position of the source in the color - color diagram .
the amplitude of the high frequency noise increases as the @xmath11 parameter decreases .
finally , we have shown that the soft color ( via the @xmath11 parameter ) correlates with the frequency of the khz qpo .
this work was supported by the netherlands foundation for research in astronomy ( astron ) under grant 781 - 76 - 017 , by the netherlands research school for astronomy ( nova ) , and the nwo spinoza grant 08 - 0 to e.p.j .
van den heuvel .
pr acknowledges support from the european union through the training and mobility research network grant erbfmrx / ct98/0195 .
mm is a fellow of the consejo nacional de investigaciones cientficas y tcnicas de la repblica argentina .
this research has made use of data obtained through the high energy astrophysics science archive research center online service , provided by the nasa / goddard space flight center .
bildsten , l. 1993 , apj , 418 , l21 .
campana s. , stella l. , mereghetti s. , colpi m. , tavani m. , ricci d. , dal fiume d. & belloni t. , 1998 , , 499 , l65 .
callanan p.j .
, filippenko a.v . &
garca m.r . , 1999 , iau circ no 7086 chevalier , c. & ilovaisky , s. a. 1991 , a&a , 251 , l11 chevalier c. , ilovaisy s.a . , leisy p. , & patat f. , 1999 , a&a , 347 , l51 cui , w , barret , d. , zhang , s. n. , chen , w. , boirin , l. & swank , j. 1998 , , 502 , l49 ford , e. c. , kaaret , p. , chen , k. , tavani , m. ; barret , d. ; bloser , p. ; grindlay , j. , harmon , b. a. , paciesas , w. s. & zhang , s. n. , 1997 , , 486 , l47 ford , e. c. & van der klis , m. 1998 , , 506 , l39 ford , e. c. , van der klis , m. , van paradijs , j. , mndez m. , wijnands , r. & kaaret p. 1998
, , 508 , l155 hasinger , g. & van der klis , m. 1989 , , 225 , 79 ( hk89 ) hjellming r.m . & han x. , 1995 , in ` x - ray binaries ' , cambridge university press , p.308 kaaret , p. , yu , w. , ford , e. c. , & zhang , n. s. 1998 , , 497 l93 kitamoto s. , tsunemi h. , miyamoto s. , roussel - dupre d. , 1993 , , 403 , 315 .
mndez , m. , van der klis , m. , ford e. c. , wijnands , r. & van paradijs , j. 1999 , , 511 , l49 mndez , m. , van der klis , m. , van paradijs , j. , lewin , w. h. g. , lamb , f. k. , vaughan , b .a . , kuulkers , e. & psaltis , d. 1997 , , 485 , l37 .
priedhorsky , w. c. & terrell , j. 1984 , , 280 , 661 prins , s. & van der klis , m. 1997 , a&a , 319 , 498 schulz , n. s. , hasinger , g. & trmper , j. 1989 , a&a , 225 , 48 shahbaz , t. , casares , j. & charles , p. a. 1997 , a&a , 326 , l5 stella , l. & vietri , m. 1998 , 492 , l59 van der klis , m. 1994 , apjs , 92 , 511 van der klis , m. 1995 , in x - ray binaries , eds . w. h. g. lewin , j. van paradijs & e. p. j. van den heuvel , ( cambridge u. press , cambridge ) p. 252 .
van der klis , m. 1999 , in proc . of the 3rd william
fairbank meeting , rome june 29 - july 4 1998 verbunt f. , belloni t. , johnston h.m .
, van der klis m. & lewin w.h.g . , 1994 ,
a&a , 285 , 903 yoshida k , mitsuda k. , ebisawa k. & done c. , 1993 pasj , 45 , 605 yu , w. , li , t. p. , zhang , w. & zhang , s. n. 1999 , , 512 , l35 zhang , s. n. , yu , w. & zhang , w. 1998 494 , l71 zhang , w. , jahoda , k. , kelley , r. l. , strohmayer , t. e. , swank , j. h. , & zhang , s. n. 1998a , , 495 , l9 cccccccccc + & & & & & + + @xmath24 & rms@xmath25 & @xmath26 & & rms@xmath27 & @xmath28 & @xmath29 & @xmath30 & count rate@xmath31 & @xmath14(dof ) + 1.1 & & & & @xmath32 & @xmath33 & @xmath34 & @xmath35 & 45 & 1.23(82 ) + 1.5 & & & & @xmath36 & @xmath37 & @xmath38 & @xmath39 & 200 & 0.79(82 ) + 1998 march 2 & @xmath20 & 1.5 ( fixed ) & & @xmath40 & @xmath41 & @xmath42 & @xmath43 & 1481 & 0.80(112 ) + + + & & & & & + + @xmath24 & rms@xmath25 & @xmath26 & & rms@xmath27 & @xmath26 & @xmath44 & & count rate@xmath31 & @xmath14(dof ) + @xmath45 & @xmath46 & @xmath47 & & @xmath48 & @xmath49 & @xmath50 & & 1166 & 1.67(81 ) + @xmath51 & @xmath52 & @xmath53 & & @xmath54 & @xmath55 & @xmath56 & & 1629 & 1.99(81 ) + @xmath57 & @xmath58 & @xmath59 & & @xmath60 & @xmath61 & @xmath62 & & 1977 & 1.11(81 ) + @xmath63 & @xmath64 & @xmath65 & & @xmath66 & @xmath67 & @xmath68 & & 2013 & 1.02(81 ) + @xmath69 & @xmath70 & @xmath71 & & @xmath72 & @xmath73 & @xmath74 & & 2169 & 1.46(81 ) + @xmath75 & @xmath76 & @xmath77 & & @xmath78 & @xmath79 & @xmath80 & & 2342 & 1.41(81 ) + | based on _ rossi x - ray timing explorer _ data , we discuss the classification of the soft x - ray transient aquila x1 in the z / atoll scheme , and the relation of its kilohertz quasi - periodic oscillations ( khz qpo ) properties to the x - ray colors .
the color - color diagram shows one elongated ( `` banana '' ) structure and several `` islands '' of data points .
the power spectra of the island are best represented by a broken power - law , whereas those of the banana by a power - law below @xmath0 1 hz plus an exponentially cut - off component at intermediate frequencies ( 3060 hz ) . the parameters of these two components change in correlation with the position of the source in the color - color diagram .
based on the pattern that the source shows in the color - color diagram and its aperiodic variability we conclude that aquila x1 is an atoll source .
we have also investigated the possible correlation between the frequency of the khz qpo and the position of the source in the color - color diagram .
the complexity seen in the frequency _ versus _ count rate diagram is reduced to a single track when the frequency is plotted against hard or soft color . # 1*#1 * |
this is an updated preprint of a paper whose journal version has already appeared in print @xcite .
the main reason for the update was to include references to the papers @xcite , which have independently obtained equivalent results using different methods .
the enumeration of pattern - avoiding permutations has received a considerable amount of attention lately ( see @xcite for a survey ) .
we say that a permutation @xmath4 of order @xmath5 _ contains _ a permutation @xmath6 of order @xmath7 , if there is a sequence @xmath8 such that for every @xmath9 $ ] @xmath10 if and only if @xmath11
. one of the central notions in the study of pattern - avoiding permutations is the _ wilf equivalence _ : we say that a permutation @xmath12 is wilf - equivalent to a permutation @xmath13 if , for every @xmath14 , the number of permutations of order @xmath5 that avoid @xmath12 is equal to the number of permutations of order @xmath5 that avoid @xmath13 . in this paper
, we consider pattern avoidance in matchings .
this is a more general concept than pattern avoidance in permutations , since every permutation can be represented by a matching .
a _ matching _ of size @xmath15 is a graph on the vertex set @xmath16=\{1,2,\dotsc,2m\}$ ] whose every vertex has degree one .
we say that a matching @xmath17 _ contains _ a matching @xmath18 if there is a monotone edge - preserving injection from @xmath19 to @xmath20 ; in other words , @xmath21 contains @xmath22 if there is a function @xmath23 such that @xmath24 implies @xmath25 and @xmath26 implies @xmath27 .
let @xmath21 be a matching of size @xmath15 , and let @xmath28 be an arbitrary edge of @xmath21 .
if @xmath29 , we say that @xmath30 is an _
l - vertex _ and @xmath31 is an _
r - vertex _ of @xmath21 . obviously , @xmath21 has @xmath15 l - vertices and @xmath15 r - vertices .
let @xmath32 and @xmath33 be two edges of @xmath21 , with @xmath34 and @xmath35 , and assume that @xmath36 .
we say that the two edges @xmath37 and @xmath38 _ cross _ each other if @xmath39 , and we say that @xmath38 is _ nested _ below @xmath37 if @xmath40 .
we say that a matching @xmath21 on the vertex set @xmath16 $ ] is _ permutational _ , if for every l - vertex @xmath30 and every r - vertex @xmath31 we have @xmath41 . there is a natural one - to - one correspondence between permutations of order @xmath15 and permutational matchings of size @xmath15 :
if @xmath4 is a permutation of @xmath15 elements , we let @xmath42 denote the permutational matching on the vertex set @xmath16 $ ] whose edge set is the set @xmath43\bigl\}$ ] . in this paper , we will often represent a permutation @xmath4 on @xmath15 elements by the ordered sequence @xmath44 .
thus , for instance , @xmath45 refers to the matching on the vertex set @xmath46 $ ] , with edge set @xmath47 .
figure [ fig - match ] depicts most of the matchings relevant for this paper .
note that a permutational matching @xmath42 contains the permutational matching @xmath48 if and only if @xmath4 contains @xmath6 .
[ cols="^,^,^ " , ] let @xmath49 be an even number . a _
dyck path _ of length @xmath5 is a piecewise linear nonnegative walk in the plane , which starts at the point @xmath50 , ends at the point @xmath51 , and consists of @xmath5 linear segments ( `` steps '' ) , of which there are two kinds : an _ up - step _ connects @xmath52 with @xmath53 , whereas a _ down - step _ connects @xmath52 with @xmath54 .
the nonnegativity of the path implies that among the first @xmath7 steps of the path there are at least @xmath55 up - steps .
let @xmath56 denote the set of all dyck paths of length @xmath57 .
it is well known that @xmath58 , where @xmath59 is the @xmath15-th catalan number .
every dyck path @xmath60 can be represented by a
_ dyck word _ ( denoted by @xmath61 ) , which is a binary word @xmath62 such that @xmath63 if the @xmath30-th step of @xmath64 is an up - step , and @xmath65 if the @xmath30-th step of @xmath64 is a down - step .
it can be easily seen that a word @xmath66 is a dyck word of some dyck path if and only if the following conditions are satisfied : * the length @xmath67 is even . *
the word @xmath68 has exactly @xmath69 terms equal to 1 . *
every prefix @xmath70 of @xmath68 has at most @xmath71 terms equal to 1 .
we will use the term dyck word to refer to any binary word satisfying these conditions .
the set of all dyck words of length @xmath57 will be denoted by @xmath72 .
let @xmath73 denote the set of all matchings on the vertex set @xmath16 $ ] .
for a matching @xmath74 , we define the _ base _ of @xmath21 ( denoted by @xmath75 ) to be the binary word @xmath62 such that @xmath63 if @xmath30 is an l - vertex of @xmath21 , and @xmath65 if @xmath30 is an r - vertex of @xmath21 .
the base @xmath75 is clearly a dyck word ; conversely , every dyck word is a base of some matching . if @xmath63 ( or @xmath65 ) we say that @xmath30 is an _
l - vertex _ ( or an _ r - vertex _ ,
respectively ) _ with respect to the base @xmath68_. let @xmath76 , let @xmath77 be an arbitrary set of matchings , and let @xmath78 ; we define the sets @xmath79 and @xmath80 as follows : @xmath81 let @xmath82 and @xmath83 denote the cardinalities of the sets @xmath73 , @xmath79 and @xmath80 , respectively . the sets @xmath80 form a partition of @xmath79 . in other words , we have @xmath84 where the union and the sum range over all dyck words @xmath78 .
if no confusion can arise , we will write @xmath85 instead of @xmath86 and @xmath87 instead of @xmath88 .
there is an alternative way to encode ordered matchings , which uses transversals of ferrers shapes .
ferrers shape _ is a left - justified array of cells , where the number of cells in a given row does not exceed the number of cells in the row directly above it .
transversal _ of a ferrers shape is a subset of its cells which intersects every row and every column exactly once .
a matching @xmath21 on the vertex set @xmath16 $ ] can be represented by a transversal of a ferrers shape as follows : first , consider a ferrers shape with @xmath15 rows and @xmath15 columns , in which the @xmath30-th row ( counted from the bottom ) has @xmath7 cells if and only if the @xmath30-th r - vertex of @xmath21 has @xmath7 l - vertices to the left of it .
next , we define a transversal of this shape : the transversal contains the cell in row @xmath30 and column @xmath31 if and only if the @xmath30-th r - vertex is connected to the @xmath31-th l - vertex by an edge of @xmath21 . this correspondence establishes a bijection between transversals and matchings , where matchings of a given base correspond to the transversals of a given shape . in the context of pattern - avoiding transversals ,
the equivalence relation @xmath89 is known as shape - wilf equivalence .
note that the permutational matchings correspond to transversals of square shapes , i.e. , to permutation matrices .
the correspondence between matchings and transversals , as well as the ( more general ) correspondence between ordered graphs and nonnegative fillings of ferrers shapes , has been pointed out by krattenthaler @xcite and de mier @xcite . using this correspondence ,
it becomes clear that some previous results on pattern - avoiding graphs @xcite are equivalent to results on pattern - avoiding fillings @xcite .
the aim of this paper is to study the relative cardinalities of the sets @xmath90 , with @xmath91 being a permutational matching with three edges .
for this purpose , we introduce the following notation : let @xmath92 be the quasiorder relation defined as follows : for two sets @xmath77 and @xmath93 of matchings , we write @xmath94 , if for each @xmath76 and each @xmath78 we have @xmath95 .
similarly , we write @xmath96 if @xmath94 and @xmath97 , and we write @xmath98 if @xmath94 and @xmath99 .
as above , we omit the curly braces when the arguments of these relations are singleton sets .
note that two permutations @xmath4 and @xmath6 are wilf - equivalent if and only if for every @xmath76 the equality @xmath100 holds , where @xmath101 is the dyck word consisting of @xmath15 consecutive @xmath102-terms followed by @xmath15 consecutive @xmath103-terms .
thus , if @xmath104 , then @xmath4 and @xmath6 are wilf - equivalent ; however , the converse does not hold in general : it is well known that all the permutations of order three are wilf - equivalent , whereas the results of this paper imply that the permutational matchings of size three fall into three @xmath89-classes .
combining the known results on ferrers transversals and the known results on matchings , the full characterization of the @xmath89 and @xmath105 relations for patterns of size three has been obtained : @xmath106 the equivalence @xmath107 follows from the results of stankova and west @xcite obtained in the context of fillings of ferrers shapes . in @xcite
, the same equivalence is proved in the context of pattern - avoiding matchings as a corollary to the result presented in this paper as theorem [ thm - bij ] .
the equivalence of @xmath108 follows from the more general result @xmath109 , which was proved by chen et al .
@xcite in the context of matchings , and by backelin et al .
@xcite in the context of transversals .
several generalizations of this result are obtained by krattenthaler @xcite ( see also @xcite ) .
the equivalence @xmath110 follows from the general results of backelin et al .
@xcite . in this paper , a different bijective argument is given ( see section [ sec-231 ] ) .
the relation @xmath111 is proved in corollary [ cor-123 - 132 ] of this paper , and a different argument is given in a forthcoming paper by stankova @xcite , which also contains the proof of @xmath112 , completing the classification .
this paper is organized as follows : in section [ sec-132 ] , we prove theorem [ thm - bij ] , which simultaneously implies @xmath107 and @xmath111 . in section [ sec-231 ]
, we present a bijective argument that implies @xmath110 .
since the matching @xmath45 is the mirror image of the matching @xmath113 , it is obvious that @xmath114 is equal to @xmath115 for each @xmath116 . however , there seems to be no straightforward argument demonstrating the stronger fact that @xmath117 . for @xmath118
, we define @xmath119 to be the matching with edge set @xmath120 .
let @xmath121 .
the goal of this section is to prove the following result : [ thm - bij ] @xmath122 .
since all the elements of @xmath123 are symmetric upon mirror reflection , this also proves that @xmath124 and @xmath117 , see corollaries [ cor - calc ] and [ cor - sym ] at the end of this section . throughout this section
, we consider @xmath76 and @xmath78 to be arbitrary but fixed , and we let @xmath49 . for the sake of brevity ,
we write @xmath125 instead of @xmath126 and @xmath127 instead of @xmath128 . for a matching @xmath129 and an arbitrary integer
@xmath130 $ ] , let @xmath131 $ ] denote the subgraph of @xmath132 induced by the vertices in @xmath133 $ ] .
there are three types of vertices in @xmath131 $ ] : * the r - vertices of @xmath132 belonging to @xmath133 $ ] . clearly
, all these vertices have degree one in @xmath131 $ ] .
* the l - vertices of @xmath132 connected to some r - vertex belonging to @xmath133 $ ] .
these have degree one in @xmath131 $ ] as well . * the l - vertices of @xmath132 belonging to @xmath133 $ ] but not connected to an r - vertex belonging to @xmath133 $ ] .
these are the isolated vertices of @xmath131 $ ] , and we will refer to them as the _ stubs _ of @xmath131 $ ] .
let @xmath132 be an arbitrary graph from @xmath134 .
the sequence @xmath135,\
g[2],\ g[3 ] , \dotsc,\ g[n-1],\ g[n]=g\ ] ] will be called _ the construction _ of @xmath132 .
it is convenient to view the construction of @xmath132 as a sequence of steps of an algorithm that produces the matching @xmath132 by adding one vertex in every step .
two graphs @xmath136 from @xmath134 may share an initial part of their construction ; however , if @xmath131\neq g'[k]$ ] for some @xmath7 , then obviously @xmath137\neq g'[j]$ ] for every @xmath138 .
it is natural to represent the set of all the constructions of graphs from @xmath134 by _ the generating tree _ of @xmath134 ( denoted by @xmath139 ) , defined by the following properties : * the generating tree is a rooted tree with @xmath5 levels , where the root is the only node of level one , and all the leaves appear on level @xmath5 . *
the nodes of the tree are exactly the elements of the following set : @xmath140 , \exists g\in{\mathcal{g}}^{m}:\ g'=g[k]\}\ ] ] * the children of a node @xmath141 are exactly the elements of the following set : @xmath142 , \exists g\in{\mathcal{g}}^{m}:\ g'=g[k ] , g''=g[k+1]\}\ ] ] it follows that the level of every node @xmath141 of the tree @xmath143 is equal to the number of vertices of @xmath141 .
also , the leaves of @xmath143 are exactly the elements of @xmath134 , and the nodes of the path from the root to a leaf @xmath132 form the construction of @xmath132 .
the generating tree of @xmath144 , denoted by @xmath145 , is defined in complete analogy with the tree @xmath143 .
our goal is to prove that the two trees are isomorphic , hence they have the same number of leaves , i.e. , @xmath146 .
we say that a graph @xmath141 on the vertex set @xmath133 $ ] is _ consistent _ with @xmath68 , if @xmath147 $ ] for some matching @xmath148 with base @xmath68 .
[ lem - t1t2 ] \1 .
a graph @xmath141 is a node of @xmath139 if and only if @xmath141 satisfies these three conditions : * @xmath141 is consistent with @xmath68 .
* @xmath141 avoids @xmath45 .
* @xmath141 does not contain a sequence of five vertices @xmath149 such that @xmath150 is a stub , while @xmath151 and @xmath152 are edges of @xmath141 .
a graph @xmath153 is a node of @xmath145 if and only if @xmath153 satisfies these three conditions : * @xmath153 is consistent with @xmath68 . *
@xmath153 avoids @xmath123 . * for every @xmath154 , @xmath153 does not contain an induced subgraph isomorphic to @xmath155 $ ] ( by an order preserving isomorphism ) .
in other words , for every @xmath154 , @xmath153 does not contain a sequence of @xmath156 vertices @xmath157 , where @xmath158 is a stub , and the remaining @xmath159 vertices induce the edges @xmath160 .
we first prove the first part of the lemma .
let @xmath141 be a node of @xmath139 .
clearly , @xmath141 satisfies conditions _ a _ and _ b _ of the first part of the lemma .
assume that @xmath141 fails to satisfy condition _
c_. choose @xmath161 such that @xmath147 $ ] for some @xmath130 $ ] .
let @xmath162 denote the r - vertex of @xmath132 connected to @xmath150 .
then @xmath163 , because @xmath150 was a stub of @xmath147 $ ] , which implies that @xmath164 and the six vertices @xmath165 induce a subgraph isomorphic to @xmath45 , which is forbidden .
this shows that the conditions _ a _ , _ b _ and _ c _ are necessary . to prove the converse , assume that @xmath141 satisfies the three conditions , and
let @xmath166 $ ] .
we will extend @xmath141 into a graph @xmath132 with base @xmath68 , by adding the vertices @xmath167 one by one , and each time that we add a new r - vertex @xmath30 , we connect @xmath30 with the smallest stub of the graph constructed in the previous steps .
we claim that this algorithm yields a graph @xmath161 . for contradiction , assume that this is not the case , and that there are six vertices @xmath168 inducing a copy of @xmath45 . by condition _
, we know that these six vertices are not all contained in @xmath141 , which means that @xmath163 . also , by condition _
, we know that @xmath169 . in the step of the above construction when we added the r - vertex @xmath170 , both @xmath150 and @xmath171 were stubs .
since @xmath170 should have been connected to the smallest available stub , it could not have been connected to @xmath171 , which contradicts the assumption that @xmath172 induce a copy of @xmath45 .
thus @xmath161 , as claimed .
the proof of the second part of the lemma follows along the same lines . to see that the conditions _ a _ , _ b _ and _ c _ of the second part are sufficient , note that every graph satisfying these conditions can be extended into a graph @xmath173 by adding new vertices one by one , and connecting every new r - vertex to the biggest stub available when the r - vertex is added .
we omit the details .
let @xmath141 be a node of @xmath139 .
we define a binary relation @xmath174 on the set of stubs of @xmath141 by the following rule : @xmath175 if and only if either @xmath176 or there is an edge @xmath177 such that @xmath178 and @xmath179 . . here
@xmath180 , assuming @xmath181 and @xmath182 are stubs . ] let @xmath153 be a node of @xmath145 .
we define a binary relation @xmath183 on the set of stubs of @xmath153 by the following rule : @xmath180 if and only if either @xmath176 or @xmath153 contains a sequence of edges @xmath184 , where @xmath185 , @xmath186 , the edge @xmath187 crosses the edge @xmath188 for each @xmath189 , and at the same time @xmath190 and @xmath191 ( see fig .
[ fig - eqst ] ; note that we may assume , without loss of generality , that the edge @xmath187 does not cross any other edge of the sequence except for @xmath192 and @xmath188 , and that no two edges of the sequence are nested : indeed , a minimal sequence @xmath193 witnessing @xmath180 clearly has these properties ) .
we remark that the relation @xmath183 has an intuitive geometric interpretation : assume that the vertices of @xmath153 are represented by points on a horizontal line , ordered left - to - right according to the natural order , and assume that every edge of @xmath153 is represented by a half - circle connecting the corresponding endpoints .
then @xmath180 if and only if every vertical line separating @xmath181 from @xmath182 intersects at least one edge of @xmath153 . using condition _
c _ of the first part of lemma [ lem - t1t2 ] , it can be easily verified that for every node @xmath141 of the tree @xmath139 , the relation @xmath174 is an equivalence relation on the set of stubs of @xmath141 .
let @xmath194 denote the block of @xmath174 containing the stub @xmath195 .
clearly , the blocks of @xmath174 are contiguous with respect to the ordering @xmath196 of the stubs of @xmath141 ; i.e. , if @xmath197 are three stubs of @xmath141 , then @xmath198 implies @xmath199 .
similarly , @xmath183 is an equivalence relation on the set of stubs of a node @xmath153 of @xmath145 ( notice that , contrary to the case of @xmath174 , the fact that @xmath183 is an equivalence relation does not rely on the particular properties of the nodes of @xmath145 described in lemma [ lem - t1t2 ] ) .
the block of @xmath183 containing @xmath195 will be denoted by @xmath200 .
these blocks are contiguous with respect to the ordering @xmath196 as well .
[ lem - child ] \1 .
let @xmath141 be a node of level @xmath201 in the tree @xmath139 .
the following holds : * let @xmath202 be an arbitrary child of @xmath141 in the tree @xmath139 .
this implies that @xmath203 $ ] .
if the vertex @xmath204 is an l - vertex with respect to @xmath68 , then @xmath202 is the only child of @xmath141 , and @xmath204 is a stub in @xmath202 . in this case ,
@xmath205 for every stub @xmath195 of @xmath141 , and @xmath206 . on the other hand ,
if @xmath204 is an r - vertex , then in the graph @xmath202 the vertex @xmath204 is connected to a vertex @xmath195 satisfying @xmath207 . in this case , we have @xmath208 whenever @xmath209 , and all the stubs @xmath210 of @xmath202 form a single @xmath211-block in @xmath202 . *
if @xmath204 is an r - vertex , then for every stub @xmath195 satisfying @xmath207 , @xmath141 has a child @xmath202 which contains the edge @xmath212 .
this implies , together with part a , that if @xmath204 is an r - vertex , then the number of children of @xmath141 in @xmath139 is equal to the number of its @xmath211-blocks .
let @xmath153 be a node of level @xmath201 in the tree @xmath145 .
the following holds : * let @xmath213 be an arbitrary child of @xmath153 in the tree @xmath145 .
this implies that @xmath214 $ ] .
if the vertex @xmath204 is an l - vertex with respect to @xmath68 , then @xmath213 is the only child of @xmath153 , and @xmath204 is a stub in @xmath213 . in this case ,
@xmath215 for every stub @xmath195 of @xmath153 , and @xmath216 . on the other hand ,
if @xmath204 is an r - vertex , then in the graph @xmath213 the vertex @xmath204 is connected to a vertex @xmath195 satisfying @xmath217 . in this case , we have @xmath218 whenever @xmath209 and @xmath219 , and all the other stubs of @xmath213 form a single @xmath220-block in @xmath213 .
* if @xmath204 is an r - vertex , then for every stub @xmath195 satisfying @xmath217 , @xmath153 has a child @xmath213 which contains the edge @xmath212 .
this implies , together with part a , that if @xmath204 is an r - vertex , then the number of children of @xmath153 in @xmath145 is equal to the number of its @xmath220-blocks .
we first prove part @xmath221 .
the case when @xmath204 is an l - vertex follows directly from the definition of @xmath174 , so let us assume that @xmath204 is an r - vertex , and let @xmath195 be the vertex connected to @xmath204 in @xmath202 . assume , for contradiction , that @xmath222 , and choose @xmath223 such that @xmath209 .
since @xmath224 , @xmath141 must contain an edge @xmath225 , with @xmath226 .
then the five vertices @xmath227 form in @xmath202 a configuration that was forbidden by lemma [ lem - t1t2 ] , part @xmath228 .
this shows that @xmath207 .
the edge @xmath212 guarantees that all the stubs larger than @xmath195 are @xmath211-equivalent in @xmath202 , whereas the equivalence classes of the stubs smaller than @xmath195 are unaffected by this edge .
this concludes the proof of part @xmath221 . to prove part @xmath229
, it is sufficient to show that after choosing a vertex @xmath195 such that @xmath207 and adding the edge @xmath212 to @xmath141 , the resulting graph @xmath202 satisfies the three conditions of the first part of lemma [ lem - t1t2 ] .
condition @xmath221 of lemma [ lem - t1t2 ] is satisfied automatically .
if @xmath202 fails to satisfy condition @xmath229 , then @xmath141 fails to satisfy one of the conditions @xmath229 and @xmath228 of that lemma , which is impossible .
similarly , if @xmath202 fails to satisfy condition @xmath228 , then either @xmath141 fails to satisfy this condition as well , or @xmath141 contains a stub @xmath230 with @xmath209 and @xmath224 , contradicting our choice of @xmath195 .
the proof of the second part of this lemma follows along the same lines as the proof of the first part , and we omit it .
we are now ready to state and prove the main theorem of this section .
[ thm - m132 ] the trees @xmath139 and @xmath145 are isomorphic .
our aim is to construct a mapping @xmath231 with the following properties : * the mapping @xmath231 maps the nodes of @xmath139 to the nodes of @xmath145 , preserving their level . *
if @xmath141 is a child of @xmath132 in @xmath143 , then @xmath232 is a child @xmath233 in @xmath234 .
furthermore , if @xmath235 and @xmath236 are two distinct children of a node @xmath132 in @xmath139 , then @xmath237 and @xmath238 are two distinct children of @xmath233 in @xmath145 .
* let @xmath132 be an arbitrary node of @xmath139 , and let @xmath239 .
let @xmath240 , @xmath241 @xmath242 be the sequence of all the distinct blocks of @xmath174 in @xmath132 , uniquely determined by the condition @xmath243 .
similarly , let @xmath244 , @xmath245 @xmath246 be the sequence of all the distinct blocks of @xmath183 in @xmath247 , uniquely determined by the condition @xmath248
. then @xmath249 and @xmath250 for each @xmath251 $ ] .
these conditions guarantee that @xmath231 is an isomorphism , because , thanks to lemma [ lem - child ] , we know that the number of children of each node of @xmath139 ( or @xmath145 ) at level @xmath7 is either equal to one if @xmath204 is an l - vertex or equal to the number of blocks of its @xmath174 relation ( or @xmath183 relation , respectively ) if @xmath204 is an r - vertex . the mapping @xmath231 is defined recursively for nodes of increasing level .
the root of @xmath139 is mapped to the root of @xmath145 .
assume that the mapping @xmath231 has been determined for all the nodes of @xmath139 of level at most @xmath7 , for some @xmath252 $ ] , and that it does not violate the properties stated above .
let @xmath132 be a node of level @xmath7 , let @xmath239 .
if @xmath204 is an l - vertex , then @xmath132 has a unique child @xmath141 and @xmath247 has a unique child @xmath153 . in this case , define @xmath253 .
let us now assume that @xmath204 is an r - vertex .
let @xmath240 , @xmath241 @xmath242 be the sequence of all the distinct blocks of @xmath174 on @xmath132 , with @xmath243 .
we may assume , without loss of generality , that @xmath254 for @xmath251 $ ] .
by assumption , @xmath183 has @xmath255 blocks on @xmath247 .
let @xmath244 , @xmath245 @xmath256 be the sequence of these blocks , where @xmath257 and @xmath258 for every @xmath259 $ ] . by lemma [ lem - child ] , the nodes @xmath132 and @xmath247 have @xmath255 children in @xmath139 and @xmath145 .
let @xmath260 be the graph obtained from @xmath132 by addition of the edge @xmath261 , let @xmath262 be the graph obtained from @xmath247 by addition of the edge @xmath263 , for @xmath259 $ ] . by lemma [ lem - child ] , the graphs @xmath264\}$ ] ( or @xmath265\}$ ] ) are exactly the children of @xmath132 ( or @xmath247 , respectively ) .
we define @xmath266 .
the @xmath211-blocks of @xmath260 are exactly the sets @xmath240 , @xmath267 @xmath268 and @xmath269 , while the @xmath220-blocks of @xmath262 are exactly the sets @xmath244 , @xmath270 @xmath271 and @xmath272 .
this implies that the corresponding blocks of @xmath260 and @xmath262 have the same number and the same size , as required ( note that if @xmath273 and @xmath274 , then the last block in the above list of @xmath211-blocks of @xmath260 is empty ; however , this happens if and only if the last entry in the list of @xmath220-blocks of @xmath262 is empty as well , so it does not violate the required properties of @xmath231 ) .
this concludes the proof of the theorem .
[ cor - calc ] @xmath275 . since @xmath276 is equal to the number of leaves of the tree @xmath139 , and @xmath277 is equal to the number of leaves of the tree @xmath145 , this is a direct consequence of theorem [ thm - m132 ] .
[ cor - sym ] @xmath117 .
let @xmath278 denote the dyck word defined by the relation @xmath279 if and only if @xmath280 . by inverting the linear order of the vertices of a matching @xmath21 with base @xmath68
, we obtain a matching @xmath281 with base @xmath278 . since every matching @xmath282 satisfies @xmath283 , we know that a matching @xmath21 avoids @xmath123 if and only if @xmath281 avoids @xmath123 , and hence @xmath284 .
note that @xmath285 .
this gives @xmath286 as claimed .
[ cor-123 - 132 ] @xmath287 .
notice that @xmath288 , and all the other graphs in @xmath123 avoid @xmath289 .
this implies that @xmath290 , and for every @xmath291 there is a @xmath78 for which this is a proper inclusion , because @xmath292 clearly belongs to @xmath293 .
the claim follows , as a consequence of corollary [ cor - calc ] .
in this section , we prove that @xmath295 .
this result is a special case of a more general theorem by backelin et al .
the new proof we present here provides a simple bijection between @xmath294-avoiding matchings of a fixed base @xmath68 and dyck paths not exceeding the path representing the dyck word @xmath68 .
we first introduce some notation : recall that @xmath296 denotes the set of all dyck paths of length @xmath57 . for two dyck paths @xmath297 and @xmath298 of length @xmath57 , we say that @xmath299 is a _ non - crossing pair _ if @xmath298 never reaches above @xmath297 .
let @xmath300 denote the set of all the non - crossing pairs of dyck paths of length @xmath57 and , for a dyck word @xmath68 of length @xmath57 , let @xmath301 be the set of all the pairs @xmath302 whose first component @xmath297 is the path represented by the dyck word @xmath68 .
recently , chen , deng and du @xcite have proved that @xmath108 by a bijective construction involving dyck paths .
their proof in fact shows that the cardinality of the set @xmath303 is equal to the number of matchings with base @xmath68 avoiding @xmath289 , and at the same time equal to the number of matchings with base @xmath68 avoiding @xmath304 . in our notation
, this corresponds to the following claim : @xmath305 in this section , we extend these equalities to the matching @xmath294 by proving @xmath306 .
this shows that @xmath307 .
we remark that the number of non - crossing pairs of dyck paths of length @xmath57 ( and hence the number of @xmath21-avoiding matchings of size @xmath15 , where @xmath21 is any of @xmath308 or @xmath294 ) is equal to @xmath309 , where @xmath310 is the @xmath15-th catalan number ( see @xcite ) .
the sequence @xmath311 is listed as the entry a005700 in the on - line encyclopedia of integer sequences @xcite .
it is noteworthy , that n. bonichon @xcite has shown a completely different combinatorial interpretation of this sequence , in terms of realizers of plane triangulations .
let us fix @xmath76 and @xmath78 .
let @xmath21 be a matching with base @xmath68 .
let @xmath312 denote the sequence of all the l - vertices with respect to @xmath68 , and @xmath313 be the sequence of all the r - vertices with respect to @xmath68 .
let @xmath314 be the neighbour of @xmath315 in @xmath21 .
an edge @xmath316 of @xmath21 is called _ short _ if @xmath317 , and it is called _ long _ if @xmath318 .
let @xmath319 and @xmath320 denote the set of the short edges and long edges , respectively , so that we have @xmath321 .
an l - vertex @xmath322 is called _ short _ ( or _ long _ ) if it is incident with a short edge ( or a long edge , respectively ) .
let @xmath21 be a matching with base @xmath68 .
@xmath21 avoids @xmath294 if and only if @xmath21 satisfies the following three conditions : * the subgraph of @xmath21 induced by the short edges avoids @xmath294 . *
the subgraph of @xmath21 induced by the long edges avoids @xmath294 .
* every short l - vertex precedes all the long l - vertices .
the first two conditions are clearly necessary .
the third condition is necessary as well , for if @xmath21 contained an edge @xmath323 and an edge @xmath324 with @xmath325 , then the six vertices @xmath326 would induce a copy of @xmath294 . to see that the three conditions are sufficient , assume for contradiction that a graph @xmath21 satisfies these conditions but contains the forbidden configuration induced by some vertices @xmath327 .
we first note that @xmath328 : indeed , it is impossible to have @xmath329 , because @xmath330 is not connected to the leftmost vertex , and the inequality @xmath331 would imply that all the three edges of the forbidden configuration are short , which is ruled out by the first condition of the lemma .
thus , the edge @xmath332 is long , and hence @xmath333 is long as well , by the third condition .
this implies that @xmath334 , hence @xmath335 as well , and all the three edges of the configuration are long , contradicting the second condition of the lemma . to construct the required bijection between @xmath336 and @xmath301
, we will use the intuitive notion of a `` tunnel '' in a dyck path , which has been employed in bijective constructions involving permutations in , e.g. , @xcite or @xcite .
let @xmath337 be a dyck path .
a _ tunnel _ in @xmath337 is a horizontal segment @xmath338 whose left endpoint is the center of an up - step of @xmath337 , its right endpoint is the center of a down - step of @xmath337 , and no other point of @xmath338 belongs to @xmath337 ( see fig .
[ fig - tunnel ] ) .
a path @xmath339 has exactly @xmath15 tunnels .
an up - step @xmath181 and a down - step @xmath340 of @xmath337 are called _ partners _ if @xmath337 has a tunnel connecting @xmath181 and @xmath340 .
let @xmath341 denote the up - steps of @xmath337 and @xmath342 denote the down - steps of @xmath337 , in the left - to - right order in which they appear on @xmath337 .
let @xmath343 be the dyck path represented by the dyck word @xmath68 , let @xmath344 be a non - crossing pair of dyck paths .
let @xmath345 be the unique matching with base @xmath68 satisfying the condition that @xmath316 is an edge of @xmath21 if and only if @xmath346 is the partner of @xmath347 . to see that this definition is valid , we need to check that if @xmath346 is partnered to @xmath347 in a path @xmath337 not exceeding @xmath348 , then @xmath349 in the matchings with base @xmath68 .
this is indeed the case , because the horizontal coordinate of @xmath350 ( which determines the position of @xmath322 in the matching ) does not exceed the horizontal coordinate of @xmath346 , while the horizontal coordinate of @xmath347 does not exceed the horizontal coordinate of @xmath351 ( note that a half - line starting in the center of @xmath346 directed north - west intersects @xmath348 in the center of @xmath350 ; similarly , a half - line starting in the center of @xmath347 directed north - east hits the center of @xmath351 ) .
see fig .
[ fig - tuhr ] . and a matching @xmath345 .
a tunnel between @xmath346 and @xmath347 corresponds to an edge @xmath316 .
the filled dots above the pair of paths represent the l - vertices of the matching , the empty dots represent the r - vertices . ] [ lem - gpm231 ] if @xmath344 , then @xmath345 avoids @xmath294 .
choose an arbitrary @xmath344 and assume , for contradiction , that there are six vertices @xmath327 in @xmath345 which induce the forbidden configuration .
let @xmath352 and @xmath353 be the tunnels corresponding to the three edges @xmath354 , @xmath355 and @xmath356 , respectively .
note that the projection of @xmath357 onto some horizontal line @xmath358 is a subset of the projections of @xmath359 and @xmath353 onto @xmath358 .
thus , the three tunnels lie on different horizontal lines and there is a vertical line intersecting all of them . since @xmath360
, the tunnel @xmath359 must lie below @xmath353 , otherwise the subpath of @xmath337 between @xmath361 and @xmath362 would intersect @xmath359 .
on the other hand , @xmath363 implies that @xmath359 lies above @xmath353 , a contradiction .
the aim of the next lemma is to show that the mapping @xmath364 can be inverted .
[ lem - indm231 ] for every @xmath365 there is a unique dyck path @xmath337 such that @xmath366 and @xmath367 .
we proceed by induction on @xmath15 .
the case @xmath368 is clear , so let us assume that @xmath369 and that the lemma holds for every @xmath370 and every @xmath371 .
let us choose an arbitrary @xmath78 , and an arbitrary @xmath365 , and define @xmath7 such that @xmath372 is an edge of @xmath21 .
let @xmath373 be the matching from @xmath374 that is isomorphic to the subgraph of @xmath21 induced by the short edges , let @xmath375 be isomorphic to the subgraph induced by the long ones , let @xmath376 and @xmath377 be the respective bases of @xmath373 and @xmath378 , and let @xmath379 and @xmath380 be the dyck paths corresponding to @xmath376 and @xmath377 . by induction , we know that @xmath381 and @xmath382 for some dyck paths @xmath383 and @xmath384 , where @xmath383 does not exceed @xmath379 , and @xmath384 does not exceed @xmath380 . let @xmath385 be the dyck word @xmath386 , and let @xmath387 be the corresponding dyck path .
note that @xmath387 does not exceed @xmath348 : assume that @xmath348 has @xmath338 up - steps occurring before the @xmath7-th down - step ; then @xmath387 is obtained from @xmath348 by omitting the @xmath388 up - steps @xmath389 , and inserting @xmath388 new up - steps directly after the @xmath7-th down - step .
let @xmath337 be the dyck path obtained by concatenating the following pieces : * an up - step from @xmath50 to @xmath390 * a shifted copy of @xmath383 from @xmath390 to @xmath391 * a down - step from @xmath391 to @xmath392 * a shifted copy of @xmath384 from @xmath392 to @xmath393 since @xmath337 clearly does not exceed @xmath387 , it does not exceed @xmath348 either .
let us check that @xmath367 : * the base of @xmath21 is equal to the base of @xmath345 .
thus , to see that @xmath21 is equal to @xmath345 , it suffices to check that @xmath373 and @xmath378 are isomorphic to the matchings induced by the short edges of @xmath345 and the long edges of @xmath345 , respectively . *
the up - step @xmath394 is clearly partnered to the down - step @xmath395 ( which connects @xmath391 to @xmath392 ) .
thus , @xmath345 contains the edge @xmath396 .
it follows that @xmath345 has @xmath397 short edges , incident to the l - vertices @xmath398 and r - vertices @xmath399 . *
the @xmath397 up - steps @xmath400 as well as the @xmath397 down - steps @xmath401 all belong to the shifted copy of @xmath383 .
since shifting does not affect the partnership relations , we see that the short edges of @xmath345 form a matching isomorphic to @xmath381 . * similarly , the up - steps @xmath402 @xmath403 are partnered to the down - steps @xmath404 , @xmath405 according to the tunnels of @xmath384 .
the corresponding long edges form a matching isomorphic to @xmath378 .
it follows that @xmath367 .
we now show that @xmath337 is determined uniquely : assume that @xmath406 for some @xmath407 .
since @xmath408 , the path @xmath409 must contain a down step from @xmath391 to @xmath392 , and this down - step must be the first down - step of @xmath409 to reach the line @xmath410 .
this shows that the subpath of @xmath409 between @xmath390 and @xmath391 is a shifted copy of some dyck path @xmath411 .
the tunnels of this path must define a matching isomorphic to @xmath412 . by induction , we know that @xmath383 is determined uniquely , hence @xmath413 . by the same argument
, we see that the subpath of @xmath409 from @xmath392 to @xmath393 is a shifted copy of @xmath384 .
this shows that @xmath414 , and @xmath337 is unique , as claimed .
we are now ready to prove the following theorem : [ thm - m231 ] for each @xmath76 and for each @xmath78 , @xmath415 is equal to @xmath416 .
putting together lemma [ lem - gpm231 ] and lemma [ lem - indm231 ] , we infer that the function that maps a pair @xmath417 to the matching @xmath345 is a bijection between @xmath301 and @xmath336 .
this gives the required result .
@xmath295 . this a direct consequence of theorem [ thm - m231 ] and the results of chen , deng and du @xcite .
we have introduced an equivalence relation @xmath89 on the set of permutational matchings , and we have determined how this equivalence partitions the set of permutational matchings of order 3 . however , many natural questions remain unanswered . for instance , it would be nice to have an estimate on the number of blocks of @xmath89 belonging to the set @xmath73 . also , is it possible to characterize the minimal and the maximal elements of @xmath418 ?
i am grateful to martin klazar for his comments and suggestions .
i would also like to thank anna de mier and zvezdelina stankova for pointing out useful references to related research . | how many matchings on the vertex set @xmath0 avoid a given configuration of three edges ?
chen , deng and du have shown that the number of matchings that avoid three nesting edges is equal to the number of matchings avoiding three pairwise crossing edges . in this paper
, we consider other forbidden configurations of size three .
we present a bijection between matchings avoiding three crossing edges and matchings avoiding an edge nested below two crossing edges .
this bijection uses non - crossing pairs of dyck paths of length @xmath1 as an intermediate step .
apart from that , we give a bijection that maps matchings avoiding two nested edges crossed by a third edge onto the matchings avoiding all configurations from an infinite family @xmath2 , which contains the configuration consisting of three crossing edges .
we use this bijection to show that for matchings of size @xmath3 , it is easier to avoid three crossing edges than to avoid two nested edges crossed by a third edge . in this updated version of this paper , we add new references to papers that have obtained analogous results in a different context . |
the da@xmath0ne @xmath8 @xcite is successfully running in frascati since 1999 .
it s particular design , with two different rings intersecting only in the two interaction regions , was intended to reach the extremely high luminosity of @xmath9 , allowing 120 bunches per ring with a lifetime of 2 hours and an average luminosity of 5 @xmath10 per year . during the year 2002 , in the kloe interaction region
, a luminosity of @xmath1 has been reached using 45 bunches per ring .
beam lifetime has reached the 40 minutes .
the total integrated luminosity , optimized by injecting new bunches during the data taking , has been of 0.3 @xmath10 in 140 days corresponding to @xmath11 , more than the double of what had been obtained in 2001 . in the last months of year 2002 , in the other interaction region ,
the dear experiment has collected 77 @xmath12 with a maximum luminosity of @xmath13 and 100 bunches circulating in each ring @xcite . in the year
2003 major changes have taken place in da@xmath0ne optics , in particular for the kloe interaction region the layout of the low beta permanent quadrupoles has been substantially modified for easier tuning of the machine . the new da@xmath0ne setup aims to reach a luminosity of @xmath14 .
the finuda experiment has been installed in the interaction region previuosly occupied by the dear experiment . in the fall of 2003 kloe
will restart taking data with the aim of collecting 2 @xmath10 before the end of 2004 .
the analysis of the data already collected by the dear and kloe experiments has produced the results that will be shown in the following .
the dear ( @xmath15 ) experiment @xcite has the goal of measuring the isospin dependent antikaon - nucleon scattering lengths , @xmath5 and @xmath16 , via the measurement of the @xmath17 line shift ( @xmath18 ) and width ( @xmath19 ) in kaonic hydrogen and deuterium . from the deser - trueman formulas @xcite one has for kaonic hydrogen : @xmath20 and for kaonic
deuterium : @xmath21 and , extracting @xmath22 from @xmath23 , one gets : @xmath24 these amplitudes give crucial informations about the meson - nucleon sigma terms and the strangeness content of the proton . the experimental setup contains a pressurized cryogenic target cell with a diameter of 11 cm and a thickness of 75 @xmath25 m of kapton where the kaons , produced in the @xmath0 decays , are stopped and captured to form kaonic atoms .
the photons produced in the de - excitation of these atoms are detected by 16 ccd-55 ( _ charged coupled devices _ ) . in a first period of data taking a nitrogen target
has been used to tune up the apparatus : the yield of kaonic nitrogen transitions is indeed 20 times higher than the kaonic hydrogen ones . using the @xmath26 collected in the october 2002 ,
the yields of the following 3 transitions have been measured for the first time ( fig.[dearkn ] ) , obtaining the following results@xcite : @xmath27 [ thb ] using the kn data the feasibility of a precision measurement of the charged kaon mass has also been proved@xcite . in december 2002 , dear has collected @xmath28 with an hydrogen target to measure the @xmath17 line of kh .
in addition , background data with no collision , equivalent to @xmath29 in collision , have been collected for the study of the background .
two different analyses based on the global fit of the kh spectrum before and after background subtraction have given as preliminary results for the @xmath17 line shift and width@xcite : @xmath30 an upgrade of the dear experiment called siddharta ( silicon drift detector for hadronic atoms resarch by timing application ) is under study .
the new setup will use fast triggerable devices , namely silicon drift detectors , the trigger being given by the charged kaons entering the cell .
a precision at the ev level is aimed on @xmath17 lines of kaonic hydrogen and deuterium .
the observation of other exotic atoms will also be possible .
kloe ( k long experiment ) is a general purpose detector @xcite whose characteristics have been optimized for the study of the cp violating decays of the neutral kaons .
the main components of the detector are ( fig.[kloedet ] ) : * a _ superconducting solenoid _ giving a magnetic field of 0.52 t. * the _ drift chamber _
@xcite : a cylindrical shape ( @xmath31=4 m , @xmath32=3.3 m ) filled with a gas mixture , 90% he+10% isobutane , instrumented with 12582 stereo sense wires .
it gives a momentum resolution @xmath33% for tracks with @xmath34 mev and @xmath35 .
the spatial resolution is 150 @xmath25 m in the radial direction and 2 mm in the longitudinal one . *
the _ electromagnetic calorimeter _
@xcite : it s a 15@xmath36 sampling calorimeter with lead and scintillating fibers read at both sides by photomultipliers . with its cylindrical shape it covers 98% of the solid angle .
the energy resolution is @xmath37 .
the time resolution is @xmath38 .
the spatial resolution for completely neutral vertices , e.g. @xmath39 , reconstructed inside the drift chamber is @xmath40 cm . *
the _ quadrupole calorimeter _ @xcite : two smaller sampling calorimeters with lead and scintillating tiles , surrounding the low @xmath41 quadrupoles to complete the calorimetric hermeticity .
[ tbh ] most of the @xmath0 mesons decay in charged kaons ( 49% ) or neutral kaons ( 34% ) , so that kaon physics is the largest part of kloe program . other particles , produced in @xmath0 decays with enough abudance to improve the existing experimental measurements , are @xmath2 ( @xmath4215% ) , @xmath3 ( @xmath421.3% ) , @xmath4 ( @xmath43 ) , @xmath5 ( @xmath43 ) and @xmath6 ( @xmath43 ) .
furthermore , thanks to the _ radiative return _ that reduces the center of mass energy of the colliding beams , kloe can measure the hadronic cross section , @xmath44 , in the crucial region from the @xmath45 threshold to the @xmath0 resonance .
results presented in the following are obtained from the analysis of 2000 data ( 25 @xmath12 ) and , in some cases , from the preliminary analysis of 2001 ( 170 @xmath12 ) and 2002 ( 280 @xmath12 ) data .
kaons are produced at a @xmath0-factory in a pure antisymmetric initial state with quantum numbers @xmath46 . at da@xmath0ne
the @xmath0 mesons are produced pratically at rest , with only a small momentum ( @xmath4213 mev / c ) in the plane of the orbit , so that the kaons are produced practically back to back . in the case of neutral kaons
one has : @xmath47 this simultaneous production allows for a clean tagging of @xmath48(@xmath49 ) looking at the @xmath49(@xmath48 ) on the other side , having also a precise information on the tagged kaon momentum .
the two main tagging algorithms in kloe are : * @xmath50 * tagged * by @xmath49 interaction in the calorimeter ( @xmath51 ) : @xmath49 interactions are identified using the time of flight and some additional request on the energy deposit@xcite . the tagging efficiency , mostly geometrical , is @xmath52% .
the @xmath48 momentum is obtained with a resolution of @xmath53 mev and an angular resolution of @xmath54 ( @xmath55 in @xmath56 ) .
* @xmath57 * tagged * by @xmath58 vertex close to the interaction point : @xmath48 decays are identified asking for the @xmath48 invariant mass and momentum .
tagging efficiency , mainly geometrical , is @xmath59% .
the resolution on @xmath49 momentum is @xmath53 mev and the angular resolution is @xmath54 .
using just the @xmath60 @xmath61 tagged @xmath48 from 17@xmath62 collected in 2000 , kloe has measured the ratio of the partial widths of the main @xmath48 decay modes@xcite : @xmath63 the measurement is fully inclusive with respect to the radiated photon energy .
the ratio in ( [ ksratio ] ) is important not only because it enters in the _
double ratio _ used to measure @xmath64 , but also because it can be used to derive the strong phases in the amplitudes @xmath65 , investigating the presence of _ isospin - breaking _ electromagnetic phase shifts .
other interesting @xmath48 branching ratios measured by kloe are : @xmath66 the results are obtained from a still preliminary analysis of 170 @xmath12 collected in 2001 and are consistent with the ones published using the 2000 data sample @xcite . combining the results in ( [ brksen ] )
one gets the first preliminary measurement of @xmath48 semileptonic asymmetry : @xmath67 this asymmetry can be compared with the value of the @xmath49 semileptonic asymmetry @xcite to perform a nice test of cpt conservation .
@xmath48 semileptonic decays can also be used to test the @xmath68 rule which forbids decays like @xmath69 .
the @xmath70 violation is commonly expressed via : @xmath71 if cpt holds , @xmath70 implies @xmath72 .
the standard model allows @xmath70 violation only at next to leading order predicting @xmath73 .
kloe preliminary result ( 2001 data sample ) is : @xmath74 in agreement with the comparabale result obtained by cplear @xcite .
using an energy scan around the @xmath0 peak , kloe has also measured the @xmath48 mass@xcite : @xmath75 by normalizing the energy scale to the precise determination of the @xmath0 mass obtained by resonant beam depolarisation at novosibirsk@xcite .
@xmath49 neutral decays are reconstructed using the momentum information obtained from the @xmath76 vertex and the time and position of the energy deposited in the calorimeter by the photons produced in the decays .
these informations are sufficient to completely reconstruct neutral vertices like @xmath77 or @xmath78 with a spatial resolution of @xmath40 cm .
exploiting the simple dynamics of the @xmath79 decay it s possible to obtain a precise measurement of the corresponding branching ratio@xcite ( 362 @xmath12 analyzed ) : @xmath80 in agreement with the recent na48 result@xcite .
preliminary results on @xmath49 charged decays are also in good agreement with the pdg values@xcite .
in particular , a final error much better than 1% is expected on @xmath81 decays , improving the error on the kaon form factors ( @xmath82 , @xmath83 ) , the @xmath49 lifetime ( @xmath84 ) and , finally , on the ckm matrix element @xmath85 .
charged kaons are tagged using the two body decays into @xmath86 and @xmath87 .
again one kaon is tagged looking at the other kaon on the opposite side .
using the momentum and time of flight informations obtained by the drift chamber and the calorimeter it s possible to isolate the different decay channels@xcite .
in particular for the @xmath88 channel one gets : @xmath89 looking at the charge asymmetries of the slopes of the dalitz plot for this channel it s possible to have an evidence of direct cp violation in charged kaon decays .
a preliminary fit of the dalitz plot with just @xmath90 has obtained : @xmath91 from the fit of the dalitz plot for the @xmath92 decays , it s possible to obtain the @xmath2-meson parameters for its three charge states and the cross section for @xmath93 with @xmath94 .
the result obtained with 17 @xmath12 from 2000 data sample is@xcite : @xmath95 using 2000 data kloe has also measured the relative branching ratio for @xmath0 decays in @xmath96 and @xmath97@xcite : @xmath98 from where one can extract the pseudoscalar mixing angle ( @xmath99 ) and the gluon content in @xmath4 ( @xmath100 ) : @xmath101 scalar mesons are detected at kloe through their decay channels : @xcite and @xmath102 @xcite . from the integration of the @xmath103 and @xmath104 mass spectra of 2001 and 2002 , these preliminary updated branching ratios are found : @xmath105 a fit of the dalitz plots to extract the @xmath5 and @xmath6 contributions is in progress . the determination of hadronic cross section @xmath106 in the energy region from @xmath45 threshold to the @xmath0 resonance is still one of the main source of error for the theoretical estimate of the muon anomaly ( @xmath107 ) @xcite .
furthermore , recent results obtained in this region from @xmath108 and @xmath109 data show an inconsistency of some percent @xcite . despite of the fact that da@xmath0ne runs at fixed energy , @xmath110 , kloe can measure the hadronic cross section via the _ radiative return_@xcite : initial state radiation ( isr ) from the beams can lower the value of @xmath111 down to the @xmath112 threshold . kloe @xmath113 events are selected asking for two charged tracks coming from the interaction region with an angle respect to the beam pipe @xmath114 .
additional cuts on the momentum , @xmath115 mev or @xmath116 mev , are used to reject tracks spiraling along the beam line . in order to suppress the events in which the @xmath45 invariant mass is reduced by the presence of a final state radiated ( fsr ) photon , a cut on the missing momentum angle , @xmath117 ,
is imposed .
background from bhabha scattering , @xmath92 decays and @xmath118 events is suppressed using a particle identification method based on the time of flight , the energy deposit in the calorimeter and the kinematic closure of the event in the hypothesis of only one photon in the final state@xcite .
residual background is evaluated from the fit of the mass distribution of the charged particles and subtracted .
the yield of the selection is @xmath119 .
the cross section is obtained by dividing for the luminosity measured with large angle bhabha events using the babayaga generator@xcite .
the @xmath7 cross section is obtained from the @xmath120 cross section by dividing for the _ radiation function _ ( fig.[shad].a ) given by the phokhara event generator@xcite used with @xmath121 and vacuum polarization off and multiplying by the appropriate kinematic factor . after correcting for the vacuum polarization the bare @xmath7 cross section shown in fig.[shad].b
is obtained .
[ thb ] the corresponding preliminary value for the @xmath45 contribution to @xmath107 in the interval @xmath122 is : @xmath123 where the theoretical error comes from the knowledge of the radiation function , of the vacuum polarization and of the bhabha cross section used to determine the luminosity .
the result ( [ amu ] ) is in good agreement with the one obtained by cmd2 with the energy scan method@xcite and confirms a discrepancy with @xmath109 data .
it s a pleasure to thank c. curceanu ( petrascu ) for the help she gave in presenting the dear results and all the da@xmath0ne team for the nice successful work they did to improve the accelerator performances .
c. curceanu ( petrascu ) _ et al . _
( dear collaboration ) , `` last results from the dear experiment at dafne '' , proceedings of `` hadron spectroscopy '' , tenth international conference on hadron spectroscopy , aschaffenburg , germany , august 31 - september 6 , 2003 .
a. aloisio _
( kloe collaboration ) , kloe - note-189 , jul 2003 . contributed to international europhysics conference on high - energy physics ( hep 2003 ) , aachen , germany , 17 - 23 jul 2003 .
_ hep - ex/0307051 _ | the da@xmath0ne frascati @xmath0 factory has continously improved its performances reaching in 2002 an instantaneous luminosity of @xmath1 .
the dear experiment , concluded in 2002 , has measured the de - excitation of kaonic atoms .
the kloe experiment , still running , has measured several branching ratios for neutral and charged kaons decays , @xmath2 , @xmath3 , @xmath4 , @xmath5 and @xmath6 mesons parameters and , via the radiative return , the @xmath7 cross section .
preliminary and final results are presented . |
since the first discovery in 2004 , eight planets have been detected by using the microlensing method @xcite .
the microlensing method is important in various aspects of exoplanet studies .
first , due to its high sensitivity to planets located in the outer region of planetary systems beyond the snow line , the microlensing method is complementary to other methods such as the radial velocity and the transit methods that are sensitive to planets orbiting close to their host stars .
in addition , the sensitivity of the microlensing method extends to very low - mass planets and earth - mass planets can be detected from ground - based observations .
furthermore , it is the only proposed method that can detect free - floating planets @xcite that are thought to be kicked out from planetary systems .
the detection rate of microlensing planets is rapidly increasing and at least five additional planet candidates were detected during the 2007 and 2008 observation seasons ( a. gould 2008 , private communication ) . the microlensing signal of a planet is a perturbation to a smooth standard light curve of a primary - induced lensing event occurring on a background star @xcite . the duration of the perturbation is short : several hours for an earth - mass planet and several days even for a gas - giant planet .
thus it is difficult to detect planetary signals from microlensing survey observations where stars are monitored on a roughly nightly basis .
currently , the observational frequency required for planet detections is achieved by employing an early - warning system and follow - up observations , where the early - warning system ( ogle : udalski et al .
1994 ; moa : bond et al .
2001 ) enables to issue alerts of ongoing events by analyzing data from survey observations real time and follow - up observations ( planet : kubas et al .
2008 ; micro - fun : dong et al .
2006 ) are focused on these alerted events .
however , the limited number of telescopes available for follow - up observations restricts the number of events that can be followed up at any given time and thus priority is given to events which will maximize the planet detection probability .
currently , the highest priority is given to high - magnification events @xcite .
these events have high intrinsic planet detection efficiency because the source trajectories always pass close to the perturbation region around the central caustic induced by the planet @xcite . despite the high chance of perturbation
, however , it is often thought that detecting low - mass planets through the channel of high - magnification events would be difficult .
this thought is based on the fact that the central caustic induced by a low - mass planet is usually smaller than the source star and thus the planetary signal would be greatly washed out by severe finite - source effect @xcite .
however , it might be that perturbations persist despite the finite - source effect and could still be detected thanks to the high photometry precision achieved by the enhanced brightness of the highly magnified source star . in this paper , we test this possibility by investigating how the pattern of central planetary perturbations is affected by the finite - source effect . the paper is organized as follows . in 2 , we briefly describe the physical properties of central caustics . in 3 , we investigate the effect of a source size on the perturbation pattern . for this
, we construct maps of perturbation pattern for planetary systems with various caustic / source size ratios and caustic shapes . based on these maps
, we search for characteristic features that may be used to identify the existence of planets .
we summarize the results and conclude in 4 .
a planetary lens corresponds to the case of a binary lens with a very low - mass companion . one important characteristic of binary lensing that differentiates it from those of single lensing is the formation of caustics , which represent the positions on the source plane where the lensing magnification of a point source becomes infinite .
then , the light curve of a lensing event resulting from the source trajectory approaching or crossing the caustic can be dramatically different from the smooth and symmetric light curve of a single - lensing event .
the set of caustics form closed curves , each of which is composed of concave curves ( fold caustic ) that meet at points ( cusps ) . for a planetary case ,
there exist two sets of caustics .
a single ` central caustic ' is located close to the primary lens and the other one or two ` planetary caustics ' are located away from the primary . compared to size of the planetary caustic , the size of the central caustic is much smaller . when the size of the central caustic is measured as the separation between the two cusps located along the primary - planet axis , it is represented by @xcite @xmath1 where @xmath2 represents the planet / primary mass ratio and @xmath3 is the primary - planet separation normalized by the einstein radius corresponding to the total mass of the planetary system . on the other hand ,
the size of the planetary caustic is ) is for caustics of planetary lenses with @xmath4 .
the planetary caustic for the case of @xmath5 has a different dependency on @xmath3 , but it has the same dependency on @xmath2 , i.e. @xmath6 @xcite . ]
@xcite @xmath7 then , the size ratio between the two types of caustic is @xmath8 considering that the size ratio is proportional to @xmath9 and the mass ratio is very small for planetary lenses , the central caustic is much smaller than the planetary caustic . in addition , the size ratio becomes even smaller for lower mass planets .
the small size of a central caustic implies that the planetary signal induced by the central caustic is more likely to be affected by larger finite - source effect .
the finite - source effect smears out the planetary signal because the magnification of a finite source corresponds to the magnification averaged over the source - star surface , i.e. @xmath10 where @xmath11 is the displacement vector of the source center with respect to the lens , @xmath12 is the vector to a position on the source star surface from the center of the source star , @xmath13 represents the surface brightness profile of the source star , @xmath14 is the point - source magnification , and @xmath15 represents the source radius normalized by the einstein radius .
while the size of the central caustic depends both on the mass ratio and the separation , the shape of the caustic is solely dependent on the separation .
when the shape is quantified as the vertical / horizontal width ratio and the vertical width @xmath16 is measured as the separation between the two off - axis cusps , the width ratio is related to the planetary separation by @xcite @xmath17 where @xmath18 \right\}. \label{eq6}\ ] ] for a given mass ratio , a pair of central caustics with separations @xmath3 and @xmath19 are identical to the first order of approximation .
the pattern of planetary perturbation is determined by two major factors .
the first factor is the caustic shape and it determines the basic pattern of the perturbation .
the second factor is the size ratio between the caustic and the source star and it characterizes how the pattern is deformed by the finite - source effect . to see how a source size affects the pattern of central perturbations , we construct maps of _ magnification excess _ in the region around central caustics with various shapes and sizes .
the magnification excess is defined by @xmath20 where @xmath21 and @xmath22 represent the lensing magnifications with without the planet , respectively .
figure [ fig : one ] show the constructed maps . in the figure ,
the panels in each column show the pattern variation depending on the caustic shape , while the panels in each row show the variation depending on the caustic / source size ratio .
we note that all caustics have an identical size as measured by the horizontal width @xmath23 . in each map ,
the regions with brown and blue - tone colors represent the areas where the planetary - lensing magnification is higher @xmath24 and lower @xmath25 than the single - lensing magnification , respectively . for each tone , the color changes into dark scales when the excess is @xmath26 , 3% , 5% , and 7% , respectively .
the coordinates @xmath27 are aligned so that @xmath28 axis matches the primary - planet axis and the center is located at the effective position of the primary lens .
the effective lens position represents the location of a single lens around which the magnification best describes that of the planetary lensing and it approximately corresponds to the center of the caustic . for a planetary case , the effective position has an offset from the original position of the primary of @xmath29 & for $ s<1 $ , \cr } \label{eq8}\ ] ] where the sign is positive when the offset vector is directed toward the planet . in the map ,
the planet is located on the right and all lengths are normalized by the einstein radius .
we take the finite - source effect into consideration by modelling the source brightness profile as @xmath30 where @xmath31 is the angle between the normal direction to the source - star surface and the line of sight .
we adopt a linear and a square - root limb - darkening coefficients of @xmath32 .
the relation between the source and caustic sizes is marked on the top of each column .
the primary - planet separation and the planet / primary mass ratio of the planetary system are marked on the right side of each row .
considering that the caustic shape is characterized solely by the primary - planet separation and for a fixed planet / primary mass ratio the caustic size is linearly proportional to the mass ratio , one can picture excess maps for other cases of planetary systems based on the presented maps in figure [ fig : one ] .
for example , a planetary system with @xmath33 has a similar perturbation pattern to that of the system with @xmath34 .
the only major difference is that the scale of the map decreases by a factor 10 because the caustic becomes smaller .
we also note that the perturbation patterns of a pair of planetary systems with @xmath3 and @xmath19 are identical and thus one can infer perturbation patterns for the cases of planetary systems with @xmath5 from those presented in the figure . from the analysis of the maps , we find that the dependency of the finite - source effect on the caustic shape is weak and perturbations persist even when the source is substantially bigger than the caustic .
specifically , we find that perturbations with @xmath35 survive when the source star ( as measured by its diameter ) is roughly 4 times bigger than the caustic , i.e. @xmath36 . based on this , @xcite derived an analytic expression for the optimal range of planetary separations ( lensing zone ) detectable through the channel of high - magnification events of @xmath37 then , the lensing zone of central perturbations ( central lensing zone ) is different for planets with different mass ratios unlike the fixed range of the classical lensing zone ( @xmath38 ) regardless of the mass ratio @xcite .
according to this expression , the central lensing zone is larger than the classical lensing zone for planets with @xmath39 despite the smaller size of the central caustic than the size of the planetary caustic . from the maps , we also find interesting features in the region near the boundary and inside of a circle with its center located at the caustic center and a radius corresponding to that of a source star .
these features are the localized arc - shaped perturbation regions with either positive or negative excess located at the edge of the circle and the region inside the circle with a very small magnification excess .
we find that these features commonly appear in the perturbation patterns affected by severe finite - source effect . for close investigation of these features ,
we construct a separate set of excess maps of a planetary lens with and without the finite - source effect .
these maps are presented in figure [ fig : two ] . in the map ,
all lengths are scaled by the source radius to better show the location of the features in units of the source radius .
another difference of the maps from those presented in figure [ fig : one ] is that the color scale changes at the levels of @xmath40 , 4% , 8% , 16,% , and 32% , respectively , to emphasize the region of very high excesses .
five circles are drawn on each map .
all circles have a common radius corresponding to the source radius .
the center of the black dashed circle is located at the caustic center .
the centers of other circles are located at the mid - points of the individual arc - shaped perturbation regions .
close inspection of the pattern leads us to find the following tendencies .
first , we find that the very small excess in the region inside the dashed circle is caused by the cancellation of the positive and negative excesses . when the center of the source star is located within this area , the amount of positive and negative excesses balances each other , resulting in a very small excess .
however , when the source center is located outside of this area , this balance breaks down and perturbation features show up .
second , we find that the arc - shaped perturbation features are caused by the localized regions of very large excesses around the caustic .
such regions include the regions just outside the cusps with very large positive excesses and the regions just outside the fold caustics with very large negative excesses .
another region of a very large negative excess is the region inside the caustic .
the localized arc - shaped perturbation regions form at the source position where the source star encompasses some of these strong perturbation regions and the resulting overall excess is seriously unbalanced .
for example , when the source star is located at the position of the blue circle , it encompasses two large positive - excess regions around two cusps but only a single large negative - excess region between the cusps . as a result ,
the overall excess is positive . by contrast ,
when the source star is located at the adjacent position of the magenta circle , it encompasses the same number of the two large positive - excess regions but now four large negative - excess regions including two additional regions around fold caustics and one inside the caustic , resulting in an overall negative excess .
the other perturbation regions can be explained in a similar way .
the existence of the characteristic features in the pattern of central perturbation region provides an important diagnostic tool that can be used to identify the existence of a companion for a high - magnification event affected by severe finite - source effect .
this is because the light curve of an event where the source crosses these features will exhibit a distinctive signal that is characterized by short - duration perturbations of either positive or negative excess and a flat residual region between these short - duration perturbations .
we present an example light curve of such an event in figure [ fig : three ] .
recently , a planetary - lensing event exhibiting a similar planetary signal was actually detected ( moa-2007-blg-400 : dong et al .
2008 ) . considering that high - magnification events are prime targets of the current microlensing follow - up observations
, we predict that more of such planetary signals would be detected .
we note , however , that the existence of these characteristic signals does not necessarily confirm the existence of a planetary companion .
this is because such signals can also be produced by a wide ( @xmath41 ) or a close ( @xmath42 ) binary companions that can also give rise to small central caustics .
therefore , detailed modeling should be done for the complete characterization of the lens system .
we investigated the effect of a finite source on the central perturbation pattern . from this
, we found that the dependency of the finite - source effect on the caustic shape is weak and perturbations survive even when the source is substantially bigger than the caustic .
specifically , we found that perturbations with fractional magnification excess @xmath0 survive when the source star is roughly 4 times bigger than the caustic .
we also found characteristic features that commonly appear in the perturbation patterns of lens systems affected by severe finite - source effect and thus can be used for the diagnosis of the existence of a companion .
these features form in and around a circle with its center located at the caustic center and a radius corresponding to that of the source star .
the light curve of an event where the source crosses these features will exhibit a distinctive signal that is characterized by short - duration perturbations of either positive or negative excess and a flat residual region between these short - duration perturbations .
this work was supported by the astrophysical research center for the structure and evolution of the cosmos ( arcsec ) of korea science and engineering foundation ( kosef ) through science research center ( src ) program . | we investigate the effect of a finite source on the planetary - lensing signals of high - magnification events . from this
, we find that the dependency of the finite - source effect on the caustic shape is weak and perturbations survive even when the source is substantially bigger than the caustic . specifically , we find that perturbations with fractional magnification excess @xmath0 survive when the source star is roughly 4 times bigger than the caustic .
we also find characteristic features that commonly appear in the perturbation patterns of planetary lens systems affected by severe finite - source effect and thus can be used for the diagnosis of the existence of a companion .
these features form in and around a circle with its center located at the caustic center and a radius corresponding to that of the source star .
the light curve of an event where the source crosses these features will exhibit a distinctive signal that is characterized by short - duration perturbations of either positive or negative excess and a flat residual region between these short - duration perturbations . |
molecular reaction dynamics studies aim at understanding chemical reactions and inelastic collisions at the atomic scale . in other words ,
this field of research draws much of the conceptual framework in which chemical reactivity , in a broad sense , can be thought @xcite . quantum state - resolved integral and differential cross sections ( icss and dcss ) , measured in supersonic molecular beam experiments , are among the most fundamental observables of molecular reaction dynamics .
this paper deals with their classical mechanical description in a semi - classical spirit .
most processes considered up to now involve three or four atoms , on purpose .
this allows both measurements at an amazing level of detail and accurate theoretical descriptions of the observables from first principles .
additionally , planetary atmospheres and interstellar clouds are mainly made of small species which dynamics should be understood . nowadays , however , much of molecular science is polarized on larger systems , like nano - objects or molecules of biological interest , and the natural trend in molecular reaction dynamics is also to move towards increasing complexity .
more and more polyatomic processes are thus under scrutiny .
state - of - the - art descriptions of state - resolved icss and dcss are in principle performed within the framework of exact quantum scattering approaches ( eqs ) @xcite . however , despite the impressive progress of computer performance achieved in the last decades , these approaches can hardly be applied to larger than three or four - atom systems as the basis sizes necessary for converging the calculations turn prohibitive .
a popular alternative is the quasi - classical trajectory method ( qctm ) @xcite .
this approach is intuitive , relatively easy to implement , much less time consuming than eqs approaches and therefore , quite appealing for studying polyatomic processes .
the price to pay is obviously a loss in accuracy as compared to eqs approaches .
nevertheless , significant advances have been made in the last few years through the replacement of the standard binning ( sb ) procedure by the gaussian weighting ( gw ) one @xcite . in the sb method , each trajectory has the same statistical weight . on the other hand ,
the gw procedure consists in weighting each trajectory by a gaussian - like coefficient such that the closer the final actions to integer values , the larger the coefficient .
this procedure proves to be especially efficient when few vibrational levels are available in the final products .
though initially proposed on the basis of rather intuitive arguments , the gw procedure can be shown to find its roots in classical @xmath0 matrix theory , the former semi - classical approach of molecular collisions pioneered by miller and marcus in the early seventies @xcite .
central quantities of chemical reaction theory are ( 1 ) the state - to - state reaction probabilities @xmath1 , where @xmath2 and @xmath3 are reagent and product quantum states , ( 2 ) the densities @xmath4 , where @xmath5 is the scattering angle or any given angle of the problem and ( 3 ) the capture probabilities @xmath6 for processes involving long - lived intermediate complexes @xcite . from these quantities , any state - resolved ics and dcs can be determined . to calculate the previous probabilities ( or density of ) , one must generate classical dynamical conditions corresponding to quantum state @xmath2 .
such a generation is readily performed in angle - action coordinates @xcite as these are in close correspondence with quantum numbers . on the other hand ,
angle - action variables should not be used to run trajectories as contrary to cartesian coordinates , they lead to strong numerical instabilities . the transformation from angle - action variables to cartesian
coordinates is therefore a crucial step of qctm . for atom - diatom ( semi ) collisions , this transformation can be found in the book by whittaker @xcite and in a paper by miller @xcite .
however , we have not been able to find in the literature the analogous transformation for a generic type of collision .
the goal of the paper is to thus to derive it .
in this work , a prototype system is presented , namely a five - atom molecule made of a triatomic ( abc ) and a diatomic ( de ) .
the former can be used as a model for a non - linear polyatomic fragment while the latter constitutes a simpler case very commonly found in practice . the transformation provided here
will therefore be relevant for a generic pair of molecular fragments , _
e.g. _ diatom + diatom , asymmetric top + diatom , asymmetric top + asymmetric top , etc after straightforward generalizations .
these , along with the transformations in @xcite allow thus to treat any case of interest .
we suppose the fragments are to be studied in the low energy regime where only the lowest vibrational states can be populated , thus the harmonic description of their vibrations is a reasonably accurate approximation .
anharmonic corrections can be introduced when necessary . throughout this work ,
the usual convention of boldfacing vector magnitudes is used .
cartesian frames centered on a generic point p are represented as @xmath7 .
a given vector @xmath8 in such a frame will be rewritten as @xmath9 if we refer it to @xmath10 instead .
calligraphic letters are used for representing matrices and second - rank tensors .
some standard transformations , _
e.g. _ that of normal modes to cartesian coordinates , are included for completeness .
finally , the two fragments , abc and de , are numbered 1 and 2 and so are their associated magnitudes . the system is schematically represented in fig .
[ diag:5atom ] .
three cartesian frames of reference are used : ( 1 ) the laboratory frame which origin is at the molecular center of mass g and is in uniform translation so that the total center - of - mass movement can be effectively removed , and ( 2 , 3 ) the two body - fixed , non - inertial reference frames with origins at each fragment s center of mass , denoted g@xmath11 and g@xmath12 .
the cartesian coordinates to which transformation from angle - actions is made are defined as the complete set of nuclei positions @xmath13 , in the @xmath14 space , plus their conjugate momenta @xmath15 , with x @xmath16 \{a , b , , e}. the total number of such coordinates yields , of course , @xmath17 . of the 30 variables chosen , 8 are not angle - actions ,
_ i.e. _ ( 1 , 2 ) the distance @xmath18 between the fragments centers of mass g@xmath11 and g@xmath12 and its conjugate momentum @xmath19 ; and ( 38 ) the position and momentum vectors for the molecular center of mass , g. the 22 angle - action variables are thus : 1.3 cm 1.2 cm 0.3 cm the vibrational phase of the @xmath20th normal mode of abc , @xmath21 . the vibrational action of the @xmath20th normal mode of abc , @xmath21 . the vibrational phase of de . the vibrational action of de .
the modulus of the total angular momentum @xmath22 .
the angle conjugate to @xmath22 . the algebraic value of the projection of @xmath22 on the laboratory @xmath23 axis .
the angle conjugate to @xmath24 .
the modulus of the orbital angular momentum @xmath25 .
the angle conjugate to @xmath25 .
the modulus of the rotational angular momentum @xmath26 of abc .
the angle conjugate to @xmath26 .
the modulus of the rotational angular momentum @xmath27 of de .
the angle conjugate to @xmath27 .
the modulus of the total rotational angular momentum @xmath28 .
the angle conjugate to @xmath29 . the algebraic value of the projection of @xmath26 on one of the three axes of inertia of abc . the angle conjugate to @xmath30 .
the six triatomic normal mode coordinates fully specify the three position vectors @xmath31 , @xmath32 and @xmath33 , in the @xmath34 plane of the body - fixed @xmath35 frame of abc .
@xmath36 is arbitrarily made to coincide with one of the abc axes of inertia when it happens to be in its equilibrium geometry .
these six normal mode coordinates also define the three momentum vectors @xmath37 , @xmath38 and @xmath39 , conjugate to the three previous position vectors , _
i.e. _ 12 coordinates as a whole .
note that these twelve cartesian coordinates are deduced from the six normal modes plus six constraints due to the fact that abc is neither in translation nor in rotation in the @xmath40 plane . the total angular momentum @xmath22 , its @xmath23-component @xmath41 , their conjugate angles @xmath42 and @xmath43 as well as the orbital @xmath25 and total rotational @xmath29 angular momenta are represented in fig .
[ diag : jlkgen ] .
the unit vectors along the @xmath44 and @xmath23 axes are respectively denoted @xmath45 and @xmath46 .
we wish to emphasize here that the three axes @xmath47 , @xmath48 and @xmath36 used at this point have nothing to do with the primed axes introduced in the previous paragraph .
several primed frames will be defined in the following which will be different from each other .
@xmath43 is the angle between @xmath45 and @xmath49 while @xmath42 is the angle between @xmath49 and @xmath50 .
, the intermolecular jacobi vector @xmath51 and its conjugate momentum @xmath52.,width=321 ] @xmath25 is represented in fig .
[ diag : lpr ] together with the jacobi vector @xmath51 between g@xmath11 and g@xmath12 .
@xmath53 is the angle between @xmath54 and @xmath51 .
the momentum @xmath52 conjugate to @xmath51 is also depicted .
like @xmath51 , @xmath52 lies in the plane orthogonal to @xmath25 .
@xmath26 is represented in fig .
[ diag : j1k1gen ] together with @xmath30 , defined as the projection of @xmath26 on the @xmath36 axis of the previously specified body - fixed frame of abc .
@xmath55 is the angle between @xmath56 and @xmath57 .
@xmath27 is represented in fig .
[ diag : j2pr ] together with the jacobi vector @xmath58 between the d and e atoms .
@xmath59 is the angle between @xmath60 and @xmath58 .
the momentum @xmath61 conjugate to @xmath58 is also represented .
both @xmath58 and @xmath61 lie in the plane orthogonal to @xmath27 . , the de interatomic jacobi vector @xmath58 and its conjugate momentum @xmath61.,width=321 ] the link between @xmath29 , @xmath26 and @xmath27 is isomorphic to the one between @xmath22 , @xmath25 and @xmath29 , as easily seen from the comparison between fig .
[ diag : kj1j2 ] and fig .
[ diag : jlkgen ] .
@xmath62 is thus the angle between @xmath63 and @xmath64 .
calling @xmath65 the unit vector along the @xmath36 axis of the abc body - fixed frame , the algebraic value @xmath66 equals plus ( minus ) @xmath67 when @xmath65 and @xmath26 make an angle lower ( larger ) than @xmath68 . finally , @xmath69 is the angle between @xmath70 and the @xmath47 axis in fig .
[ diag : j1k1gen ] .
the algorithm for computing initial conditions from the title transformation will vary slightly according to the specific application ( _ e.g. _ unimolecular dissociation , bimolecular collision ) and/or the experimental conditions to be reflected .
the transformations , however , are intrinsically general so we assume in what follows that all angle - action variables , as well as @xmath18 and @xmath19 , are either known or can be computed by the time they are referred to during the process .
the transformation can be decomposed in 11 steps , each making the subject of one of the following sections .
it is important to note that the ordering given here is somewhat arbitrary and need for reordering may arise in specific applications . in fig .
[ diag : jlk ] , the vectors @xmath22 , @xmath25 and @xmath29 are represented in the plane @xmath71 as deduced from fig .
[ diag : jlkgen ] .
the relation between these angular momenta can be written as @xmath72 squaring each side of the previous equality and rearranging leads to @xmath73 @xmath74 , equal to @xmath75 , is thus given by @xmath76 @xmath77 , equal to @xmath78 , _
i.e. _ , to @xmath79 ( given the convention adopted , @xmath77 is necessarily positive ) , is therefore given by @xmath80^{1/2}. \label{eq : t4}\ ] ] at last , @xmath81 is zero . , orbital @xmath25 and rotational @xmath29 angular momenta.,width=226,height=245 ] @xmath25 is deduced from @xmath82 by the standard euler rotation @xmath83 where , for a given angle @xmath84 , @xmath85 and @xmath86 indeed , fig .
[ diag : jlkgen ] shows that one goes from @xmath87 to @xmath14 by a rotation of @xmath88 around the @xmath36 axis followed by a rotation of @xmath89 around the resulting , ` new ' @xmath47 axis and a final rotation of @xmath90 around the ` new ' @xmath36 axis .
one may easily check that these transformations are achieved by the @xmath91 and @xmath92 matrices combined as in eq .
[ eq : t5 ] .
@xmath93 is given by @xmath94 and @xmath95 , necessarily positive as @xmath96 $ ] , is given by @xmath97^{1/2}. \label{eq : t9}\ ] ] , its cartesian components and conjugate angles.,width=321 ] from fig .
[ diag : lpr ] and following the same reasoning as above , @xmath51 can be shown to satisfy @xmath98 where @xmath99 represents the vector @xmath100 .
[ diag : lblthl ] shows how the angles @xmath101 and @xmath102 relate to @xmath25 .
@xmath103 is given by @xmath104 and @xmath105 , necessarily positive , by @xmath106^{1/2}. \label{eq : t12}\ ] ] @xmath107 is given by @xmath108 and @xmath109 by @xmath110 where @xmath111 is the modulus of the projection of @xmath25 on the @xmath112 plane , as depicted in fig .
[ diag : lblthl ] . since @xmath113
, @xmath52 lies in the plane @xmath114 of fig .
[ diag : lpr ] , @xmath115 has already been denoted @xmath19 , @xmath116 equals @xmath117 and @xmath118 is zero .
@xmath52 is then obtained with @xmath119 we still consider fig .
[ diag : jlk ] and rewrite the relation between @xmath22 , @xmath25 and @xmath29 as @xmath120 squaring each side of the previous equality and rearranging leads to @xmath121 @xmath122 , equal to @xmath123 , is thus given by @xmath124 @xmath125 , equal to @xmath126 , _
i.e. _ , to @xmath127 ( given the convention adopted , @xmath125 is necessarily negative ) , is therefore given by @xmath128^{1/2 } \label{eq : t20}\ ] ] ( one may check that @xmath125 is the just the opposite of @xmath77 ) .
at last , @xmath129 is zero .
@xmath29 is then obtained from @xmath130 by the same transformation that relates @xmath25 to @xmath82 ( see eq .
[ eq : t5 ] ) @xmath131 as already seen , the determination of @xmath26 and @xmath27 is in complete analogy with that of @xmath25 and @xmath29 ( compare fig .
[ diag : kj1j2 ] and fig .
[ diag : jlkgen ] ) . following the developments in sections [ sec3:l ] and [ sec3:k ]
, we then arrive at @xmath132 where @xmath133 , @xmath134^{1/2 } , \label{eq : t23}\ ] ] @xmath135 @xmath136 and @xmath137 in addition , @xmath138 is given by @xmath139 and @xmath140 by @xmath141^{1/2}. \label{eq : t28}\ ] ] @xmath142 is given by @xmath143 and @xmath144 by @xmath145 where @xmath146 is the modulus of the projection of @xmath29 on the @xmath112 plane . in the harmonic limit , the de bond length @xmath147 is given in terms of @xmath148 and @xmath149 by the expression @xmath150^{1/2}\sin{q_4}. \label{eq : t32}\ ] ] here
, @xmath151 is the equilibrium bond length of the diatomic , @xmath152 its reduced mass and @xmath153 its vibrational frequency ( which is readily determined from a quadratic fitting of its interaction potential ) .
although @xmath154 is sometimes called action , _ stricto sensus _
, this is only true in @xmath155 units .
the problem of the determination of @xmath58 is then analogous to that of @xmath51 . from fig .
[ diag : j2pr ] and following section [ sec3:r ] , we find @xmath156 where @xmath157 represents the vector @xmath158 .
@xmath159 is given by @xmath160 and @xmath161 by @xmath162^{1/2}. \label{eq : t35}\ ] ] @xmath163 is given by @xmath164 and @xmath165 by @xmath166 where @xmath167 is the modulus of the projection of @xmath27 on the @xmath112 plane .
again , the problem of the determination of @xmath61 is analogous to that of the determination of @xmath52 . following section [ sec3:p ] ,
we arrive at @xmath168 where @xmath169 and @xmath170^{1/2}\cos{q_4 } \label{eq : t40}\ ] ] in the harmonic approximation . and inertial - axis component @xmath30 angular momenta.,width=226,height=245 ] @xmath26 and @xmath30
are represented in fig .
[ diag : j1k1gen ] and fig .
[ diag : j1k1 ] .
the coordinates of @xmath30 in @xmath35 are given by @xmath171 , @xmath172^{1/2 } \label{eq : t41}\ ] ] and @xmath173 ( the last equation comes from the fact that the cosine of the angle between @xmath30 and @xmath26 is equal both to @xmath174 and @xmath175 , as is obvious from fig .
[ diag : j1k1 ] ) . proceeding as previously
, we find @xmath176 @xmath177 is given by @xmath178 and @xmath179 by @xmath180^{1/2}. \label{eq : t45}\ ] ] @xmath181 is given by @xmath182 and @xmath183 by @xmath184 where @xmath185 is the modulus of the projection of @xmath26 on the @xmath186 plane .
we start by determining the position vectors @xmath190 for in the @xmath35 frame ( fig .
[ diag:5atom ] ) . within the harmonic approximation ,
this task is accomplished by the standard normal mode analysis @xcite ( a generalization of the procedure used in the diatomic case ; compare this and the following with sections [ sec3:rp ] and [ sec3:prp ] ) .
first , the eigenvalues @xmath191 and eigenvectors @xmath192 of the hessian matrix @xmath193 are determined . for an @xmath194-atom molecule ,
six of the former correspond to the center - of - mass movement and overall rotation and thus are theoretically zero ( negligibly small in practice ) .
the @xmath195 non - zero eigenvalues , associated with the molecule internal vibrational modes , relate to their angular frequencies simply by @xmath196 .
quasi - classical normal mode energies are then computed from the corresponding vibrational actions as @xmath197 which allows the calculation of the normal mode displacements @xmath198 cartesian mass - weighted displacements are determined with @xmath199 , where @xmath200 is the eigenvector matrix and @xmath201 that of normal mode coordinates .
the position vectors @xmath190 are thus @xmath202 where @xmath203 are the equilibrium position vectors , @xmath204 is the mass of atom x and @xmath205 is extracted from @xmath206 according to the location given to the x - atom coordinates in @xmath193 . from fig .
[ diag : k1angles ] , we have @xmath207 which holds for x = a , b or c. and some angles.,width=321 ] when @xmath66 is positive , the @xmath35 frames in fig .
[ diag : k1angles ] and fig .
[ diag:5atom ] exactly coincide .
therefore , the dependence of @xmath208 on @xmath190 is of the same kind as in the previous sections . if , on the other hand , @xmath66 is negative , @xmath35 in fig .
[ diag : k1angles ] is different from its equivalent in fig .
[ diag:5atom ] .
in fact , in this case the @xmath48 and @xmath36 axes are oriented in the exact opposite directions as in the previous one .
the term @xmath209 , which equals @xmath210 , takes this difference into account by flipping the vector @xmath211 before it is identified as @xmath208 . in eq .
[ eq : t52 ] , @xmath212 is given by @xmath213 and @xmath214 by @xmath215^{1/2}. \label{eq : t54}\ ] ] @xmath216 is given by @xmath217 and @xmath218 by @xmath219 where @xmath220 is , as usual , the modulus of the projection of @xmath30 on the @xmath186 plane .
the momenta @xmath224 , with x = a , b or c , can be decomposed into a _ purely _ translational ( vibrational ) and a rotational components . based on the very definition of the body - fixed @xmath225 frame ( fig .
[ diag:5atom ] ) , the former is directly related to @xmath226 . to calculate these ,
normal - mode velocities are first computed using the conservation of energy @xmath227 the sign being selected according to the value of the vibrational phase @xmath228 .
cartesian mass - weighted velocities are thus @xmath229 from which @xmath230 it is important to stress that the anharmonicity of the real potential energy has been deliberately neglected within the normal - mode approximation .
to correct for its possible spurious consequences , relatively sophisticated recipes can be used at this stage .
the reader is thus referred to the available literature , _ e.g. _
@xcite , as it is not our objective to reproduce them here .
the rotational component is determined in the standard fashion .
the triatomic angular velocity is computed as @xmath231being @xmath232 the inertia tensor of abc , which can be calculated at this point since its configuration has been determined from which , the corresponding linear velocities are given by @xmath233 finally , the transformation relating @xmath224 and @xmath226 is isomorphic to eq .
[ eq : t52 ] , so the desired general expression for computing the former reads @xmath234 at this point it is a simple task to finally express all cartesian vectors in the laboratory frame . for x
= a , b or c , @xmath13 is given by the general expression @xmath235 while if x = d or e , @xmath236 in these equations , @xmath237 stands for the mass of fragment @xmath20 and @xmath238 is the system total mass .
similar relations hold for the cartesian momenta . for x
= a , b or c , these are computed using the general expression @xmath239 at last , the diatomic momenta are given by @xmath240 and @xmath241
the photo - fragmentation of ketene ( ch@xmath12co ) has been intensively investigated for over two decades , both experimental and theoretically ( _ e.g. _ @xcite ) .
following photo - excitation to the @xmath242 states , the molecule undergoes either intersystem crossing or fast internal conversion to the low lying triplet and singlet electronic states . from these , dissociation into methylene and carbon monoxide occurs . despite the triplet threshold lies @xmath2433150 @xmath244 below the singlet , the fact that it presents a small barrier to dissociation of a few cents of inverse centimeters makes the singlet channel statistically dominant from excess energies as low as @xmath243100200 @xmath244 .
such conditions make the system an effective prototype for a _ barrierless polyatomic _
unimolecular reaction on a _ single _ potential energy surface ( pes ) . in direct correspondence with the model transformation we introduced above
, the molecule constitutes a five - atom system which dissociates into a triatomic and diatomic fragments .
additionally , the experimental excitations are compatible with the harmonic normal mode
approximation for the ch@xmath12 and co products . in what follows
we briefly report on the application of the title transformation to the study of this process .
full details and results will be given in a separate work so we simply introduce it here as a corroboratory test case .
, after excitation with a 308 nm laser .
comparison with the experiment.,width=321 ] in fig .
[ fig : petjc4 ] we compare our calculations with the most recent experimental results @xcite for the products translational energy distributions , in correlation with the rotational state of co. a 308 nm laser is used in the experiment , corresponding to an excess energy of 2350 @xmath244 .
the theoretical results are obtained using the so - called _ exit - channel corrected phase - space theory _ , proposed by hamilton and brumer @xcite .
this method basically consist in generating microcanonical initial conditions _ at the products _ and then propagate the trajectories backwards in time , the statistics being performed with those reaching the inner transition state ( ts ) .
the photo - excited ketene molecule is supposed to be long lived prior to its fragmentation , thereby justifying the use of a microcanonical distribution .
we employed the high - level _ ab initio _ pes and transition state locations recently reported @xcite .
the theoretical predictions are in very good agreement with the experiment , as can be seen in fig .
[ fig : petjc4 ] .
the curve has been artificially smoothed by using a convolution with an ` apparatus ' function , _
i.e. _ @xmath245 to recover the experimental tails .
the two peaks correlating with the @xmath246 and @xmath247 ch@xmath12 scissor - mode states are fairly well reproduced . in order to further verify the validity of the transformation provided ,
we have calculated the determinant of the jacobian matrix .
the original is not a square matrix , with dimensions 30@xmath24824 .
therefore , for being able to calculate the determinant we introduced an additional transformation to a set of jacobi coordinates , from which the ( null ) center - of - mass coordinates and momenta are later removed .
the calculation starts with the transformation from angle - actions to cartesian and from these to jacobi coordinates .
the center - of - mass jacobi vectors are then removed and the determinant of the resulting 24@xmath24824 jacobian matrix , from angle - actions to ( reduced ) jacobi coordinates , is computed .
we confirmed that it yields 1 within numerical accuracy .
we have presented the transformation from angle - action to cartesian coordinates , for polyatomic systems . in the quasi and semi - classical approaches ,
this provides an expeditious way to generate initial conditions in close correspondence with nowadays experiments and yet , solve the equations of motion using the ` ideal ' cartesian coordinates .
the methodology and expressions provided here can either be directly used or straightforwardly generalized to deal with any case of interest , ranging from the study of bimolecular collisions to polyatomic unimolecular dissociations .
preliminary results of the particular application to the study of the unimolecular dissociation of ketene in the singlet electronic state , have been discussed .
a very good agreement is observed between the experimental values and theoretical predictions for correlated translational energy distributions .
the validity of the transformation have been further verified by numerical computation of the determinant of the jacobian matrix , which yields unity within reasonable accuracy .
support from an inter - university agreement on international joint doctorate supervision between the instituto superior de tecnologas y ciencias aplicadas , cuba and the universit bordeaux 1 , france , as well as the pnap/7/3 project of the cuban institution , are gratefully acknowledged . | the transformation from angle - action variables to cartesian coordinates is a crucial step of the ( semi ) classical description of bimolecular collisions and photo - fragmentations .
the basic reason is that dynamical conditions corresponding to experiments are ideally generated in angle - action variables whereas the classical equations of motion are ideally solved in cartesian coordinates by standard numerical approaches . to our knowledge
, the previous transformation is available in the literature only for triatomic systems .
the goal of the present work is to derive it for polyatomic ones . |
frequency analysis of signals @xcite is a classical problem that has broad applications ranging from communications , radar , array processing to seismology and astronomy .
grid - based sparse methods have been vastly studied in the past decade with the development of compressed sensing ( cs ) which exploit signal sparsity
the number of frequency components @xmath0 is small but suffer from basis mismatches due to the need of gridding of the frequency interval @xcite .
its research has been recently advanced owing to the mathematical theory of super - resolution introduced by cands and fernandes - granda @xcite , which refers to recovery of fine details of a sparse frequency spectrum from coarse scale time - domain samples only .
they propose a gridless atomic norm ( or total variation norm ) technique , which can be cast as semidefinite programming ( sdp ) , and prove that a continuous frequency spectrum can be recovered with infinite precision given a set of @xmath1 regularly spaced samples . the technical method and theoretical result were then extended by tang _
@xcite to the case of partial / compressive samples , showing that only a number of @xmath2 random samples are sufficient for the recovery with high probability via atomic norm minimization ( anm ) .
moreover , yang and xie @xcite study the multiple - measurement - vector ( mmv ) case , which arises naturally in array processing applications , with similar results proven using extended mmv atomic norm methods .
however , a major problem of existing atomic norm methods is that the frequency spectrum can be recovered only when the frequencies are sufficiently separated , prohibiting commonly known high resolution the capability of resolving two closely spaced frequency components
. a sufficient minimum separation of frequencies is @xmath3 in theory .
empirical evidences in @xcite suggest that this number can be reduced to @xmath4 , while according to @xcite it also depends on @xmath0 , @xmath5 and the number of measurement vectors . in this paper
, we attempt to propose a high resolution gridless sparse method for super - resolution to break the resolution limit of existing atomic norm methods .
our method is motivated by the formulations and properties of atomic @xmath6 norm and the atomic norm in @xcite .
in particular , the atomic @xmath6 norm has no resolution limit but is np hard to compute .
to the contrary , as a convex relaxation the atomic norm can be efficiently computed but suffers from a resolution limit as mentioned above .
we propose a novel sparse metric and theoretically show that the new metric fills the gap between the atomic @xmath6 norm and the atomic norm .
it approaches the former under appropriate parameter setting .
with the sparse metric we formulate a nonconvex optimization problem and present a locally convergent iterative algorithm for super - resolution .
the algorithm iteratively carries out anm with a sound reweighting strategy , which determines preference of frequency selection based on the latest estimate and enhances sparsity and resolution , and is termed as reweighted atomic - norm minimization ( ram ) . to the best of our knowledge , ram implements the first reweighting strategy in the continuous dictionary setting while existing reweighted @xmath7 algorithms ( see , e.g. , @xcite ) are for the discrete setting .
extensive numerical simulations are carried out to demonstrate the high resolution performance of ram with application to doa estimation compared to existing arts .
we consider the super - resolution problem in the most general case with partial samples and mmvs .
in particular , we observe the samples of the data matrix @xmath8 on the rows indexed by @xmath9 of size @xmath10 , denoted by @xmath11 .
the @xmath12th element of @xmath13 is ( corrupted by noise in practice ) where @xmath14 denotes a discrete complex sinusoid with frequency @xmath15 $ ] , and @xmath16 is the coefficient vector of the @xmath17th sinusoid .
that is , each column of @xmath13 is superimposed by @xmath0 discrete sinusoids .
we are interested in recovering the frequencies @xmath18 given @xmath11 .
meanwhile , it is also of interest to recover the complete data matrix @xmath13 .
the resulting problem is known as continuous / off - grid cs according to @xcite , which differs from existing cs framework in the sense that every frequency @xmath19 can take any continuous value in @xmath20 rather than constrained on a finite discrete grid .
the single - measurement - vector ( smv ) case where @xmath21 is known as line spectral estimation .
the mmv case where @xmath22 is common in array processing .
therein the sampling index set @xmath23 refers to sensor placement of a linear sensor array and a smaller sample size means use of less sensors .
@xmath11 consists of measurements of the sensor array and each column vector corresponds to one data snapshot .
each frequency corresponds to the direction of one source .
therefore , the frequency estimation problem is known as direction of arrival ( doa ) estimation .
the super - resolution or continuous cs problem is tackled from the perspective of signal recovery .
the frequencies are then retrieved from the computational result . in particular , we seek a _ frequency - sparse _
candidate @xmath24 , which is composed of a few frequency components , in a feasible domain defined by the observed samples . to do this
, we first define a sparse metric of @xmath24 and then optimize the metric over the feasible domain .
a direct sparse metric is the smallest number of frequency components composing @xmath24 , known as the atomic @xmath6 norm and denoted by @xmath25 . according to @xcite @xmath25
can be characterized as the following rank minimization problem : the first constraint in ( [ formu : atom0norm ] ) imposes that @xmath24 lies in the range space of a ( hermitian ) toeplitz matrix @xmath26 whose first row is specified by the transpose of @xmath27 .
the frequencies composing @xmath24 are encoded in @xmath28 .
once an optimizer of @xmath29 , say @xmath30 , is obtained the frequencies can be retrieved from @xmath31 according to the vandermonde decomposition lemma ( see , e.g. , @xcite ) , which states that any positive semidefinite ( psd ) toeplitz matrix @xmath31 can be decomposed as @xmath32 , where the order @xmath33 and @xmath34 ( see a method for realization of the decomposition in ( * ? ? ?
* appendix a ) ) .
the atomic @xmath6 norm directly enhances sparsity , however , it is nonconvex and np - hard to compute and encourages computationally feasible alternatives . in this spirit , the atomic ( @xmath7 ) norm , denoted by @xmath35 ,
is introduced as a convex relaxation of @xmath25 and has the following semidefinite formulation @xcite : from the perspective of low rank matrix recovery ( lrmr ) , ( [ formu : an_sdp ] ) attempts to recover the low rank matrix @xmath28 by relaxing the pseudo - rank norm in ( [ formu : atom0norm ] ) to the nuclear norm or equivalently the trace norm for a psd matrix .
the atomic norm is advantageous in computation compared to the atomic @xmath6 norm , however , it suffers from a resolution limit due to the relaxation which is not shared by the latter @xcite .
inspired by the link between continuous cs and lrmr demonstrated above , we propose the following sparse metric of @xmath24 : where @xmath36 is a regularization parameter that avoids the first term being @xmath37 when @xmath28 is rank deficient .
note that the log - det heuristic @xmath38 is a common smooth surrogate for the matrix rank ( see , e.g. , @xcite ) . from the perspective of lrmr ,
the atomic @xmath6 norm minimizes the number of nonzero eigenvalues of @xmath28 while the atomic norm minimizes the sum of the eigenvalues .
in contrast , the new metric @xmath39 penalizes @xmath40 , where @xmath41 denotes the eigenvalues .
we plot the function @xmath42 with different @xmath43 s in fig .
[ fig : sparsity ] , according to which we expect that the new metric @xmath39 bridges @xmath35 and @xmath25 when @xmath43 varies from @xmath44 to @xmath45 .
formally , we have the following results and we provide their proofs in an extended journal paper @xcite . with different @xmath43 .
the plotted curves include the @xmath6 and @xmath7 norms corresponding to @xmath46 and @xmath47 respectively , and @xmath48 corresponding to @xmath49 with @xmath50 and @xmath51 .
@xmath48 is translated and scaled such that it equals 0 and 1 at @xmath52 and 1 respectively for better illustration.,width=278 ] let @xmath53 . then , i.e. , they are equivalent infinitesimals .
[ thm : epsilontoinf ] let @xmath54 .
then , we have the following results : 1 . if @xmath55 , then i.e. , they are equivalent infinities . otherwise , @xmath39 is a positive constant depending only on @xmath24 ; 2 .
let @xmath56 be the optimizer of @xmath29 to the optimization problem in ( [ formu : nonconvexrelax ] ) .
then , the smallest @xmath57 eigenvalues of @xmath58 are either zero or approach zero as fast as @xmath43 ; 3 . for any cluster point of @xmath56 at @xmath59 , denoted by @xmath60 , there exists an atomic decomposition @xmath61 of order @xmath62 such that @xmath63 .
[ thm : epsilontozero ] theorem [ thm : epsilontoinf ] shows that the new metric @xmath39 plays the same role as @xmath35 in the limiting scenario when @xmath53 , while theorem [ thm : epsilontozero ] says that it is equivalent to @xmath25 as @xmath54 .
consequently , it fills the gap between @xmath35 and @xmath25 and enhances sparsity and resolution compared to @xmath35 as @xmath43 gets small .
moreover , theorem [ thm : epsilontozero ] characterizes the properties of the optimizer @xmath56 as @xmath54 including the convergent speed of the smallest @xmath64 eigenvalues and the limiting form of @xmath31 via the vandermonde decomposition .
in fact , we always observe via simulations that the smallest @xmath64 eigenvalues of @xmath31 become zero once @xmath43 is modestly small .
with the proposed sparse metric @xmath39 , we solve the following optimization problem for signal and frequency recovery : or equivalently , where @xmath65 denotes the feasible domain of @xmath24 .
for example , in the noiseless case , it is the set @xmath66 .
since the log - det term @xmath67 is a concave function of @xmath29 , the problem is nonconvex and no efficient algorithms can guarantee to obtain the global optimum . a majorization - maximization ( mm )
algorithm is introduced as follows .
let @xmath68 denote the @xmath69th iterate of the optimization variable @xmath29 .
then , at the @xmath70th iteration we replace @xmath67 by its tangent plane at the current value @xmath71 . as a result ,
the optimization problem at the @xmath70th iteration becomes since @xmath67 is strictly concave in @xmath29 , at each iteration its value decreases by an amount greater than the decrease of its tangent plane .
it follows that the objective function in ( [ formu : problem ] ) monotonically decreases at each iteration and converges to a local minimum . to interpret the optimization problem in ( [ formu : problem_j ] ) , let us define a _ weighted continuous dictionary _ w.r.t . the original continuous dictionary @xmath72 , where @xmath73 is a weighting function .
for @xmath74 , we define its _ weighted atomic norm _
@xmath75 as its atomic norm induced by @xmath75 : according to the definition above , @xmath76 specifies preference of the atoms @xmath77 . to be specific ,
an atom @xmath78 , @xmath79 , is more likely selected if @xmath80 is larger .
moreover , the atomic norm is a special case of the weighted atomic norm with a constant weighting function ( i.e. , without any preference ) according to @xcite .
suppose that @xmath81 with @xmath82 .
then , [ thm : weightan ] let @xmath83 and @xmath84 . by theorem [ thm : weightan ] we can rewrite the optimization problem in ( [ formu : problem_j ] ) as the following _ weighted atomic norm minimization _ problem : as a result
, the proposed iterative algorithm can be interpreted as _ reweighted atomic - norm minimization _ ( ram ) . if we let @xmath85 be a constant function or equivalently , @xmath86 , such that there is no preference of the atoms at the first iteration , then the first iteration coincides with the anm . from the second iteration ,
the preference is defined by the weighting function @xmath87 specified above .
note that @xmath88 corresponds to the power spectrum of capon s beamforming ( see , e.g. , @xcite ) if @xmath89 is interpreted as the covariance of the noiseless data and @xmath43 as the noise variance .
therefore , the reweighting strategy makes the frequencies around those estimated by the current iteration preferable at the next iteration and thus enhances sparsity . at the same time
, the preference leads to finer details of the frequency spectrum in that area and enhances resolution .
since the `` noise variance '' @xmath43 can be translated as the confidence level in the current estimate , from this perspective we should gradually decrease @xmath43 and correspondingly increase the confidence in the solution during the algorithm .
in this subsection , we study the success rate of ram in super - resolution compared to anm . in particular , we fix @xmath90 and @xmath91 with the sampling index set @xmath23 being generated uniformly at random .
we vary the duo @xmath92 and at each combination we randomly generate @xmath0 frequencies such that they are mutually separated by at least @xmath93 .
we randomly generate the amplitudes @xmath94 independently and identically from a standard complex normal distribution . after obtaining the noiseless samples , we carry out super - resolution using anm and ram , both implemented by an off - the - shelf sdp solver sdpt3 @xcite .
the recovery is called successful if both the relative mse of signal recovery and the mse of frequency recovery are less than @xmath95 . at each combination @xmath92 , the success rate
is measured over 20 monte carlo runs . in ram , we first scale the measurements such that @xmath96 and compensate the recovery afterwards .
we start with @xmath86 and @xmath97 as default .
we halve @xmath43 when beginning a new iteration until @xmath98 . we terminate ram
if the relative change ( in the frobenius norm ) of the solution @xmath99 at two consecutive iterations is less than @xmath100 or the maximum number of iterations , set to 20 , is reached .
we plot the success rates of anm and ram with @xmath101 in fig .
[ fig : phasetrans_1 ] , where it is shown that successful recovery can be obtained with more ease with a smaller @xmath0 and a larger frequency separation @xmath93 , leading to a phase transition in the sparsity - separation domain .
it is shown that ram significantly enlarges the success phase and hence enhances sparsity and resolution compared to anm .
at @xmath102 we did not find a single failure in our simulation whenever @xmath103 and @xmath104 .
the phase transitions of both anm and ram are not sharp since the frequencies are separated by _ at least _
@xmath93 and a set of _ well separated _ frequencies can be possibly generated at a small value of @xmath93 .
it is also observed that ram tends to converge in less iterations with a smaller @xmath0 and a larger @xmath93 .
( top ) and @xmath102 ( bottom ) , @xmath90 and @xmath91 .
the grayscale images present the success rates , where white and black colors indicate complete success and complete failure , respectively.,title="fig:",width=158 ] ( top ) and @xmath102 ( bottom ) , @xmath90 and @xmath91 .
the grayscale images present the success rates , where white and black colors indicate complete success and complete failure , respectively.,title="fig:",width=156 ] ( top ) and @xmath102 ( bottom ) , @xmath90 and @xmath91 .
the grayscale images present the success rates , where white and black colors indicate complete success and complete failure , respectively.,title="fig:",width=158 ] ( top ) and @xmath102 ( bottom ) , @xmath90 and @xmath91 .
the grayscale images present the success rates , where white and black colors indicate complete success and complete failure , respectively.,title="fig:",width=156 ] we apply the proposed ram method to doa estimation .
in particular , we consider a 10-element sparse linear array ( sla ) with sensors positions indexed by @xmath105 , where the distance between the first two sensors is half the wavelength .
hence , we have that @xmath106 and @xmath107 .
we consider that @xmath108 narrowband sources impinge on the sensor array from directions corresponding to frequencies @xmath109 , @xmath110 , @xmath111 and @xmath112 , and powers @xmath113 , @xmath113 , @xmath114 and @xmath115 , respectively . it is challenging to separate the first two sources which are separated by only @xmath116 .
complex normal noise is added to the samples with variance @xmath117 and @xmath65 is defined as @xmath118 , where @xmath119 ( mean + twice standard deviation ) upper bounds the noise energy with high probability .
we consider both the cases of uncorrelated and correlated sources while the later case is usually considered to be more difficult with existing methods such as music ( see , e.g. , @xcite ) . in the latter case ,
sources 1 and 3 are set to be coherent ( completely correlated ) .
assume that @xmath120 data snapshots are collected which are corrupted by i.i.d .
gaussian noise of unit variance .
we propose a dimension reduction technique to reduce the order of the sdp matrix from @xmath121 to @xmath122 and accelerate the computational speed , which is detailed in @xcite .
we terminate ram within maximally 10 iterations and consider music and anm for comparison .
our simulation results of 100 monte carlo runs are presented in fig .
[ fig : noisy_noiselevel ] ( only the first 20 runs are presented for music for better illustration ) . in the absence of source correlations
, music has satisfactory performance in most scenarios .
however , its power spectrum exhibits only a single peak around the first two sources ( i.e. , the two sources can not be separated ) in at least 3 out of the first 20 runs ( indicated by the arrows ) .
moreover , music is sensitive to source correlations and can not detect source 1 when it is coherent with source 3 .
anm can not separate the first two sources in the uncorrelated source case and always produces many spurious sources .
in contrast , the proposed ram always correctly detects 4 sources near the true locations , demonstrating its capabilities in enhancing sparsity and high resolution .
anm and ram take @xmath123s and @xmath124s on average , respectively , while these numbers can be greatly decreased with more sophisticated algorithms ( see @xcite ) .
in this paper , we studied the spectral super - resolution problem with partial samples and mmvs .
motivated by its connection to the topic of lrmr , we proposed reweighted atomic - norm minimization ( ram ) for achieving high resolution compared to currently prominent atomic norm minimization ( anm ) and validated its performance via numerical simulations .
z. yang and l. xie , `` on gridless sparse methods for line spectral estimation from complete and incomplete data , '' revised version submitted to _ ieee transactions on signal processing _ , _ available online at http://arxiv.org/abs/1407.2490_ , 2014 .
m. fazel , h. hindi , and s. p. boyd , `` log - det heuristic for matrix rank minimization with applications to hankel and euclidean distance matrices , '' in _ american control conference _ , vol .
3.1em plus 0.5em minus 0.4emieee , 2003 , pp . | the super - resolution theory developed recently by cands and fernandes - granda aims to recover fine details of a sparse frequency spectrum from coarse scale information only .
the theory was then extended to the cases with compressive samples and/or multiple measurement vectors .
however , the existing atomic norm ( or total variation norm ) techniques succeed only if the frequencies are sufficiently separated , prohibiting commonly known high resolution . in this paper , a reweighted atomic - norm minimization ( ram )
approach is proposed which iteratively carries out atomic norm minimization ( anm ) with a sound reweighting strategy that enhances sparsity and resolution .
it is demonstrated analytically and via numerical simulations that the proposed method achieves high resolution with application to doa estimation . |
the high - frequency radio sky ( @xmath0 ghz ) has been much less widely studied than the population at lower radio frequencies ( e.g. 1.4 ghz ) , mainly due to the increased time required to survey an area to an equivalent depth at higher frequencies . in recent years
several high - frequency surveys have been conducted , albeit with higher flux density limits than surveys at lower frequencies .
@xcite carried out the ninth cambridge ( 9c ) survey at 15 ghz using the ryle telescope .
this survey covers 520 deg@xmath1 to a completeness limit of @xmath2 25 mjy , and was the first high - frequency survey to cover a significant proportion of the sky .
a series of deeper regions were also observed @xcite , with 115 deg@xmath1 complete to @xmath2 10 mjy and 29 deg@xmath1 complete to @xmath2 5.5 mjy .
the whole southern sky has been surveyed by the australia telescope 20 ghz ( at20 g ) survey @xcite , which has a flux density limit of 40 mjy and is 93 per cent complete above 100 mjy .
this survey is complementary to the 9c survey as it covers a larger area at a shallower flux density .
more recently , the at20g - deep pilot survey @xcite surveyed 5 deg@xmath1 to a completeness level of 2.5 mjy . the tenth cambridge ( 10c )
survey was observed with the cambridge arcminute microkelvin imager ( ami ; @xcite ) at 15.7 ghz .
the observations , mapping and source extraction are described in @xcite ( hereafter paper f ) and the source counts and spectral properties are presented in @xcite ( hereafter paper d ) .
the 10c survey is complete to 1 mjy in ten different fields covering a total of @xmath3 ; deep areas covering @xmath4 , contained within these fields , are complete to 0.5 mjy , making the 10c survey the deepest high - frequency radio survey published to date .
@xcite used data at a range of frequencies to study the spectral indices of 10c sources in the lockman hole .
they found a significant change in spectral index with flux density ; the median spectral index @xmath5 ( where @xmath6 for a source with flux density @xmath7 at frequency @xmath8 ) calculated between 1.4 and 15.7 ghz changes from 0.75 for sources with @xmath9 mjy to 0.08 for sources with @xmath10 mjy .
this shows that there is a population of flat - spectrum sources emerging below 1 mjy ; @xcite suggest that this may be due to the cores of fanaroff and riley type i ( fri ; @xcite ) sources becoming dominant at 15.7 ghz .
there have been several attempts to model the high - frequency radio sky , often extrapolating from lower frequencies .
early evolutionary models of radio sources @xcite successfully fitted the available data at frequencies @xmath11 10 ghz down to flux densities of a few mjy .
more recently , produced a model of the radio source counts at frequencies @xmath12 5 ghz ( up to 30 ghz ) which successfully fitted the data available at the time .
the model splits the sources into flat and steep - spectrum populations , with the flat - spectrum population further divided into flat - spectrum radio quasars and bl lacs , and determines the epoch - dependent luminosity function for each population .
starforming galaxies , ghz - peaked spectrum ( gps ) sources and objects in the late stages of agn evolution are also included in the model . a more recent model by used physically grounded models to extrapolate the 5-ghz source count , which is well known observationally , to higher frequencies .
they focus on the spectral behaviour of blazars and compare three different models which treat flat - spectrum sources differently .
this scheme is successful at high flux densities but does not accurately reproduce the observed 15 ghz source count below @xmath2 10 mjy . the number of sources is significantly underestimated ( by a factor of @xmath13 ) , indicating that the properties of these sources are not well understood , largely due to the complexity and diversity of the high - frequency spectra of individual sources .
@xcite have produced a semi - empirical simulation of the extragalactic radio continuum sky ( the skads simulated sky ; s@xmath14 ) .
the simulation splits the radio sources into separate populations : frii sources , fri sources , radio - quiet agns and starforming galaxies , which are split into quiescent starforming and starbursting galaxies . the observed ( and extrapolated ) radio continuum luminosity functions
are used to generate a catalogue of @xmath2 320 million simulated sources .
this simulation covers 20 @xmath15 20 deg@xmath1 out to a cosmological redshift of @xmath16 and down to a flux density of 10 njy at 151 , 610 mhz , 1.4 , 4.86 and 18 ghz .
@xcite showed that the simulation fails to reproduce the spectral index distribution of 10c sources , dramatically under - predicting the number of flat spectrum sources .
thus , although good progress has been made in recent years in modelling aspects of the extragalactic radio source population , the high - frequency radio source population at low flux densities is poorly described . to understand the nature of the faint , high - frequency population and constrain the models better observations of the faint ,
high - frequency sky are required . in this paper
we present new , deep observations at 15.7 ghz in two 10c fields , which when combined with existing 10c data enable the source counts to be constrained down to 0.1 mjy , a factor of five deeper than the 10c source count .
this paper is laid out as follows .
section [ section : obs ] describes the observations and data reduction , and section [ section : source_catalogue ] details the methods used to produce the source catalogue .
section [ section : checks ] discusses the effects of flux density variability .
the catalogue is used to derive the source counts in section [ section : source_counts ] and these results are discussed in section [ section : sc_discussion ] before some brief conclusions are presented in section [ section : conclusions ] .
the 15.7-ghz observations were conducted between august 2008 and july 2014 using the arcminute microkelvin imager large array ( ami la ) , located near cambridge , uk .
the ami la consists of eight 13-m antennas with baselines of 18 to 110 m , giving it a primary beam of 5.5 arcmin and a resolution of 30 arcsec . for the 10c survey
the la observations were made of 10 fields of different sizes , covering a total of 27 deg@xmath1 .
each field consists of an outer region complete to 1 mjy beam@xmath17 , and an inner deeper region complete to 0.5 mjy beam@xmath17 .
here we describe further observations of two of the 10c fields : the ami001 ( j0024 + 3152 ) field and the lockman hole field ( j1052 + 5730 ) , details of which are given in table [ tab : regions ] . from these two fields we selected sub - fields that we observed further , using a rastering technique in the same way as the 10c survey observations , with hexagonal rasters of 37 pointings
each separated by 4 arcmin . in the ami001 field ,
450 hours of new observations have been made in several hexagons ( each consisting of 37 pointings ) , and the rms noise in the centre of the two deepest hexagons is @xmath18 . the noise map for the ami001 field is shown in fig .
[ fig : ami1_noise ] . in the lockman hole field ,
300 hours of additional observations were carried out in one 37-pointing hexagon .
the rms noise in this hexagon is @xmath19jy beam@xmath17 , as shown by the noise map in fig .
[ fig : lh_noise ] .
lllld + region & comments & area & observing time & + & & / deg@xmath1 & / hours & + + 10c shallow & complete to 1 mjy & 3.56 & & 150 + 10c deep & complete to 0.5 mjy & 1.69 & & 45 + 10c ultra - deep & this work deepest obs . in two hexagons centred on & 0.38 & additional 450 & 16 + & @xmath20 and @xmath21 & & & + + 10c shallow & complete to 1 mjy & 1.78 & & 120 + 10c deep & complete to 0.5 mjy & 0.54 & & 45 + 10c ultra - deep & this work
hexagon centred on @xmath22 & 0.18 & additional 300 & 21 + all raw data files were reduced with the up - to - date reduction pipeline in the ami in - house software package reduce ( paper f ) .
this new pipeline was developed from that used for the 10c survey with enhanced methods for flagging interference and produces consistent results .
@xmath23-fits files are written out from reduce and concatenated to produce a single @xmath23-fits file for each field .
mapping is carried out in aips using the same imaging pipeline as that for the 10c survey , with automated clean procedures and the imean task used to estimate the noise .
the pixel size in the image is @xmath24 arcsec .
details of the data reduction and imaging procedure can be found in paper f.
the source fitting was performed using the cambridge in - house software source_find in the same way as for the 10c survey in order to make the new catalogue as comparable as possible to the 10c catalogue . the procedure implemented in source_find
is described briefly here , full details being given in paper f. the images were used in conjunction with the noise maps shown in figs .
[ fig : ami1_noise ] and [ fig : lh_noise ] to identify component ` peaks ' above a given signal - to - noise ratio @xmath25 ( in this case @xmath26 , to take account of the 1.082 phase error correction factor applied to the flux densities ( see section [ section : flux_values ] for details ) . ] ) .
initially , all pixels greater than @xmath27 ( where @xmath28 is the local noise ) were identified to ensure all peaks greater than @xmath29 after interpolation are found . the position and flux density of each peak ( ra@xmath30 , dec@xmath30 and @xmath31 )
was then found by interpolation between the pixels , and any peaks less than @xmath29 were discarded .
the error on the peak flux density ( @xmath32 ) is taken to be the thermal noise error @xmath33 combined in quadrature with the five per cent calibration error , @xmath34 .
the integration area , consisting of contiguous pixels down to a lowest contour value of 2.5@xmath28 , was then calculated for each component .
if more than one peak lay inside the same integration area the sources were classified as a ` group ' .
the aips task jmfit was then used to fit a 2d elliptical gaussian to each component .
this was used to estimate the integrated flux density , position and angular size ( ra@xmath35 , dec@xmath35 , @xmath36 and @xmath37 ) for each component .
the error on the integrated flux density ( @xmath38 ) was estimated as the error due to thermal noise ( @xmath33 ) combined in quadrature with the five per cent calibration error , @xmath39 .
the maps display an increase in noise close to bright ( @xmath40 mjy ) sources due to amplitude , phase and deconvolution errors . because this noise is non - gaussian
it is often not included in the noise maps , leading to spurious detections close to bright sources .
for this reason , an ` exclusion zone ' was defined around any source with a peak flux density greater than 15 mjy during the source fitting process , and any source detected within this zone was excluded from the catalogue . as derived empirically in paper d ,
each exclusion zone is a circle centred on the bright source , with a radius defined by : @xmath41 where @xmath31 is the peak flux density of the bright source .
the positions and radii of these exclusion zones are shown in table [ tab : zones ] .
.the positions of the centres of the exclusion zones around bright sources and their radii . [ cols="^,^,^",options="header " , ] jy , showning the regions in the ami001 field with rms noise @xmath42 and 20 @xmath43jy . ]
@xcite investigated the effects on deep source counts at 1.4 ghz of sample variance induced by source clustering by extracting a series of independent samples from the s@xmath14 catalogue and comparing to observations .
they used this to present a method for estimating the uncertainty in the source count caused by sample variance for an arbitrary radio survey . in the ami001 field
there is 0.23 deg@xmath1 with rms noise @xmath44jy and 0.42 deg@xmath1 with rms noise @xmath45jy , as shown in the contour plot in fig .
[ fig : noise_kntrs_ami001 ] .
it should therefore be possible to detect a source with @xmath46 mjy in an area of 0.23 deg@xmath1 ; reading off fig .
2 in @xcite for this survey limit and area gives an uncertainty due to cosmic variance of @xmath47 per cent . at higher flux densities
, this uncertainty decreases as the area over which a source could be detected ( the effective survey area ) increases .
for example , for a survey limit of 0.125 mjy ( equivalent to an rms noise of 25@xmath48jy ) and an area of 0.42 deg@xmath1 the uncertainty is 12 per cent .
we therefore expect cosmic variance to affect the faintest flux density bin by @xmath47 per cent .
several effects which could bias the source counts are considered here ; however for the reasons discussed no corrections are made for these biases .
1 . resolution bias . to calculate the source counts a sample which is complete in terms of integrated flux density is required , but the sources are detected according to their peak flux densities .
this means that a resolved source of a given total flux density is more likely to fall below the peak flux density detection threshold than a point source with the same total flux density . due to the relatively large beam size of ami ,
only 7 per cent of sources in the ami001 field and 4 per cent in the lockman hole field are extended , so resolution is not expected to have a significant effect on these source counts .
2 . eddington bias @xcite . statistical fluctuations due to thermal noise can alter the flux density of sources and can therefore cause some sources to be put into the wrong bins .
given the shape of the source count , if the true number of sources in one bin is larger than in adjacent bins , more sources will be scattered into that bin than out of it
. the observed number of sources will therefore be biased high .
in paper d it is estimated that this effect will increase the number of sources in the faintest 10c flux density bin by @xmath49 per cent . as incompleteness is expected to reduce the number of sources in this bin by a similar amount they do not correct for this effect .
no correction is applied here for the same reason .
variability in flux densities can cause sources near the edge of flux density bins to move between bins .
the shape of the source counts means that at the bottom of the bin , the number of sources in the positive phase of variability which are included in the bin will be marginally higher than the number of sources in the negative phase of variability which are excluded .
the opposite effect occurs at the other end of the bin but will not be enough to offset this effect .
therefore , variability will boost the number of sources in a bin , causing a constant shift in the observed source count .
the long timescale over which the observations were made means that the effects of short - term variability are generally averaged out , and as discussed in section [ section : checks ] , we do not find any evidence for significant variability on longer timescales .
figs . [ fig : counts ] and [ fig : de - zotti - counts - combined ] show that the new deeper counts are consistent with the extrapolated 10c fit .
there is no sign of the upturn observed in the 1.4-ghz source counts at @xmath50 mjy ( e.g. ) ; there is thus no evidence for a new population ( e.g. of starforming galaxies ) contributing to the 15.7-ghz source population above 0.1 mjy .
this is not surprising , as a source with @xmath51 , typical for a starforming galaxy @xcite , with @xmath52 mjy will have @xmath53 mjy , and would therefore only have appeared at the bottom of the faintest flux density bin here . a model of the high - frequency ( @xmath54 ) source counts
was produced by , as described in section [ section : intro ] .
the new 15.7-ghz source counts presented here are compared to the latest version of the model , extracted from their website , in fig .
[ fig : de - zotti - counts - combined ] .
paper d showed that the model under - predicts the number of sources in the 10c survey below @xmath55 mjy ; it is clear in fig .
[ fig : de - zotti - counts - combined ] that the model continues to under - predict the number of sources observed by a factor of two as flux density decreases .
@xcite studied the spectral indices of a sample of 10c sources and showed that the proportion of flat - spectrum sources in particular is too low in the model below @xmath56 mjy ; the model predicts that at @xmath57 mjy steep - spectrum sources outnumber flat - spectrum source by a factor of three , while the observations show that there are twice as many flat - spectrum sources as steep - spectrum sources .
it is likely that this under - prediction of the number of flat - spectrum sources in the sub - mjy population is responsible for the discrepancies between the model and the observed count seen here at @xmath58 . above 5 mjy the model over - predicts the number of sources observed , this is discussed further in paper d. the higher - flux density end ( 0.5 mjy to several jy ) of the 15-ghz source counts is studied by combining the 9c and 10c counts with the at20 g survey by @xcite , who investigated the source counts of the steep and flat - spectrum populations separately .
these counts are compared to the and @xcite models , and they find that both underestimate the number of flat - spectrum sources below 5 mjy .
the new 15.7-ghz source count is also compared to the s@xmath14 catalogue in fig .
[ fig : de - zotti - counts - combined ] .
all sources with @xmath59 mjy were selected from the simulation , and the source count was calculated in the same bins as for the observed count .
the simulation under - predicts the observed number of sources in a similar way to the model .
it is likely that this under - prediction is again due to a lack of flat spectrum sources in the model , as @xcite showed that the s@xmath14 and 10c spectral index distributions are significantly different , with the simulation missing almost all the flat spectrum sources .
( note that the s@xmath14 and models are not entirely independent as they are both extrapolations from models constructed using low - frequency data . )
@xcite used multi - wavelength data to show that these flat - spectrum 10c sources are probably the result of emission from the cores of radio galaxies , which therefore have a far greater contribution than predicted by either of the models discussed here . below @xmath60 mjy there is a better agreement between the observed and simulated counts due to a slight flattening in the s@xmath14 counts .
this flattening is due to the greater contribution of starforming sources ( both quiescent and starbursting ) to the simulated catalogue below @xmath60 mjy ; starforming sources comprise 21 per cent of the simulated sources with @xmath61 , compared to only 7 per cent of sources with @xmath62 mjy .
however , given that there is no flattening in the new ami count , it is not clear what contribution , if any , a population of starforming sources is making to the observed counts at @xmath46 mjy . a study of the multi - wavelength properties of 10c sources in the lockman hole @xcite used radio - to - optical ratios to show that at least 94 per cent of the 10c sample ( @xmath63 mjy ) are radio loud , demonstrating that there is no significant population of starforming galaxies present at these flux density levels .
given that there is no change in the slope of the source count for @xmath64 , it seems unlikely that starforming galaxies are making a significantly greater contribution in this flux density range than at higher flux densities .
therefore , the population of starforming galaxies predicted to be making a significant contribution ( @xmath65 per cent ) in this flux density range by the s@xmath14 simulation do not appear to be present .
this is consistent with the results of several recent studies of the faint ( @xmath66 mjy ) source population at lower frequencies @xcite which have also found fewer starforming galaxies than predicted by the simulation .
this paper presents new very deep ( best rms noise = 16 @xmath43jy beam@xmath17 ) 15.7-ghz observations in two fields .
these are the deepest high - frequency radio observations to date , and enable us to calculate the source counts down to @xmath67 mjy .
this is a factor of five deeper than previously achieved with the 10c survey .
the source counts are consistent with the extrapolated fit to the 10c count .
there is thus no evidence for a new population of objects contributing to the 15.7-ghz source counts above 0.1 mjy , suggesting that the high - frequency radio sky continues to be dominated by radio galaxies down to at least this flux density level .
we do not observe the population of starforming galaxies predicted to begin to contribute @xmath64 by the s@xmath14 simulation .
comparisons with the model and s@xmath14 simulation show that both these underestimate the observed number of sources at low flux densities by a factor of two .
this is probably due to the flat - spectrum cores of radio galaxies contributing more significantly to the counts than predicted by the models .
we thank the staff of the mullard radio astronomy observatory for maintaining and operating ami .
ihw and cr acknowledge science and technology facilities council studentships .
ihw acknowledges support from the square kilometre array south africa project and the south african national research foundation . this research has made use of nasa s astrophysics data system .
we thank the referee for their careful reading of this manuscript . | we present deep 15.7-ghz observations made with the arcminute microkelvin imager large array in two fields previously observed as part of the tenth cambridge ( 10c ) survey .
these observations allow the source counts to be calculated down to 0.1 mjy , a factor of five deeper than achieved by the 10c survey .
the new source counts are consistent with the extrapolated fit to the 10c source count , and display no evidence for either steepening or flattening of the counts .
there is thus no evidence for the emergence of a significant new population of sources ( e.g. starforming ) at 15.7 ghz flux densities above 0.1 mjy , the flux density level at which we expect starforming galaxies to begin to contribute .
comparisons with the model and the skads simulated sky show that they both underestimate the observed number of sources by a factor of two at this flux density level .
we suggest that this is due to the flat - spectrum cores of radio galaxies contributing more significantly to the counts than predicted by the models .
[ firstpage ] galaxies : active radio continuum : galaxies catalogues surveys |
spontaneous collective motion , a coordinated motion of an assembly of moving entities which interact locally without any leader , has recently drawn a lot of attention from the statistical physics community @xcite . such a phenomenon is present both in biological systems like motility assays @xcite and bacterial colonies @xcite or on a larger scale insects swarms @xcite and flocks of birds @xcite , as well as in engineered systems like driven colloids @xcite , droplets @xcite , or grains @xcite . from a theoretical perspective ,
such moving individuals are represented as polar self - propelled particles , that is , particles set in motion by a driving force directed along the heading vector of the particle .
this driving force being balanced by a friction force , a constant speed is reached in the absence of interaction with obstacles or other particles .
assuming , as a simplification , that the speed is always constant leads to minimal models like the vicsek model @xcite , where particles also interact with their neighbors so as to align their velocity vectors , up to some noise .
in other words , vicsek - type models can be thought of as spin - models @xcite , where particles move along the direction of their spin instead of remaining fixed on the node of a lattice .
this analogy with spin models is important because in two dimensions ( the dimension in which the vicsek model is usually defined ) , the mermin - wagner theorem @xcite prevents the existence of long - range order for equilibrium models of spins that are invariant under global rotation of the spins , due to the presence of low - energy excitations called spin - waves @xcite .
the existence of long - range order in the two - dimensional vicsek model @xcite thus reveals the intrinsic non - equilibrium character of the model , and to some extent , of the phenomenon of collective motion itself .
the presence of long - range order in the vicsek model can be understood as resulting from the continuous evolution of the neighborhood of a given particle , which successively interact with different particles . as mentioned above
, an important simplification of vicsek - type models is to identify the direction of motion with that of the heading vector of the particles . at low enough density , when the relaxation time of the velocity is short as compared to the typical time between successive interactions , this approximation is well - justified . in a denser regime however , as observed in experiments on shaken polar grains @xcite , velocity and heading vectors may have different directions , and it is a priori relevant to consider them as distinct dynamical variables , as done in @xcite , assuming a suitable coupling between velocity and heading .
in this situation , we are thus dealing with a fluid of particles carrying a ( classical ) spin .
when considering both velocity and heading ( or spin ) variables , the possible existence of collective motion at equilibrium can not be immediately ruled out by standard arguments if spins and velocities are coupled .
the first argument ( in two dimension ) is the mermin
wagner theorem , but its applicability is not granted if spins interact with velocities , because the system is then no longer invariant under a rotation of the spins alone .
the second argument is that for an isolated system at equilibrium the momentum is conserved and no spontaneous global motion can emerge .
alternatively , boundaries or a substrate may break momentum conservation , but they act as a momentum sink , also preventing collective motion .
these arguments are however valid only if the standard relation @xmath0 between momentum and velocity holds .
although very general , it is well - known that such a relation breaks down when the ` potential energy ' ( the potential term in the lagrangian ) depends on the velocities as is the case for charges in magnetic field .
a third argument related to the second one , is that at equilibrium the center - of - mass velocity is conjugated , in a thermodynamic sense , to momentum just like temperature is conjugated to energy and is thus equal to that of the surrounding medium @xcite
. invoking galilean invariance , one then usually sets the center - of - mass velocity to zero .
however galilean invariance does not hold in those above situations where the potential term in the lagrangian depends on the velocities , so that the center - of - mass velocity should be a relevant parameter in this case .
a natural question is thus to investigate whether a coupling between velocity and spin variables could break the galilean invariance , modify the standard definition of momentum , and as a result allow for the possibility of collective motion at equilibrium equilibrium being understood in the generic sense of the statistical steady - state of a hamiltonian system , in the absence of external forcing . in this paper
, we investigate this issue by considering a simple two - dimensional model of point - like particles carrying a spin and evolving according to a conservative dynamics coupling spin and velocity in a minimal way .
the dynamics is originally defined in a lagrangian formalism , from which a hamiltonian formulation is derived .
a non - standard expression of the momentum of each particle in terms of its velocity and spin is obtained .
we study the statistical mechanics of such a system , and show that in the fully connected ( or mean - field ) case a transition to collective motion occurs .
the velocity of the center of mass , which in the absence of galilean invariance , is a relevant variable , also feeds back on the magnetization properties : it acts as an external magnetic field that smoothens the transition and stabilizes non trivial local minima of the free energy .
molecular dynamics ( md ) simulations of finite size systems confirm the existence of a homogeneous polar phase . increasing sizes
, the system organizes into domains of collectively moving particles structured around topological defects .
we start by considering a liquid of xy - spins in two dimensions .
the lagrangian of such a model is described by @xmath1 with : [ eq : lagrangienxy ] @xmath2 where @xmath3 , with @xmath4 , @xmath5 and @xmath6 , denoting the position , the angle and the spin of each particle respectively ( the dot on top of a variable denotes time derivative ) .
note that the spin @xmath6 is defined as @xmath7 .
the other parameters appearing in the lagrangian are the mass @xmath8 of the particles , their moment of inertia @xmath9 associated to spin rotation , the coupling @xmath10 between spins , the external field @xmath11 , and the interaction potential @xmath12 between particles ( e.g. , hard sphere or lennard - jones potential ) . with the model as it stands , there is coordination of spins at low temperature .
however there is no coupling between spin and particle motion , so that no collective motion can emerge . to couple the motion of the particles to the spin we add to the lagrangian a term @xmath13 where @xmath14 is a coupling constant between spin and velocity , which is expected to favor the alignment of the velocity of particle @xmath15 to its _ own _
spin @xmath6 .
apart from the fact that it should contribute to align velocities when spins align although we shall see that the effect is really indirect and counter - intuitive this term is motivated by the observation of such self - alignment in real systems of self - propelled grains @xcite and has been identified as a key ingredient for the dynamics of self - propelled discs @xcite . leaving aside the potential term @xmath16 , although we shall reintroduce it when moving to the molecular dynamics simulations , our starting point is thus the following dimensionless lagrangian : @xmath17 where we used the following redefinitions @xmath18 the parameter @xmath14 now controls the strength of the alignment between spin and velocity .
the euler
lagrange equations then read [ eq : lagrangegroupreq ] @xmath19 where @xmath20 and @xmath21 is a unit vector perpendicular to the spin , defined as @xmath22 . finally , we reformulate the dynamics in the hamiltonian formalism .
the momenta @xmath23 and @xmath24 conjugate to the positions @xmath4 and angles @xmath5 read : [ eq : conjmoment ] @xmath25 the spin of the particle acts as an internal potential vector and only for @xmath26 is the momentum equal to the velocity
. the hamiltonian is obtained by the usual legendre transform of the lagrangian @xmath27 expressing @xmath28 and @xmath29 in terms of the conjugate variables @xmath23 and @xmath24 , one finds , discarding an irrelevant constant term : @xmath30 the equations of motion follow : [ eq : hamiltongroupreq ] @xmath31 let us note that , up to a constant the hamiltonian can be formally written as @xmath32 although this is not a proper formulation in terms of the canonical variables , it shows that written in terms of kinetic and potential energy , the hamiltonian takes a form similar to the one of a standard liquid of particles carrying an xy - spin .
the subtlety hidden in this misleadingly simple formulation is that @xmath28 is not proportional to @xmath23 .
taking advantage of the lagrangian formulation , we apply noether s theorem to obtain the conserved quantities of the dynamics , associated to the continuous transformation under which the lagrangian is invariant .
first of all , the lagrangian is invariant under time and space translation , so that the dynamics conserves the energy @xmath33 and the _ generalized _ linear momentum @xmath34 obtained by summing over all particles : @xmath35 where @xmath36 is the center of mass velocity and @xmath37 is the total magnetization of the system .
this simple but essential formula relates the total momentum of the system to the collective motion , described by the velocity of the center of mass , and the total magnetization of the system .
although the generalized linear momentum considered here is not proportional to the velocity of the center of mass , this conservation law is in stark contrast with most two - dimensional models for collective motion , for which momentum is not conserved .
the equations of motion are not invariant under a rotation of the system ( i.e. , a global rotation of the position of particles ) , as can be seen from , where the scalar product is not invariant , as the spins do not rotate , and also because of the presence of the external field @xmath11 .
, the invariance is however recovered under a transformation in which the spins are rotated together with the positions of the particles . in that case
a supplementary conserved quantity is @xmath38 with @xmath39 the usual angular momentum . in practice , even when the field @xmath40 , this rotational invariance is broken because of boundary conditions .
] finally , let us stress a significant difference with the standard xy - model @xcite .
in the present case , there is no symmetry under rotation of the spins alone .
accordingly spin waves are not slow modes of the dynamics and the mermin - wagner theorem ( at least in its form ) does not apply , opening the way for possible long - range order .
interestingly the dynamics is not time reversible .
applying the transformation @xmath41 : @xmath42 to the euler - lagrange equations , they transform into @xmath43 which differ from the original ones [ eq . ] , unless @xmath26 .
note that although the hamiltonian is symmetric under time reversal , it does not imply time reversibility of the trajectories because @xmath23 transforms under time reversal into @xmath44 , as can also be checked directly from eq . .
the invariance of the trajectories can be recovered using the more general transformation @xmath45 and @xmath46 . in this case
, one has @xmath47 and both the hamiltonian and the trajectories are invariant .
the invariance under this more general transformation can be interpreted as a generalized form of microreversibility . at a heuristic level
, this suggests that a generalized form of detailed balance , associated to a uniform measure in the micro - canonical ensemble , may hold .
also central in classical mechanics is the galilean invariance , which states that the equations of motion are identical in different coordinate systems moving with constant velocity with respect to each other . applying a galilean transformation to the euler - lagrange equations @xmath48
while keeping @xmath5 and its derivatives unchanged , one finds in the new frame , dropping primes to lighten notations , @xmath49 where the extra term @xmath50 in the last equation breaks galilean invariance .
this absence of galilean invariance is actually connected to the fact that the total momentum @xmath34 is not proportional to the velocity @xmath36 of the center of mass , as can be seen from the following argument .
the term responsible for the broken galilean invariance is the term @xmath51 coupling spin and velocity [ eq . ] .
let us momentarily consider a general coupling term @xmath51 , assuming simply that it depends only on the spins @xmath52 and the velocities @xmath28 , @xmath53 .
then the total momentum reads @xmath54 and is thus generically not proportional to the velocity of the center of mass .
however , if the coupling term @xmath51 satisfies galilean invariance , one has @xmath55 , where @xmath56 is the velocity shift associated to an infinitesimal galilean transformation , resulting in @xmath57 so that one recovers @xmath58 . in the following , when discussing the role of the broken galilean invariance , we may thus consider the following coupling term , @xmath59 for @xmath60 , this term corresponds to the coupling defined in eq . ,
while for @xmath61 , galilean invariance is recovered and @xmath58 . to gain some intuition on particle motion , it is useful to analyze the equations of motion for a single and free particle @xmath62 .
the individual momentum is conserved , @xmath63 .
taking the temporal derivative of eq . and using eq .
, one gets @xmath64 where the coordinate system has been set such that @xmath65 , and @xmath66 .
one recognizes the equation of motion of a pendulum : the spin @xmath67 oscillates around the direction @xmath68 with a frequency @xmath69 .
the velocity of the particle follows from eq . , as illustrated on fig .
[ fig : dirspin ] in the case @xmath70 .
( top ) and @xmath71 ( middle ) .
the momentum @xmath72 is conserved , the spin @xmath6 oscillates around the direction given by @xmath23 with a frequency @xmath69 .
the velocity @xmath28 is then entirely set by eq . .
the plots in the bottom row show that only the @xmath73-components of the trajectories differ , whereas the @xmath74-components oscillate on the same trajectory . ] using the conservation of the angular momentum @xmath75 ( see footnote 1 ) , one finds @xmath76 .
the motion of the particle , perpendicular to the direction of its momentum , follows the periodic motion of @xmath77 .
the dynamics along the direction of the momentum is more complicated . from eq .
one has @xmath78 . for small enough @xmath14 , @xmath79
, the particle always moves in the direction of its momentum .
however , for larger k , the motion strongly depends on the dynamics of @xmath80 ; and , for small @xmath80 , that is when the spin points to a direction close to that of the momentum , the particle moves into the opposite direction ! this counter - intuitive behavior is deeply rooted in the conservation of @xmath81 . in the following
we shall see that the constraints imposed on the dynamics by the conservation laws , together with the coupling between the spins and the velocities , lead at the mean - field level to the onset of collective motion .
starting from the hamiltonian formulation of the model a liouville equation can be written for the probability density function of the phase - space point @xmath82 .
it follows that at equilibrium the probability distribution is a function of the conserved quantities of the dynamics , namely the energy @xmath83 and the linear momentum @xmath34 . in the microcanonical ensemble , the conserved quantities
can not be exchanged with the environment ; they have constant values @xmath84 and @xmath85 respectively , and the distribution @xmath86 is given by @xmath87 where @xmath88 is the microcanonical partition function , defined by normalizing @xmath89 to @xmath90 .
in other words , all configurations with the same values of the conserved quantities are equally probable . in practice ,
it is more convenient from a computational viewpoint to work in the canonical ensemble where the conserved quantities are exchanged with a reservoir . the probability measure in the canonical ensemble is then given by @xmath91 where @xmath92 and @xmath93 are intensive control parameters of the reservoir that determine the thermal averages of @xmath83 and @xmath34 ( @xmath92 denotes , as usual , the inverse temperature ) . the distribution of the microscopic configurations can be written more explicitly as @xmath94 with @xmath95\\ \psi(r_{ik } , \theta_{ik } ) & = \exp\bigl(\beta j(r_{ik } ) \cos\theta_{ik}\bigr ) \
, , \end{aligned}\ ] ] and where @xmath96 is the partition function , with @xmath97 from the distribution , one can easily compute the first and second moments of the momenta @xmath23 and @xmath24 .
for the first moments , one finds @xmath98 and @xmath99 using eq . in eq .
yields @xmath100 .
taking the sum over all particles and dividing by @xmath101 then leads to @xmath102 this relation , which shows that the intensive parameter @xmath93 associated to the conservation of momentum is the averaged velocity of the center of mass is not specific to the present model .
it is a general property of equilibrium systems @xcite .
however , in the presence of galilean invariance one can arbitrarily set @xmath103 and safely ignore @xmath93 in the partition function . here because of the broken galilean invariance we shall on the contrary keep it as a true and independent intensive thermodynamic parameter . inserting eq . into eq .
, we end up with @xmath104 which can be seen as a ` local ' counterpart of eq . .
for the second moment , one finds a generalization of the equipartition relations , with @xmath105 , @xmath106 hence not only are the average values of momentum and spin related , as could have been anticipated from eq . ,
but so are also their fluctuations .
note however that the temperature is _ not _ proportional to the total kinetic energy because of the coupling between the translational velocities and the spins .
we now study the behavior of the mean magnetization as a function of temperature and center of mass velocity ( or equivalently , @xmath93 ) . to keep the discussion at a simple enough level , we focus on the fully connected geometry , which is akin to a mean - field approximation ( note that the momenta @xmath107 and spins @xmath52 however remain two - dimensional vectors ) . in this case , the interaction amplitude @xmath108 is simply a constant , independent of the distance between particles . to ensure that energy remains extensive , we take this constant to be equal to @xmath109 .
the interaction term can then be rewritten as follows @xmath110 where @xmath111 is the magnetization per spin . disregarding the constant term ( which amounts to a shift in the energy reference ) , the integrations of eqs . , and over @xmath112 , @xmath113 and @xmath114 are readily computed , yielding @xmath115 where @xmath116 is the volume occupied by the system .
the integral that remains to be computed is @xmath117 where one recognizes the mean field partition function of the conventional xy - model in the presence of an external field @xmath118
@xcite . following standard techniques
( see appendix a ) , one finds @xmath119 where the function @xmath120 is given by @xmath121\ ] ] with @xmath122 , the modified bessel function of order @xmath123 and @xmath124 . in the large @xmath101 limit ,
the integral in eq . can be computed using the saddle point approximation , yielding to exponential order @xmath125 where the saddle - point @xmath126 is given by @xmath127 with @xmath128 the unit vector @xmath129 . using eq .
and stirling s approximation , the free energy density @xmath130 then reads @xmath131 it is a function of intensive variables only .
finally the average magnetization @xmath132 per particle is : @xmath133 since @xmath134 , we end up with @xmath135 . combining this last result with the definition of @xmath136 and eqs . and
, we get that @xmath137 .
hence @xmath132 can be obtained from the self - consistent equation , simply replacing @xmath138 by @xmath132 . , but with non - zero effective field @xmath139 .
solid curves are obtained from numerically minimizing @xmath120 .
parameters are @xmath70 , @xmath140 . ] using @xmath141 , we also obtain the average energy per particle @xmath142 : @xmath143 eq . can be solved numerically , the result being depicted in fig .
[ fig : meanfieldanal ] for the case @xmath70 in the absence of external field @xmath144 . without loss of generality
we choose @xmath93 in the @xmath73-direction . when @xmath145 , there is a phase transition from an isotropic to a magnetized phase at @xmath146 . for nonzero values of @xmath93
, we find in the upper half of the figure smooth magnetization curves which demonstrate that the coupling between the spins and the particle velocities is encompassed into an effective field @xmath147 : as long as @xmath148 , the average velocity of the center of mass acts as an external field , polarizing the spins .
this nontrivial effect of the center - of - mass velocity is directly related to the loss of galilean invariance . actually , starting from the more general coupling term given in eq .
, one finds that the effective external field @xmath149 is changed into is also changed into @xmath150\ ] ] with @xmath151 and @xmath152 . ]
@xmath153 so that the contribution from @xmath93 to the effective field vanishes for @xmath61 , when galilean invariance is recovered . in its lower half , figure
[ fig : meanfieldanal ] shows further solutions .
these are also obtained from minimizing @xmath120 , but here the minima are local ones . in these solutions ,
the magnetization @xmath154 is _ opposite _ to the center - of - mass velocity @xmath93 .
we can already anticipate that these local minima are essential to the low temperature physics of the model in the microcanonical ensemble : for a system with vanishing fixed total momentum , eq .
( [ eq : pvgks ] ) imposes @xmath155 and the system will select the minima for which @xmath93 and @xmath132 are anti - aligned .
although observing an ordering transition in a mean - field framework usually does not come as a surprise , let us emphasize that the present transition is non - standard even at mean - field level , in the sense that collective motion can not be observed in equilibrium systems where momentum is either conserved or exchanged with a substrate , as long as the relation @xmath156 is valid .
the transition clearly relies on the coupling between spin and velocity , and disappears for @xmath26 .
also , the above results only hold for the fully - connected model , that is at the mean - field level . in the case of the standard two - dimensional xy - model
, it is very well known that mean - field approximations erroneously predict a transition towards a true long - range ordered phase at low temperature , which in two dimensions is replaced by the celebrated berezinskii kosterlitz
thouless transition towards a quasi - long - range ordered phase , with zero magnetization , but infinite correlation length of its fluctuations @xcite .
physically , long - range order in the two - dimensional xy - model is destroyed by the low - energy spin - wave excitations associated with the invariance of the dynamics under a continuous rotation of the spins . in the present case ,
we have seen in section [ sec : symm ] that this symmetry is absent for @xmath157 when globally rotating the spins alone .
one can however argue that a more general transformation , rotating both the spins and the velocities , should be applied to restore the mermin - wagner theorem .
when a reservoir imposes @xmath93 , or when the conserved momentum has a fixed value @xmath158 , isotropy is actually broken . yet ,
when @xmath159 , the system is isotropic and long wave - length excitations involving spins and velocities may be expected to destroy long - range order .
in order to perform numerical simulations of the dynamics described by the equations of motion , one needs to specify the spatial dependence of the ferromagnetic interaction .
doing so , one notices that the interaction not only leads to alignment of the spins but also to attraction / repulsion of interacting spins , depending on their relative orientations .
typically , spins of similar direction cluster together while oppositely pointed spins repel each other .
an aligned phase is thus expected to end up with all particles being very close together . in order to avoid this undesired behavior ,
we add a repulsive potential @xmath12 as introduced in eq . , which translates into a repulsive force term in eq .
. in the following
, we set @xmath160 for @xmath161 and @xmath162 for @xmath163 .
the interaction range is thus one unit length and the two interactions are of the same magnitude for @xmath164 . .
( b d ) candidates for ground states for @xmath165 .
black dots indicate particle positions in the @xmath166-plane ; left : blue arrows are @xmath28 ; right : red arrows are @xmath6 . @xmath167 . ]
we perform simulations in the microcanonical ensemble in square boxes of size @xmath168 , with periodic boundary conditions and a fixed density of particles @xmath169 ; @xmath170 $ ] .
the equations of motion are integrated using a standard fourth - order runge
kutta approximation .
the total energy @xmath84 and momentum @xmath171 are set by the initial condition . in the following
, we shall always start from initial conditions where all positions and spins orientations are random , and all velocities , both @xmath28 and @xmath24 , are zero .
hence in the initial conditions the total magnetization @xmath172 and the velocity of the center of mass @xmath173 , so that the total momentum @xmath174 . typically , starting from such an initial condition , the system reaches a disordered steady state .
however , from the phase transition diagram in fig .
[ fig : meanfieldanal ] , one would expect that a system prepared with low enough energy will also be at low temperature and will thus evolve towards an ordered state with nonzero magnetization @xmath132 .
if this were the case , because @xmath175 is conserved , the system would acquire a finite velocity of its center - of - mass @xmath176 : _ collective motion would set in spontaneously_. we first look for the ground state for different values of @xmath14 .
we perform slow annealing of the system by removing rotational kinetic energy at a constant rate : every @xmath177 integration steps , a factor @xmath178 is applied to all @xmath24 .
we checked that the annealing rate was small enough to ensure that we observed ( statistically ) the same result for different rates .
we also checked that the total momentum @xmath34 remained null during the annealing procedure .
figure [ fig : gs ] displays the final states obtained from this annealing procedure in a system of @xmath179 particles .
one observes quite a rich phase behavior : for @xmath180 all spins are indeed aligned as in panel ( a ) and , as anticipated , all velocities align in the opposite direction , so that collective motion is present . for @xmath181 , @xmath182 , @xmath183 , and @xmath90
we observe three kinds of different final states occurring at random , which are depicted in panels ( b ) to ( d ) .
they have large - scale structure in their magnetization field , such that the total magnetization vanishes ; as a result @xmath184 : the velocities fluctuate independently , and no global motion takes place .
the three states in fig.[fig : gs]b d can be seen as candidates for the ground state , representing local minima in a free - energy landscape .
the frequency at which one of these states is picked by the annealing process depends on @xmath14 and on the number of particles @xmath101 . for @xmath179 , we could only find the magnetized ground state for @xmath180 .
we now investigate how magnetization resists thermal fluctuations .
figure [ fig : kescan ] displays the phase diagram for systems with zero total momentum @xmath175 . to obtain it , we slowly anneal the system from a disordered initial state with energy @xmath84 , but stop the annealing at a beforehand chosen value of @xmath83 , which is then conserved .
we then wait for the system to relax and start to average the magnetization .
one can see that for sufficiently small @xmath14 magnetization survives on a finite range of energy . .
values of @xmath14 run from @xmath185 to @xmath182 in equal steps . in all simulations @xmath175 , @xmath186 , @xmath187 , @xmath188 . ]
( square ) , @xmath189 ( circle ) , @xmath190 ( triangle ) . in all simulations @xmath180 , @xmath191 , @xmath188 . ] in the fully connected model .
@xmath180 , @xmath192 , @xmath140 , @xmath193 . ] focusing on the case @xmath180 , for which collective motion is indeed observed ( at least in the finite size system ) , we investigate the transition towards the disordered state at high energy .
the simulation data is plotted in fig .
[ fig : compare_fc ] for different system sizes and different momenta . here , for the purpose of comparison with the canonical case
, we choose to plot the magnetization as a function of @xmath194 , where the temperature is measured using the rotational velocity fluctuations , as prescribed by eq . .
a crossover from low to high values of the magnetization is clearly visible when temperature is decreased .
to better compare it with the canonical calculation of sec .
[ sec : fc ] above , although no quantitative agreement is to be expected due to the finite range of spin interactions @xmath195 and the addition of the repulsive potential @xmath196 , we invert the canonical relations @xmath197 and @xmath198 from eqs . and .
we thus obtain in fig .
[ fig : constp ] the magnetization computed in the canonical ensemble as a function of ( @xmath83 , @xmath199 or more conveniently ( @xmath92,@xmath34 ) . for finite @xmath34 continuous branches of solution relate the disordered state at high temperature to the homogeneous magnetized state at low temperature .
these branches resemble very much those obtained from the microcanonical simulations ( fig .
[ fig : compare_fc ] ) .
the zero - momentum case is very peculiar : there is a finite range of energy , or @xmath194 , where no solutions with homogeneous magnetization exist . because of finite - size effects
, this peculiarity could not be captured in the microcanonical simulations . also , from the data in fig .
[ fig : compare_fc ] it is not yet clear whether the magnetization at @xmath200 survives the limit of large systems , @xmath201 .
the transition shifts more and more to the left , but more systematic investigation would be necessary to know whether it converges to a nonzero critical temperature . as a first step we looked at finite size effects , focusing on the ground state .
we ran the annealing protocol on ten systems of size @xmath202 , for respectively @xmath180 and @xmath203 . by increasing the system size , the fully ordered state is now replaced by an inhomogeneous state with zero total magnetization , in a way similar to the effect of increasing @xmath14 at a given system size ( fig . [ fig : gs]b d ) .
the crossover size is larger for smaller @xmath14 . given that for @xmath26 no long - range order is expected , these observations suggest that long - range order may not survive in the thermodynamic limit for all @xmath14 . obtaining the precise scaling in @xmath14 , temperature and system size would require a deeper analysis , both theoretically and numerically , which is beyond the scope of the present work .
in order to discuss the above observations , let us recall that the hamiltonian can be interpreted as the sum of the kinetic and potential energies , in which the peculiarity of the present model is entirely encoded in the unusual relation @xmath204 ( see eq . in section [ definition ] ) . to obtain low energy states ,
the system should ( i ) minimize its kinetic energy , thus decrease all @xmath205 and generate small @xmath206 states ; ( ii ) align its spins , thus favoring states with large @xmath207 .
however this is incompatible with the constraint @xmath208 . in particular , because the constraint imposes latexmath:[$n|{\mathbf{v}}| = k energy .
this dynamical frustration is all the more important that @xmath14 is large .
this is why decreasing @xmath14 is a way to favor ordered states at low energy . at larger @xmath14 ,
the system selects zero magnetization states , at the price of generating bend and splay in the spin field .
also the larger the system , the smaller are these distortions .
this is why only small @xmath14 and small size system exhibit homogeneously ordered phases . in summary ,
we have proposed in this paper a conservative model of particles in which velocities and spins are coupled .
starting from a lagrangian formulation and deriving from it the hamiltonian one , we obtained using the symmetries of the problem that the total ( generalized ) linear momentum is conserved but is no longer proportional to the center - of - mass velocity .
we then studied the effect of this important change on spin statistics and on collective motion .
our main findings are that ( i ) collective motion sets in starting from an immobile system , provided that the spins are able to align ferro - magnetically as observed in particular in the fully connected geometry , or in small enough systems , and ( ii ) the parameter @xmath210 thermodynamically conjugated to linear momentum acts as an external magnetic field on the magnetization .
more generally , the main interest of the present model is to show that collective motion , albeit perhaps of a nonstandard type , is possible even for conservative models , provided that spins and velocities are coupled .
although it is hard to imagine a physical realization of the present model , the latter however has the virtue of showing that , at least at a conceptual level , energy dissipation is not a necessary ingredient for collective motion . also , in a spirit similar to that of @xcite it brings the transition to collective motion on a theoretical playground where a number of tools have been developed to characterize phase transitions . the present paper aimed at introducing the model ,
discuss its symmetries and the crucial role of the broken galilean invariance .
as such , it remains very preliminary .
we have only looked at the mean field scenario , and illustrated the model behavior on a few md simulations .
a number of perspectives can be mentioned .
obviously one would like to investigate more systematically the phase diagram of the system and the existence of a phase transition in the infinite size limit .
also , it would be interesting to make progress in the direction of the theoretical analysis of the model in finite dimension .
in particular two limits of interest are @xmath211 , which corresponds to the intricate physics of the xy - model and its berezinskii - kosterlitz - thouless transition , and @xmath212 , where galilean invariance is recovered .
perturbative approaches using @xmath14 and/or @xmath213 as small parameters could be a way to tackle this theoretical analysis . besides
, we have not yet explored the possibility that the model might be invariant under a more complicated transformation than the galilean one , in analogy to the lorentz transform that arises in the context of electromagnetism .
whether such a transformation exists is however far from clear ; in electromagnetic systems , the invariance under the lorentz transform is only obtained when considering relativistic mechanics , while our model remains within the realm of newtonian mechanics .
interesting discussions with j .- l .
barrat on the mermin - wagner theorem are acknowledged .
in this appendix , we provide the detailed derivation of the mean - field free energy @xmath214 given in eq . .
starting from eq . , we need to compute the following integral , @xmath215 to write the exponential function as a linear function of @xmath216 we use the hubbard
stratonovich transformation , @xmath217 considering for instance the @xmath73-component , and setting @xmath218 and @xmath219 , one gets @xmath220 combining the two directions @xmath73 and @xmath74 yields for the partition function , using the notation @xmath221 , @xmath222 where @xmath223 up to a shift on the variable @xmath5 ( that is irrelevant since integration is on the circle ) , the scalar product @xmath224 can be expressed as @xmath225 where @xmath226 .
we end up with @xmath227 denoting @xmath228 the modified bessel function of order @xmath229 : @xmath230 we finally get eqs . and of the main text . | we propose a conservative two - dimensional particle model in which particles carry a continuous and classical spin .
the model includes standard ferromagnetic interactions between spins of two different particles , and a nonstandard coupling between spin and velocity of the same particle inspired by the coupling observed in self - propelled hard discs . because of this coupling galilean invariance is broken and the conserved linear momentum associated to translation invariance is not proportional to the velocity of the center of mass .
also , the dynamics is not invariant under a global rotation of the spins alone .
this , in principle , leaves room for collective motion and thus raises the question whether collective motion can arise in hamiltonian systems .
we study the statistical mechanics of such a system , and show that , in the fully connected ( or mean - field ) case , a transition to collective motion does exist in spite of momentum conservation .
interestingly , the velocity of the center of mass , which in the absence of galilean invariance , is a relevant variable , also feeds back on the magnetization properties , as it acts as an external magnetic field that smoothens the transition .
molecular dynamics simulations of finite size systems indeed reveal a rich phase diagram , with a transition from a disordered to a homogeneous polar phase , but also more complex inhomogeneous phases with local order interrupted by topological defects .
= 1 _ keywords _ collective motion , xy - model , hamiltonian dynamics , statistical mechanics , numerical simulations |
one of the many suprising applications of shared entanglement @xcite is superdense coding introduced by bennett and wiesner @xcite . in the simplest example of this protocol
, two people ( alice and bob ) share a pair of entangled qubits ( spin half particles or any other two state systems ) in a bell state .
alice can then perform any of the four unitary operations given by the identity @xmath0 or the pauli matrices @xmath1 and @xmath2 on her qubit .
each of these four unitary operations map the initial state of the two qubits to a different member of the bell state basis .
clearly , these four orthogonal and therefore fully distinguishable states can be used to encode two bits of information . after encoding her qubit
, alice sends it off to bob who can extract these two bits of information by performing a joint measurement on this qubit and his original qubit .
this apparent doubling of the information conveying capacity of alice s qubit because of prior entanglement with bob s qubit is referred to as superdense coding ( which we shall henceforth refer to as just dense coding ) .
dense coding has been implemented experimentally with polarization entangled photons @xcite .
some generalizations of the scheme to pairs of entangled n - level systems in non - maximally entangled states ( as opposed to qubits in bell states ) @xcite and to distributed multiparticle entanglement @xcite have also been studied . in the simplest example
, the power of the method stems from the accessibility of all the four possible bell states through local operations done by alice alone on her qubit @xcite .
obviously this accessibility is related to the fact that the qubits shared by alice and bob are in an _ entangled _ state .
so an important question to ask is : _ in what way does the capacity of conveying information through dense coding depend upon the degree of entanglement of the initial shared pair of qubits _ ?
barenco and ekert @xcite and hausladen _ et al .
_ @xcite have investigated this question when the initial shared entangled state is a pure state .
their analysis shows that the amount of information conveyed by the dense coding procedure decreases monotonically from its maximum value ( two bits per qubit ) with the decrease of the magnitude of the shared entanglement .
it becomes one bit per qubit when the entanglement becomes zero .
recently , there has been a lot of work in quantifying the entanglement of systems in mixed states @xcite .
these measures of entanglement have their physical interpretation in the process of entanglement purification @xcite . a natural question to ask
is : _ in what way does the capacity to do dense coding depend on the degree of entanglement as specified by these measures ?
_ answering such a question , in effect , would mean _ linking up the apparently disconnected concepts of purification and channel capacity_. in this paper we take a step in this direction by obtaining bounds on the number of bits of information conveyed per qubit ( let us call this * c * after capacity ) during dense coding in terms of the various measures of entanglement
. it should be noted that * c * is a classical capacity as it quantifies number of bits of information , but the carriers of the information are quantum ( qubits ) .
we also invert our results so that the value of * c * gives us information about the range within which the initial shared entanglement lies . in other words ,
bounds on entanglement measures ( and hence on purification procedures ) can be expressed in terms of the classical capacity * c*. this might be helpful , because , as we shall show , the classical capacity for dense quantum coding is readily calculable for certain special classes of dense coding protocols .
suppose , alice has a set of mixed quantum states @xmath3 at her disposal to convey some classical information ( sequences of ones and zeros ) to bob .
each state @xmath3 can be regarded as a separate letter .
also suppose that alice sends the state @xmath3 to bob with an _ a priori _
probability @xmath4 .
the ensemble that alice uses to communicate with bob is therefore given by @xmath5 in the above case , the average number of bits of information that alice can convey to bob per transmission of a letter state is bounded from above by the holevo function @xcite @xmath6 where @xmath7 denotes the von neumann entropy of the state @xmath8 ( here , and throughout the rest of the paper , @xmath9 stands for @xmath10 ) . using a different notation ,
the holevo function can be rewritten as @xmath11 where @xmath12 stands for @xmath13 and is known as the quantum relative entropy between the states @xmath14 and @xmath8 @xcite .
it was recently shown that this bound can be achieved in the limit of an infinite ensemble by appropriate block coding ( grouping together and pruning long strings of letter states to represent messages ) at alice s end and appropriate measurements at bob s end @xcite .
now consider the case of dense coding .
alice and bob initially share an entangled pair of qubits in some state @xmath15 , which may be mixed .
alice then performs local unitary operations on her qubit to put this shared pair of qubits in either of the states @xmath16 or @xmath17 . in general
, alice may use a completely arbitrary set of unitary operations to generate these states : @xmath18 in the above equation , @xmath19 acts on alice s qubit and @xmath0 acts on bob s qubit . by sending her
encoded qubit to bob , alice is essentially communicating with bob using the states @xmath16 and @xmath17 as separate letters . the number of bits she can communicate to bob using this procedure is thus bounded by the holevo function @xmath20 given in eq.([holv ] ) .
moreover , if some block coding is done on a large enough collection of qubits in addition to the dense coding , then the number of bits of information communicated is equal to the holevo function .
we will thus take @xmath21 assuming that any additional necessary block coding will automatically be done to supplement the dense coding .
exactly the same assumption has been used in ref.@xcite to calculate the capacity for dense coding in the case of pure letter states .
eqs.([cg ] ) and ( [ see ] ) define the most general version of dense coding and we shall refer to this as completely general dense coding ( cgcd ) .
a simpler example of dense coding is the case when the letter states are generated from the initial shared state @xmath15 by @xmath22 in the above set of equations , the first operator of the combination @xmath23 acts on alice s qubit and the second operator acts on bob s qubit .
we shall refer to this case ( i.e when the letter states are generated by eqs.([w0])-([w1 ] ) ) as simply general dense coding ( gdc ) .
the generality present in gdc is that alice is allowed to prepare the different letter states with unequal probabilities . in other words
, one has to use eqs.([avg])-([see ] ) to estimate the capacity * c*. in the more special case when alice not only generates the four letter states according to eqs.([w0])-([w1 ] ) ) but also with equal probability , the ensemble is given by @xmath24 and the capacity becomes @xmath25 we shall call this simplest case special dense coding ( sdc ) . among all the possible ways of doing gdc , sdc is the optimal way to communicate when @xmath15 is a pure state ( as we shall show in the next section ) or a bell diagonal state .
however , we do not know the optimal way to communicate when @xmath15 is a completely general state and cgcd is allowed . for most of our paper
, we shall obtain bounds on the classical capacity * c * for sdc only .
but we shall point out those results which are valid for gdc and cgdc as well .
though the main aim of this paper is to establish bounds on the capacity @xmath26 when the letters are mixed states , we shall begin with a calculation of @xmath26 for pure letter states .
consider the initial shared pure state @xmath15 to be , @xmath27)-([w1 ] ) , the other letter states are given by @xmath28 from which we obtain @xmath29 . as all @xmath3 are pure states we have @xmath30 thus from eqs.([holv1 ] ) and ( [ see ] ) we have @xmath31 we will consider only the case of sdc .
thus the ensemble used is obtained from eq.([avg1 ] ) to be @xmath32 thus from eq.([sw ] ) for the capacity * c * , we get @xmath33 now we should recall that a good measure of entaglement for a pure state of a system composed of two subsystems a and b is given by the von neumann entropy of the state of either of the subsystems @xcite . let us call this measure the von neumann entropy of entanglement and label it by @xmath34 .
thus @xmath35 where @xmath36 stands for partial trace over states of system a. therefore , for all the states @xmath3 , @xmath37 thus , @xmath38 we now prove that for pure states , sdc ( using all alphabet states with equal _ a priori _ probability ) is the optimal way to communicate among all possible ways of doing gdc ( i.e when the letter states are generated by eqs.([w0])-([w1 ] ) ) . consider the general case when the states @xmath3 are sent with probabilities @xmath4 . then from eq.([avg ] ) we have @xmath39 where @xmath40 and @xmath41 therefore , @xmath42 and @xmath43 form separate blocks inside the matrix @xmath44 and @xmath45 eq.([sw ] ) indicates that we need to choose the probabilities @xmath4 in such a way that @xmath46 is maximized .
density matrices with same diagonal elements have the highest von neumann entropy when the nondiagonal elements are zero .
applying this fact to eqs.([ro1 ] ) and ( [ ro2 ] ) we get @xmath47 using eqs.([nond1 ] ) and ( [ nond2 ] ) in the expressions for @xmath42 and @xmath43 and calculating the entropy @xmath46 gives @xmath48 where the normalizations of the state amplitudes and the probabilities @xmath4 have been used . from analysis of von neumann entropies it is well known that the expression @xmath49 has a maximum value when both @xmath50 and @xmath51 are equal .
thus @xmath46 , and hence the classical capacity * c * is maximized when @xmath52 thus , among all the possible ways of performing gdc , sdc is the optimal way to communicate when pure states are being used as letters . from the above result and eq.([pure1 ] ) we can conclude that in the case of gdc with pure letter states we have @xmath53 this result ( eq.([pure2 ] ) ) had also been obtained in ref.@xcite from logical arguments . following a procedure analogous to the above proof
it can be shown that for bell diagonal letter states sdc is again the optimal way to communicate among all the possible ways of doing gdc .
in order to derive a lower bound on the classical capacity * c * for sdc we need to prove a crucial lemma concerning the nature of the ensemble used in sdc : _ for states @xmath3 defined in accordance to eqs.([w0])-([w1 ] ) , the ensemble @xmath44 for sdc as defined in eq.([avg1 ] ) is a disentangled state irrespective of the nature of @xmath15_. to prove this , we have to start by assuming @xmath15 to be a most general state of two qubits .
this is given by @xcite @xmath54,\end{aligned}\ ] ] where the indeces @xmath55 and @xmath56 take on values from @xmath57 to @xmath58 . substituting @xmath15 from eq.([wo1 ] ) into eqs.([w0])-([w1 ] ) we get @xmath59 , \\
w_2 & = & \frac{1}{4}[i\otimes i + r_2 \sigma_2 \otimes i- \sum_{m \neq 2 } r_m \sigma_m \otimes i + i\otimes \sum_m s_m \sigma_m \nonumber \\&+ & \sum_{n } t_{2n } \sigma_2 \otimes
\sigma_n -\sum_{m\neq 2,n } t_{mn } \sigma_m \otimes \sigma_n ] , \\
w_3 & = & \frac{1}{4}[i\otimes i + r_3 \sigma_3 \otimes i- \sum_{m \neq 3 } r_m \sigma_m \otimes i + i\otimes \sum_m s_m \sigma_m \nonumber \\&+&\sum_{n } t_{3n } \sigma_3 \otimes \sigma_n - \sum_{m\neq 3,n } t_{mn } \sigma_m \otimes \sigma_n ] .
\label{w12}\end{aligned}\ ] ] using expressions for @xmath3 from eqs.([wo1])-([w12 ] ) in eq.([avg1 ] ) we get @xmath60.\ ] ] this is very clearly a disentangled state .
a plausible physical argument to support this result can be as follows .
we know that an equal mixture of the four bell states is a disentangled state .
so it seems highly likely that an equal mixture of less entangled states will be disentangled as well .
this result is , of course , valid only for sdc , where the equal probabilities of the letter states result in the careful cancellation of all the entanglement carrying terms .
it allows us to easily derive a lower bound on * c * for sdc in terms of one of the measures of entanglement .
to investigate quantitatively the relationship between entanglement and * c * for arbitrary mixed letter states , it is necessary to use some measure of entanglement for mixed states .
one such measure of entanglement , the relative entropy of entanglement , has been introduced in ref.@xcite .
it has been shown to have a statistical interpretation @xcite as well as a physical interpretation @xcite in forming bounds on the process of entanglement purification @xcite .
we will use the symbol @xmath61 to represent this measure of entanglement . for an arbitrary mixed state @xmath14 it is given by @xmath62 where @xmath63 is the set of disentangled states . a property which has to be satisfied by any legitimate entanglement measure
is its invariance under local unitary operations @xcite .
@xmath64 is thus invariant under local unitary operations of the state @xmath14 .
as the state @xmath44 of the ensemble used for sdc has been shown to be a disentangled state @xmath65 in the previous section , we have @xmath66 using the above inequality in eq.([c ] ) for * c * of sdc we get @xmath67 as each of the states @xmath3 are derived from @xmath15 via local unitary operations only , we have @xmath68 for all values of @xmath69 . combining eqs.([eq ] ) and ( [ zeq ] ) we get @xmath70 this means that the classical capacity * c * for sdc is bounded from below by the relative entropy of entanglement of the initial shared mixed state .
this result is however , not generalizable to gdc as it relies crucially on @xmath44 for sdc being a disentangled state .
in this section , we are going to divert a little bit from the main theme linking * c * to entanglement measures and point out another interesting bound on * c * stemming from the mutual distinguishability of the letter states @xmath3 .
it is known that the relative entropy @xmath71 is a kind of statistical measure of the distinguishability between the quantum states @xmath3 and @xmath72 @xcite .
thus for gdc one can define an average mutual distinguishability function as @xmath73 we show below that this average mutual distinguishability function forms an upper bound on * c*. putting @xmath44 from eq.([avg ] ) into eq.([see ] ) we obtain @xmath74 where the factor @xmath75 has been inserted before @xmath3 in the second step as its value equals unity .
we now use the joint convexity property of the relative entropy @xmath76 to expand the right hand side of eq.([c1 ] ) and obtain @xmath77 note that the above result is completely general because neither does it require the special class of local unitary operations given by eqs.([w0])-([w1 ] ) , nor does it require the probabilities to be uniform . thus the classical capacity for cgdc
is bounded from above by the average mutual distinguishability function .
note that we are pointing this out just as an interesting bound and it is not linked to the main theme of the paper ( relating entanglement measures stemming from purification procedures to dense coding capacities ) .
in this section we find an upper bound on * c * in terms of yet another measure of entanglement , namely , the entanglement of formation @xmath78 . consider a decomposition of an arbitrary mixed state @xmath14 in terms of pure states @xmath79 : @xmath80 then the entanglement of formation of this state is defined as @xcite @xmath81 where the minimum is taken over all decompositions of @xmath14 of the type given by eq.([dec ] ) .
the initial shared state @xmath15 used in dense coding , will , in general , have several decompositions in terms of pure states .
let the particular decomposition from which its entanglement of formation @xmath82 is calculated ( referred to as the entanglement minimizing decomposition @xcite ) be @xmath83 where @xmath84 are pure states and @xmath85 are probabilities . as eq.([decw ] ) gives the entanglement minimizing decomposition , we have @xmath86 while the normalization of the probabilities imply @xmath87 as each of the signal states @xmath3 are derived from @xmath15 by local unitary operations , they can be decomposed as @xmath88 where each pure state @xmath89 is connected to the pure state @xmath84 by exactly the same local unitary operation as that which connects @xmath3 to @xmath15 . as any legitimate measure of entanglement has to remain invariant under local unitary operations , we have @xmath90 and @xmath91 > from eqs.([im ] ) and ( [ i ] ) we have @xmath92 now , in the case of sdc , the capacity ( from eq.([c ] ) ) is @xmath93 where we have put @xmath94 using joint convexity ( eq.([convex ] ) ) to expand the right hand side of eq.([grhl ] ) , we get @xmath95 now , for each value of @xmath55 , the expression @xmath96 can be regarded as the classical capacity * c * for sdc with the states @xmath89 as the four letter states . as each of the states @xmath89 ,
are pure , eq.([pure1 ] ) implies @xmath97 putting eq.([onee ] ) into eq.([less ] ) we get @xmath98 using eqs.([norm ] ) and ( [ equa ] ) to simplify the right hand side of eq.([qm1 ] ) we get @xmath99 the above bound is valid for gdc as well . to see this one just has to repeat the above proof starting with @xmath100 and replace eq.([ensm ] ) with @xmath101 eq.([less ] ) then gets replaced by @xmath102 now note the fact that @xmath103 is the expression for the classical capacity * c * for cgdc with @xmath89 being the letter states . when the states @xmath89 are generated from @xmath104 according to eqs.([w0])-([w1 ] ) ( i.e. when gdc protocol is being followed ) , then the purity of the states @xmath89 guarantees that * c * is less than @xmath105 ( as shown in section [ purer ] ) .
using this fact in eq.([ii0 ] ) we again end up with eq.([upbnd ] ) .
thus even in the case of gdc , the capacity * c * is bounded by @xmath106 .
having analytically proven that @xmath106 is an upper bound on * c * for gdc , we now proceed to check whether the even smaller ( as proved in ref.@xcite ) quantity @xmath107 is also an upper bound . however , we do not attempt to prove this analytically for a completely general initial shared state @xmath15 .
instead we calculate the capacity * c * ( for sdc only ) for those specific classes of the initial shared state @xmath15 whose relative entropy of entanglement @xmath108 is already known .
we then plot this capacity * c * as a function of the relative entropy of entanglement @xmath61 for each of these classes of states and check whether the curve @xmath109 lies below the plot of @xmath110 .
at first we have a look at mixed states of the type @xcite @xmath111 where @xmath112 is one of the four bell states which are defined by @xmath113 ) as lambda states of the type a. for them , the relative entropy of entanglement is @xcite @xmath114 while the * c * for sdc is @xmath115 the plot of the classical capacity * c * for these states as a function of @xmath61 of the state has been shown in figure [ lam1 ] and it is indeed found that @xmath116 the equality holds true only at the two ends of the graph , namely at maximal entanglement @xmath117 ( when @xmath118 approaches a bell state ) and zero entanglement @xmath119 ( when @xmath118 approaches @xmath120 ) .
next we look at states of the type @xcite @xmath121 which we call lambda states of the type b. for these @xcite @xmath122 where @xmath123 are the eigenvalues of @xmath124 .
the * c * for sdc in this case is given by @xmath125 as is clear from a simple comparison of eqs.([b1 ] ) and ( [ b2 ] ) , for lambda states of the type b , @xmath126 thus , a curious feature of lambda states of type b is that the capacity for sdc is actually always _ equal _ to @xmath110 .
we will discuss a bit more about this curious aspect later in the section .
now we consider another special class of states called the werner states @xcite parameterized by a number f called the fidelity @xcite and given by @xmath127 the relative entropy of entanglement of these states is given by @xcite @xmath128 and the * c * for sdc is calculated to be @xmath129 the plot * c * versus @xmath61 has been drawn in fig.[wer ] and it is found that even in this case , @xmath116 now consider the case of a general bell diagonal state @xmath130 it has been proved in ref.@xcite that when all @xmath131 $ ] then @xmath132 while when any of the @xmath133 ( say @xmath134 ) @xmath135 , then @xmath136 the * c * for sdc always turns out to be @xmath137 first consider the case when all @xmath131 $ ] . from eqs.([zero ] ) and ( [ ccc ] ) we find that @xmath138 where @xmath139 has been used to proceed from the first to the second step and @xmath140 ( because @xmath131 $ ] ) has been used to proceed from the second to the third step . now consider the complementary case ( @xmath141 ) . from eqs.([fint ] ) and ( [ ccc ] ) we find that @xmath142 where in order to proceed from the third to the fourth step we have used the simple fact that @xmath143 is greater than either of the terms @xmath144,@xmath145 or @xmath146 .
thus for all bell diagonal states we have @xmath116 now we will point out a curious fact about the situation when any two of the eigenvalues of a bell diagonal state ( say @xmath147 and @xmath148 ) are zero .
when @xmath149 ( which means @xmath119 ) , we have * c*@xmath150 ( using eq.([ccc ] ) ) . in all the other cases ( for which we use eqs.([fint ] ) and ( [ ccc ] ) ) we have @xmath151 where we have used the simple fact that @xmath152 .
thus for all bell diagonal states with only two nonzero eigenvalues we have @xmath153 on the basis of the results obtained in this section let us conjecture : * conjecture * : _ the * * for sdc with completely general ( possibly mixed ) states is bounded from above by @xmath110_. a great deal of empirical evidence has been presented in this section in support of the conjecture . in the next section
we proceed to give a heuristic justification in support of our conjecture .
in fact , we will try to justify an even stronger upper bound on * c*.
to justify the conjecture of the previous section , we will have to examine the following interesting question : how does the capacity * c * change if alice and bob first locally purify their ensemble and distill bell states ( following the optimal purification procedure ) and follow this up by sdc ?
various purification procedures have been described in refs.@xcite .
here we assume that alice and bob follow the optimal purification process : one which helps them to distill the maximum fraction of bell states from the initial ensemble .
they will , after optimal purification , have a fraction @xmath154 ( where @xmath154 is called the _ entanglement of distillation _ ) shared pairs in bell states and a fraction @xmath155 pairs in a disentangled state .
they now complete their purification process by converting the final subensemble of disentangled pairs to pure states by projective measurements .
we refer to such a protocol as _
complete purification_. after a complete purification , alice can use the fraction @xmath154 of bell pairs to send bob information at the rate of 2 bits / pair and the fraction @xmath155 of pure disentangled pairs to send information at the rate of 1 bit / pair . thus if alice and bob initially shared @xmath56 pairs , the classical capacity after a complete and optimal purification procedure is @xmath156 note that the above result is only asymptotically true ( @xmath157 ) .
naively , one might expect this * c * to be lower than the * c * before purification .
this is because , as mentioned in ref.@xcite , entanglement concentration is a more destructive process than quantum data compression .
some amount of shared entanglement is destroyed in the process of purification .
so it might be expected that the final ensemble after a purification will be able to convey less classical information than the original ensemble . on the contrary , as we will justify , the capacity for sdc with the purified ensemble is greater than that with the unpurified ensemble when an optimal and complete purification protocol is used .
the gain comes from the fact that now there are two separate sub - ensembles instead of a single ensemble . in this section
, we will try to justify , albeit heuristically , the following statement ( as a more fundamental conjecture than the one presented in the previous section ) : + * conjecture * : _ the capacity for sdc is more when it is preceeded by a complete and optimal purification procedure_. we will first consider the special case of bell diagonal mixed states , for which the proof of the increase in the sdc capacity on a complete and optimal purification ( our conjecture ) can be rigorously proved . for bell diagonal states with entropy @xmath158
, there is a purification protocol called hashing ( with the distillable fraction being @xmath159 ) @xcite .
> from eq.([ccc ] ) we see that for these bell diagonal states , the * c * for sdc is equal to @xmath160 . as hashing may not necessarily be the optimal protocol , we have @xmath161 .
this immediately implies @xmath162 . for the complementary case of bell diagonal states with @xmath163
, we have ( from eq.([ccc ] ) for sdc ) , @xmath164 .
this is obviously less than @xmath165 for any finite value of the entanglement of distillation ( i.e for all inseperable states @xcite ) .
thus for all bell diagonal states the * c * for sdc can be improved by a prior optimal and complete purification of the ensemble of shared states . for the more general case of arbitrary mixed entangled states , the proof will , essentially , be heuristic
. our approach will be to split the change in the classical information capacity due to complete and optimal purification into two parts .
the first part is positive ( an increase in capacity ) and due to the addition of classical side channels during the purification procedure .
these side channels are used by alice ( a ) to communicate the results of her measurements to bob ( b ) , or vice versa .
as this communication during purification is already conveying information from a to b ( or vice versa ) , the channel capacity of the classical side channels should be directly added to the classical capacity ( of course , for this , we have to implicitly assume the additivity of classical capacities ) .
the information conveyed from a to b ( or vice versa ) during purification is actually used by the two parties to precisely identify their entangled and disentangled subensembles .
there is also a negative contribution ( decrease in capacity ) during the purification as a fraction @xmath166 of shared pairs loose all their entanglement .
of course a part of this entanglement is pumped into the entangled subensemble , but the remaining is lost .
our job will be to argue that the positive contribution in channel capacity due to addition of classical side channels outweighs the negative contribution due to loss of entanglement in an optimal and complete purification procedure .
consider the process of purification as a process of information gain by a and b. before the purification , both a and b have an equal amount of knowledge about their shared system ( they both know the full density matrix ) .
finally , each of their shared states are pure ( because of the completeness of the purification procedure ) and both of them have equal knowledge ( i.e each know which of the pairs form the disentangled subensemble and which of the pairs form the maximally entangled subensemble ) .
thus they have both gained an equal amount of information @xmath167/pair about their shared pairs during the process of purification .
now , acquiring a part of this information may not cause any destruction of shared entanglement ( lets call it @xmath168/pair ) , while the remaining part ( lets call it @xmath169/pair ) does cause a lowering of shared entanglement .
we now contend , that this part , @xmath169/pair has to be equal to the number of classical communication channels used in the optimal purification procedure .
the logic follows from the fact that the optimal strategy would be for a and b to collaborate in such a way that both of them gain the entire information @xmath169/pair with the least destruction of entanglement .
they must then each acquire only a fraction of the information complementary to the fraction acquired by the other . in this way ,
the total information acquired by a and b from the shared pairs through direct entanglement degrading measurements is @xmath169/pair .
they can then use a minimum of @xmath169 classical side channels per pair to communicate their fraction of the acquired information to the other party . on the other hand , if each wanted to acquire a fraction of the total information by direct measurements which was not entirely complementary to the part acquired by the other , they would destroy more entanglement than is really necessary .
thus we would expect that the classical side channels contribute to boosting the capacity up by at least an amount @xmath169/pair in the optimal purification procedure .
now consider how much the capacity decreases due to the degradation of shared entanglement during the purification procedure .
the classical capacity of a fraction @xmath166 of the pairs drops from _ at most _ 2 bits / pair to 1 bit / pair ( due to the _ completeness _ of the purification procedure , the final capacity can not be lower than 1 bit / pair ) . thus the drop in classical capacity of each of the finally disentangled pairs
is not more than 1 bit .
this implies that the net decrease in classical capacity of the entire ensemble due to loss of entanglement hasnt been more than @xmath166 bits / pair .
when the information @xmath169 is greater than 1 bit / pair , the proof of our statement is straightforward . the increase in capacity due to classical side channels ( i.e @xmath169/pair ) is more than 1 bit / pair , while the decrease in capacity due to loss of entanglement is less than @xmath166 bits / pair .
as @xmath166 is a fraction ( i.e @xmath170 ) , the increase in capacity on purification overrides the inevitable decrease in capacity due to degradation of shared entanglement .
clearly , thus the capacity to do sdc increases on prior purification when @xmath171 bit .
now , we have to show that no matter what the value of @xmath169 is , it is _ possible _ to find a purification protocol for which @xmath169 is greater than @xmath166 bits .
in other words , if initially there are @xmath56 shared pairs ( with @xmath56 being large ) , then the number of shared pairs which loose all their entanglement on gaining an information of @xmath169/pair can be made less than @xmath172 by an appropriate choice of the purification protocol . from such a statement
it directly follows that the increase in capacity on purification overrides the inevitable decrease in capacity due to degradation of shared entanglement .
a simple way in which one may try to justify the above proposition is from the fact that one pair generally allows the extraction of upto two bits of information ( one bit from each of a s and b s qubit ) .
so , it should be possible , in general , to extract @xmath172 amount of information from measurements on less than @xmath172 shared pairs ( whose entanglement is degraded ) .
this intutive understanding leaves open the question as to whether the information @xmath169/pair _ relevant _ to purification can be acquired in this way . for
that we resort to explicit exemplification .
consider the class of mixed states @xmath8 ( having nonzero entropy @xmath173 ) and a diagonal decomposition @xmath174 in terms of the pure states @xmath175 .
also suppose that this diagonal decomposition does not coincide with the entanglement minimizing decomposition @xcite of @xmath8 .
for such states , acquiring the information about the entropy ( @xmath173/pair ) is essential during purification , as the final ensemble is pure .
we now contend that acquiring this @xmath173 /pair of information _ necessarily _ destroys some amount of shared entanglement .
we will justify this by the method of contradiction .
suppose , it was really possible to acquire the information about the entropy @xmath173 without destroying any entanglement . on acquiring an information of the amount equal to @xmath173/pair , a and b will be able to divide their initial ensemble to four seperate pure subensembles , each being comprised of one of the states @xmath175 .
the weight of each of these subensembles will be equal to the eigenvalues @xmath133 .
the average entanglement shared by a and b after extracting the information @xmath173/pair is thus @xmath176 .
however , for the class of states @xmath8 that we are considering , @xmath176 is necessarily greater than the initial shared entanglement ( as quantified by the entanglement of formation @xmath78 ) . thus by purely local actions and classical communications , a and b have been able to increase the shared entanglement for these classes of mixed states .
this contradicts the definition of entanglement as a quantity that can not be increased by local actions and classical communications .
mixed states with the diagonal decomposition not coincident with the entanglement minimizing decomposition can not be locally purified without necessarily destroying some shared entanglement_. thus the information about the entropy of these states is of the type @xmath169 ( i.e acquiring the information necessarily causes degradation of entanglement ) .
a corollary which immediately follows ( though not directly relevant to the main point of the paper ) , is : _ the optimal purification of mixed states with the diagonal decomposition not coincident with the entanglement minimizing decomposition necessarily has a nonzero @xmath169 and thus necessarily requires the use of classical side channels_. now consider such states when their entropy , and thereby @xmath169 , exceeds @xmath57 bit .
thus the total information @xmath172 needed to be acquired is greater than @xmath56 bits .
however , purifiability ( nonzero @xmath154 ) implies that the total entanglement loss can be entirely concentrated to the entanglement loss of @xmath177 pairs which is less than @xmath172 pairs .
therefore , it is _ possible _ to gain the relevant @xmath172 bits of information by destroying the entanglement of less than @xmath172 pairs .
now we really need an unproven extension of the above statement : _ even when _
@xmath178 bit and _ irrespective of the nature of the decomposition of _
@xmath8 , it is _ possible _ to gain the relevant @xmath172 amount of information by destroying the entanglement of less than @xmath172 pairs .
though this may seem a large extrapolation , it seems highly _ plausible _ because one shared pair allows the extraction of upto two bits of information .
thus we would expect that , no matter what the value of @xmath169 is , the increase in capacity on purification overrides the inevitable decrease in capacity due to degradation of shared entanglement .
thus the capacity to do sdc with the two pure subenembles ( maximally entangled and fully disentangled ) after a complete and optimal purification , is expected to be greater than the initial mixed ( and uniformly entangled ) ensemble .
we thus conjecture , with the help of eq.([cppd ] ) , the following bound on the capacity of doing sdc with any mixed entangled state @xmath179 this is an even stronger upper bound on the classical capacity than the bound conjectured in the previous section as the relative entropy of entanglement @xmath61 is necessarily greater than @xmath154 @xcite . from eq.([bnded ] ) , it directly follows that @xmath180 ( our previous conjecture ) .
note that in this section , this conjecture has been rigourously proved for all bell diagonal states , heuristically proved for all states with entropy @xmath181 bit ( i.e @xmath171 bit ) and shown to be highly plausible for other types of states .
in this paper we have obtained bounds on the capacity of doing dense coding in terms of the different measures of entanglement stemming from purification procedures .
the rigorously proved part of the results of this paper is that for sdc one always has @xmath182 on the other hand if we allow for a conjecture ( well supported by examples and heuristically justified in section.[pff ] ) , we have the stronger result @xmath183 what is the significance of bounds on the * c * for sdc when it can be readily calculated ?
the importance can be realized when we invert the above equation and write @xmath184 thus by calculating ( or measuring ) the channel capacity for sdc , we can draw an inference about the range in which the entanglement of the shared states ( as quantified by @xmath61 ) lies .
also one of the bounds , namely , @xmath185 continues to hold even when we are allowed to vary the a priori probability of the various signal states ( gdc ) .
thus , our inequality allows one to impose a readily calculable ( from the expression for @xmath78 in ref.@xcite ) limit on the capacity for gdc , without having to optimize gdc over all possible a priori signal probabilities .
the physical interpretation of the upper bounds becomes clear from the considerations given in section.[pff ] . during optimal and complete entanglement purification procedures one
generally enhances the classical capacity much more due to added classical side channels than the inevitable decrease in capacity brought by the loss of entanglement .
the information transferred through these classical side channels during purification helps to identify the entangled and disentangled subensembles .
it is through this identification that they directly play the role of boosting the capacity . as after an optimal and complete entanglement purification procedure ,
the capacity is @xmath165 , this post - purification capacity is an upper bound on the pre - purification capacity . as both the entanglement measures @xmath78 and @xmath61 are upper bounds on @xmath154 , all our upper bounds follow immediately .
it is interesting to note that the upper and the lower bounds are trivial in complementary regimes . when @xmath186 , the upper bounds are trivial , but the lower bound ( @xmath187 ) is non trivial . on the other hand when @xmath188 , the upper bounds are non trivial while the lower bound is trivial .
we believe that this paper imparts physical significance to the various measures of entanglement from the viewpoint of forming bounds on certain kinds of dense coding procedures .
this direction of physical interpretation of the entanglement measures is very different from the standard interpretations which stem from entanglement dilution and distillation processes . in a sense
, this paper _ links up the apparently disconnected notions of entanglement purification and dense coding_. the considerations of the previous section implies the following lower bound on the distillable entanglement : @xmath189 .
this is important because as yet there is no explicit formula for @xmath154 for a general mixed entangled state .
thus the readily calculable quantity @xmath190 ( in the case of sdc ) may offer a convinient lower bound on the distillable entanglement of a given mixed state .
the interesting question to examine is what apart from entanglement is involved in determining the dense coding capacity .
this may be some measure of distinguishability between the letter states ( hinted by the fact that the average mutual distinguishability function forms an upper bound on the classical capacity * c * ) .
an aim of further research should be to work towards a complete formula for the capacity of dense coding . moreover , working with a similar motivation as this paper
, one can also try to relate entanglement measures stemming from purification procedures to the other uses of shared entanglement such as teleportation @xcite and secret key distribution @xcite .
we would like to thank vladimir buzek , daniel jonathan and peter knight for valuable discussions .
this work is supported by the inlaks foundation , elsag - bailey , hewlett - packard , the european tmr networks erb 4061pl951412 and erbfmrxct96066 , the uk engineering and physical sciences research council , the european science foundation and the leverhume trust .
99 for a review see : plenio , m. b. , and vedral , v. , 1998 , _ cont .
_ , * 39 * , 431 .
bennett , c. h. , and wiesner , s. , 1992 , _ phys .
_ , * 69 * , 2881 .
mattle , k. , weinfurter , h. , kwiat , p. g. , and zeilinger , a. , 1996 , _ phys .
_ , * 76 * , 4656 .
barenco a. , and ekert , a. k. , 1995 , _ j. mod . opt .
_ , * 42 * , 1253 .
hausladen , p. , jozsa , r. , schumacher , b. , westmoreland , m. , and wootters , w. k. , 1996 , _ phys . rev .
a _ , * 54 * , 1869 .
bose , s. , vedral , v. , and knight , p. l. , 1998 , _ phys .
a _ , * 57 * , 822 .
bennett , c. h. , divincenzo , d. p. , smolin , j. a. , and wootters , w. k. , 1996 , _ phys .
a _ , * 54 * , 3824 .
vedral , v. , plenio , m. b. , rippin , m. a. , and knight , p. l. , 1997 , _ phys .
_ , * 78 * , 574
. vedral , v. , plenio , m. b. , jacobs , k. , and knight , p. l. , 1997 , _ phys .
a _ * 56 * , 4452 .
vedral , v. , plenio , m. b. , 1998 , _ phys .
_ , a * 57 * , 1619 .
bennett , c. h. , brassard , g. , popescu , s. , schumacher , b. , smolin , j. a. , and wootters , w. k. , 1996 , _ phys .
_ , * 76 * , 722 .
deutsch , d. , ekert , a. , jozsa , r. , macchiavello , c. , popescu , s. , and sanpera , a. , 1996 , _ phys .
_ , * 77 * , 2818 .
bennett , c. h. , bernstein , h. j. , popescu , s. , and schumacher , b. , 1996 , _ phys .
a _ , * 53 * , 2046 .
kholevo , a. s. , 1973 , _ probl .
peredachi inf _ , * 9 * , 177 [ kholevo , a. s. , 1973 , _ problems of information transmission _ , * 9 * , 177 ] ; the holevo bound was first conjectured by : gordon , j. p. , 1964
, quantum electronics and coherent light , _ proceedings of the international school of physics `` enrico fermi , '' course xxxi _ , edited by p. a. miles
( academic , new york ) , 156 - 181 ; levitin , l. b. , 1964 , information , complexity and control in quantum physics , edited by p. a. miles , ( academic , new york ) , 111 - 115 .
ohya , m. , 1989 , _ rep . math .
_ , * 27 * , 19 ; hiai , f. , and petz , d. , 1991 , _ comm .
_ , * 143 * , 99 ; donald , m. j. , 1986 , _ comm . math .
_ , * 105 * , 13 ; donald , m. j. , 1987 , _ math .
_ , * 101 * , 363 .
popescu , s. , and rohrlich , d. , 1997 , _ phys .
a _ , * 56 * , r3319 .
horodecki , r. , horodecki , p. , and horodecki , m. , 1995 , _ phys .
a _ , * 200 * , 340 . rains , e. m. , entanglement purification via separable superoperators , 1997 , lanl e - print quant - ph/9707002 .
wootters , w. k. , 1998 , _ phys .
_ , * 80 * , 2245 .
werner , r. f. , 1989 , _ phys .
a _ , * 40 * , 4277 .
horodecki , m. , horodecki , p. , and horodecki , r. , 1997 , _ phys .
_ , * 78 * , 574 .
bennett , c. h. , brassard , g. , crepeau , c. , jozsa , r. , peres , a. , and wootters , w. k. , 1993 , _ phys .
_ , * 70 * , 1895 .
ekert , a. k. , 1991 , _ phys .
_ , * 67 * , 661 . | ideal dense coding protocols allow one to use prior maximal entanglement to send two bits of classical information by the physical transfer of a single encoded qubit .
we investigate the case when the prior entanglement is not maximal and the initial state of the entangled pair of qubits being used for the dense coding is a mixed state .
we find upper and lower bounds on the capability to do dense coding in terms of the various measures of entanglement .
our results can also be reinterpreted as giving bounds on purification procedures in terms of dense coding capacities . |
over the years there have been various mathematical incarnations of the operations of string theory .
these come in many different flavors as we recall in 2 . here
we will view moving strings as naturally defining surfaces with partially measured foliations . on this geometric / topological level
the string interactions are described by operadic structures .
these are the mathematical structures that describe gluing operations where both the surfaces and the foliations are glued .
the result which is joint with bob penner is a combinatorially defined open / closed cft with partial compactification @xcite .
this approach is very powerful as it gives the right phenomenology .
first on the degree zero level , that is restricting to connected components , one obtains a new proof of the axioms of open / closed tft .
secondly we obtain the correct bv formalism in the closed sector and interesting new behavior in the open / closed sector . passing to the homology or operator level
, we were able to give an action on the hochschild co chains of a frobenius algebra which had been expected from string topology , @xmath0branes and purely on mathematical grounds @xcite .
finally , we give applications in the semi simple case by defining a quantum loop space and considering stabilizations of the moduli spaces .
it is a pleasure to thank our collaborator bob penner .
we also thank the organizers of the isqs xvi for providing this wonderful forum .
the basic picture we have in mind is that strings are either parameterized circles or intervals . as such they have a measure sitting on them and special points , namely @xmath1 in the closed case or the endpoints of the interval in the open case .
now as the strings move , split and recombine they sweep out a surface . depending on our point of view
, we could stop here and we would be considering tft .
we could also include more data as discussed below .
the second fundamental aspect of this _ a posteriori _ actually not
so nave picture is that we can let the string system run for a bit , then stop and then let it run again .
the final result will be a surface with has been glued from the surface swept out during the first time period and the surface swept out during the second .
this is what gives rise to the operadic structure . in order to be more precise ,
we need to fix the geometry we are talking about . in the table
below , we fix the data we are gluing which will always be surfaces together with extra data and specify the theory we are encoding mathematically . in order to simplify things we will for the moment restrict to the closed sector .
this means there will be only one distinguished point per boundary component .
[ cols="<,<,<",options="header " , ] we will be concerned with the last entry . in order do the open / closed version we will have to add more points on the boundary .
these points correspond to the ends of open strings and are hence labelled by a set of @xmath0brane labels chosen from a set of elementary labels @xmath2 . to give the gluing structure we will pass to the power set @xmath3 as the labelling set for the points .
the label @xmath4 will mean `` closed string '' .
if a point has a collection of labels , we think of this as the intersections of relevant branes . it is allowed that these can all be empty in a realization .
adopting the point of view that the strings have measures , we see that the surface they sweep out actually has a foliation , which has a transversal measure , see figure 1 .
we can think of each point of the string as giving a leaf of this foliation .
the strings themselves will also give a foliation transversal to the previous one .
each string has a measure and hence the first foliation has a transversal measure .
going on , we can `` squeeze '' the leaves together like a curtain to fit into bands on the surface .
these bands will now end on an interval between two marked points called windows .
as the bands also come with a transversal measure , they have a width .
thus the whole data can be thought of as combinatorially given by a graph on the surface one edge for each band whose edges are labelled by positive real numbers that indicate the width .
we will enlarge the picture by allowing the graphs to degenerate .
this means that the underlying ribbon graph can have a different topological type from the surface it lies on , see @xcite for details and examples .
we call a foliation admissible if its graph as a graph on the surface has no parallel ( homotopic ) edges and none of the edges are parallel to a window , where in the homotopy the endpoints are not allowed to pass each other or the marked points .
we also exclude the cases in which an edge is parallel to a part of the boundary including a point marked with @xmath4 .
finally , we require that there is at least one edge .
simply looking at the string interaction , we stay on the level spaces of foliations of surfaces . here
the string interaction is captured by gluing the surfaces along the interacting strings / boundary components as depicted in figure 1 .
the point of view of @xcite is that we only glue if the widths are the same .
this corresponds to the dynamic picture mentioned above .
the important thing to notice is that in the open gluing , the marked points become punctures , while in the closed gluing they are erased after the gluing .
it is this fundamental difference which leads to the correct algebra of interactions .
we can go one step further and consider parameterized families of these operations .
what this basically means is that we consider varying weights on the bands . here
the topology of the space is given by varying the weights and erasing a band as its width tends to zero . just as the gluing space level operations above there are family gluing operations as well . with a lot of work , @xcite we could show that there are operations induced on the homology level as well .
what this means is that we obtain vector spaces of operators .
these are graded by their degree which is the number of parameters .
these operators then define a version of open / closed cft , see below .
the lowest degree @xmath1 corresponds to the connected components and hence to tft .
using this description gives a new proof for minimal axioms of open / closed tft @xcite .
we will call @xmath5 the space of classes of admissible foliations under orientation - preserving homeomorphisms of brane labelled surfaces with @xmath6 active closed window and @xmath7 active open windows . here
we call a window `` open ''
if both endpoints ( which may coincide ) are labelled by a nonempty @xmath0brane label and a window `` closed '' if it is on a boundary component with one marked point that is marked by @xmath4 .
being active means that the window is hit by leaves of the foliation .
theorem the spaces @xmath5 of surfaces with partially measured foliations form an @xmath8 graded c / o structure .
this is a new type of operad like structure defined in @xcite .
being a c / o structure roughly means that we can associatively glue a closed window to a closed one or an open window to an open one under the condition that the weights agree . here
gluings of surfaces are allowed , that is the two windows that are glued can either lie on two different surfaces or on the same surface .
one of the really surprising results which requires hard work is : theorem the operations on the space level induce the structure of a bi modular bi
operad on the homology level .
this is the gold standard for mathematical descriptions of cft and we were not able to show this using the previous approach of @xcite where we allowed more gluings on the space level using an overall projective scaling of the weights .
this in turn however prevented us from defining self gluing .
of course on the homology level the operations of @xcite and @xcite agree .
another important result is that we can even lift the operations above to the chain level , which is important for applications to string topology .
this fact is actually needed in the proof of the theorem above which consists of a careful analysis of flows of foliations on the respective surfaces .
it should be remarked that on the homology level there is no condition on weights anymore .
the homology is linear and hence we can look for ( operadic ) algebra representations of it .
i.e. each element of homology defines an operator in such a way that the relations are preserved under gluing .
this means roughly that gluing two families representing homology classes corresponds to concatenating the relevant operators .
notice that homology is graded and these degrees should be preserved .
a degree @xmath9 class can be represented by a @xmath9dimensional family and this should correspond to an operator of degree @xmath9 .
the degree @xmath1 classes correspond to points up to homotopy and are therefore represented by the connected components .
disregarding the data of the parameterized families of graphs , each surface with marked points on the boundary and a foliation can be decomposed into bi - gons , triangles , annuli with one point on each of the boundaries , once punctured annuli of the same type , discs with two punctures and one marked point on the boundary and pairs of pants with one marked point on each boundary .
the surfaces without punctures are exactly the surfaces of figure 2 .
moreover the indecomposable families of degree @xmath1 and @xmath10 without punctures are exactly the ones depicted in figure 2 .
different decompositions of surfaces with families of foliations give rise to relations .
different pieces can glue together to form the same surface and therefore the composition of the respective operators should agree . in this way one obtains all the expected relations in degree @xmath1 .
the relations of degree zero involving the open sector and the open / closed interaction are for example depicted in figure 3 . for the standard frobenius relations on the closed sector and
the relations of degree @xmath10 which establish that @xmath11 is a bv operator that is compatible with a natural gerstenhaber bracket we refer the reader to @xcite .
theorem an algebra over the modular bi - operad @xmath12 is a pair of vector spaces @xmath13 with the following properties : @xmath14 is a commutative frobenius bv algebra @xmath15 , and @xmath16 is a @xmath17-colored frobenius algebra ( see e.g. , @xcite for the full list of axioms of such a structure ) . in particular , there are multiplications @xmath18 and a non - degenerate metric on @xmath19 which makes each @xmath20 into a frobenius algebra . furthermore , there are morphisms @xmath21 which satisfy the following equations : letting @xmath22 denote the dual of @xmath23 , @xmath24 the morphism permuting two tensor factors , and letting @xmath25 be arbitrary non - empty brane - labels , we have @xmath26 these constitute a spanning set of operators and a complete set of independent relations in degree zero .
all operations of all degrees supported on indecomposable surfaces are generated by the degree zero operators and @xmath11 . here
the intriguing equation is the bv vanishing .
this is a new feature , which has to do with the partial compactification . using the results of transitivity of the four moves on decompositions of constant foliations , see figure 3 , and the theorem above
we also obtain the restriction to an open / closed tft : theorem an algebra over the degree @xmath1 part of the operad , that is an open / closed tft is precisely given by the data @xmath13 which have the following properties : @xmath14 is a commutative frobenius algebra , and @xmath16 is a @xmath17-colored frobenius algebra which satisfies the cardy and center equation .
hence , as the degree @xmath1 part , we recover the cardy / lewellen axiomatic picture ( @xcite , see also @xcite ) from the point of view of strings yielding foliations .
this includes the non commutative frobenius algebras of the open string sector and the easy description of the cardy equation of figure 2 . in this section
, we will restrict ourselves to the closed string sector .
as we stated above , we augmented the admissible foliations , by allowing the graphs to be degenerate that is of a `` smaller '' topological type than the surface . to be precise a graph of a foliation
is called quasi filling if the complementary regions are either polygons or once punctured polygons .
let @xmath27 denote the space of quasi filling graphs of closed sector surfaces of genus @xmath28 with @xmath29 punctures and @xmath6 marked points , that is surfaces whose boundaries each only have one marked point labelled by @xmath4 .
theorem @xcite @xmath27 is homotopy equivalent to the decorated moduli space of surfaces of genus @xmath28 with @xmath29 punctures and @xmath6 marked points .
in particular , if there are no punctures @xmath30 that is it is isomorphic to the moduli space of genus @xmath28 curves with @xmath6 punctures and one tangent vector at each puncture . in other words , we recover moduli space and hence restricting the operad structure to this subspace we obtain a combinatorial version of cft .
this is however a bit subtle , since on the topological level the statement is true only _ cum grano salis_. the mathematically correct statement is theorem using the gluing of @xcite the subspaces @xmath31 form a rational ( i.e. densely defined ) cyclic operad which induces a cyclic operad structure on the relative chain complex of open cells indexed by marked ribbon graphs ( see @xcite for full details ) .
there are three sources of motivation to look for actions of a chain model of moduli space .
perhaps the most intriguing come from using the logic of kontsevich - kapustin - rozansky @xcite , which we can rephrase as follows .
if the closed string states are thought of as deformations of the open string states and the open string states are represented by a category of @xmath0-branes , then the closed strings should be elements of the hochschild co chains of the endomorphism algebra of this category . now thinking on the worldsheet , we can insert closed string states .
that is , for a world sheet , we should get a correlator by inserting , say @xmath6 closed string states .
this is what we will have done , if one simplifies to a space filling @xmath0-brane and twists to a tcft .
the second motivation is from string topology @xcite , where surfaces should act on the homology of the loop space of a manifold .
now it is well established that if the manifold is simply connected , then the homology of the loop space is calculated by the hochschild complex of co - chains of the manifold .
going one step in the spectral sequence and supposing that the manifold is compact and hence has poincar duality , we again expect a cell level action of moduli space .
lastly just looking at deligne s conjecture and its generalizations ( see @xcite for a full list of references and proofs thereof ) we are motivated to include all surfaces .
as expected there is indeed such an action on the hochschild co chains , which can be understood as an action of the discretization of the foliations .
theorem there is an action on the hochschild co chains of a frobenius algebra by the relevant chain complex of ribbon graphs which calculates the co homology of moduli space . restricting to a particular subpart and partially compactifying
, we obtain an action which is the one of string topology of chas and sullivan ( possibly up to lower order terms ) @xcite .
fix a commutative unital frobenius algebra @xmath32 with multiplication @xmath33 and paring @xmath34 .
set @xmath35 and let @xmath36 be the euler element of @xmath32 that is @xmath37 where @xmath11 is the adjoint of @xmath33 . using dualization , we need to define correlators : @xmath38 for any cell given by a surface @xmath39 with boundary and marked points as before and a ribbon graph @xmath40 representing a foliation @xmath41 $ ] with varying weights . here
we think of @xmath42 where @xmath43 is the tensor algebra .
the action is now roughly given as follows ( full details are given in @xcite ) : ( 1 ) duplicate edges so that the number of incoming edges at the vertex @xmath44 .
we sum over all possibilities to do this , if this is not possible then the operation is zero .
( 2 ) assume the @xmath45 are pure tensors .
pull apart the edges and decorate the pieces of the boundary with the elements of @xmath46 .
cut along all the edges of the graph and call the set of disjoint pieces of surface @xmath47 .
let @xmath48 be the index set of the components @xmath49 of the @xmath45 decorating edges belonging to a piece @xmath50 and let @xmath51 be the euler characteristic of the surface @xmath52 .
notice that the pieces @xmath52 which possibly have non trivial topology are a subset of the pieces of the surfaces , that are obtained by cutting along the original edges before duplication .
( 3 ) set @xmath53 where @xmath54
we wish to point out that the topology of the surface pieces @xmath52 enters only through the factor @xmath55 .
this factor is invertible for @xmath56 precisely if the algebra @xmath32 is semi simple .
moreover if @xmath32 is semi - simple with idempotents @xmath57 , set @xmath58 then @xmath59 if all @xmath60 .
we will call such an algebra a normalized semi simple frobenius algebra .
all semi simple frobenius algebras can be obtained from a normalized one by scaling the metric . moreover ,
any semi simple finite dimensional algebra can be endowed with a metric that makes it into a normalized semi
simple frobenius algebra .
for such an algebra the action of a cell @xmath61 equals to the action of the surface where the pieces @xmath52 are replaced by discs .
we call this operation stabilization . here
the boundary pieces of @xmath52 are cut and glued to an @xmath62 in a fixed fashion , which keeps the relative order of the pieces intact .
theorem for a normalized semi
simple frobenius algebra the action factors through the stabilization .
moreover the stabilized surfaces form an operad and contain an @xmath63 suboperad .
hence the hochschild co - chains of a normalized semi
simple frobenius algebra have the structure of an @xmath63 algebra .
the first part of this is immediate . for the second we construct the stabilization via a colimit whose
maps actually add topology to the pieces @xmath52 and explicitly give the @xmath63 operations as the geometrization of the operations of @xcite .
if @xmath32 is not semi simple the action does not pass through the stabilization .
we can however flow the metric to a normalized one which has the effect of changing the co multiplication .
also , in the semi simple case the operations are shifted from the stabilized ones by invertible elements .
it is well known that there is a cyclic model for the chains of loop space of a simply connected manifold @xmath64 given by @xmath65 . calculating the homology we are lead to the @xmath66 term of the spectral sequence which is isomorphic to @xmath67 , with @xmath68 .
now say @xmath64 is a smooth projective variety , we can quantum deform @xmath32 to @xmath69 , where @xmath70 is the quantum cohomology at the point @xmath71 .
this gives us a model for _ quantum chains on the loop space _ : @xmath72 .
as such we have the operations above and in particular the string bracket induced by the chas sullivan prop . now if @xmath73 is semi
simple this is expected generically if @xmath64 is a fano variety that has a system of exceptional sheaves of appropriate length , e.g. @xmath74 then we can actually use gravitational descendants to flow to a normalized semi simple frobenius algebra @xcite theorem given a semi simple point , we can flow to a normalized point ( essentially by coupling to gravity ) at which the quantum chains on the loop space are an @xmath63 algebra .
in this situation the quantum deformed string topology bracket vanishes .
the correlators on @xmath75 for a frobenius algebra define chain level family field theories with partition functions @xmath76 .
one result is that when @xmath32 is semi simple , the operations on the degenerate higher genus penner
boundary actually come from lower genus moduli space via an invertible morphism .
we expect to show that we can identify our stabilization with that of tilmann , segal and madsen @xcite and prove that the partition function is actually fixed by its genus zero contribution . for this we can pull back co - chains along our stabilization maps of the colimit which induces push forward on the @xmath77duals .
this means that we can effectively push the operations to infinite genus .
if we have an identification with the usual stabilization we can apply the mumford conjecture proved in @xcite . the upshot will be a classification result for these theories in terms of @xmath78 classes along the lines of @xcite . in order to compare to gw
invariants , we need to `` go to the boundary '' of moduli space in the deligne
mumford sense .
one approach is using frames of graphs la fulton macpherson ( see also @xcite ) another is to cut the surface along a curve that does not intersect any arcs and then contracting these curves to points .
this essentially yields kontsevich s compactification .
one indication that this approach is natural is given by the observations in appendix b of @xcite . the next step is then to lift the @xmath79 from the kontsevich compactification to the dm compactification .
going beyond this we wish to point out that the underlying algebra @xmath80 is the algebra governing the quantum cohomology of the symmetric products @xcite .
these in turn define the genus @xmath28 gw invariants as discussed in @xcite .
this would provide another approach to teleman s classification result @xcite and give further insight to givental s `` loop space approach '' @xcite .
we have discussed how to regard moving strings as providing surfaces with foliations . using this approach ,
we have recovered the axiomatics of open / closed tft , given a new combinatorial cft and a partial compactification of it which relates to / defines string topology operations and lends itself to give actions on the hochschild complex of a frobenius algebra .
we also obtained a partial list of axioms for the cft and partially compactified cft . in the future
we hope that we can get a full axiomatic description .
furthermore , we were able to give a chain level action of the relevant structures generalizing deligne s conjectures .
along the same lines of reasoning it should be possible to get open / closed operations on the hochschild complex of a category or a pair of algebras in the space filling @xmath0brane situation , say in a landau ginzburg model .
interesting future directions include the study of mapping spaces to varieties / orbifolds .
an even more ambitious project is to try to connect our constructions to the chiral derham complex or other bundle theories .
g. moore and g. b. segal .
_ lectures on branes , @xmath81-theory and @xmath82 charges .
_ lecture notes from the clay institute school on geometry and string theory held at the isaac newton institute , cambridge , uk . | we give a brief summary of algebraic aspects of string theory arising in the noncommutative geometry setting of foliations called string diagrammatics which we introduced jointly with bob penner .
we furthermore discuss how this gives rise to actions on the hochschild complex of a frobenius algebra .
we then explain how this leads to new quantum chains for loop spaces and a stabilization in the semi simple case . |
lorentzian wormholes were originally found as solutions of the einstein field equations with non - trivial topology . in their more simple examples
they present two mouths in asymptotically flat spacetimes connected by a throat @xcite .
more general and purely geometric definitions allow to remove the requirement for asymptotically flat regions and the introduction of trivial topologies in wormhole spacetimes , both static @xcite and dynamic @xcite .
all these wormholes , nonetheless , must be threaded by matter that violates the null energy condition if they are going to be traversable by material systems @xcite .
+ the requirements for exotic matter that violates the energy conditions can be reduced introducing scalar fields or electric charges . very recently
kim and lee @xcite have presented exact solutions for a static charged wormhole .
these spacetimes are free of horizons , so the charge , although having effects on the stability , does not seriously affect the spacetime itself .
+ despite nothing is known about the equation of state of exotic matter , a stability analysis is possible through linearized perturbations around static solutions of the einstein field equations .
this imposes constraints onto the state equation and the stability domain of the solutions .
this approach has been implemented by poisson and visser for spherically symmetric thin - shell wormholes @xcite . for surgical techniques to construct thin - shell wormholes ,
the reader is referred to the papers by visser @xcite .
+ in this paper we extend the stability analysis to the case of charged wormholes with the aim to establish whether the presence of charge effectively can increase the stability domain of this type of spacetimes .
we do not intend to provide any explanation on the mechanisms that might allow the wormhole to acquire or maintain its charge , but we rather focus on the consequences of the charge on the global stability of the system .
we use geometrized units ( c = g=1 ) . the reissner - nordstrm metric is the unique spherically symmetric solution of the vacuum einstein - maxwell coupled equations . in schwarzschild coordinates @xmath0 it takes the form @xmath1 where @xmath2 is the mass and @xmath3 is the sum of the squares of the electric ( @xmath4 ) and magnetic ( @xmath5 ) charges , measured by distant observers . in a local orthonormal frame , the nonzero components of the electromagnetic field tensor are @xmath6 and @xmath7 . if @xmath8 , this geometry represents a black hole with an event horizon with radius @xmath9 if @xmath10 it is a naked singularity .
+ from the reissner - nordstrm geometry we can take two copies of the region with @xmath11 : @xmath12 where @xmath13 if @xmath8 and @xmath14 if @xmath15 and glue them together at the hypersurface @xmath16 to make a charged spherically symmetric thin shell wormhole @xmath17 .
this construction creates a geodesically complete manifold @xmath18 with two asymptotically plane regions connected by a throat .
on @xmath18 we can define a new radial coordinate @xmath19 , where the positive and negative signs correspond , respectively , to @xmath20 and @xmath21 .
@xmath22 represents the proper radial distance to the throat , which is placed in @xmath23 .
distant observers in @xmath18 will see the wormhole as having mass @xmath2 and charges @xmath4 and @xmath5 .
to study this traversable wormhole we use the standard darmois - israel formalism @xcite . for a modern treatment of this formalism
see for example @xcite .
+ the wormhole throat is placed at @xmath24 .
this shell is a synchronous timelike hypersurface .
we can adopt coordinates @xmath25 in @xmath24 , with @xmath26 the proper time on the shell . in order to analyze the stability under spherically symmetric perturbations ,
we let the radius of the throat be a function of the proper time , @xmath27 .
then @xmath28 the second fundamental forms associated with the two sides of the shell are : @xmath29 where @xmath30 are the unit normals ( @xmath31 ) to @xmath24 in @xmath18 : @xmath32 in the orthonormal basis @xmath33 ( @xmath34 , @xmath35 , @xmath36 , @xmath37 ) we have that @xmath38 and @xmath39 where the dot means @xmath40 . + the einstein equations on the shell reduce to the lanczos equations : @xmath41+kg_{\hat{\imath}\hat{\jmath}}=8\pi s_{\hat { \imath}\hat{\jmath } } , \label{e10}\ ] ] where @xmath42\equiv k_{_{\hat{\imath}\hat{\jmath } } } ^{+}-k_{_{\hat{\imath}\hat{\jmath}}}^{-}$ ] , @xmath43=[k_{\hat{\imath}}^{\hat{\imath}}]$ ] and @xmath44 is the surface stress - energy tensor , with @xmath45 the surface energy density and @xmath46 the surface tensions . then replacing eqs .
( [ e8 ] ) and ( [ e9 ] ) in eq .
( [ e10 ] ) we obtain @xmath47 @xmath48 where @xmath49 is the surface pressure .
the surface energy density is negative , indicating the presence of _ exotic matter _ in the throat .
+ the dynamical evolution of the wormhole throat is governed by the einstein equations plus the equation of state @xmath50 that relates @xmath51 and @xmath52 . replacing the equation of state in eq .
( [ e12 ] ) and using eq .
( [ e11 ] ) , a second order differential equation is obtained for @xmath53 .
this equation has an unique solution for given initial conditions @xmath54 and @xmath55 , where @xmath56 is an arbitrary ( but fixed ) time .
+ from the eqs .
( [ e11 ] ) and ( [ e12 ] ) above , it is easy to verify the energy conservation equation : @xmath57 where @xmath58 is the area of the wormhole throat . the first term of eq .
( [ e13 ] ) represents the internal energy change of the throat and the second the work done by the internal forces of the throat . as in the case of uncharged wormholes
@xcite , the conservation equation can be written in the form @xmath59 which can be integrated to give @xmath60 this can be formally inverted to obtain @xmath61 . thus , replacing @xmath62 in eq .
( [ e11 ] ) and regrouping terms , the dynamics of the wormhole throat is completely determined by a single equation : @xmath63 ^{2}=-1 , \label{e16}\ ] ] with the initial conditions @xmath54 and @xmath55 .
) .,width=604,height=850 ] .,width=604,height=850 ] eq .
( [ e16 ] ) can be rewritten in the form : @xmath64 with @xmath65^{2}. \label{e18}\ ] ] let us consider now a static solution with radius @xmath66 and the corresponding values of energy density @xmath67 and pressure @xmath68 : @xmath69 @xmath70 to study the stability of this static solution under radial perturbations , we carry out a linearized analysis , as it is done for the noncharged case in @xcite . a taylor expansion to second order of the potential @xmath71 around the static solution yields : @xmath72 where the prime means @xmath73 . from eq .
( [ e18 ] ) the first derivative of @xmath71 is @xmath74 since @xmath75 , using the eq .
( [ e14 ] ) , we have that @xmath76
. then @xmath77 the second derivative of the potential is @xmath78 .
\label{e24}\ ] ] since @xmath79 , defining the parameter @xmath80 , we have that @xmath81 . using again that @xmath82 , we obtain @xmath83 .
\label{e25}\ ] ] replacing in @xmath84 , we have that @xmath85 , so @xmath86 , \label{e26}\ ] ] with @xmath87 } { a_{0}^{2}-2ma_{0}+q^{2 } } + 2\left ( a_{0}^{2}-3ma_{0}+2q^{2}\right ) \beta _ { 0}^{2}\right ] , \label{e27}\ ] ] where @xmath88 . +
the wormhole is stable if and only if @xmath89 . the curves where @xmath90 correspond to @xmath91 } { 2\left ( a_{0}^{2}-2ma_{0}+q^{2}\right ) \left ( a_{0}^{2}-3ma_{0 } + 2q^{2}\right ) } , \label{e28}\ ] ] and @xmath92 @xmath93 when @xmath94 .
+ for the analysis of the stability regions it is convenient to consider five cases accordingly to the value of charge : 1 .
case @xmath95 .
there are two regions of stability , + i ) @xmath96 if @xmath97 , and + ii ) @xmath98 if @xmath99 .
2 . case @xmath100 .
there are two regions of stability , + i ) @xmath96 if @xmath101 , and + ii ) @xmath98 if @xmath102 .
3 . case @xmath103 .
there are two regions of stability , + i ) @xmath104 + and + ii ) @xmath98 if @xmath99 .
4 . case @xmath105 .
there are two regions of stability , + i ) @xmath98 if @xmath106 , and + ii ) @xmath98 if @xmath107 .
case @xmath108 .
there is one region of stability , + @xmath98 if @xmath109 .
+ the regions of stability for different values of charge are shown in figs .
[ fig1 ] and [ fig2 ] . for ordinary matter
, @xmath110 represents the velocity of sound , so it should satisfy @xmath111 . if the matter is exotic , as it happens to be in the throat ,
@xmath110 is not necessarily the velocity of sound and it is not clear which values it can take ( see discussion in @xcite ) . + for @xmath112 the stability regions are similar to the case of uncharged wormholes .
the _ tongue _ shaped region with @xmath113 moves slowly to lower values of @xmath114 and @xmath115 as the charge increases .
values of @xmath114 in the interval @xmath116 $ ] are not included in the stability region unless @xmath117 is very close to one .
the other region of stability only moves slowly to lower values of @xmath115 for higher values of charge .
+ the situation changes dramatically if @xmath118 . in this case , for every value of the parameter @xmath114 there are values of the radius @xmath115 that makes the wormhole stable .
when @xmath119 and @xmath120 , it is stable regardless of the equation of state of the exotic matter ( which determines @xmath114 ) .
if @xmath121 , besides the large negative @xmath114 region of stability , it includes also a small one with positive @xmath114 , for radii of the throat in the range @xmath122{\frac { q^{2}}{m^{2}}-1}$ ] .
we have applied a linearization stability analysis to thin - shell wormholes endowed with charge and we have found that the presence of large charges ( @xmath123 ) , significantly increases the possibility of obtaining stable wormhole spacetimes , imposing less severe restrictions on the equation of state of the exotic matter that must be present in the wormhole throat . although it is not possible to give a full physical meaning to the parameter @xmath124 in the absence of a detailed microphysical model for the exotic matter , our analysis shows that , for a given radius @xmath66 , it is _ always _ possible to find stable traversable wormhole solutions for _ any _ value of @xmath124 by taking adequate values of charge and mass .
this work has been supported by universidad de buenos aires ( ubacyt x-143 , efe ) , conicet ( pip 0430/98 , ger ) , anpct ( pict 98 no .
03 - 04881 , ger ) , and fundacin antorchas ( ger ) .
ger thanks the max - planck - institut fr kernphysik for kind hospitality and dr .
diego f. torres for discussions .
some calculations in this paper were done with the help of the package grtensorii @xcite .
n. sen , ann .
( leipzig ) * 73 * , 365 ( 1924 ) ; k. lanczos , _ ibid . _ * 74 * , 518 ( 1924 ) ; g. darmois , memorial de sciences mathematiques , fasticule xxv ch v ( 1927 ) ; w. israel , nuovo cimento * 44b * , 1 ( 1966 ) ; * 48b * , 463(e ) ( 1967 ) . | the linearized stability of charged thin shell wormholes under spherically symmetric perturbations is analyzed .
it is shown that the presence of a large value of charge provides stabilization to the system , in the sense that the constraints onto the equation of state are less severe than for non - charged wormholes .
+ keywords : lorentzian wormholes ; exotic matter ; einstein - maxwell spacetimes |
psr j1846@xmath00258 is a young ( @xmath2800yr ) , 326ms pulsar , discovered in 2000 with the _ rossi x - ray timing explorer _
( _ rxte _ ; * ? ? ? * ) .
no radio pulsations have been detected despite deep searches @xcite .
psr j1846@xmath00258 has a large inferred magnetic field of @xmath3 g , above the quantum critical limit , and a braking index of @xmath4 ( * ? ? ?
* hereafter lkgk06 ) .
the pulsar was therefore believed to be simply a high - magnetic field rotation - powered pulsar ( rpp ) , with radio pulsations that do not cross our line of sight .
however , in 2006 may , the source experienced a series of x - ray bursts and a sudden increase in x - ray flux , which strongly suggest magnetic activity @xcite .
the source also showed spectral variations and changes in the surrounding nebula @xcite .
a magnetar is a neutron star whose magnetic energy powers the bulk of its emission @xcite , while a rpp produces radiation via the loss of rotational kinetic energy ( e.g. , * ? ? ?
psr j1846@xmath00258 is a unique transition object between these two source classes .
the x - ray luminosity of psr j1846@xmath00258 can be entirely accounted for by the spin - down power of the pulsar , however , the observed x - ray bursts and flux flare are phenomena only seen thus far from magnetars @xcite .
a neutron star glitch is defined as a sudden , usually unresolved , increase in spin frequency , @xmath5 .
glitches are often accompanied by an increase in the frequency derivative , @xmath6 , and are sometimes followed by an exponential decay on timescale @xmath7 , where some of the initial jump in @xmath5 is recovered . in general , a glitch at time @xmath8 can be modeled as : @xmath9 where @xmath10 is the frequency of the pulsar prior to the glitch and @xmath11 is the initial frequency jump , which can be decomposed into the part of the glitch that is permanent , @xmath12 , and that which decays , @xmath13 .
the recovery fraction is defined as @xmath14 .
the variety of observed pulsar glitches is interesting .
the fractional magnitude of the observed change in @xmath5 ranges over 6 orders of magnitude from @xmath15 @xcite .
some glitches are characterized only by a @xmath5 increase ( e.g. , psr b1758@xmath023 ; * ? ? ?
* ) , while some glitches ( typically in young pulsars ) are dominated by an increase in the magnitude of @xmath6 ( e.g. , the crab pulsar ; * ? ? ?
some glitches decay on timescales of @xmath2days ( e.g. , * ? ? ?
* ) , while others display very long decay time scales ( @xmath2 hundreds of days ; * ? ? ?
the fraction of the glitch that decays is also highly variable . in older pulsars
the amount of glitch recovery is typically small ( @xmath16 , * ? ? ?
* ) while in the crab pulsar the recovery can be nearly complete ( e.g. , @xmath17 , * ? ? ?
however , in many glitches , no recovery is detected .
glitches are now known to be ubiquitous in magnetars as well as rotation - powered pulsars @xcite .
what remains to be seen is whether glitches have the same physical origin in both types of objects , or if the super - critical magnetic fields of magnetars are responsible for different glitch origins and evolutions .
while some magnetar glitches are indistinguishable from those observed in rpps , others occur contemporaneously with radiative changes such as bursts , flux enhancements , and spectral or pulse profile variations , such as in the anomalous x - ray pulsar ( axp ) 1e 2259 + 586 @xcite .
no radiative changes have been observed with rpp glitches ( e.g. , * ? ? ?
rpp glitches are believed to arise from a sudden unpinning of vortices in the superfluid interior crust of the pulsar ( e.g. , * ? ? ?
magnetar glitches , on the other hand , may be triggered by strong internal magnetic fields as the crust is deformed , either plastically or cracked violently @xcite . in this paper
, we discuss the timing behavior of psr j1846@xmath00258 prior to , during , and following the period of magnetic activity observed in 2006 .
we show that a large glitch occurred contemporaneous with the x - ray bursts and onset of the flux flare , accompanied by an unusual increase in the timing noise of the pulsar .
we show that the glitch recovery is very unusual for a rpp but is reminiscent of timing behavior observed from magnetars and is further evidence of magnetic activity in psr j1846@xmath00258 .
observations of psr j1846@xmath00258 were made using the proportional counter array ( pca ; * ? ? ? * ; * ? ? ? * ) on board _
rxte_. the pca consists of an array of five collimated xenon / methane multi - anode proportional counter units ( pcus ) operating in the 2 60kev range , with a total effective area of approximately @xmath18 and a field of view of @xmath19 fwhm .
our entire _ rxte _
data set spans 9.7yr from 1999 april 18 through 2008 december 10 ( mjd 51286 54810 ) .
data from 2000 january 31 - 2005 july 27 ( mjd 51574 53578 ) were reduced and analyzed previously and details can be found in lkgk06 .
eleven observations taken in 1999 april 18 - 21 are of limited use for the current analysis since they can not be unambiguously phase connected to the rest of the data .
analysis of data spanning 2005 july 27 - 2008 december 10 ( mjd 53578 54810 ) is presented here .
data were collected in `` goodxenon '' mode , which records the arrival time ( with 1-@xmath20s resolution ) and energy ( 256 channel resolution ) of every unrejected event . typically , two to three pcus were operational during an observation .
we used the first xenon layer of each operational pcu and extracted events in channels 4 48 ( approximately 2
20kev ) , as this maximizes the signal - to - noise ratio for this source .
observations were downloaded from the heasarc archive and data from each active pcu were merged and binned at ( 1/1024)s resolution .
photon arrival times were converted to barycentric dynamical time ( tdb ) at the solar system barycenter using the j2000 source position ra = @xmath21 , decl @xmath22 @xcite and the jpl de200 solar system ephemeris .
the known ephemeris from lkgk06 was used to fold each time series with 16 phase bins .
resulting profiles were cross - correlated in the fourier domain with a high signal - to - noise ratio template created by adding phase - aligned profiles from all observations .
the cross - correlation process assumes that the pulse profile is stable ; indeed , we found no evidence for variability that could bias toa measurement , as shown in figure [ fig : profiles ] , and confirmed by @xcite .
the figure shows average profiles from before , after , and throughout the outburst and unusual timing behavior . for each individual observation ,
the cross - correlation yielded the time of arrival ( toa ) of phase - zero of the average pulse profile at the fold epoch .
the toas were fitted to a timing model ( see section [ sec : timing ] ) using the pulsar timing software package tempo .
after phase - connecting the data , we merged observations occurring on a single day and used the ephemeris to re - fold the data in order to obtain more precise toas .
this process produced 199 toas with a typical uncertainty @xmath29ms ( @xmath22.7% of the pulse period ) .
further details of the observation and analysis are given in lkgk06 .
phase - coherent timing is a powerful method for obtaining accurate pulsar parameters , but can only be used when timing noise and glitches are relatively small ( e.g. , * ? ? ? * ) . to obtain a phase - coherent timing solution , each turn of the pulsar is accounted for by fitting toas with a taylor expansion of the pulse phase @xcite . phase - coherent timing for psr j1846@xmath00258 spanning 2000 january 1 2005
july 27 ( mjd
5157453578 ) is discussed in lkgk06 .
phase coherence was maintained for the next 308 days without incident .
phase coherence was lost with the observation occurring on 2006 may 31 ( mjd 53886 ) which contained 4 x - ray bursts and a pulse flux increase @xcite . for the following 32 observations spanning 192 days , no unambiguous phase coherent timing solution was possible .
instead , we performed periodograms to determine the pulse frequency .
uncertainties were determined from a monte carlo simulation , where noise was added to simulated sinusoidal pulses and the frequency for each trial was determined in the same way as for the real data ( see lkgk06 for further details ) .
phase coherence was once again obtained starting 2007 january 26 ( mjd 54126 ) with closely spaced bootstrapping observations after the source reappeared from behind the sun after 48 days .
a single phase coherent timing solution was obtained spanning 2007 january 26 2008 december 10 ( mjd 5412654810 ) .
this timing solution is severely contaminated by long - term glitch recovery and timing noise , so the fitted parameters for the global solution are of limited value . to analyze the long - term rotational behavior of the pulsar , we created short phase - coherent timing solutions from 2000 until the onset of bursts , and from 2007 and 2008 .
each timing solution included only @xmath5 and @xmath6 , and included as much data as possible while requiring the reduced @xmath23 value of the fit to be @xmath21 .
this resulted in 11 measurements of @xmath5 and @xmath6 pre - glitch and 11 measurements post - glitch . in order to better utilize the available data in the post - glitch period
, we created short overlapping timing solutions , each of which uses approximately half the data from two of the above described short data subsets , and has the same fitted parameters and @xmath23 requirements .
this can be useful because the short coherent timing solutions result in parameter fits that are dominated by the end points , which can be problematic when @xmath5 is varying rapidly from glitch recovery as in this case , or when timing noise is a significant effect .
this produces an additional 9 post - glitch measurements of @xmath5 and @xmath6 .
coherent frequency measurements ( crosses ) , overlapping frequency measurements ( filled circles ) and frequency measurements obtained via periodograms ( open circles ) , are plotted in the top panel of figure [ fig : freq ] , with the pre - burst @xmath5 , @xmath6 , and @xmath24 removed .
the middle panel of figure [ fig : freq ] shows measurements of @xmath6 from the short coherent timing solutions as crosses , with the overlapping @xmath6 measurements in filled circles .
in addition , open circles show three measurements of @xmath6 that are calculated from weighted least - squares fits to nine periodogram measurements of @xmath5 .
a frequency increase is apparent in the frequency residual plot ( top panel and inset , figure [ fig : freq ] ) , indicating that a glitch occurred ( also noted by * ? ? ?
two measurements of @xmath5 are larger than the pre - glitch predicted value .
we calculated the average @xmath25 at mjd 53890 ( 4 11days after the glitch occurred ) .
the initial value of @xmath11 at the time of the glitch was presumably larger than this value , as indicated by the exponential fit to the data discussed below .
we fitted the measured frequencies of psr j1846@xmath00258 ( fitting only 3 pre - glitch @xmath5 values ) with an exponential recovery glitch model ( eq .
1 ) , the results of which are shown in figure [ fig : glitchresids ] .
the top panel of the figure shows frequency measurements with the pre - glitch ephemeris ( @xmath5 , @xmath6 and @xmath24 ) removed ( as in figure [ fig : freq ] ) , while the bottom panel shows residuals from the glitch fit .
the uncertainties from the periodogram measurements of @xmath5 are @xmath26 orders of magnitude larger than those from short coherent fits to the data and thus contribute minimally to the overall @xmath23 value .
thus the glitch residuals ( bottom panel , figure [ fig : glitchresids ] ) are shown to highlight the deviation from the fit of the coherent frequencies , which dominate the @xmath23 .
significant deviations from the fit can be seen during the period of glitch recovery , giving rise to a large reduced @xmath23 value of @xmath27 for 45 degrees of freedom for the best fit ( bottom panel , figure [ fig : glitchresids ] ) .
given the poor fit , the formal uncertainties underestimate the true values .
to determine more reasonable uncertainties on the glitch parameters , we multiplied the uncertainties on the coherent @xmath5 measurements by a factor until the reduced @xmath23 of the fit @xmath28 .
we applied a multiplicative factor only to the phase coherent @xmath5 measurements because these uncertainties were @xmath29 orders of magnitude smaller than those from periodogram @xmath5 measurements , determined in a different way , and the contribution to @xmath23 was mainly from the coherent @xmath5 values .
quoted uncertainties on the glitch parameters are from @xmath30 contours from the fit with multiplied uncertainties .
the fitted value of the initial fractional frequency increase is @xmath31 , very large for such a young pulsar , where a typical value is @xmath32 .
more remarkable yet , however , is the amount by which the frequency recovers .
we find @xmath33 , corresponding to a net decrease in frequency of @xmath34hz .
all fitted glitch parameters are given in table [ table : glitch ] . while the deviation from the exponential fit is very significant for several months ( bottom panel , figure [ fig : glitchresids ] ) the overall evolution after the glitch
is dominated by the exponential recovery : the deviation from the fit is @xmath22 orders of magnitude smaller than the overall post - glitch decrease in @xmath5 .
the deviation from exponential recovery may well have been larger in the period just following the glitch , however , the large uncertainties on the periodogram measurements of @xmath5 ( @xmath35hz ) prevent any firm conclusion .
however , since the corresponding pulse toas can not be unambiguously phase - connected , it is likely that large variations in @xmath5 and @xmath6 occurred during the 240-day period between the glitch epoch and when we regained phase - coherence .
it is also possible that a second , smaller glitch ( @xmath36 ) occurred during this period .
the observed deviation from the exponential recovery decreases as the glitch recovers .
thus , in the closing months of 2008 , the pulsar was rotating very regularly again , similar to its pre - glitch behavior .
figure [ fig : nudot ] shows post - glitch phase - coherent measurements of @xmath6 ( a subset of the @xmath6 measurements shown in the middle panel of figure [ fig : freq ] ) .
the pre - glitch measurements are excluded here for clarity .
the solid line is the derivative of the glitch model fitted to the @xmath5 measurements , clearly showing that there is significant deviation .
the overall effect of the glitch recovery on @xmath6 is clear , however , from mjd 54100 - 54300 the @xmath6 measurements deviate from the exponential recovery by @xmath37% .
the effect of this anomalous change in @xmath6 is not directly evident in the measurements of @xmath5 ( which are dominated by the exponential recovery ) but does help explain why the exponential glitch fit is not a satisfactory description of the data , and is clear in the residuals of the glitch fit in the bottom panel of figure [ fig : freq ] .
the glitch during psr j1846@xmath00258 s outburst was accompanied by a major pulsed flux enhancement @xcite . in order to quantify the radiative properties of the source during the glitch recovery , we extracted its pulsed flux using all available _ rxte _ observations .
first , we generated separate event lists for each pcu in fits format using the standard ` ftools ` .
we then filtered our event lists such that we only preserved photons in the 2
20kev band , from the first xenon layer .
the photon arrival times were then barycented using the source position and the jpl de200 solar system ephemeris .
we folded the filtered barycented photon arrival times using the ephemeris determined in our phase - coherent timing analysis using 16 phase bins . using the folded profiles , we calculated the rms pulsed flux in each pcu using the fourier method described by @xcite keeping only the contribution from the 1st harmonic given the source s roughly sinusoidal profile .
not all the observations were pointed at psr j1846@xmath00258 , therefore we corrected for the reduced efficiency in each pcu due to the offset pointing using the collimator response of each pcu and the instrument attitude files . finally , we averaged the pulsed flux in each pcu weighted by the fractional exposure of each pcu .
we excluded the contribution from pcu 0 because of the loss of its propane layer and the numerous detector breakdown events .
our pulsed flux time series is presented in the bottom panel of figure [ fig : freq ] .
the event lists for each pcu created for the pulsed flux analysis were binned into 31.25 ms lightcurves and were searched for bursts using the burst search algorithm introduced in @xcite .
no additional bursts were found other than the 5 reported in @xcite .
we checked for a correlation between torque and pulsed flux by plotting the pulsed flux against the spin frequency derivative in log - log space , shown in figure [ fig : torqueflux ] .
all measurements of @xmath6 before and after the glitch and bursts are shown , including 3 measurements of @xmath6 obtained from weighted least - squares fits to @xmath5 measurements obtained from periodograms .
these 3 @xmath6 measurements have uncertainties @xmath26 orders of magnitude larger than those for @xmath6 measurements from coherent timing .
the plot shows a possible correlation between flux and torque when both parameters are extreme . fitting a power law to the data gives a power law index of @xmath28 .
however , this value is constrained by a single measurement of @xmath6 when the flux , and timing noise , are both very large . excluding this point ,
there is no significant correlation between flux and @xmath6 .
the observed variations in @xmath6 during the later phase connected period , are at the @xmath38 level , while the variation in @xmath6 immediately following the glitch are at the @xmath39 level ( figure [ fig : nudot ] ) .
correspondingly small fluctuations in pulsed flux are not detectable in these data .
four x - ray bursts in psr j1846@xmath00258 coincided with the onset of a flux flare on 2006 may 31 ( mjd 53886 ) .
the pulsed flux decayed over @xmath22 months and reached quiescence around the time of the fifth burst on 2006 july 27 ( mjd 53943 ) @xcite .
significant spectral changes also occurred @xcite and flux enhancement up to 300kev was observed @xcite .
contemporaneous with the sudden change in the x - ray emission from psr j1846@xmath00258 , we observed a large glitch with an initial frequency increase of @xmath40 .
the glitch decayed over @xmath41days , with a recovery fraction of @xmath33 , resulting in a net decrease of the pulse frequency of @xmath42hz .
furthermore , the timing behavior during the period of recovery is not well modeled by a simple exponential function and measurements of @xmath6 in particular are suggestive of a high level of timing noise for several hundred days following the glitch .
this glitch and subsequent recovery reinforces that psr j1846@xmath00258 underwent a period of magnetic activity in 2006 .
this glitch is entirely different from the previous glitch in this source , which was radiatively silent , small in magnitude ( @xmath43 ) , dominated by a change in @xmath6 ( @xmath44 ) , and had no measurable recovery ( lkgk06 ) .
this small glitch is similar to those observed in the crab pulsar ( e.g. , * ? ? ?
* ) and other very young rotation - powered pulsars such as psr b0540@xmath069 @xcite .
it is unusual to have two such disparate initial @xmath11 magnitudes in the same source , particularly in such a young pulsar , though this has been seen in some older pulsars such as psr b1737@xmath030 , which has glitch magnitudes spanning four orders of magnitude @xcite .
in fact , the glitch reported here is the largest glitch ever observed in any of the pulsars with characteristic ages less than @xmath45kyr ( the crab pulsar , b0540@xmath069 , b1509@xmath059 , and j1119@xmath06127 ) , none of which have experienced glitches with fractional magnitudes larger than @xmath32 .
the glitch reported here has a recovery fraction of @xmath46 .
the measured value of @xmath47 implies that the net frequency change after the glitch recovery is negative , @xmath48 , as shown in figure [ fig : freq ] .
a similar effect was recently observed in the axp 4u 0142 + 61 , though with much smaller magnitude of @xmath49 @xcite .
the negative change in @xmath5 resulting from the over - recovery of the psr j1846@xmath00258 glitch is similar in magnitude to the unresolved timing event seen in the magnetar sgr 1900 + 14 in the months before the giant flare in 1998 @xcite . in that case , an enhanced spin - down of the magnetar was observed over @xmath23 months .
well spaced timing observations around the time of the timing event were not available , so no glitch could be resolved , if indeed one occurred .
@xcite attribute the observed behavior in sgr 1900 + 14 to one of two possibilities .
the first is an increase in the magnitude of @xmath6 by a factor of @xmath50 , persisting for @xmath51days .
the second is that a negative glitch , that is , a sudden spin - down occurred , with magnitude @xmath52 .
however , a timing event similar to that observed in psr j1846@xmath00258 could also describe the data .
it is curious , however , that such similar fractional changes in @xmath5 should occur in two sources that experienced such disparate radiative changes , with the energy output from sgr 1900 + 14 being several orders of magnitude larger than from psr j1846@xmath00258 .
axp 1e 2259 + 586 experienced a glitch contemporaneous with 80 x - ray bursts , a flux flare and pulse profile changes in 2002 @xcite .
neither this glitch nor the large psr j1846@xmath00258 glitch can be described with a single exponential decay model .
the addition of an exponential growth component better describes the 1e 2259 + 586 glitch , however , the addition of a similar component for the psr j1846@xmath00258 glitch does not provide a significant improvement to the fit .
interestingly , the flux enhancement observed in 1e 2259 + 586 lasted much longer ( @xmath53yr ; * ? ? ? * ) than the glitch recovery time scale ( @xmath54days ) , whereas the reverse is true for psr j1846@xmath00258 , with @xmath55days and a flux decay timescale of @xmath56days @xcite .
the 1e 2259 + 586 event can also be distinguished from the psr j1846@xmath00258 glitch in that its recovery fraction is much smaller , with @xmath57 .
in 2001 , the axp 1rxs j170849.0@xmath0400910 also experienced a glitch with recovery that was not well described by a simple exponential , and not improved with the addition of a second exponential term @xcite .
@xcite argued that it is unlikely that 1rxs j170849.0@xmath0400910 experienced bursts or a pulsed flux flare associated with this glitch because a flux flare would have had to decay on a time scale less than the glitch decay time scale .
however , this is exactly the behavior observed from psr j1846@xmath00258 , albeit with a much longer time scale .
long - term spectral changes and flux variations have been claimed in 1rxs j170849.0@xmath0400910 @xcite .
an interesting characteristic of some axp glitches is a period of enhanced spin - down immediately following the glitch , as observed in 1e 1841@xmath0045 , 1rxs j170849.0@xmath0400910 , and 1e
2259 + 586 @xcite .
the additional instantaneous spin - down at the glitch epoch owing to the exponential recovery can be quantified as @xmath58 ( found by taking the derivative of the exponential term in eq . 1 and setting @xmath59 ) .
for the 1e 2259 + 586 glitch , @xmath60 , while a typical value for a rpp is @xmath61 ( see * ? ? ?
* and references contained therein ) . for psr j1846@xmath00258 the instantaneous spin - down is @xmath62 , larger than for any rpp glitch , but not as large as those measured for axps .
it should be noted , however , that an enhanced spin - down is not observed in every axp glitch . in psr j1846@xmath00258
, there is evidence for a @xmath2200-day interval where @xmath6 deviates significantly from an exponential glitch recovery , and it is possible that further significant deviation occurred during the 240-day period of unconnected data directly following the glitch . perhaps @xmath6 is varying in a stochastic fashion similar to that observed in the axp 1e 1048.1@xmath05937 @xcite .
in this axp , a @xmath2year - long period of rapid @xmath6 variations followed a large pulsed flux flare and a possible glitch in 2002 .
another glitch in 2007 coincided with the onset of a pulsed - flux flare , again followed by stochastic variations in @xmath6 .
another similarity between these sources is that the flux enhancement decayed away long before the timing variations subsided .
alternatively , perhaps the variations are more simply attributed to timing noise , as is seen in many young rpps .
the behavior remains unusual however , since this is qualitatively very different from the mild timing noise observed in psr j1846@xmath00258prior to magnetic activity , and such a large change in timing noise behavior is unprecedented among rpps .
a simple estimate of the transfer of rotational kinetic energy at the time of the glitch can be obtained by treating the star as a solid rotating body .
assuming the canonical neutron star moment of inertia of @xmath63g@xmath64 , the energy deposited in the solid crust at the time of the glitch owing to the increase in @xmath5 is @xmath65ergs .
however , it is well established that neutron stars are comprised of at least two components , a solid crust coupled to the core and a loosely coupled superfluid component in the crust .
taking the two - component model into account but not making any further assumptions about the nature of the two components , we can calculate the energy transferred between the components at the time of the glitch .
the first constraint is that angular momentum is conserved , that is , @xmath66 where @xmath11 is the observed increase in spin - frequency , and @xmath67 is the unknown change in spin - frequency of the superfluid , and @xmath68 and @xmath69 are the moment of inertia of the solid crust and core , and superfluid , respectively .
an estimate of the energy contained in the glitch can then be calculated as @xmath70 the frequency lag between the crust and superfluid can be estimated from the glitch decay time , @xmath7 , as @xmath71 where @xmath72 is the characteristic age of the pulsar @xcite . for psr j1846@xmath00258 ,
the frequency lag is @xmath73hz , giving a glitch energy of @xmath74ergs for the measured glitch parameters ( see table [ table : glitch ] ) . if the measured glitch decay timescale , @xmath7 , in this case is dominated by processes external to the neutron star ( as discussed in section [ sec : magglitchdisc ] ) , any other glitch decay timescale would occur on a shorter timescale than the one observed , so this estimate of the frequency lag and thus glitch energy can be considered an upper limit . for a distance to the pulsar of 6kpc @xcite the energy estimated to have been released in the bursts and flux
flare , assuming isotropic emission , is @xmath75ergs ( 2
a new estimate of the distance to the pulsar of @xmath210kpc @xcite increases the amount of energy contained in the radiative outburst to @xmath76ergs ( 2
for either distance , the energy contained in the radiative outburst is several orders of magnitude larger than that contained in the glitch .
this is similar to the bursts and glitch from 1e 2259 + 586 , for which the energy contained in the glitch was @xmath22 orders of magnitude less than the energy contained in the bursts and flux flare @xcite .
this is suggestive that the glitch alone is not responsible for the radiative outburst , in contrast to the argument put forward by @xcite .
moreover , there is no evidence for this event ( or in any other similar axp event ) that the glitch preceded the radiative event .
this could be tested only by sensitive continuous x - ray monitoring of this and other similar sources .
rotation - powered pulsar glitches are thought to arise from differential rotation in the neutron star , where the crust contains superfluid neutrons rotating more rapidly than the surrounding matter ( e.g. , * ? ? ?
* ; * ? ? ?
the angular momentum of a rotating superfluid is quantized in vortices which are thought to become pinned to nuclei in the star s crust .
the vortex lines are therefore under extreme stresses due to the differential rotation between the crust and the superfluid .
it is thought that the vortex lines experience sudden unpinning , resulting in the transfer of angular momentum to the crust , observed as an increase in the spin - frequency of the neutron star .
magnetar glitches may instead be triggered by strong internal magnetic fields as the crust is deformed , either plastically or cracked violently @xcite .
this idea is supported by the large number of glitches now observed to occur at the same epoch as magnetically powered radiative events , such as bursts and flares ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the physics underlying glitches with @xmath47 is unclear
. the classical glitch model of vortex unpinning in the superfluid crust of the neutron star does not readily produce such dramatic glitch recoveries .
one possibility is that some parts of the superfluid are in fact rotating more slowly than the crust . then the initial @xmath5 increase would be from a transfer of angular momentum from a more rapidly rotating region of the superfluid to the crust , which is then followed by a transfer of angular momentum from the crust to the more sluggish region of the superfluid .
this is the explanation put forward by @xcite to explain the net spin - down event in sgr 1900 + 14 .
they argue that regions of slowly rotating superfluid can occur in magnetars because vortex motion is dominated by advection across the neutron star surface by the deforming crust and that gradual plastic deformation of the neutron star crust will cause the superfluid to rotate more slowly than the crust .
however , it is not clear whether such behavior is expected in a neutron star with a magnetic field of @xmath77 g , spinning relatively rapidly compared to the magnetars . that the recovery of the psr j1846@xmath00258 glitch so far overshoots the initial frequency increase is suggestive of an external torque following the glitch .
previous evidence for an external torque includes the large fraction of @xmath78 implied to have decoupled in the 1e 2259 + 586 glitch .
however , this may also be explained with a core glitch , that is , where the core of the pulsar ( and thus a large fraction of @xmath78 ) decouples temporarily from the crust @xcite . in psr
j1846@xmath00258 , however , the post - glitch relaxation amplitude is much greater than the initial glitch amplitude , offering support for the idea that the post - glitch spin - down behavior results from an external source .
one possibility is that a magnetic field twist responsible for the x - ray bursts and flux enhancement also affects the spin - down of the pulsar . in this case
the observed recovery is driven by the propagation of magnetic field untwisting ( similar to a shock wave ) through the magnetosphere . during this process , the spin - down of the star may increase because the effective magnetic field has increased .
when the `` shock '' reaches the light cylinder , which can take place on few month time scales , the spin - down should return to its pre - burst value @xcite .
this theory also allows for non - monotonic behavior in the spin - down after an event , as observed in both axp 1e 1048.1@xmath05937 @xcite and psr j1846@xmath00258 .
this model allows for a delay between flux variations and the onset of timing variability , as was observed in 1e 1048.1@xmath05937 , though not observed in psr j1846@xmath00258 .
however , as in the model of variably rotating superfluid , it is not immediately clear how the more rapid rotation and smaller magnetic field affect the relevance of this model . in particular ,
the different light cylinder radii of psr j1846@xmath00258 compared to the axps should result in different shock propagation times .
alternatively , it has been proposed that fallback disks from the supernova explosion creating the neutron star could be interacting with magnetars and be responsible for some of the observed emission ( e.g. , * ? ? ?
* ; * ? ? ?
in this case , the initial x - ray bursts could irradiate a fossil disk commencing a period of disk activity .
the interaction between the neutron star and a disk could cause the enhanced spin - down , which would decay as the disk cooled .
however , in the framework of this model , it is difficult to understand how accretion causing variations in the spin - down rate could continue for so much longer than the pulsed - flux enhancement .
another rpp , psr j1119@xmath06127 , has very similar spin properties to psr j1846@xmath00258 .
its @xmath79 , @xmath80 , and @xmath72 are all similar , and notably , it has a similarly large magnetic field of @xmath81 g @xcite .
this pulsar has shown some indication of unusual x - ray emission .
no magnetospheric x - ray emission is detected , but thermal pulsations with a @xmath82 pulsed fraction and a large surface temperature are detected @xcite .
no direct evidence of magnetic activity ( i.e. bursts or flux enhancements ) is present in this pulsar . given the similarities between psr j1119@xmath06127 and psr j1846@xmath00258 , both sources , as well as other high b - field rpps are currently being monitored with _ rxte _
for similar magnetic activity .
psr j1846@xmath00258 may be related to the transient axps ( taxps ) .
psr j1846@xmath00258appears to be a typical rpp for @xmath83 of the time , with brief periods of magnetic activity occurring approximately once a decade .
the taxp xte j1810@xmath0197 increased in brightness by a factor of @xmath2100 and was subsequently visible for several years as a pulsed x - ray source @xcite .
another taxp , 1e 1547@xmath05408 was detected as a pulsed radio source after an x - ray outburst in which the flux increased by at least a factor of 16 @xcite .
interestingly , both xte j1810@xmath0197 and 1e 1547@xmath05408 , two _ bona fide _ taxps , are the only two magnetars with detected radio pulsations .
by contrast , no radio pulsations have been detected from psr j1846@xmath00258despite extensive searches both before and after the magnetic activity @xcite .
in _ rxte _ observations of psr j1846@xmath00258 , we have observed a large glitch with an unusual quasi - exponential over - recovery of @xmath5 and substantial timing noise contemporaneous with x - ray bursts and a flux increase .
these observations strengthen the tie between magnetic activity in neutron stars and unusual glitch activity , as has been previously noted ( e.g. , * ? ? ?
a glitch with recovery fraction @xmath47 has never before been observed from a rotation - powered pulsar and is not compatible with the standard model of pulsar glitches .
the unusually large glitch recovery reported here for psr j1846@xmath00258 , as well as the radiative changes occurring contemporaneously with the psr j1846@xmath00258 glitch and several axp glitches , together provide the best evidence that there are physical differences between typical rpp glitches and some glitches observed in magnetars .
ongoing timing observations of psr j1846@xmath00258 are required to obtain a deterministic post - outburst braking index measurement that is uncontaminated by long - term glitch recovery .
a measurement of the braking index for this source after outburst is of considerable interest as a change would strongly suggest that a magnetic reconfiguration occurred at the time of the outburst .
we thank an anonymous referee for helpful comments that greatly improved the text .
we thank a. beloborodov , d. eichler , e.v .
gotthelf , and b. link for useful discussions relating to the text .
this research made use of data obtained from the high energy astrophysics science archive research center online service , provided by the nasa - goddard space flight center .
mal is an national science and engineering research council ( nserc ) pgs - d fellow .
vmk holds the lorne trottier chair in astrophysics and cosmology and a canada research chair in observational astrophysics .
funding for this work was provided by nserc discovery grant rgpin 228738 - 03 , fqrnt , cifar and cfi .
lc parameter & value + epoch ( modified julian day ) & 53883.0(3.0 ) + @xmath84 & @xmath85 + @xmath86 & 0.0041(2 ) + @xmath7 ( days ) & 127(5 ) + @xmath87 & 8.7(2.5 ) + @xmath88 & @xmath89 + @xmath90 & @xmath91 + + figures in parentheses are uncertainties in the last digits quoted and are the estimated @xmath92 uncertainties from the fitted exponential glitch recovery model using @xmath30 contours .
the model is fitted to data with uncertainties that are increased by a factor until @xmath93 ( see section [ sec : timing ] for details ) . | we report a large spin - up glitch in psr j1846@xmath00258 which coincided with the onset of magnetar - like behavior on 2006 may 31 .
we show that the pulsar experienced an unusually large glitch recovery , with a recovery fraction of @xmath1 , resulting in a net decrease of the pulse frequency .
such a glitch recovery has never before been observed in a rotation - powered pulsar , however , similar but smaller glitch over - recovery has been recently reported in the magnetar axp 4u 0142 + 61 and may have occurred in the sgr 1900 + 14 .
we also report a large increase in the timing noise of the source .
we discuss the implications of the unusual timing behavior in psr j1846@xmath00258 on its status as the first identified magnetically active rotation - powered pulsar . |
the catalan numbers @xmath7 are well - known integers that arise in many combinatorial problems .
stanley @xcite , gives @xmath8 combinatorial interpretations of these numbers . in 1874
e. catalan @xcite observed that the numbers @xmath9 are integers , and their number - theoretic properties were studied by several authors ( see dickson @xcite ) . for @xmath1 , is the middle binomial coefficient @xmath10 , and for @xmath3 it is @xmath11 . except for @xmath12 , these integers are even , and it is convenient for our purposes to divide them by 2 , so we consider the numbers @xmath13 some properties of these numbers are given in @xcite , where they are called super catalan numbers " .
an intriguing problem is to find a combinatorial interpretation to the super catalan numbers .
the following identity ( * ? ? ?
* equation ( 32 ) ) , together with the symmetry property @xmath14 and the initial value @xmath15 , shows that @xmath16 is a positive integer for all @xmath17 and @xmath18 .
@xmath19 formula allows us to construct recursively a set of cardinality @xmath16 but we have not found any natural description of it for @xmath20 .
shapiro @xcite gave a combinatorial interpretation to in the case @xmath3 , which is the catalan number identity @xmath21 a similar interpretation works for the case @xmath1 of ( when multiplied by 2 ) , which is the identity @xmath22 another intriguing formula for @xmath16 , which does not appear in @xcite , is @xmath23 although suggests a combinatorial interpretation for @xmath16 based on a decomposition of pairs of objects counted by catalan numbers , we have not found such an interpretation .
in this paper , we give a combinatorial interpretation for @xmath24 for @xmath25 and for @xmath26 for @xmath27 .
the first few values of @xmath16 for @xmath4 and @xmath28 are as follows : @xmath29 we show that @xmath30 counts pairs of dyck paths of total length @xmath31 with heights differing by at most 1 .
we give two proofs of this result , one combinatorial and one using generating functions .
the combinatorial proof is based on the easily checked formula @xmath32 which we interpret by inclusion - exclusion .
our interpretation for @xmath33 is more complicated , and involves pairs of dyck paths with height restrictions .
although we have the formula @xmath34 analogous to @xmath35 , we have not found a combinatorial interpretation to this formula , and our proof uses generating functions . interpretations of the number @xmath30 in terms of trees , related to each other , but not , apparently , to our interpretation , have been found by schaeffer @xcite , and by pippenger and schleich @xcite .
all paths in this paper have steps @xmath36 and @xmath37 , which we call _ up steps _ and _ down steps_. a step from a point @xmath38 to a point @xmath39 is denoted by @xmath40 .
the _ level _ of a point in a path is defined to be its @xmath41-coordinate .
dyck path _ of _ semilength _
@xmath18 ( or of length @xmath31 ) is a path that starts at @xmath42 , ends at @xmath43 , and never goes below level @xmath44 .
it is well - known that the number of dyck paths of semilength @xmath18 equals the catalan number @xmath45 .
height _ of a path @xmath46 , denoted by @xmath47 , is the highest level it reaches .
every nonempty dyck path @xmath48 can be factored uniquely as @xmath49 , where @xmath50 is an up step , @xmath51 is a down step , and @xmath46 and @xmath52 are dyck paths .
thus the map @xmath53 is a bijection from nonempty dyck paths to pairs of dyck paths .
let @xmath54 be the set of pairs of dyck paths @xmath55 of total semilength @xmath18 .
this bijection gives @xmath56 , so by , we have @xmath57 . our interpretatoin for @xmath30 is a consequence of the following lemma [ l - main ] .
we give two proofs of this lemma , one combinatorial and the other algebraic .
the algebraic proof will be given in the next section .
[ l - main ] for @xmath58 , @xmath45 equals the number of pairs of dyck paths @xmath55 of total semilength @xmath18 , with @xmath46 nonempty and @xmath59 .
let @xmath60 be the set of dyck paths of semilength @xmath18 , and let @xmath61 be the set of pairs of dyck paths @xmath55 of total semilength @xmath18 , with @xmath46 nonempty and @xmath59 .
for a given pair @xmath55 in @xmath61 , since @xmath46 is nonempty , the last step of @xmath46 must be a down step , say , @xmath40 . by replacing @xmath40 in @xmath46 with an up step @xmath62
, we get a path @xmath63 . now raising @xmath52 by two levels ,
we get a path @xmath64 .
thus @xmath65 is a path that ends at level @xmath66 and never goes below level @xmath44 .
the point @xmath67 belongs to both @xmath63 and @xmath64 , but we treat it as a point only in @xmath64 , even if @xmath64 is the empty path . the condition that @xmath68 yields @xmath69 , which implies that the highest point of @xmath70 must belong to @xmath64 .
see figure [ f - pic1 ] below .
now let @xmath41 be the leftmost highest point of @xmath70 ( which is in @xmath64 ) , and let @xmath71 be the step in @xmath70 leading to @xmath41 . then @xmath71 is an up step . by replacing @xmath71 with a down step @xmath72 , and lowering the part of @xmath64 after @xmath41 by two levels
, we get a dyck path @xmath73
. see figure [ f - pic2 ] below . with the following two key observations , it is easy to see that the above procedure gives a bijection from @xmath61 to @xmath60 .
first , @xmath74 in the final dyck path @xmath51 is the rightmost highest point .
second , @xmath38 in the intermediate path @xmath70 is the rightmost point of level @xmath75 in both @xmath70 and @xmath63 .
[ t - main ] for @xmath58 , the number @xmath30 counts pairs of dyck paths @xmath55 of total semilength @xmath18 with @xmath76 .
let @xmath70 be the set of pairs of dyck paths @xmath55 with @xmath68 , and let @xmath77 be the set of pairs of dyck paths @xmath55 with @xmath78 . by symmetry , we see that @xmath79 .
now we claim that the cardinality of @xmath70 is @xmath11 .
this claim follows from lemma [ l - main ] and the fact that if @xmath46 is the empty path , then @xmath59 for every @xmath80 .
clearly we have that @xmath81 , and that @xmath82 is the set of pairs of dyck paths @xmath55 , with @xmath76 .
the theorem then follows from the following computation : @xmath83
in this section we give an algebraic proof of lemma [ l - main ] .
let @xmath84 be the generating function for the catalan numbers , so that @xmath85 then @xmath84 satisfies the functional equation @xmath86 .
let @xmath87 and let @xmath88 be the generating function for dyck paths of height at most @xmath89 . although @xmath88 is a rational function , a formula for @xmath88 in terms of @xmath90 will be of more use to us than the explicit formula for @xmath88 .
[ l - g ] for @xmath91 , @xmath92 let @xmath46 be a path of height at most @xmath93 . if @xmath46 is nonempty then @xmath46 can be factored as @xmath94 , where @xmath50 is an up step , @xmath95 is a dyck path of height at most @xmath96 ( shifted up one unit ) , @xmath51 is a down step , and @xmath97 is a dyck path of height at most @xmath89 .
thus @xmath98 , so @xmath99 . equation clearly holds for @xmath100 and @xmath101 .
now suppose that for some @xmath93 , @xmath102 then the recurrence , together with the formula @xmath103 , gives @xmath104^{-1 } = \left[1-\frac c{1+c}\,\frac{1-c^{k } } { 1-c^{k+1}}\right]^{-1}\\ & = \left[\frac{1-c^{k+2}}{(1+c)(1-c^{k+1})}\right]^{-1 } = ( 1+c)\frac{1-c^{k+1}}{1-c^{k+2}}.\end{aligned}\ ] ] we can prove lemma [ l - main ] by showing that @xmath105 ; this is equivalent to the statement that the number of pairs @xmath55 of dyck paths of semilength @xmath106 with @xmath59 is @xmath11 .
[ t - firstsum ] @xmath107 let @xmath108 thus if @xmath109 then @xmath110 we have @xmath111 and @xmath112 let @xmath113 and @xmath114 then @xmath115 .
we have @xmath116 so @xmath117 , and @xmath118 so @xmath119 therefore @xmath120 thus @xmath121 . by similar reasoning , we could prove theorem [ t - main ] directly : the generating function for pairs of paths with heights differing by at most 1 is @xmath122 where we take @xmath123 , and a calculation like that in the proof of theorem [ t - firstsum ] shows that this is equal to @xmath124 although the fact that the series in theorem [ t - firstsum ] telescopes may seem surprising , we shall see in theorem [ t - general ] that it is a special case of a very general result on sums of generating functions for dyck paths with restricted heights .
[ l - general ] let @xmath125 be a rational function of @xmath126 and @xmath90 of the form @xmath127 where @xmath128 is a polynomial in @xmath126 of degree less than @xmath17 , with coefficients that are rational functions of @xmath90 , and the @xmath129 are distinct positive integers .
let @xmath130 .
then @xmath131 where @xmath132 is a rational function of @xmath90 .
first we show that the lemma holds for @xmath133 . in this case ,
@xmath134 and @xmath135 now we consider the general case .
since @xmath136 is a proper rational function of @xmath126 , it has a partial fraction expansion @xmath137 for some rational functions @xmath138 , so @xmath139 the general theorem then follows by applying the special case to each summand
. [ t - general ] if @xmath140 are distinct integers , then @xmath141 is a rational function of @xmath90 .
apply lemma [ l - general ] to @xmath142 and @xmath143
it is natural to ask whether there are combinatorial interpretations to @xmath16 for @xmath144 similar to theorem [ t - main ] .
it is straightforward ( with the help of a computer algebra system ) to evaluate sums like that count @xmath17-tuples of paths with height restrictions , and we find that sums like that in theorem [ t - general ] involving products of @xmath17 path generating functions may generally be expressed in the form @xmath145 where @xmath146 and @xmath147 are rational functions of @xmath74 .
however , for general @xmath17 we have not found a set of @xmath17-tuples of paths counted by @xmath16 . we have found a set of paths counted by @xmath33 , though it is not as simple as one would like . in the previous section
all our paths had an even number of steps , so it was natural to assign a path with @xmath18 steps the weight @xmath148
. we shall continue to weight paths in this way , even though some of our paths now have odd lengths .
the case @xmath153 is lemma [ l - g ] .
now let @xmath154 be a ballot path counted by @xmath149 , where @xmath155 , so that @xmath154 is of height at most @xmath89 and @xmath154 ends at level @xmath150 .
then @xmath154 can be factored uniquely as @xmath156 , where @xmath157 is a path of height at most @xmath89 that ends at level 0 and @xmath158 is a path from level 1 to level @xmath150 that never goes above level @xmath89 nor below level 1 . using @xmath159
, we obtain @xmath160 and follows by induction .
although we will not need it here , there is a similar formula for the generating function @xmath163 for paths of height at most @xmath89 that start at level @xmath164 , end at level @xmath150 , and never go below level 0 : for @xmath165 we have @xmath166 with @xmath167 for @xmath168 .
it is well known that @xmath163 is @xmath170 times a rational function of @xmath41 , and it is useful to have an explicit formula for it as a quotient of polynomials .
( see sato and cong @xcite and krattenthaler @xcite . )
let us define polynomials @xmath171 by @xmath172 the first few values are @xmath173 these polynomials can be expressed in terms of the chebyshev polynomials of the second kind @xmath174 by @xmath175 it is not difficult to show that @xmath176 and thus we obtain @xmath177 for @xmath178 , and in particular , @xmath179 and @xmath180 we can now describe our combinatorial interpretation of @xmath33 : @xmath33 counts triples of ballot paths whose heights are @xmath89 , @xmath181 , and @xmath182 for some @xmath89 , ending at levels 4 , 3 , and 2 , together with some additional paths of height at most 5 .
( note that if a path of height @xmath182 ends at level 2 , then @xmath89 must be at least 6 . ) more precisely , let @xmath183 be the generating function for ballot paths of height @xmath89 that end at level @xmath150 , so that @xmath184
. then we have : using the method described in lemma [ l - general ] and theorem [ t - general ] , we find , with the help of maple , that the sum @xmath190 is equal to @xmath191 then follows from , the formula @xmath192 , and the formulas for @xmath193 , @xmath194 | it is well known that the numbers @xmath0 are integers , but in general there is no known combinatorial interpretation for them .
when @xmath1 these numbers are the middle binomial coefficients @xmath2 , and when @xmath3 they are twice the catalan numbers
. in this paper , we give combinatorial interpretations for these numbers when @xmath4 or @xmath5 .
partially supported by nsf grant dms-0200596 ] @xmath6= |
elastic and inelastic @xmath0 and @xmath1 @xmath2interactions play an important role in understanding of structure of hadrons .
variety of scattering processes , differentiation them into hard and soft ones , single and double diffraction and central hadron production tell us that this structure is rather complicated .
what is the source of complexity ?
if hard processes are the area of application of current quark consideration then soft ones are described in constituent quark approach .
interplay between hard and soft interactions in high energy experiments strictly relates to interconnection between small and large size quark configurations inside a hadron .
we think that this interplay is a manifestation of fluctuations of hadronic matter distributions inside hadrons and these fluctuations in turn are results of correlated motion of valence quarks inside hadrons .
the question arises : is it possible to construct the dynamical system of quarks which can be observed at one instant of time as constituent ( dressed ) quarks and at the another one - as current ( bare ) quarks ? proposed by author the semiclassical model of strongly correlated quarks , scqm , demonstrated how these configurations could be realized inside hadrons@xcite . in this paper
we give further elaboration of the model whichpossesses the features of both constituent and current quark models and allows one to extract confining potential and force(section 2 ) . in section 3
we apply this model for description of diffractive processes and show that explicit manifestation of quark correlation inside hadrons is a single diffractive scattering .
let us imagine the following hypothetical picture : single , colored quark imbedded into vacuum .
because of vacuum fluctuations one can observe two competing processes : first , polarization of @xmath3 sea by the color field of valence quark ( vq ) and second , the tend of vacuum fluctuations to destroy this polarization . as a result
, one can say about a vacuum pressure on a single , colored quark .
this effect can be interpreted as the response of the vacuum on the presence of , say , point defect or dislocation like in solid state physics .
what happens if we place in vicinity of this quark the corresponding antiquark ? by virtue of opposite color signs
their polarization fields interfere in the overlapping region *
destructively*. so the pressure of vacuum on quark ( antiquark ) from outside exceeds that one going from inner space between quark and antiquark .
this results in an * attractive * force between quark ( dislocation ) and antiquark ( antidislocation ) .
the density of the remaining part of polarization field around quark ( antiquark ) is identified with hadronic matter distribution . at maximum displacement in @xmath4 system , that corresponds to small overlapping of polarization fields ,
hadronic matter distributions have maximum values .
so quark and antiquark located nearby start moving towards each other .
the closer they to one another , the larger destructive interference effect and the smaller hadronic matter distributions are around vqs and the larger their kinetic energies . for such interacting @xmath4 system
the total hamiltonian is @xmath5 were @xmath6 , @xmath7 masses of valence antiquark and quark , @xmath8 their velocity depending on displacement @xmath9 and @xmath10 quark antiquark potential energy with separation @xmath11 it can be rewritten as @xmath12 + \left [ \frac{m_{q}}{(1-\beta^{2})^{1/2}}+u(x)\right ] = h_{\overline{q}}+h_{q},\ ] ] were @xmath13 is potenial energy of quark or antiquark .
therefore , keeping in mind that quark and antiquark are strongly correlated we consider each of them separately as undergoing oscillatory motion in 1 + 1 dimension .
generalization to three
quark system in baryons is performed according to @xmath14 symmetry : an antiquark is replaced by two correspondingly colored quarks to get color singlet baryon and destructive interference takes place between color fields of three valence quarks .
putting aside the mass and charge differences of valence quarks we may say that inside baryon three quarks oscillate along the bisectors of equilateral triangle .
hereinafter we consider that axis @xmath15 is perpendicular to the plane of oscillation @xmath16 .
vq with its polarized surroundings ( hadronic matter distribution ) form constituent quark .
according to our approach potential energy of valence quark , @xmath17 corresponds to the mass @xmath18 of constituent quark : @xmath19 with @xmath20 the knowledge of the mechanism and structure of vacuum polarization around valence quark would give the information about the confining potential .
we can not say at the moment for sure what is the microscopical mechanism of interaction of valence quark with vacuum .
it could be instanton induced interactions , excitation of fractal structure of space time , etc .
so we assume , as a first approximation , that the polarization field can be taken in gaussian form:@xmath21 where exponent is written in quadratic form .
the same is@xmath22for @xmath23 we define the mass of constituent quark at maximum displacement @xmath24 where @xmath25 and @xmath26 are masses delta isobar and nucleon correspondingly .
therefore , the parameters of the model are the masses of vqs,@xmath27 which are chosen to be @xmath28 @xmath29 , maximum displacement , @xmath30 and parameters of gaussian function , @xmath31 as shown below , @xmath32 and @xmath33 are adjusted by comparison of calculated and experimental values of inelastic cross sections , @xmath34 for @xmath0 and @xmath35 collisions . using ( 3)(5 )
we can calculate the confining potential @xmath36 and force @xmath37 .
they are shown in fig .
as one can see the confining potential is essentially nonlinear .
the behavior of potential evidently demonstrates the relationship between constituent and current quark states inside a hadron . at maximum displacement quark
is nonrelativistic , constituent one ( vq surrounded by polarized sea ) , since according to ( 3 ) the confining potential corresponds to the mass of constituent quark . at the origin of oscillation ,
@xmath38 antiquark quark in mesons and 3 quarks in baryons , being close to each other , have maximum kinetic energy and correspondingly minimum potential energy and mass : they are relativistic , current quarks ( bare vqs ) .
intermediate region corresponds to increasing ( decreasing ) of quark s mass by dressing ( undressing ) of quarks due to vacuum polarization .
this mechanism agrees with local gauge invariance principle .
indeed , phase rotation of wave function of single quark in color space @xmath39 on angle @xmath40 depending on displacement @xmath9 of the quark in coordinate space @xmath41 results in it s dressing ( undressing ) by quark and gluon condensate that corresponds to the transformation of gauge field @xmath42 @xmath43 here we dropped color indices and took into account only scalar component , @xmath44 , of gauge field .
thus gauge transformation maps internal ( isotopic ) space of colored quark onto coordinate space . on
the other hand this dynamical picture of vq dressing ( undressing ) corresponds to chiral symmetry breaking ( restoration ) .
the behaviour of field @xmath44 and hadronic matter distribution,@xmath45 , for quark antiquark system during their oscillations is shown in fig .
2 . , ( eq . ( 5 ) ) , and hadronic matter distribution @xmath45 , ( eq .
( 4 ) ) , in quark antiquark system during one half of the period of oscillations ; @xmath46 distance in _ fermi _ between quark and antiquark depicted as dots.,width=451 ] due to this mechanism of vqs oscillations nucleon runs over the states corresponding to the certain terms of the infinite series of fock space @xmath47 confining force drastically differs from the one given by string models .
when vqs are close each other it is very weak and fulfills the asymptotic freedom
behavior of quarks of qcd . at larger distances between vqs
it starts growing rapidly , then reaching maximum value goes down , asymptotically approaching zero .
thus at large distances inside hadrons quarks being in a constituent state are almost free .
hence , it is clear why additive quark model , where quarks are treated as massive , almost unbound and extended objects , works well .
we must emphasize that interaction between vqs is not direct but a result of polarization of surrounding vacuum combined with destructive interference .
attractive force between vqs in ground state hadrons does not appear as gluon string but goes from vacuum suppression that predominates from outside .
therefore , our approach reflects the features of bag models , as well .
the model is in agreement with the experiments ( fig .
3 ) for description of vq structure function inside a nucleon @xmath48 .\ ] ] of cause , our description is classical and we must take into account quantum corrections
. it will be the subject of forthcoming papers .
nevertheless , classical consideration of vq oscillations is justified by e. schrodinger s paper@xcite where he , analyzing the motion of wave packet solution of time dependent schrodinger equation for harmonic oscillator , demonstrated that this wave packet moves in exactly the same way as corresponding classical oscillator . in our model
vq with its surroundings can be treated as ( nonlinear ) wave packet . in forthcoming paper
we ll show that these wave packets possess soliton like features .
different configurations of quark contents in colliding hadrons realized at the instant of collision result in different types of reactions .
the probability of finding any quark configuration inside a hadron is defined by the probability of vq s displacement in proper frame of a hadron : @xmath49 with @xmath50 configurations with nonrelativistic constituent quarks ( @xmath51 ) in both colliding hadrons lead to multiparticle production in central and fragmentation regions .
hard scattering with jet production and large angle elastic scattering take place when configurations with current vqs ( @xmath52 ) in both colliding hadrons are realized .
intermediate configurations inside one ( both ) of colliding hadron are responsible for single ( double ) diffraction processes and semihard scattering . using impact parameter representation , namely inelastic overlap function ( iof )
, we can calculate total , inelastic , elastic and single diffractive cross sections for @xmath0 and @xmath1 collisions . in impact parameter representation iof
can be specified via the unitarity equation @xmath53 where @xmath54elastic scattering amplitude and @xmath55 is iof .
iof is connected with inelastic differential cross sections in impact parameter space : @xmath56 and @xmath35 collisions at @xmath57 and @xmath58 @xmath59 ; triangles are henzi and valin approsimation@xcite.,width=215 ] then inelastic , elastic and total cross sections can be expressed via iof as @xmath60@xmath61 ^{2}% d^{2}\mathbf{b},\]]@xmath62 d^{2}\mathbf{b}.\ ] ] and @xmath35 collisions ; data points are compilations of experimental data taken from electronic data base hepdata@xcite.,width=451 ] since iof relates to the probability of inelastic interaction at given impact parameter , ( 12 ) , we carried out monte carlo simulation of inelastic nucleon nucleon interactions .
inelastic interaction takes place at definite impact parameter @xmath63 if at least one pion is produced in the region where hadronic matter distributions of colliding protons overlap @xmath64 where indices @xmath65 and @xmath66 refer to quarks from different nucleons and @xmath67 , @xmath68 , @xmath69 masses of hadronic matter composed in constituent quarks @xmath65 and @xmath66 , @xmath70 , @xmath71 their @xmath72-factors ; intergrand expression is convolution of hadronic matter density distributions of quarks @xmath65 and @xmath73 this condition corresponds hisenberg picture@xcite with modified right hand side : @xmath74 in the original heisenberg inequality is replaced by @xmath75 it is justified by the fact that the average transverse momentum of produced particles increases with energy . specifying the quark configurations in each colliding nucleons by ( 9 ) and ( 3)(5 ) we calculate @xmath55 for certain values of impact parameter @xmath63 and then cross sections @xmath76 and @xmath77 the values of adjusted parameters of the model , @xmath78 @xmath79 @xmath80 @xmath81 and @xmath82 @xmath79 are chosen by comparison of calculated iof with so called `` bel''parametrization@xcite , ( fig .
4 ) , and calculated inelastic cross sections with experimental ones , @xmath34 in @xmath0 and @xmath35 collisions at @xmath83 @xmath84 fig . 5 shows the result of calculation for @xmath0 total cross section at wide range of collision energies .
one can see that the model with parameters fixed at one energy ( @xmath58 @xmath85 describes the energetic behavior of @xmath77 the growth of cross sections with energy is due to the continuous tails of polarization fields around vqs not compensated by destructive interference with fields of two other vqs . with rising collision energy
these tails result in the increasing effective size of hadronic matter distribution inside nucleons and correspondingly the increasing radius of interactions .
the model gives for cross sections linear logarithmic energy dependence and the curve deviates from data points at very high energies .
the reason of this is that in our geometrical approach , ( ( 12 ) ( 16 ) ) , we did nt take into account the real part of scattering amplitude . at energies
@xmath86 @xmath59 calculated cross sections were corrected on contributions of regge poles exchange by using donnachie and landshoff parametrization@xcite:@xmath87 processes .
extended sizes of colliding nucleons are drawn on impact parameter plane : big grey circles are constituent quarks of dissociating nucleon , small black circles
( semi)bare quarks of the another nucleon.,width=384 ] an oscillatory motion of vqs appearing as interplay between constituent and bare ( current ) quark configurations results in fluctuations of hadronic matter distribution inside colliding nucleons .
the manifestation of these fluctuations is a variety of scattering processes , hard and soft , in particular , the process of single diffraction ( sd ) .
we select sd
events among inelastic @xmath88 events with the criterion @xmath89 , where @xmath90 ( fig . 6 ) . here @xmath91 @xmath92 @xmath93 are momenta of quarks forming the final state proton .
as one can see , sd events correspond to constituent quark configuration inside one colliding nucleon and ( semi)bare quark configuration inside another one .
7 shows that calculated sd cross section slightly depends on energy ; our calculation for lhc energy gives @xmath94
@xmath95 and @xmath35 collisions ; data points are compilations of experimental data taken from electronic data base hepdata@xcite.,width=451 ]
proposed dynamical model of hadron structure , scqm , possesses some important features .
parameters of the model , maximal displacement of vq , @xmath96 , and extension of quark and gluon condensate around vqs , @xmath97 and @xmath98 characterize extended sizes of hadron .
owing to noncompensated tails of vacuum condensates at outer sides of vqs and condition ( 18 ) , the model describes the energy dependence of inelastic and total cross sections as a result of increasing effective sizes of colliding hadrons . on the other hand
, there is no overlap of hadronic matter distributions in space between vqs of each interacting hadron even at very high energy collisions because destructive interference of their polarization fields reduces resulting hadronic matter at the center of quark system to zero value .
so , the model keeps unitarity and does not need of inclusion of such a questionable effect as antishadowing
according to our model diffraction dissociation is not shadowing effect of nondiffractive inelastic process both compose inelastic process , they merely differ by configurations of quarks inside interacting hadrons .
because of plane oscillations of vqs and flattened ( perpendicular to the plane of oscillation ) form of hadronic matter distributions nucleons are deformed , non - spherical objects . in paper @xcite , assuming that spin of vqs are perpendicular to the plane of oscillations , we showed that such a deformation of proton could manifest itself in total cross section differences between longitudinal and transversal polarization states in proton proton collisions .
apparently experimental evidence of such a deformation are quadrupole transition amplitudes in @xmath99 electroproduction which are sensitive to quadrupole deformation of the nucleon@xcite .
furthermore hadronic matter distribution inside hadrons and in turn the sizes of hadrons are fluctuating quantities
. fluctuations of extended sizes of nucleons allows one to understand why black disk
limit is not saturated in @xmath0 and @xmath35 collisions up to very high energies .
we must mention the paper@xcite of s. barshay with co - authors where they incorporated fluctuations in the eikonal into geometric picture for explaining of diffractive processes .
there is a connection between eikonal and inelastic overlap function used in our calculations of cross sections through the expression @xmath100 were @xmath101 is eikonal .
it is obvious that fluctuations of hadronic matter distributions caused by interplay between constituent and current quark configurations inside colliding hadrons lead to fluctuation of eikonal . in their approach
the fluctuations of eikonal were controlled by phenomenological parameter , depending on energy .
they predicted that single diffraction cross section increases reaching 1415 mb at cms energies 20@xmath10240 tev then starts to decrease .
contrary to this prediction there is no apparent reason for any maximum , according to our approach , and this cross section increases steadily , very slightly up to asymptotic energies .
our unified geometrical explanation of diffractive and nondiffractive processes could give an answer on long standing question : what is pomeron ?
historically the concept of `` pomeron '' , starting from simple regge pole with intercept @xmath103 , transformed to a rather complicated object with relatively arbitrary features and smooth meaning . to produce rising cross sections
it must have intercept such that @xmath104 .
the fact that the parameter @xmath105 is universal , independent of particles being scattered in hadronic and dis interactions , could say us that the nature of the cross section growth is the same for all processes .
our interpretation of pomeron is geometrical one .
both diffractive and nondiffractive particle production emerge from disturbance ( excitation ) of overlapped continuous vacuum polarization fields ( gluon and @xmath3 condensate ) around valence quarks of colliding hadrons followed by fragmentation process .
the type of interaction depends on quark configurations inside both colliding hadrons occurring at the instant of interaction and the value of impact parameter .
so , what we used to call `` pomeron '' in @xmath106 channel is solely continuum states in @xmath107 channel and we claim that pomeron is unique in elastic , inelastic ( diffractive and nondiffractive ) and dis
. 99 g. musulmanbekov , _ proc . of xvii int .
kazimierz meeting on particle theory and phenomenology , iowa , 1995 _ , world scientific , 1996 , p. 347 - 353 . ; _ proc . of xxvi int.symp .
on multiparticle dynamics , sept . 1 - 5 , faro , 1996 _ , world scientific , 1997 , p. 357363 ; nucl .
suppl . * b71*(1999 ) 117 .
benvenuti et al . , phys .
lett . , * 237b * ( 1990 ) 599 .
a.donnachie and p.v .
landshoff , cern th 6635/92 .
troshin and n.e .
tyurin , hep - ph/9810495 .
g.musulmanbekov , proc .
vii workshop on high energy spin phys .
july 712 1997 , dubna , dubna 1997 , p. 165 - 170 .
a.feassler , prog .
* 44 * ( 2000 ) 197 .
s.barshay , p. heiliger , d. rein , z. phys . , * 56c * ( 1992 ) 77 . | hadron structure is considered in the frame of strongly correlated quark model ( scqm ) .
it is shown that quark correlations result in fluctuations of hadronic matter distributions and single diffractive dissociation processes in hadronic collisions , hard or soft , are manifestation of these fluctuations inside colliding hadrons . |
the study of proton - nucleus @xmath1 collisions is important because they are tractable intermediaries between @xmath2 and @xmath3 collisions , when intense interest exists in discovering the extent to which the dense medium created in an @xmath3 collision differ from that of linear superpositions of @xmath2 collisions .
one of the properties of @xmath0 collisions is that the momentum of the leading baryon in the projectile fragmentation region is degraded , a phenomenon commonly referred to , somewhat inappropriately , as baryon stopping .
such a transference of baryon number from the fragmentation to the central region contributes to the increase of matter density at mid - rapidity , thereby raising the likelihood of the formation of quark - gluon plasma .
thus it is important to understand the process of baryon momentum degradation and its dependence on nuclear size or centrality .
since it is not feasible to perform first - principle parameter - free calculations of the momentum degradation at this point , experimental guidance is of crucial importance .
recently , several experiments have produced useful data on the subject , in particular , e910 and e941 at the ags , and na49 at the sps @xcite .
it is the @xmath4 dependences of the distributions of @xmath5 and @xmath6 that we shall focus on ; moreover , their dependences on centrality will guide us in our determination of the nuclear effect on baryon momentum degradation .
the degradation of baryon momentum has been studied in various approaches before @xcite-@xcite . in refs .
@xcite-@xcite the investigations are done at the nucleon level , while in refs .
@xcite-@xcite the string model is the basis . since it is questionable that the concept of color strings can be relevant in heavy - ion collisions where the abundance of color charges in the overlap region renders unlikely the development of constricted color flux tubes @xcite , our approach in this paper will be on the various levels of the constituents of the nucleon that are consistent with the parton model .
more specifically , we shall use the valon model @xcite-@xcite to keep track of the momenta of the constituents and the recombination model @xcite to describe the hadronization of the partons .
the valons play a role in the collision problem as the constituent quarks do in the bound - state problem .
thus a nucleon has three valons which carry all the momentum of the nucleon , while each valon has one valence quark and its own sea quarks and gluons .
although soft processes are non - perturbative , the valon model ( including recombination ) nevertheless provides a systematic way of calculating all subprocesses that contribute to a particular inclusive process .
some of the subprocesses can be identified with certain diagrams in other approaches , e.g. , baryon junction and diquark breaking terms @xcite .
the valon distributions in the proton will be determined by fitting the parton distributions at low @xmath7 .
the effect of the nuclear medium on the valon distribution of the proton projectile will involve two parameters , one characterizing the momentum degradation and the other flavor flipping .
the color indices are all averaged over , since multiple gluon interactions , as the projectile traverses the target nucleus , are numerous and uncomputable .
their effects are , however , quantified in terms of the number , @xmath8 , of target nucleons that participate in the @xmath0 collision .
for that reason it is important that the experimental data must have centrality selection expressible in terms of the average @xmath9 . the inclusive process @xmath10 has been studied before in the valon model @xcite .
however , the pertinent data available at that time had no centrality cuts and were for fixed @xmath11 @xcite .
now , na49 has @xmath12-@xmath13 data on the production of @xmath14 and @xmath15 in the proton fragmentation region ( with contribution from the nuclear target subtracted ) for various values of @xmath9 @xcite . to extract more information from these refined data
we have improved our formulation of the valon model , allowing for momentum degradation of each of the valons independently and for changes in their flavors .
the effect on the nucleon momentum distributions can be calculated and the effective degradation length deduced . in sec .
ii we determine the valon and parton distribution functions from the published distribution functions at low @xmath7 . the valon model for @xmath0 collisions
is then discussed in sec .
iii , in which momentum degradation is formulated . in sec .
iv we consider in detail the projectile fragmentation process and the subsequent process of quark recombination so that the inclusive distribution of nucleons can be calculated .
the net proton and neutron distributions are then calculated and compared to the experimental data in sec .
v. predictions for pion production in the proton fragmentation region are made in sec .
the conclusion is given in sec .
the distribution functions of valons and partons have been considered in refs .
@xcite . they were , however , determined by fitting the muon and neutrino scattering data of the late 70s .
we now have modern parton distributions from various groups at various values of @xmath7 .
it is therefore appropriate for us to revisit the problem and determine the valon distribution functions in light of the new parton distribution functions . in the valon model a proton
is considered to consist of three valons ( @xmath16 ) , which have the same flavors as the valence quarks ( @xmath17 ) that they individually contain .
thus a valon may be regarded as a parton cluster whose structure can be probed at high @xmath7 , but the structure of a nucleon itself in a low-@xmath11 scattering problem is described in terms of the valons . as in the parton model we work in a high - momentum frame so that it is sensible to use the momentum fractions of the constituents . reserving @xmath18 for the momentum fraction of a quark , we use @xmath19 to denote the momentum fraction of a valon . in this paper @xmath19 never denotes rapidity .
let the exclusive valon distribution function be @xmath20 where @xmath21 and @xmath22 refer to the @xmath23 valons and @xmath24 the @xmath25 valon .
the delta function ensures that the three valons exhaust the momentum of the proton .
the exponents @xmath26 and @xmath27 will be determined by the parton distribution functions .
the normalization factor @xmath28 is determined by requiring that the probability of finding these three valons in a proton be one , i. e. @xmath29 thus we have @xmath30^{-1 } , \label{2.3}\end{aligned}\ ] ] where @xmath31 is the beta function .
the single valon distributions are obtained by appropriate integrations @xmath32 @xmath33 the two - valon distributions are trivial because of the @xmath34-function in eq .
( [ 2.1 ] ) . in a deep inelastic scattering the structure of the proton is probed to reveal the parton distributions , which in the valon model are convolutions of @xmath35 with the evolution functions that describe the valon structure .
the latter have two varieties : @xmath36 for the favored partons and @xmath37 for the unfavored partons .
that is , @xmath38 is favored in @xmath23 and @xmath39 is favored in @xmath25 , but they are unfavored in @xmath25 and @xmath23 , respectively . at high @xmath7 we may ignore the influence of the spectator valons when one valon is probed ( the usual impulse approximation ) , so we can write for the @xmath38 and @xmath39 quark distribution functions as @xmath40 , \label{2.6}\end{aligned}\ ] ] @xmath41 .
\label{2.7}\end{aligned}\ ] ] it should be emphasized that @xmath42 , @xmath43 and @xmath44 are noninvariant distributions defined in the phase space @xmath45 and @xmath46 , while @xmath47 and @xmath48 are invariant distributions defined in the phase space @xmath49 .
the favored distribution @xmath50 has two parts , valence and sea , while the unfavored distribution @xmath51 has only sea .
the valence part is also referred to as the non - singlet component , @xmath52 , so we may write @xmath53 at high @xmath7 both @xmath54 and @xmath55 can separately be determined in perturbative qcd .
however , since low-@xmath7 parton distributions are now available by extrapolation , and especially since we shall apply the valon model to low-@xmath11 processes ( and therefore at low @xmath7 ) , we use phenomenological forms for @xmath54 and @xmath55 at low @xmath7 with parameters to be determined by the low-@xmath7 parton distributions .
we adopt the forms @xmath56 @xmath57 where eq .
( [ 2.9 ] ) satisfies the requirement that there is only one valence quark in a valon , i. e. @xmath58 eq .
( [ 2.10 ] ) has the usual sea - quark distribution , also used previously @xcite .
the parton distributions that we use to fit are the ones determined by the cteq collaboration @xcite .
in particular , they have the distributions at low @xmath7 , labeled cteq4lq , posted on the web @xcite .
we choose the one at @xmath59 evolved from @xmath60 .
we fit the @xmath61 and @xmath62 distribution functions using eqs .
( [ 2.4])-([2.10 ] ) by varying @xmath63 and @xmath64 .
the results are shown in fig .
1 , where the cteq functions are in solid lines , and our fitted curves are in dotted lines .
we have attempted to fit the @xmath38-quark distribution as perfectly as possible ; the fit of the @xmath39-quark distribution turns out to be good only at high @xmath18 . missing
the normalization of @xmath62 at low @xmath18 is a blemish , but is acceptable since we shall use the valence quark distributions mainly in the large-@xmath18 region . besides
, the reliability of an extrapolation of high-@xmath7 deep inelastic scattering data to low @xmath7 can always be called into question .
the parameters of the fit are @xmath65 @xmath66 the values of @xmath26 and @xmath27 are not very different from the ones determined in ref .
@xcite based on muon - scattering data at @xmath67 analyzed by duke and roberts @xcite ; there we had @xmath68 and @xmath69 .
as can be seen from eqs .
( [ 2.4 ] ) and ( [ 2.5 ] ) , our present result means that @xmath70 and @xmath71 at large @xmath19 , as opposed to @xmath72 and @xmath73 , respectively , for the previous result .
the values of @xmath74 and @xmath75 in eq .
( [ 2.13 ] ) suggest that @xmath52 is highly peaked near @xmath76 , according to eq .
( [ 2.9 ] ) .
that is as it should be , since @xmath7 is only about twice the value of @xmath77 and thus not much evolution .
if there were no evolution , i. e. , @xmath78 , then @xmath79 would be @xmath80 and @xmath81 , a situation indicating no probing of the valons .
what we have at @xmath82 gives only a modest degree of resolution of the valon structure , and @xmath83 is changed from @xmath80 to @xmath84 in the large @xmath19 region , while @xmath85 becomes non - vanishing .
these are the valence and sea quark distributions that we shall use for nucleon production at low @xmath11 in the following .
the valon model for inclusive reactions is basically a @xmath86-channel description of particle production in contradistinction from the regge - gribov approach @xcite which is essentially based on cutting @xmath87-channel exchanges in elastic amplitudes .
the two approaches are roughly complementary in that the former can best describe the fragmentation region , while the latter is more suitable for the central region .
it has been known for a long time that the pion inclusive cross section in the proton fragmentation region has a @xmath4 dependence that is very similar to the quark distribution in the proton @xcite .
that similarity has been reconfirmed more recently in @xmath0 collisions by the e910 experiment @xcite even at the relatively low energy of @xmath88 .
it suggests that the proton structure is highly relevant to the spectra of particles produced in the fragmentation region .
the connection between the quark distribution and the pion inclusive distribution was first put on a calculable basis by the recombination model @xcite , which was subsequently improved in the framework of the valon model @xcite .
the idea of recombination as a basis for hadronization in hadronic collisions is eminently reasonable , since for a pion to be detected at @xmath89 , say , it is far less costly to have a quark at @xmath90 coalescing with an antiquark at @xmath91 , each @xmath92 , to form a pion at @xmath93 , as compared to the fragmentation process for which a quark or diquark must first have @xmath94 . in quark jets ,
fragmentation is reasonable because hadronization is initiated by a leading quark , but it is less persuasive when adapted to hadronic fragmentation unless the projectile momentum resides entirely in a quark or diquark @xcite , contrary to ochs observation of the relevance of the canonical quark distributions .
the @xmath86-channel treatment of relating quark distributions to inclusive hadron distributions in @xmath2 collisions essentially regards the effect of the opposite - going initial proton as unimportant , an approximation that can only be justified in the fragmentation regions due to short - range correlation in rapidity .
thus the valon model , as it has been developed up to now , is not expected to be applicable to the central regions , where the interaction between the two incident particles is of paramount importance . for @xmath0 collisions
even the factorization of the fragmentation properties is not entirely valid , since the hadron distributions are known to depend on centrality or target size .
this is the problem that we shall treat in this paper in the framework of the valon model .
more specifically , we shall consider the problem of momentum degradation of the valons due to the nuclear medium .
recombination occurs outside the nucleus and is therefore unaffected .
the na49 data on @xmath12-@xmath13 collisions are presented in terms of two mean values , @xmath9 , of the number of participating nucleons in the target : @xmath95 for central collisions and @xmath96 for non - central collisions @xcite .
we assume a poissonian fluctuation from that mean with the distribution @xmath97 which is normalized by @xmath98 where we have excluded the @xmath99 term , since it is necessary for @xmath100 in order to have a collision .
thus @xmath8 is the number of nucleons in the nucleus that suffer inelastic collisions in any given event , @xmath8 being an integer .
that is the counting on the target side , while on the projectile side we count in terms of the valons .
let the @xmath101th valon encounter @xmath102 collisions . in general
, the total valonic collisions is bounded by @xmath103 the upper bound occurs only when all three valons participate in each of the struck nucleon , while the lower limit is for only one valon per struck nucleon . in the appendix
we shall show that the data favor the lower bounds , so for simplicity we proceed in the following with the assumption @xmath104 it is useful to interpret this in the @xmath87-channel picture .
the incident proton is represented by three constituents , each of which exchanges @xmath102 ladders with @xmath102 target nucleons so that the overall diagram is highly non - planar .
cutting the ladders gives rise to the particles produced in the @xmath86 channel , mostly in the central region .
focusing on the evolution of the valons , we first assume that the three valons interact with the target independently , since they are loosely bound to form the proton , just as the nucleons are in a deuteron .
if we denote the degradation effect of the nucleus on the @xmath101th valon by @xmath105 , then we can write the evolution equation on the valon distribution as @xmath106 where @xmath107 is the valon distribution function after @xmath108 interactions with the nucleus .
as in eqs .
( [ 2.6 ] ) and ( [ 2.7 ] ) , @xmath109 and @xmath107 are noninvariant distributions defined in the phase space @xmath110 , etc . , while @xmath111 is an invariant distribution defined in @xmath112 .
we postpone our discussion of what @xmath111 is until sec .
v. for an event with @xmath8 collisions the evolved valon distribution function @xmath113 with @xmath8 partitioned as in eq .
( [ 3.4 ] ) is given by the multinomial formula @xmath114}{\nu !
\over \nu_1 ! \nu_2
! \nu_3 ! }
g'(y'_1 , y'_2 , y'_3;\nu_1 , \nu_2 , \nu_3 ) , \label{3.6}\end{aligned}\ ] ] where @xmath115 $ ] implies that the summation is over @xmath116 , and @xmath117 subject to the constraint of eq .
( [ 3.4 ] ) .
note that if the nucleus had no effect on the valons , i.e. , @xmath118 independent of @xmath102 , then @xmath119 becomes @xmath120 in eq .
( [ 3.5 ] ) and so does @xmath121 in eq.([3.6 ] ) , as it should . to relate to the experimental @xmath9 , we have @xmath122 where @xmath123 is given in eq .
( [ 3.1 ] ) .
it is useful to introduce the notion of an effective nucleon after @xmath8 collisions by giving it a momentum fraction @xmath124 and defining the probability of finding it at @xmath124 by @xmath125 conservation of baryon number requires that @xmath126 the possibility that flavor can change is a secondary issue that will be discussed later ; here , the issue is that the baryon should be somewhere in the interval @xmath127 . the second half of eq .
( [ 3.10 ] ) puts a constraint on @xmath128 , since we can obtain from eqs .
( [ 3.5 ] ) and ( [ 2.2 ] ) @xmath129 the likelihood that the three evolved valons will in reality reconstitute a nucleon is extremely low , but the fictitious nucleon that they form carries a baryon number that is conserved , and a momentum that is not conserved .
indeed , we expect the average @xmath124 to decrease with @xmath8 , i. e. , @xmath130 that is commonly referred to as stopping . in sec .
v we shall infer from the data what the stopping power is . even without stopping , such as in @xmath2 collisions ,
it does not mean that the real proton produced can not have @xmath131 .
it is known that in @xmath2 collisions the proton inclusive cross section @xmath132 is nearly flat in @xmath4 .
stopping goes on top of that distribution , making it roughly exponential decrease in @xmath4 . how to proceed from the valon distribution @xmath107 to
the detected proton distribution @xmath133 is the subject of the next section .
the convolution equation ( [ 3.5 ] ) can be simplified when expressed in terms of the moments on account of the convolution theorem .
thus let us define @xmath134 @xmath135 g(y_1 , y_2 , y_3 ) \label{3.14}\end{aligned}\ ] ] and similarly for @xmath136 and @xmath137 .
it then follows from eqs .
( [ 3.5 ] ) and ( [ 3.6 ] ) that @xmath138 } { \nu
! \over \nu_1 ! , \nu_2 ! , \nu_3 ! } \ , \tilde{g}(n_1 , n_2 , n_3 ) \prod ^3_{i=1 } \tilde{d}(n_i,\nu_i ) .
\label{3.15}\end{aligned}\ ] ] now , @xmath139 itself can be described by a convolution equation @xcite .
if instead of the discrete @xmath102 we use a continuous variable @xmath55 that denotes the length of the nuclear medium a valon traverses , we can express the change on @xmath140 for an incremental distance @xmath141 in the form @xcite @xmath142 with some reasonable kernel @xmath143 . in terms of the moments with @xmath144 eq .
( [ 3.16 ] ) becomes @xmath145 whose solution is @xmath146 .
\label{3.19}\end{aligned}\ ] ] the constraint ( [ 3.11 ] ) implies @xmath147 since @xmath102 is proportional to @xmath55 , let us now revert @xmath148 to @xmath149 and write eq.([3.19 ] ) as @xmath150 where @xmath151 is trivially related to @xmath152 with a power exponent whose detail need not be specified here . from eq .
( [ 3.20 ] ) follows @xmath153 we now can use eq .
( [ 3.21 ] ) in ( [ 3.15 ] ) and obtain @xmath154^{\nu } .
\label{3.23}\end{aligned}\ ] ] what we have derived here is that the dependence on @xmath8 is in the exponent , implying that the effects of the successive collisions with the nucleons in the target nucleus are multiplicative , as is reasonable .
we can now go back to eq .
( [ 3.12 ] ) and calculate the average @xmath155 after @xmath8 collisions . using eq . (
[ 3.9 ] ) we get @xmath156 \left\{[2 + d(2 ) ] /3\right\}^{\nu } \label{3.24}\end{aligned}\ ] ] with the help of eqs .
( [ 3.22 ] ) and ( [ 3.23 ] ) .
the first factor in the square brackets is just 1 , since it is @xmath157 which is guaranteed by the @xmath34-function in eq.([2.1 ] ) .
hence , we have @xmath158/3 < 1 .
\label{3.26}\end{aligned}\ ] ] averaging over @xmath123 , defined in eq.([3.1 ] ) , yields @xmath159 this is the average momentum fraction of the effective nucleon after @xmath9 collisions but before fragmentation into final - state particles in the fragmentation region .
we now consider the problem of how a projectile proton fragments and how the quarks recombine to form the detected nucleon , thereby specifying how the inclusive distribution can be calculated . to give an overview of the procedure ,
let us summarize the two steps above by the two invariant distributions : @xmath160 , the probability of finding a @xmath38 quark at @xmath90 , another @xmath38 quark at @xmath91 , and a @xmath39 quark at @xmath161 , and @xmath162 , the recombination function , which specifies the probability that those three quarks coalesce to form a proton at @xmath18 .
how these distributions are related to the valon distributions will be discussed later .
but first we state that the invariant distribution function for the detection of a proton at @xmath18 is @xmath163 where the normalization factor @xmath164 will be given below .
( [ 4.1 ] ) is the essence of the recombination model @xcite-@xcite . for meson production ,
only the distributions for @xmath165 and @xmath166 need be considered and eq .
( [ 4.1 ] ) can be simplified accordingly . as in all distributions considered in this paper ,
color and spin components are averaged over in the initial state and summed in the final state so that we are not concerned explicitly with such degrees of freedom .
flavor , however , is different , since we identify the final - state particles by their flavors ; that problem will be treated presently . before proceeding we
emphasize what has been mentioned in the preceding section already , namely : eq .
( [ 4.1 ] ) is expected to be valid in the proton fragmentation region only , if @xmath160 is to be determined from the projectile valon distributions with no quarks originating from the target nucleons . in this @xmath86-channel approach
, the factorization of the projectile and target fragmentations , apart from the momentum - degradation effect studied in the previous section , can be justified only if the two fragmentation regions are well separated . at ags that is not the case .
even at sps the central region in rapidity can encompass sizable portions of the positive and negative @xmath4 variables .
the application of the valon model to the analysis of the data therefore needs some help from the experiments .
fortunately , the na49 collaboration has treated their data in such a way as to eliminate the contribution of the target fragmentation from the projectile fragmentation region . from their data on
the net proton produced , @xmath167 , for which we use the abbreviated notation @xmath168 , they subtract the distribution for @xmath169 , which is @xmath170 $ ] . by charge conjugation symmetry
, @xmath169 should have no projectile fragmentation , only target fragmentation .
thus the difference @xmath171 should have no target fragmentation @xcite . with those data as our goal for analysis , eq .
( [ 4.1 ] ) is then particularly suitable .
the normalization factor @xmath164 in eq .
( [ 4.1 ] ) is determined by requiring that the proton distribution in the absence of any target , @xmath172 , satisfies the sum rules @xmath173 which follow from the condition that the number of proton and its momentum fraction be 1 . without any collision
the quarks are identified with the unresolved valons , so @xmath174 becomes @xmath175 the recombination function is the time - reversed form of the valon distribution , so @xmath176 putting these in eq .
( [ 4.1 ] ) , we obtain in view of ( [ 2.1 ] ) @xmath177 because of the presence of @xmath178 , the two integrals in eq .
( [ 4.2 ] ) are identical , and we get @xmath179 the factor @xmath180 , although known from eq .
( [ 2.3 ] ) , will cancel the similar factor that will emerge from the integral in the numerator of eq .
( [ 4.1 ] ) , just as they appear explicitly in eq .
( [ 4.5 ] ) .
the identification of the recombination function with the invariant form of the valon distribution in eq .
( [ 4.4 ] ) is the principle characteristic of the valon model . on the one hand
, it recognizes the role of the wave function of the proton both in a projectile and in a produced proton . on the other hand
, the momentum fractions @xmath181 of the outgoing valons can add up to a proton at @xmath18 , so there is no need for any constituent in the process to have a momentum fraction greater than @xmath18 , as would be necessary in a quark fragmentation model .
one may then question how in a collision process can the quarks at @xmath182 , and @xmath161 in eq .
( [ 4.1 ] ) become the valons of the outgoing proton .
the answer is that hadronization occurs outside the target , and that the quarks moving downstream dress up themselves and become the valons of the produced proton without any change in the net momentum of each quark / valon , which is all that matters in the specification of @xmath183 .
we now consider the quark distribution @xmath184 in eq .
( [ 4.1 ] ) in @xmath0 collision before recombination . in the preceding section
we have formulated the procedure to calculate the effect of the nuclear medium on the momenta of the valons as they traverse the target .
momentum degradation is , however , only one of the effects of valon - nucleon interaction .
if such an interaction is represented by a regge exchange , we should also consider the possibility of flavor changes of the valons due to non - vacuum exchanges at non - asymptotic energies . in the spirit of the @xmath86-channel approach that we have taken , in which the probabilities of occurrences at various stages are assembled multiplicatively
, we assume that the flavor changes at each of the @xmath102 collisions are incoherent so that the net probability of a flavor change after @xmath185 interactions is also multiplicative .
let @xmath165 be the probability of a flavor change from @xmath23 to @xmath25 , or from @xmath25 to @xmath23 , at one of the @xmath102 interactions .
furthermore , let @xmath186 be the probability of flavor change after @xmath102 interactions
. then @xmath186 satisfies the recursion relation @xmath187 where the first term on the rhs denotes no change in the last step from @xmath102 to @xmath188 , while the second term denotes a change in the last step .
the solution of eq .
( [ 4.7 ] ) is @xmath189 .
\label{4.8}\end{aligned}\ ] ] we may now write what a @xmath23 and a @xmath25 valon become after @xmath102 interactions in obvious notation : @xmath190 @xmath191 where @xmath192 .
this regeneration process depends on one parameter @xmath165 .
we expect @xmath165 to decrease with energy @xmath193 . here
we treat it as one free parameter to fit the na49 data at one energy @xmath194 . for the quark distribution in a valon we have the favored and unfavored types discussed in sec .
ii , and denoted by @xmath47 and @xmath48 , respectively .
we drop the @xmath7 dependence , since we now consider low-@xmath11 hadronic processes for which there is no precise @xmath7 .
nevertheless , we shall use the parameterization in eqs .
( [ 2.9 ] ) , ( [ 2.10 ] ) and ( [ 2.13 ] ) , from the @xmath59 cteq parton distributions .
we use the following notation for the invariant distributions of quarks in valons with superscript @xmath195 signifying ` favored ' and @xmath38 ` unfavored ' : @xmath196 : favored quark at @xmath197 in valon at @xmath198 , @xmath199 : two favored quarks at @xmath197 and @xmath200 in valon at @xmath198 .
other distributions involving unfavored quarks are similarly defined .
examples of @xmath196 , @xmath201 , @xmath199 , and @xmath202 are depicted by diagrams in fig .
2 . in view of eq .
( [ 4.9 ] ) and eq .
( [ 4.10 ] ) we can write by definition @xmath203 @xmath204 @xmath205 @xmath206 @xmath207 where @xmath208 . \label{4.16}\end{aligned}\ ] ] in terms of these @xmath209 and @xmath210 distributions we can now write out , by inspection , all possible contributions to the @xmath17 quarks , shown in fig . 3 , for the production of a proton @xmath211 where @xmath212 denotes symmetrization of @xmath90 and @xmath91 .
we have ignored the contributions corresponding to all three quarks coming from the same valon .
the quark distribution from the proton source for proton production is then @xmath213 where @xmath214 is given in eq.([3.8 ] ) . for @xmath215 production
we need only change all @xmath216 functions in eqs .
( [ 4.11 ] ) - ( [ 4.15 ] ) to @xmath55 functions , since all anti - quarks are in the sea . denoting the corresponding quantities in eqs . ( [ 4.17 ] ) and ( [ 4.18 ] ) by @xmath217 and @xmath218 , we have finally for net proton production @xmath219 where the convolution is defined by the integral in eq.([4.18 ] ) .
it should be recognized that @xmath214 involves a summation of @xmath220 over @xmath8 , which in turn involves a summation @xmath221 over @xmath222 , @xmath223 and @xmath117 that appear in the @xmath209 and @xmath210 distributions .
we are now ready to substitute @xmath160 and @xmath224 , defined in eq .
( [ 4.4 ] ) , into eq.([4.1 ] ) to calculate @xmath225 .
nine convolution integrals are involved : @xmath226 , and @xmath181 .
obviously , we should go to the moments and reduce them to products .
first , using eqs .
( [ 2.1 ] ) in ( [ 4.4 ] ) , we have @xmath227 for convenience , let us leave out the known factors and define @xmath228 then define the moments @xmath229 @xmath230 thus it follows @xmath231 } { n ! \over
n_3 ! } \tilde{f } ( n_1 + \alpha + 3 , n_2 + \alpha + 3 , n_3 + \beta + 3 ) . \label{4.24}\end{aligned}\ ] ] from eq .
( [ 4.18 ] ) we expect from the convolution theorem to have @xmath232 where @xmath233 is given by eq.([3.15 ] ) , apart from the poissonian sum of ( [ 3.8 ] ) .
however , because of the ordering of @xmath197 relative to @xmath198 in eq .
( [ 4.17 ] ) , the simple form of ( [ 4.25 ] ) is valid only for the first term of @xmath234 .
for that first term , which we denote by @xmath235 , we have @xmath236 , \label{4.26}\end{aligned}\ ] ] where the subscript @xmath9 has been omitted .
the summation over @xmath102 in eq .
( [ 3.15 ] ) for @xmath237 should extend over @xmath238 and @xmath186 in eq . ( [ 4.26 ] ) .
the moments of @xmath239 and @xmath240 are defined as usual for the invariant distributions as @xmath241 for all other terms in eq .
( [ 4.17 ] ) there are @xmath201 and @xmath210 functions , and the simple form of eq .
( [ 4.25 ] ) must be modified , especially when only two valons contribute . for notational simplicity
let us define @xmath242 @xmath243 @xmath244 @xmath245 @xmath246 where @xmath247 is defined by @xmath248 and similarly for @xmath249 .
these moments arise whenever two quarks are from the same valon , as they do for all @xmath210 distributions .
take , for example , the third term in eq .
( [ 4.17 ] ) ; we have @xmath250 2\left\ { p_{\nu_1}p_{\nu_2 } k({x_1 \over y'_1})k({x_2 \over y'_2})l({x_3 \over y'_2-x_2 } ) + \cdots\right\}\nonumber\\ & = & \int^1_0 dy'_1 \int^{1-y'_1}_0 dy'_2 g'_{uu } ( y'_1 , y'_2 ) y'^{n_1 - 1}_1 y'^{n_2+n_3 - 2}_2 2p_{\nu_1}p_{\nu_2 } \nonumber\\ & & \cdot \int^1_0 dz_1 z_1^{n_1 - 2}k(z_1)\int^1_0 dz_2 z_2^{n_2 - 2}(1-z_2)^{n_3 - 1}k(z_2 ) \int^1_0 dz_3 z_3^{n_3 - 2}l(z_3 ) + \cdots
\nonumber\\ & = & 2 \{\tilde{g}'_{uu } ( n_1 , n_2 + n_3 - 1 ) \tilde{v}^f_{11}\tilde{w}^{fu}_{2,23 } \}_{12 } , \label{4.34}\end{aligned}\ ] ] where @xmath251 here means symmetrization of @xmath252 and @xmath253 , and @xmath254 performing the same type of operations on all terms we obtain @xmath255 substituting this in eq .
( [ 4.24 ] ) , we have the final form for the moments @xmath256 . for @xmath215 production
we need only change all @xmath257 in eqs.([4.28 ] ) - ( [ 4.32 ] ) to @xmath258 . for the production of neutron
the recombination function is @xmath259 where we use @xmath21 and @xmath22 to refer to @xmath25 and @xmath24 to @xmath23 , so the dependence of @xmath260 on @xmath261 is the same as that of @xmath262 given in eq .
( [ 2.1 ] ) .
the quark distribution @xmath263 with the corresponding identification of @xmath90 and @xmath91 with the @xmath39 quark , and @xmath161 with the @xmath38 quark , also has eight terms , as shown in fig .
we can write , by inspection , the moments @xmath264 similar to eq .
( [ 4.35 ] ) @xmath265 substituting this in eq . ( [ 4.24 ] ) gives @xmath266 .
for @xmath15 production simply replace all @xmath257 in @xmath266 by @xmath258 . finally for net nucleon production
we have @xmath267 @xmath268 we shall use these in the next section to determine the @xmath18 dependences to be compared with the data .
before we can compute the distributions @xmath225 , we need to specify quantitatively the degradation function @xmath269 introduced in eq .
( [ 3.5 ] ) and discussed between eq .
( [ 3.16 ] ) and ( [ 3.22 ] ) .
we proposed an evolution equation for @xmath140 in eq .
( [ 3.16 ] ) but left the kernel @xmath270 unspecified . now , to proceed we must specify @xmath271 , which is uncalculable because it represents the non - perturbative effect of the nuclear medium on a valon as it propagates an incremental distance . we shall use an one - parameter description of the effect , so we shall be approximate by assuming that the effect is like a one - gluon exchange @xcite , i. e. , @xmath272 where the singularity at @xmath273 is regularized by the subtraction @xmath274 evidently , eq .
( [ 5.1 ] ) satisfies the condition @xmath275 which is required by the constraint @xmath276 stated in eq .
( [ 3.20 ] ) that follows from baryon conservation . from the definition of the moments given in eq . ( [ 3.17 ] ) we find , using eq .
( [ 5.1 ] ) , @xmath277 , \label{5.4}\end{aligned}\ ] ] where @xmath278 is the digamma function and @xmath279 is the euler s constant , 0.5722 . substituting @xmath280 into eq .
( [ 3.19 ] ) , and changing @xmath55 to @xmath185 that involves a constant factor , thereby effecting a change from one unknown parameter @xmath281 to another , @xmath282 , we obtain the form for @xmath149 in eq.([3.21 ] ) with @xmath283\}. \label{5.5}\end{aligned}\ ] ] this is a one - parameter description of the effect of momentum degradation . we shall vary @xmath282 to fit the data .
( [ 5.5 ] ) is a rigorous consequence of a simple form for @xmath271 given in eq .
( [ 5.1 ] ) , whose reliability is unknown .
the validity of @xmath151 as expressed in eq.([5.5 ] ) can only be inferred _ a posteriori _ from the fit of the data .
the exponential dependence on the degradation strength @xmath282 follows only from the linear dependence of @xmath271 on @xmath281 , and is sensible .
we have only two free parameters , @xmath282 and @xmath165 , to vary to fit the na49 data on @xmath12-@xmath215 and @xmath284-@xmath15 @xcite .
recall that @xmath165 is introduced in eq .
( [ 4.7 ] ) in connection with flavor changes .
we repeat that the data do not include target fragmentation because @xmath285 and @xmath286 have been subtracted out .
thus the data represent only proton fragmentation and are ideal for our analysis by @xmath287 and @xmath288 .
since the data are in the @xmath18 variable , we must make the inverse transformation from our moments to @xmath225 . instead of making the inverse mellin transform ,
let us exploit the orthogonality of the legendre polynomials and shift the variable to the interval @xmath289 .
thus , define @xmath290 so that @xmath291 if we expand the distribution @xmath292 in terms of @xmath293 @xmath294 then the inverse is @xmath295 these @xmath296 can be expressed in terms of the moments @xmath297 if we express @xmath293 as a power series in @xmath18 @xmath298 where @xmath299 are known from the properties of @xmath300 .
thus from eq .
( [ 5.9 ] ) we have @xmath301 where @xmath302 is defined in eq .
( [ 4.22 ] ) .
it is now clear that our theoretical results in @xmath302 can be transformed to @xmath303 through eqs.([5.8 ] ) and ( [ 5.11 ] ) once we have the coefficients @xmath299 .
furthermore , if @xmath302 becomes unimportant for @xmath304 , then the sum in eq .
( [ 5.8 ] ) can terminate at @xmath164 . to determine @xmath299 , we make use of the recursion formula @xmath305 to infer through eqs .
( [ 5.6 ] ) and ( [ 5.10 ] ) @xmath306 , \label{5.13}\end{aligned}\ ] ] @xmath307 , \label{5.14}\end{aligned}\ ] ] where @xmath308 , and @xmath309 . for @xmath310 and/or @xmath311
, we have @xmath312 with these we can generate all @xmath299 , so @xmath296 can be directly computed . summarizing our procedure ,
we calculate @xmath313 and @xmath314 , substitute them in eq.([5.11 ] ) and then ( [ 5.8 ] ) , and then use eq.([4.21 ] ) to determine @xmath287 and @xmath288
. we vary @xmath282 and @xmath165 to fit the data of na49 shown in fig .
the inclusive distribution @xmath315 corresponds to our @xmath316 .
the solid lines are our results for @xmath5 and the dashed line @xmath6 for both @xmath317 and @xmath318 .
the values of the parameters adjusted are @xmath319 the most striking aspect of our result is that the normalization of the calculated distributions turns out to be correct , even though we have no free parameter to adjust that . the degradation strength @xmath282 affects the shape of the distributions and the flavor - flip probability @xmath165 affects the difference between @xmath5 and @xmath6 .
the agreement between theory and experiment is fairly good , considering that we have only two free parameters and that the experimental errors ( a typical size of which is shown in the figure ) are large , especially at large @xmath18 .
the shapes of the distributions for @xmath5 are reasonably well reproduced and so is the dependence on @xmath9 . for @xmath6
the calculated curves are somewhat steeper than the data
. however , there exist some data points for @xmath6 above @xmath320 that are significantly lower , though with much larger errors . taken as a whole
the agreement is satisfactory .
thus we conclude that the physical process of proton fragmentation and the nucleon momentum spectra are well understood in the framework of the valon model .
we can determine from the value of @xmath282 what the momentum degradation length is .
first , we get from eq.([5.5 ] ) @xmath321 for @xmath322 , since @xmath323 . by definition eq .
( [ 3.26 ] ) we obtain @xmath324 . inserting this in eq .
( [ 3.27 ] ) we obtain a dependence of @xmath325 on @xmath9 that can be well approximated by @xmath326 for @xmath327 .
this gives the fractional momentum loss per collision @xmath328 if we related @xmath9 to average nuclear path length @xmath55 by @xmath329 and @xmath330 , where @xmath331 , then for @xmath332 mb we have @xmath333 with @xmath55 in fm .
thus , if we define the degradation length @xmath334 by @xmath335 then @xmath336^{-1 } \simeq 17 \ , { \rm fm } .
\label{5.21}\end{aligned}\ ] ] this gives an estimate of how far a proton must travel in a nuclear medium in order to lose its momentum by a factor of @xmath337 .
the value of @xmath338 for flavor - flip probability may at first sight appear to be surprisingly large .
however , if it is regarded as an effective way of accounting for resonance production , it becomes quite acceptable . to see that , we first state that resonance production , which we have not taken into account explicitly , can easily produce neutron from @xmath339 through @xmath340 .
the process can be depicted by a dual diagram , as shown in fig .
6(a ) . such a leakage of @xmath341 charge through the emission of @xmath342 is equivalent to a flavor flip , which changes @xmath16 to @xmath343 , symbolized by a square box in fig .
6(b ) , and the favored process , for example , of having valence quarks changes from @xmath17 to @xmath344 .
since a consideration of resonance production would involve masses , threshold , polarization , decay distribution , and other complications , our method of using flavor changes to account for the effect presents considerable technical economy . even though resonance production can occur only at hadronization in the end charge leakage
can take place at any point where a projectile valon interacts with the target ; hence , the consideration in sec . iv leading up to eq .
( [ 4.10 ] ) is an effective way to take such subprocesses into account .
having successfully computed the nucleon distributions , resulting in the determination of the only two free parameters in the model , we are now able to predict the pion distributions without any further ambiguities . the na49 data presented in ref . @xcite do not include the pion spectra .
the e910 data @xcite do have the pion distributions in the proton fragmentation region ; however , being at @xmath345 the energy is too low to avoid substantial spill - over of quarks and produced hadrons from target fragmentation into the @xmath346 region . without the target fragmentation being subtracted , as is done for @xmath5 in the na49 data , the e910 data can not be compared to the predictions from our model .
we present our result below for a future comparison .
the invariant distribution for pion production is @xmath347 where @xmath348 is the invariant distribution for finding a quark at @xmath90 and an antiquark at @xmath91 , and @xmath349 is the corresponding recombination function to form a pion .
an important aspect about pion production concerns the role of gluons , a subject that will be discussed below in connection with @xmath48 . for
now , we consider the appropriate forms for @xmath350 and @xmath349 .
we begin with the valon distribution @xmath351 , as given by eq.([3.8 ] ) . as before
, we shall omit the subscript @xmath9 , and replace it by the valon labels so that for the two - valon distributions we have @xmath352 and similarly for @xmath353 by integrating out @xmath354 .
the single - valon distributions are involved in the 5 subprocesses that contribute to @xmath355 shown in fig . 7 for the production of @xmath342 .
of course , because of the flavor changes , the valon labels are only indicative of the unchanged components
. more precisely , we can express @xmath350 in the moment form , as in eq .
( [ 4.35 ] ) , for @xmath342 @xmath356\{\tilde{v}^f_{11 } \tilde{v}^u_{22 } \}_{12}\nonumber\\ & & + 2 \tilde{g}'_{ud } ( n_2 , n_1)\{\tilde{v}^u_{12 } \tilde{v}^u_{21 } \}_{12 } + 2 \tilde{g}'_u ( n_1 + n_2 - 1)\tilde{w}^{fu}_{1 , 12}\nonumber\\ & & + 2 \tilde{g}'_d ( n_1 + n_2 - 1)\tilde{w}^{uu}_{3 , 12 } , \label{6.3}\end{aligned}\ ] ] and for @xmath357 @xmath358 for the recombination function we have @xmath359 where @xmath360 is the valon distribution in a pion . for the latter we adopt the same form derived in ref .
@xcite @xmath361 which satisfies @xmath362 consequently , the structure function of the pion @xmath363 , that is related to the single - valon distribution @xmath364 @xmath365 behaves at large @xmath18 as @xmath366 , in agreement with the counting rule @xmath367 , where @xmath368 is the number of residual quarks ( @xmath369 for pion and @xmath370 for proton ) . using eqs .
( [ 6.5 ] ) and ( [ 6.6 ] ) in ( [ 6.1 ] ) yields @xmath371 by defining @xmath372 we then have for the moments [ see eq .
( [ 4.22 ] ) ] @xmath373 } { n ! \over n_1 !
\tilde{f}_{\pi}(n_1 , n_2 ) , \qquad n = n_1 + n_2 , \label{6.11}\end{aligned}\ ] ] whereupon eqs .
( [ 6.3 ] ) and ( [ 6.4 ] ) can be used for the production of @xmath342 and @xmath357 , respectively .
the inverse transform can be done as before , using eqs .
( [ 5.8 ] ) and ( [ 5.11 ] ) .
having formulated the procedure to calculate the pion distribution in the projectile fragmentation region , we must confront one final issue on the role of the gluons .
although the gluons carry roughly half the momentum of a proton , no glueball has ever been seen .
they therefore hadronize by converting to @xmath374 pairs , which subsequently form pions .
we take them into account by enhancing the sea to saturate the momentum sum rule @xcite .
that is , for the purpose of pion production we revise the normalization of the quark distribution @xmath48 such that @xmath374 in the sea carry all the momentum of the incident proton apart from the momenta of the valence quarks , leaving nothing for the gluons .
the average momentum fraction carried by the valence quark in a valon is @xmath375 where eq .
( [ 2.9 ] ) has been used . if we denote the saturated sea distribution by @xmath376 where only the constant factor @xmath377 has changed from eq .
( [ 2.10 ] ) , then each sea quark carries on average a momentum fraction of @xmath378 .
that is to be identified with @xmath379 , where @xmath195 is the number of flavors . from eq .
( [ 2.13 ] ) we have @xmath380 . setting @xmath381
, we get @xmath382 with this value in @xmath383 , which is then used in eqs.([4.11 ] ) - ( [ 4.15 ] ) in place of @xmath48 , we obtain the appropriate @xmath209 and @xmath210 functions that should be used for the calculation of @xmath384 in eqs.([6.3 ] ) - ( [ 6.4 ] ) . evidently , there are no parameters to adjust for the calculation of the pion distributions .
the results are shown in fig . 8 for @xmath342 and @xmath357 separately at @xmath96 and @xmath318 .
for ease of comparison between @xmath342 and @xmath357 the same curves are replotted in fig . 9 , where the charge and @xmath9 dependences are grouped differently .
no data points are included because none correspond to proton fragmentation only and at the @xmath9 considered , as discussed at the beginning of this section .
nevertheless , if one compares our results to the data of e910 shown in @xcite , there is rough agreement , both in normalization and in shape .
generally speaking , the difference between @xmath342 and @xmath357 diminishes as @xmath9 is increased , though not as rapidly as in the e910 data at @xmath385 .
we are confident that our predictions will agree well with the data when they become available , since the pion distributions have always demonstrated the reliability of the recombination model @xcite .
we have formulated the projectile fragmentation problem in @xmath0 collisions in the valon model . despite the non - perturbative aspect of the problem , the formulation results in a well - defined procedure of calculating various contributions to the nucleon and meson distributions in the proton fragmentation region .
the nuclear target effects , which present the only unknown in the model , are summarized by two parameters @xmath282 and @xmath165 .
they are determined by fitting @xmath5 and @xmath6 distributions for two values of @xmath9 . with
those parameters fixed , the predictions for the pion distributions can be calculated without any other adjustable parameters .
the results of our calculations have shown that the na49 data can be well described by the valon model .
the normalization of the nucleon spectra turns out to be correct without any freedom for adjustment .
the shapes of the distributions are also acceptably reproduced .
the inferred value of @xmath282 that gives a quantitative measure of momentum degradation can be translated to a degradation length @xmath334 in the form @xmath386 for the degraded momentum fraction with @xmath387 fm .
while a value for @xmath334 can serve as a succinct numerical summary of the stopping effect of the nucleus , we admit that it can not be inferred directly from the @xmath0 collision data without a detailed analysis in the framework of the valon model .
this aspect of the problem is worthy of further attention in the hope that a degradation length can be extracted by an appropriate model - independent analysis of the data on nucleon inclusive cross section .
it should be noted that since the valon model does not make explicit use of regge exchanges , it is not capable of predicting the energy dependence .
it is a model that should be applied only at asymptotic energies where scaling behavior prevails . to compensate for that drawback
, it makes possible the @xmath86-channels approach to the calculation of the fragment distributions .
there are obvious directions into which this work can be extended .
one is to incorporate strangeness and study the distributions of hyperons and kaons .
another is to generalize from @xmath0 to @xmath3 collisions . from the properties of degradation that this work has revealed ,
one is better positioned to assess the extent to which nuclear matter can be compressed in @xmath3 collisions . furthermore , with some knowledge about strangeness production in @xmath0 collisions , one can determine for @xmath3 collisions how much strangeness enhancement is normal and how much anomalous .
although the valon model represents an approach to multiparticle production that appears to be orthogonal to most other approaches based on strings , it should be recognized that the @xmath86-channel and @xmath87-channel approaches are complementary , not contradictory .
one may be able to identify diagrams in figs . 3 and 4 that correspond to baryon junction or diquark breaking .
just as the notion of duality has benefitted the investigation of high - energy processes many years ago that found its idealization in the form of veneziano amplitude , which has perfect @xmath86- and @xmath87-channel symmetry , so also here in nuclear processes the exploration of complementary descriptions of common as well as unusual phenomena can help to elucidate the underlying dynamics responsible for them .
we are grateful to g. veres for a helpful communication .
this work was supported , in part , by the u. s. department of energy under grant no .
de - fg03 - 96er40972 .
99 b. cole , talk given at quark matter 2001 , 15th int .
conf . on ultra - relativistic nucleus - nucleus collision ,
stony brook , ny january 2001 .
r. c. hwa , phys .
lett . * 52 * , 492 ( 1984 ) .
l. p. csernai , and j. i. kapusta , phys .
* 29 * , 2664 ( 1984 ) .
j. hfner and a. klar , phys .
b*145 * , 167 ( 1984 ) . c. y. wong , phys .
lett . * 52 * , 1393 ( 1984 ) .
b. z. kopeliovich and b. g. zakharov , z. phys .
c * 43 * , 241 ( 1989 ) ; a. capella and b. z.kopeliovich , phys .
b*381 * , 325 ( 1996 ) .
d. kharzeev , phys .
b*378 * , 238 ( 1996 ) .
s. e. vance , m. gyulassy and x .-
wang , phys .
b*443 * , 45 ( 1998 ) .
r. c. hwa , nucl .
b ( proc . suppl . )
* 92 * , 348 ( 2001 ) .
r. c. hwa , phys .
d*22 * , 759 ( 1980 ) .
r. c. hwa , and m. s. zahir , phys .
rev.d*23 * , 2539 ( 1981 ) .
r. c. hwa , phys . rev .
d*22 * , 1593 ( 1980 ) .
k. p. das and r. c. hwa , phys .
* 68b * , 459 ( 1977 ) .
r. c. hwa , and m. s. zahir , phys .
rev.d*31 * , 499 ( 1985 ) .
d. s. barton _ et .
rev.d*27 * , 2580 ( 1983 ) .
cteq collaboration , h. l. lai , _ et .
d*51 * , 4763 ( 1995 ) .
cteq4lq , http://zebu.uoregon.edu/@xmath388parton / partongraph.html .
d. duke and r. g. roberts , phys .
lett . , 289 ( 1979 ) .
v. n. gribov , sov .
phys.- jetp * 30 * , 709 ( 1970 ) . w. ochs , nucl .
* b118 * , 397 ( 1977 ) .
b. andersson , g. gustafson , and c. peterson , phys .
l. * 60b * , 221 ( 1977 ) .
a. capella , u. sukhatme , c. i. tan and j.tran thanh van , phys .
l. * 81b * , 68 ( 1979 ) .
r. c. hwa , j. pit , and n.pitov , phys .
lett . * 85 * , 4008 ( 2000 ) .
we thank g. veres for private communication on this point .
d. w. duke and f. e. taylor , phys.rev .
d*17 * , 1788 ( 1978 ) .
iii we have considered the problem of describing @xmath8 collisions with the target nucleons in terms of the number of collisions that each valon experiences . with the _
i_th valon encountering @xmath102 collisions , the sum @xmath389 satisfies the bounds given in eq .
( [ 3.3 ] ) . in this appendix
we investigate the phenomenological preference for that sum within that range .
let us define the integer @xmath390 by @xmath391 so that @xmath390 is bounded by @xmath392 since at least one valon must interact with the nucleus , let that valon be @xmath393 .
let @xmath12 be the probability that either one of the other two valons also interacts .
furthermore , let @xmath394 denote the probability that out of @xmath8 independent collisions the target nucleons encounter , @xmath390 valonic collisions occur . for @xmath395
, we have @xmath396 and @xmath397 , which count the probabilities that valons @xmath398 and @xmath399 interact in addition to the @xmath393 valon . generalizing that to @xmath8 collisions , we have @xmath400 which is a binomial distribution of having @xmath401 valonic collisions by the @xmath398 and @xmath399 valons out of a maximum of @xmath402 possible such collisions .
we can define a generating function from @xmath394 : @xmath403^{\nu}. \label{a4}\end{aligned}\ ] ] the above consideration can now be applied to eq .
( [ 3.23 ] ) where @xmath404 is related to @xmath405 under the assumption of eq . ( [ 3.4 ] ) , i. e. , @xmath406 .
that assumption is now liberated by eq .
( [ a1 ] ) and ( [ a2 ] ) . the factor @xmath407^{\nu}$ ] in eq .
( [ 3.23 ] ) should thus be replaced by @xmath408^{\nu}$ ] , which allows @xmath390 to vary between @xmath8 and @xmath409 . in the limit @xmath410 becomes @xmath411 and we recover the earlier result , if we identify @xmath412 using ( [ a4 ] ) in eq .
( [ 3.23 ] ) , we finally have , with the help of eq.([3.8 ] ) @xmath413 the summation over @xmath102 is included in the sum in eq .
( [ 3.15 ] ) but without the restriction of ( [ 3.4 ] ) .
we have used the @xmath233 given in ( [ a6 ] ) to calculate @xmath292 for @xmath5 and @xmath6 , as in secs .
iv and v. the only difference from before is that we now have one additional parameter @xmath12 in ( [ a2 ] ) to adjust , which controls the number of valonic collisions above @xmath414 .
our best fit of the data , as in fig . 5 , yields @xmath415 with the values of @xmath282 and @xmath165 being essentially the same as in eq.([5.16 ] ) .
since @xmath12 is so small , any contribution from @xmath390 different from @xmath8 can be neglected .
thus it is the data that instruct us on the suitable range of values of @xmath102 , namely : @xmath416 is dominantly at its lower bound @xmath8 . | the proton and neutron inclusive distributions in the projectile fragmentation region of @xmath0 collisions are studied in the valon model . momentum degradation and flavor changes due to the nuclear medium
are described at the valon level using two parameters .
particle production is treated by means of the recombination subprocess .
the centrality dependences of the net proton and neutron spectra of the na49 data are satisfactorily reproduced .
the effective degradation length is determined to be 17 fm .
pion inclusive distributions can be calculated without any adjustable parameters .
pacs number(s ) : 25.75.dw , 13.85.ni oits 706 + |
precise timing of astrophysical events is one of the fundamental tools of astronomy , and is an important component of essentially every area of study .
there are two basic sources of uncertainty in timing : the astrophysical data characterizing the event , and the time stamp with which the event is referenced .
unfortunately , since the accuracy of the time stamp is something that is often taken for granted , the improvements in data are sometimes not accompanied ( or are not uniformly accompanied ) by the requisite improvements in accuracy of the time stamp used .
this situation can lead to confusion , or even spurious inferences .
timing plays a particularly important role in the study of exoplanets .
indeed , many of the ways in which exoplanets are discovered involve the detection of transient or time - variable phenomena , including the radial velocity , transit , microlensing , and astrometry techniques .
furthermore , in some cases much can be learned about planetary systems from the precise timing of these phenomena . as examples ,
the measurement of terrestrial parallax in microlensing allows one to infer the mass of the primary lens and so the planetary companion ( e.g. , * ? ? ?
* ) , and one can constrain the eccentricity of transiting planets by comparing the times of primary transits and secondary eclipses ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
possibly the most promising application of timing in exoplanets , however , comes from transit timing variations ( ttvs ) . with an exquisitely periodic phenomena like transiting planets
, we will be able to measure many effects using the departures from strict periodicity , such as the gravitational perturbations from additional planets @xcite , trojans @xcite , and moons @xcite , stellar quadrupoles @xcite , tidal deformations , general relativistic precession @xcite , orbital decay ( e.g. , * ? ? ?
* ) , and proper motion @xcite .
because of their great potential , ttvs have become the focus of many groups .
the typical data - limited transit timing precisions of most observations are around 1 minute , with the best transit time precision yet achieved of a few seconds @xcite . however , as discussed above , accuracies of transit times are limited not only by the data themselves , but also by the time stamp used . in order to make these difficult measurements useful , it is critical that a time stamp be used that is considerably more accurate than the uncertainty due to the data themselves . furthermore , since thorough characterization of ttvs will require the use of all available data spanning many years from several groups , this time stamp must be stable in the long term , and all groups must clearly convey how it was calculated .
unfortunately , we have discovered that , in the exoplanet community , the julian date ( jd ) and its geocentric ( gjd ) , heliocentric ( hjd ) , and barycentric ( bjd ) counterparts are currently being quoted in several different , and often unspecified , time standards . in addition , the site arrival time and its time standard is not quoted .
this general lack of homogeneity and specificity leaves quoted time stamps ambiguous at the 1 minute level .
more alarmingly , the most commonly - used time standard , the coordinated universal time ( utc ) , is discontinuous and drifts with the addition of each leap second roughly each year .
the pulsar community has solved the problem of precise timing well beyond the level that is currently necessary for exoplanet studies , and we can benefit from the techniques they have developed over the past 40 years .
in particular , their current state of the art program ( tempo2 ) models pulsar arrival times to 1 ns precision @xcite . this program is highly specialized and generally can not be applied outside of pulsar timing observations , but many of the effects they consider are relevant to optical observers in the exoplanet community . in this article , we summarize the effects one must consider in order to achieve timing accuracy of 1 @xmath1s well beyond the accuracy that will likely be required by the exoplanet community for the foreseeable future .
section [ theory ] provides the background required to understand each of the effects that could change the arrival time of a photon .
they are listed in order of decreasing magnitude , so latter subsections can be ignored for low - precision measurements .
section [ practice ] discusses the practical limitations to achieving high - precision timing .
we begin with the effects which may cause errors that are comparable to or exceed the bjd correction .
these should be read and understood by everyone .
we continue with remaining effects , in order of decreasing magnitude , which can be ignored for low - precision ( @xmath2 ms ) measurements .
we conclude 3 by listing additional effects , the errors due to which are negligible ( @xmath3s ) .
we begin [ sec : calculating ] by detailing the procedure one must follow in order to calculate the bjd@xmath0 , which is designed to be a useful reference for those already familiar with the concepts of precision timing . in the latter part of this section , we describe our particular idl and web - based implementation of this procedure .
lastly , in the appendix , we discuss some of our specific findings about the time stamps currently in use and how these are calculated throughout the exoplanet community .
while we focus on the effects of timing on the optical / infrared exoplanet community , timing precision of order 1 minute is necessary for many other areas , such as the study of rapidly rotating white dwarfs @xcite .
this article should be equally applicable in such cases .
the biggest source of confusion comes from the fact that time standards and reference frames are independent from one another , even though there are many overlapping concepts between the two .
we will use the following terminology : `` reference frame '' will refer to the geometric location from which one could measure time
different reference frames differ by the light - travel time between them ; `` time standard '' will refer to the way a particular clock ticks and its arbitrary zero point , as defined by international standards ; and `` time stamp '' is the combination of the two , and determines the timing accuracy of the event .
the bjd@xmath0 , the time stamp we advocate , can be calculated using the equation : @xmath4 where jd@xmath5 is the julian date in coordinated universal time ; @xmath6 is the rmer delay , discussed in [ sec : roemer ] ; @xmath7 is the clock correction discussed in [ sec : clock ] ; @xmath8 is the shapiro delay discussed in [ sec : shapiro ] ; and @xmath9 is the einstein delay , discussed in [ sec : einstein ] .
the order of these terms is such that they are of decreasing magnitude , so one need only keep the terms up to the precision required .
the timing precision required by current exoplanet studies ( @xmath10 1 s ) requires only the terms up to and including @xmath7 . because future solar system ephemerides may enable more precise calculations of the arrival time at the barycenter , or in order to allow others to check that the original conversion was done accurately enough for their purpose , the site arrival time ( e.g. , the jd@xmath5 ) should always be quoted in addition to the bjd@xmath0 . due to the finite speed of light ,
as the earth travels in its orbit , light from an astrophysical object may arrive early or be delayed by as much as 8.3 minutes from the intrinsic time of the extraterrestrial event .
this is called the rmer delay , @xmath11 , in honor of ole rmer s demonstration that the speed of light is finite .
since most observers can not observe during daylight , a bias is introduced and in practice the delay ( as distinct from the early arrival time ) is only as much as 7 minutes , for a peak - to - peak variation of 15 minutes .
figure [ fig : bjdvjd ] shows an example of this effect for a maximally affected object on the ecliptic . in order to show the observational bias , our example assumes the object is at 0@xmath12 right ascension and 0@xmath13 declination .
this curve shifts in phase with ecliptic longitude and in amplitude with ecliptic latitude .
we also place our observer at the earth s equator , but note that the asymmetry will be larger at different latitudes . and the uncorrected jd@xmath14 ( see text for definitions ) over the course of a year .
we plot the correction for a maximally - affected object on the ecliptic for an observer at latitude of zero degrees .
we exclude all points where the object has an airmass greater than three and the sun is higher than -12@xmath13 in order to highlight observing biases.,width=312 ] the solution to this problem is to calculate the time when a photon would have arrived at an inertial reference frame .
this time delay is the dot product of the unit vector from the observer to the object , @xmath15 , and the vector from the origin of the new reference frame to the observer , @xmath16 @xmath17 where @xmath18 is the speed of light and @xmath15 can be written in terms of its right ascension ( @xmath19 ) and declination ( @xmath20 ) , @xmath21
this equation is general as long as @xmath19 , @xmath20 , and @xmath16 are in the same coordinate system ( e.g. , earth mean equator j2000 ) and the object located at ( @xmath19 , @xmath20 ) is infinitely far away . other forms of this equation in the literature assume that we have the angular coordinates of the new origin or that the earth and the new origin are in the same plane ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but we explain in [ sec : calculating ] why this form is most practical for calculating the delay .
the hjd , which uses the sun as the origin of the new reference frame , is only accurate to 8 s because of the acceleration of the sun due primarily to jupiter and saturn ( fig .
[ fig : bjdvhjd ] ) .
it was popular when people first began considering this effect because it is relatively simple to calculate from tables without a computer @xcite , and remains popular because self - contained algorithms exist to approximate it without any external tables ( e.g. , * ? ? ? * ) .
however , because the hjd is not useful when accuracies of better than 8 s are needed , most of the algorithms in use today use approximations that are only precise at the 1 s level , and it becomes impossible to back out the original jd from the hjd unless we know the exact algorithm used . and the hjd@xmath0 for a maximally - affected object on the ecliptic .
the primary periodicity is due to jupiter and the secondary periodicity is due to saturn.,width=312 ] because of these problems , the hjd was formally deprecated by international astronomical union ( iau ) resolution a4 in 1991 , in favor of the bjd , a time referenced to the solar system barycenter ( ssb ) . the analogous correction to the rmer delay in our solar system
can also be significant in the target system .
we refer to this as @xmath11 . for example
, for transiting planets with @xmath22 au , @xmath11 can be as large as 30 s. in general , the position of the planet during primary transit has become the unspoken standard reference frame for transiting planets , while the host star s photosphere is the unspoken standard for radial velocity ( rv ) planets . in theory , the timing would be much more stable in the target s barycentric reference frame , but the accuracy with which we can convert to this frame depends on the measurements of the system . since different observers may use different values as measurements improve , quoting the jd in the frame of the target s barycenter may obfuscate the long term reliability of timing .
therefore , we argue it is better to quote julian date in the ssb reference frame , and correct for @xmath11 only when comparing observations at different phases in the planet s orbit .
this @xmath11 correction is not necessary for ttvs of the primary transit , since the planet is always in the same phase .
nevertheless , we should explicitly state the object s reference frame to avoid any potential ambiguity , particularly when comparing any combination of primary transits , secondary transits , rvs , and another primary transit of a different planet in the same system , when it may not be obvious which origin is being used . for rv measurements , which are taken at many different phases ,
the effect is much smaller and can generally be ignored because the star s orbit around the barycenter is small . for a typical hot jupiter , ( i.e. , a jupiter mass planet in a 3 day orbit around a solar mass star ) ,
the maximum time difference in the rv signal ( for an edge - on orbit ) is 20 ms , which would change the measured rv by @xmath23 m s@xmath24 .
while planets farther out will cause a larger timing offset , the difference in the measured rv is even smaller . to be clear
, the jd can be specified in many time standards @xcite , and while the iau has made no explicit statement regarding the allowed time standards of the gjd , hjd , or bjd , their meaning in any given time standard is unambiguous .
unfortunately , they have been specified in many standards , usually implicitly
. however , the particular time standard used affects how useful the time stamp is as an absolute reference . we must be careful not to directly compare bjds or hjds in different time standards , as each has different offsets , periodic terms , and/or rates , which can introduce systematic errors of over 1 minute .
for this reason , it is critical that any stated bjd or hjd also specify the time standard used when one - minute accuracies are important , and the uncertainty of a time that is quoted without a standard should be assumed to be at least 1 minute .
first , it may be useful to summarize the relevant standards of time : universal time , ut1 : : defined by the mean solar day , and so drifts forward and backward with the speeding and slowing of the earth s rotation .
generally , it slows due to the tidal braking of the moon , though changes in the earth s moment of inertia and complex tidal interactions make its exact behavior unpredictable .
it is rarely used directly in astronomy as a time reference , but we mention it for context .
international atomic time , tai : : based on an average of atomic clocks all corrected to run at the rate at the geoid at 0 k , with 1 s equal to `` 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom , '' as defined by resolution 1 of the thirteenth meeting of the confrence gnrale des poids et mesures ( cgpm ) in 1967 .
this definition is based on the duration of the ephemeris time second , which was previously defined as 1/31,556,925.9747 of the tropical year for 1900 january 0 at 12 hours ephemeris time by resolution 9 of the eleventh cgpm in 1960 .
tai is the fundamental basis for many other time standards , and is the default time standard of the sloan digital sky survey .
coordinated universal time , utc : : runs at the same rate as tai , except that it is not allowed to differ from ut1 by more than 0.9 s. every 6 months , at the end of 31 december and 30 june , the international earth rotation and reference systems service ( iers ) may elect to add ( or subtract ) a leap second to utc in order to keep it within 0.9 s of ut1 .
utc is therefore discontinuous and drifts relative to tai with the addition of each leap second , which occur roughly once per year . as of january 2009 , the current number of leap seconds , @xmath25 , is 34 .
the full table of leap seconds is available online and is typically updated several months in advance of when an additional leap second is to be added .
utc is the current international standard for broadcasting time . as a result ,
when a modern , network - connected computer s clock is synchronized to a network time protocol ( ntp ) server , it will be in utc .
thus , this is the system of time most familiar to astronomers and non - astronomers alike ( modulo time zones and daylight savings time ) .
universal time , ut : : an imprecise term , and could mean ut1 , utc , or any of several other variations .
in general , such imprecise language should be avoided , as the potential ambiguity is up to 1 s. in the context of a time stamp , it is likely utc , but some people may intentionally use ut to imply 1 s accuracy . while explicitness is preferred ( i.e. , utc @xmath26 1 s ) , any time stamp quoted in ut should be assumed to be uncertain at the 1 s level unless the time standard has been independently verified .
terrestrial time , tt(tai ) : : a simple offset from tai of 32.184 s released in real time from atomic clocks and never altered .
this offset is to maintain continuity between it and its predecessor , the ephemeris time ( et ) .
terrestrial time , tt(bipm ) : : a more precise version of tt(tai ) . the international bureau of weights and measures ( bipm )
reanalyzes tt(tai ) and computes a more precise scale to be used for the most demanding timing applications .
the current difference between tt(tai ) and tt(bipm ) is @xmath27s , and must be interpolated from a table maintained by the bipm and published online with a 1 month delay .
terrestrial time , tt : : sometimes called terrestrial dynamical time ( tdt ) , can refer to either tt(tai ) or tt(bipm ) . from this point on , we will not make the distinction , but when accuracies of better than 30 @xmath1s are required , tt(bipm ) must be used .
barycentric dynamical time , tdb : : corrects tt for the einstein delay to the geocenter , @xmath9 , which is the delay due to time dilation and gravitational redshift from the motions of the sun and other bodies in the solar system .
the conversion from tt to tdb can not be written analytically , but is usually expressed as a high - order series approximation @xcite . the difference is a predominantly a periodic correction with a peak - to - peak amplitude of 3.4 ms and a period of 1 yr .
tdb was slightly modified by iau resolution b3 in 2006 , converging on the same definition as the jpl ephemeris time , t@xmath28 , also called coordinate time ( ct ) in the jpl ephemerides of solar system objects .
barycentric coordinate time , tcb : : physically and mathematically equivalent to the tdb as defined in 2006 @xcite , and differs only by an offset and rate of about 0.5 s yr@xmath24 due primarily to time dilation in the sun s gravitational potential .
tdb and tcb were roughly equal at 1977 jan 1.0 tai , and now differ by about 16 s. caution must be always be exercised , however , as these definitions are subject to change at any point , though usually with a few year s notice . assuming the time is measured according to current definition of utc , then the clock correction , from utc to tdb , can be written as the sum of the corrections from utc to tai , tai to tt , and tt to tdb : @xmath29 of course , if one wishes to express the bjd in another time standard ( or start with something other than utc ) , the clock correction would change accordingly . however , not every time standard is well - suited to precise , astrophysical time stamps , and the use of any time standard other than tdb should be viewed simply as an adequate approximation to tdb .
most readily available programs that calculate the time stamp assume that the user has already applied the utc - to - tt part of this clock correction , which is often not true .
we feel this assumption has contributed the widespread confusion regarding time stamps .
as this last point is our primary motivation for writing this article , we elaborate here on the effects of time standards on the reliability of time stamps . for the sake of simplicity
, we only discuss the effects of time standards on the bjd .
each of these effects also applies to the hjd , though the improvement in accuracy of the time stamp is negligible compared to the accuracy of the hjd reference frame for all but the utc time standard .
the least preferred , though most commonly used , time standard for the bjd is utc ( bjd@xmath5 ) , and is equivalent to ignoring @xmath7 altogether . because utc is discontinuous and drifts with the addition of each leap second ,
comparing two bjd@xmath5 time stamps could result in spurious differences if any leap seconds have been introduced between observations .
therefore , 1 s timing accuracies can not be achieved using the bjd@xmath5 over a span that straddles the addition of one or more leap seconds ( roughly 1 yr ) .
figure [ fig : tdbvutc ] shows the difference between the bjd@xmath5 time stamp with the uniform bjd@xmath0 time stamp ( described below ) from 1961 jan 1 , when utc was defined ( though its definition has evolved over the years ) , to 2010 december 31 , the furthest future date for which the value of utc can be accurately predicted at this writing . and the bjd@xmath5 .
it shows the discontinuities and slow drift in bjd@xmath5 due to the addition of leap seconds . without correcting for these , relative timing between two reported values of bjd@xmath5
can only be trusted over short time scales.,width=312 ] bjd in tt ( bjd@xmath14 ) , which is equivalent to ignoring the ( tdb - tt ) term in equation ( [ eq : utc2tdb ] ) , corrects for the discontinuity and drift introduced by leap seconds and is appropriate for timing accurate to 3.4 ms .
bjd in tdb ( bjd@xmath0 ) is usually the best time stamp to use in practice , as it further corrects the bjd@xmath14 for all known effects on the motions , and therefore rates , of our atomic clocks . while bjd@xmath0 is not perfect , any more accurate time stamp is unique for each target .
bjd in tcb ( bjd@xmath30 ) corrects for the gravitational potential , primarily from the sun , which causes the clock to run slower than it otherwise would .
however , if one is concerned about effects of this magnitude , the analogous correction of putting it in the gravitational potential of each observed object is also required .
since these two rates are small ( @xmath31 ) , and opposite in sign , we believe it is best to ignore bjd@xmath30 except perhaps as an intermediate step in calculating the target - specific frame . technically , the use of tcb was recommended by iau resolution b1.3 in 2000 .
however , because of the greater practicality of using tdb ( see [ sec : calculating ] ) , and the drifting difference between tcb and tdb and tt , we believe its use will only lead to confusion without any foreseeable benefit to the exoplanet community .
the shapiro delay , @xmath32 @xcite , is a general relativistic effect in which light passing near a massive object is delayed . for an object at an angle @xmath33 from the center of the sun ,
the shapiro delay is @xmath34 this can be as large as 0.1 ms for observations at the limb of the sun , but for objects more than 30@xmath13 from the sun , the correction is less than 20 @xmath1s .
there is also the analogous correction , @xmath32 , for the target system .
similar to @xmath11 , @xmath32 depends on the measurements of the target , which may be refined over time .
therefore , the time should generally be quoted without @xmath32 , but include it when comparing times where this could be significant . as we discussed in [ sec : clock ] , relativity dictates that the motion of the observer influences the rate at which the observed clock ticks .
the use of tdb corrects for an observer moving with the geocenter , but in reality we observe from the surface of the earth or from a satellite , for which there is an additional term to @xmath9 : @xmath35 here , @xmath36 is the location of the observer with respect to the geocenter , and @xmath37 is the velocity of the geocenter .
again , there is an analogous correction , @xmath38 , for the target system , which should be ignored when quoting the time but included when comparing times if necessary .
of course , the accuracy of the output time stamp is only as good as our assumptions and the inputs needed to specify the time standard and reference frame .
here we discuss their effects on the accuracy of the time stamps .
the first four subsections , through [ sec : planewave ] , have no reasonable upper bound . in each case , the accuracy of the inputs must be evaluated depending on the accuracy of time stamp required .
the later three subsections through [ sec : double ] are organized in decreasing magnitude , all of which can be ignored for accuracies no better than 21.3 ms . finally ,
[ sec : negligible ] discusses effects of less than 1 @xmath1s that we have ignored .
an error of 1 in the position of the target amounts to a timing error of as much as 0.28 s in the bjd ( fig .
[ fig : bjdra ] ) .
such error would be common if the coordinates of the field center are used instead of the specific object s coordinates . in particular , if doing a survey , one may wish to assign the same bjd to all objects in a given frame .
however , with a 10@xmath13 offset , which is possible with some wide - field transit surveys , the error can be as large as 200 s for objects at the edge of the field .
an error of 0.25 will yield 1 ms timing offsets , and 0.25 mas accuracy is necessary for 1 @xmath1s timing .
calculated for an object at r.a .
= 0@xmath39 and an object at r.a .
= 0@xmath39 0@xmath40 4@xmath41 observed at the same time .
this difference is as large as 200 s for a 10@xmath13 offset.,width=312 ] the accuracy of a typical computer clock depends on its intrinsic stability , the computer workload , its operating system , and the reliability of the network connection .
older computers with a parallel port ccd interface may produce unreliable timing because the clock may slow or stop completely during ccd readout .
without any special effort , a modern windows machine with a network connection is accurate to @xmath42 seconds , and with third - party software like dimension 4 , we have found it to be stable to 0.1 s. an ntp - synchronized linux machine is typically accurate to @xmath43 ms . of course , the stability of the clock only sets a lower limit on the absolute accuracy of the time recorded in the fits image header .
ntp synchronization attempts to measure and compensate for network latency , but the accuracy of time stamps also depends on the particular software package taking the image and the hardware it uses , which is difficult to calibrate . unless independently verified , the time recorded in image headers should not be trusted to better than 0.25 s. however , various solutions exist for higher precision timing , such as gps - triggered shutters . in particular , it is worth emphasizing the 1 s error in the _ hubble space telescope _ ( hst ) clock and potential 6.5 s error in the kepler clock , described in more detail in the appendix , both of which have already achieved data - limited transit timing precisions of that order ( e.g. , * ? ? ?
* ; * ? ? ?
when calculating the time of exposure of an image , we typically use the time at midexposure . however , the precise time of exposure that should be used is the flux - weighted mean time of exposure .
the magnitude of this error depends on the intrinsic stability of source , the stability of the atmosphere , and the exposure time . in the diabolical case of a cloud completely covering the object during one half of the exposure
, the error could be as large as half the exposure time . for a typical hot jupiter that dims by 1% over the course of a 15 minute ingress ,
the error introduced into the time stamp during ingress or egress by using the mid exposure time is 0.25% of the exposure time
150 ms for a 1 minute exposure .
near the peak of some high - magnification microlensing events ( i.e. , * ? ? ?
* ) , the flux may double in as little as 6 minutes . during such instances , using the mid exposure time will result in a time stamp error of as much as 1/4 of the exposure time
15 s for a 1 minute exposure .
equation ( [ eq : roemer ] ) assumes the object is infinitely far away , and therefore the incoming wavefronts are plane waves .
in reality , the wavefronts are spherical , which introduces a distance - dependent , systematic error .
the maximum error introduced by the plane - wave approximation is 1000 s for the moon , 100 s for the main asteroid belt , 5 s for the kuiper belt , 1 ms at the distance of proxima centauri , and 150 ns at the distance to the galactic center . the fully precise equation , assuming spherical wavefronts , is @xmath44 where @xmath45 is the distance from the observer to the target . in such instances ,
the distances must be derived from precise ephemerides for both the target and observer .
although this form is generally applicable , it is not generally practical because at large distances , double precision floating point arithmetic can not reliably recover the small difference between @xmath46 ( the distance from the barycenter to the target ) and @xmath45 . to solve this problem , the pulsar community ( e.g. , * ? ? ?
* ) uses the two - term taylor expansion of equation ( [ eq : exactroemer ] ) about @xmath47 : @xmath48 one may recognize the first term as plane wave approximation ( eq . [ [ eq : roemer ] ] ) . in practice , the accuracy of this two - term approximation exceeds the accuracy of the `` exact '' calculation using double precision at a distance of 10,000 au ( 0.05 pc ) . at the distance of proxima centauri ,
the accuracy of this approximation is at worst 1 ns .
however , this taylor expansion is divergent when @xmath49 .
it should never be used for objects inside 1 au and may still be inadequate for other objects inside the solar system .
it has a maximum error of 1 day for the moon , 20 s for the main asteroid belt , and 40 ms for the kuiper belt . in these cases , using the exact formula ( eq . [ [ eq : exactroemer ] ] ) may be easiest .
therefore , we recommend that for precise calculations of any solar system body , equation ( [ eq : exactroemer ] ) should be used . for better than 1 ms timing of any object outside the solar system , equation ( [ eq : roemercorr ] ) should be used .
most readily available time stamp calculators use the position of the geocenter , rather than the location of the observer on the surface of the earth . neglecting the light - travel time from the surface of the earth to the center introduces a 21.3 ms amplitude variation with a period of 1 sidereal day . in practice
, most observers can only observe their targets at night , creating a systematic bias of between 8 ms and 21.3 ms ( fig .
[ fig : geocenter ] ) .
calculated at the geocenter and at the precise location of the observer on the surface of the earth . while geometrically , this effect will oscillate between @xmath50 ms with a period of 1 sidereal day , we exclude points when the sun is above -12@xmath13 and object is at @xmath51 , which introduces a large observational bias.,width=312 ] usually , when people calculate the bjd@xmath5 , they neglect @xmath7 , and input jd@xmath5 for algorithms designed to take jd@xmath14 .
this effectively uses the positions of the earth , sun , and planets offset in time by 32.184 +
n s to calculate the correction . when the bjd@xmath5 is calculated in this manner , we denote it as bjd@xmath52 . the correct way to calculate the
bjd@xmath5 would be to first calculate the bjd@xmath0 , then subtract the @xmath7 correction is a poor approximation to bjd@xmath0 and should not be used . ] .
figure [ fig : utcvutc ] shows , for a maximally affected object on the ecliptic , an example of the difference between the bjd@xmath52 and the fully correct bjd@xmath5 .
fortunately , this amounts to at most a 13.4 ms difference ( though growing with the utc - tt difference ) , which is below the precision of most clocks and the geocentric correction that is usually ignored .
therefore , to an accuracy of @xmath43 ms , one can safely say bjd@xmath0@xmath53
bjd@xmath14@xmath53 bjd@xmath52@xmath54 , making it easy to convert currently published values of bjd@xmath52 to the superior bjd@xmath0 . and the commonly calculated bjd@xmath52 using the positions of the earth , planets , and sun delayed by 32.184 + n
s. it shows that the difference can safely be ignored for @xmath55 ms precision , and therefore the approximate bjd@xmath14 or bjd@xmath0 can be recovered from currently published bjd@xmath5 simply by adding 32.184 + n s.,width=312 ] representing jd as a double precision floating - point number limits the accuracy to about 1 ms , and any operation done on the full jds will be even less accurate .
many programs require the use of a reduced or modified julian date , and/or can return the jd to bjd offset in seconds , but care must be taken at every step of the way never to store the full jd as a double precision number if 1 ms precision is required .
the shapiro delay occurs for other bodies as well , but observations at the limb of jupiter only delay light by 200 ns .
typical modern , commercial gps units use the world geodetic system ( wgs84 ) , which is referenced to the international terrestrial reference system ( itrs ) with an error of about 15 m , which amounts to a 50 ns error in the time stamp .
the index of refraction of the atmosphere is not exactly 1 and changes with its composition , temperature , and pressure , which changes the speed of light .
however , the largest reasonable deviation due to this effect is only tens of ns .
the pulsar community must specify a frequency - dependent dispersion measure delay . at radio wavelengths ( 21 cm )
, the delay can be as much as 1 s , but the dispersion delay contributes at less than 1 @xmath1s shortward of @xmath57 m .
the most practical way to precisely calculate the bjd@xmath0 time stamp is using jpl s de405 ephemeris .
it contains the position of thousands of bodies in the solar system , including the sun , planets , spacecraft , moons , asteroids , and comets .
it is oriented to the international celestial reference frame ( icrf ) , which is consistent with the fk5 system at j2000.0 within the 50 mas error of fk5 , and has its origin at the ssb with its axes fixed with respect to extragalactic objects @xcite .
therefore , it is recommended to use the 3-space cartesian coordinates retrieved from the jpl de405 ephemeris directly with the j2000 object coordinates in equation ( [ eq : roemer ] ) .
the following is an outline of the steps required to properly calculate the bjd@xmath0 . 1 .
calculate the midexposure time in jd@xmath5 .
most fits image headers give date - obs in utc at the beginning of exposure .
if high precision is required ( @xmath58 s ) , read the caveats about clock precision in [ sec : clockacc ] carefully ; depending on the sky conditions or variability of the object , one may need to account for the flux - weighted mean time of exposure .
2 . convert the midexposure time to jd@xmath0 by applying @xmath7 ( eq . [ [ eq : utc2tdb ] ] ) . for times accurate to 3.4 ms , one can use the simpler jd@xmath14 and calculate bjd@xmath59 ( using the positions of planets delayed by the tt - tdb offset ) . the difference between bjd@xmath59 and bjd@xmath14
is no more than @xmath10 200 ns , which is well below the precision of the bjd@xmath14 .
if better than 30 @xmath1s precision is required , the tt(bipm ) - tt(tai ) offset must be applied .
3 . retrieve the jpl ephemeris of the observing station for the times spanning the observing window .
jpl s horizons system is designed for this . to return inputs for use with equation ( [ eq : roemer ] ) and j2000 target
coordinates : select `` vector table '' , the ssb as the coordinate origin , and `` earth mean equator and equinox of reference epoch '' for the reference plane .
this will return the cartesian coordinates of the observing station with respect to the ssb in the j2000 earth mean equator reference frame at steps as small as 1 minute in ct , which is the same as tdb .
if the observing station is on earth and better than 20 ms timing is required , another ephemeris must be generated ( from horizons ) for the observer s position with respect to the non - rotating geocenter , and added to the geocentric positions .
we note that the precise conversion from latitude , longitude , and elevation to the cartesian coordinates with respect to the non - rotating geocenter is not trivial , and requires tables of measured precession and nutation of the earth .
interpolate the positions of the observing station to each midexposure time jd@xmath0 .
input the interpolated x , y , and z positions of the observing station , and the target s j2000 earth mean equator coordinates into equation ( [ eq : roemer ] ) . depending on the distance to the target and the precision required , equation ( [ eq : exactroemer ] ) ( and the target s ephemeris ) or equation ( [ eq : roemercorr ] ) may be required
. one must be careful to use sufficiently accurate target coordinates . 7 .
if greater than 0.1 ms precision is required , apply the shapiro correction ( equation [ [ eq : shapiro ] ] ) .
an ephemeris of the sun is required , which can be generated from horizons .
if greater than 1 @xmath1s precision is required , apply the additional einstein correction for the observing station s position with respect to the geocenter .
the geocentric velocity is required , which can also be given by horizons .
our idl code implementing this procedure is available online .
it requires the jd@xmath5 at midexposure and target coordinates ( @xmath19 , @xmath20 ) in j2000 as inputs .
we outline its procedure here .
more explicit details , as well as the calling procedure and dependencies , are commented inside the code .
we compute @xmath7 using craig markwardt s tai_utc program to read the leap second table , and his tdb2tdt program to compute the tt - tdb correction , which uses a 791-term fairhead and bretagnon analytical approximation to the full numerical integration , with an error of 23 ns @xcite .
our code will automatically update its leap second table the first time it runs after every january 1 or july 1 , but this requires a periodic internet connection and the use of the wget program .
it will terminate on failure to update , but this protection can be bypassed for those that elect to ( or have to ) update their table by hand . by default , we ignore the @xmath27s tt(bipm ) - tt(tai ) correction , which would require a constant internet connection , would not apply to data acquired in the previous month , and is likely negligible for most applications .
however , our code can optionally correct for it if an up - to - date file is supplied . to read and interpolate the ephemeris from jpl , we use craig markwardt s routines jplephread and jplephinterp for the earth , sun , and other planets .
if the observing station is space - borne , the smaller ephemeris used with those programs does not include satellites , so we use an expect script to automate a telnet session to the horizons system and automatically retrieve the ephemeris , which we quadratically interpolate to the desired times using idl s interpol .
the accuracy of this interpolation depends on how quickly the object s position is changing and the step size of the ephemeris .
horizons can only return @xmath60 data points per query , so the smallest step size ( 1 minute ) limits the calculation to a range of 60 days . for the geocenter ,
a 100 minute step size is sufficient for 60 ns accuracy , but , for example , a 2 minute step size is required for 1 @xmath1s accuracy for the _ hst _ ( though it is still limited by its clock accuracy ) .
we have found that a 10 minute step size is adequate for 1 ms timing for most objects and allows a range of nearly 2 yr .
if the observer is on the earth , and the coordinates ( latitude , longitude , and elevation ) are given , we correct for the additional delay . if no observer - specific information is given , we assume the observer is at the geocenter , and the result will be biased by @xmath61 ms ( fig . [
fig : geocenter ] ) . if the target s ephemeris can be returned by horizons and its unique name
is given , we use our expect script to generate its ephemeris too , and calculate the exact @xmath6 ( eq .
[ [ eq : exactroemer ] ] ) .
if not , and instead the distance is given , we use the two - term approximation to the spherical wave solution ( eq . [ [ eq : roemercorr ] ] ) .
otherwise , we use the plane wave approximation ( eq . [ [ eq : roemer ] ] ) .
lastly , we include the shapiro correction and the additional einstein correction due to the position of the observer with respect to the geocenter , either from the surface of the earth ( if given the coordinates ) , or the spacecraft . in the geocentric case , our code agrees with barycen to 200 ns ( peak to peak ) and the authors of barycen report that their code agrees with fxbary to 1 @xmath1s .
the ephemeris we generate for a location on the surface of the earth agrees with horizons to 20 nano - lt - s , and the geocentric bjds we calculate from horizons ephemeris agree with the bjds we calculate using craig markwardt s routines within 10 ns .
the near - exact agreements between these methods is not surprising , and do not necessarily indicate that they are accurate to better than 1 @xmath1s .
our code was inspired by barycen and both rely on craig markwardt s routines ( the difference comes from the fact that we index the jpl ephemeris with jd@xmath0 instead of jd@xmath14 ) , and all methods use jpl s de405 ephemeris .
the primary advantage of our code is that it includes the jd@xmath5 to jd@xmath14 correction ( but can optionally ignore it ) .
the choice of starting with jd@xmath5 is a departure from what is typically done with such time stamp calculators , but we feel this is a far more robust starting point . the current confusion has shown that many assume jd@xmath5 as the starting point , which is likely due to a lack of explicitness in the programs and/or unfamiliarity with various time standards .
our hope is that people are unlikely to make the opposite mistake ( assume the input should be jd@xmath14 instead of jd@xmath5 ) since our code is very explicit and calculating the jd@xmath5 is almost always a trivial calculation from the date - obs fits header keyword .
additionally , our code can easily correct for the observer s position on the earth or from a spacecraft , and can include the spherical wave correction . in order to schedule observations ,
even @xmath26 10 minute precision is generally good enough , and one can approximate bjd@xmath0 @xmath53 jd@xmath5 ; for more demanding observing schedules , we provide software to iteratively calculate the reverse correction . along with the idl source code
, we provide a web - based interface to our codes , though not every feature is enabled .
specifically , it is limited to 1 ms precision , can only do one target at a time , only does the plane wave approximation , and is limited to 10,000 jds at a time .
those with applications for which these features are too limited should download our source code and run it locally .
timing of transient events is a powerful tool for characterizing many astronomical phenomena . in the field of exoplanets in particular ,
the search for variations in the times of primary transits and secondary eclipses , or transit timing variations , is one of the most promising new techniques for studying planetary systems . the accuracy with which transit times , and indeed any transient phenomenon , can be measured is limited not only by the data themselves , but by the time stamp to which the transit time is referenced .
as the quality of transit timing data crosses the threshold of 1 minute precision , the precise time standard and reference frame in which event times are quoted becomes important .
achieving uniform and accurate time stamps with accuracies of better than 1 minute that can reliably be compared to one another requires extraordinary care in both our techniques and our terminology .
we have found that the time standards adopted by various groups that measure transit times can differ by as much as a minute , and are typically left unspecified . as these ambiguities can be significant compared to the timing precisions that are quoted
, they may therefore lead to spurious detections of transit timing variations or biased eccentricity measurements .
here we have summarized the effects one must consider in order to achieve timing precision of 1 @xmath1s .
we argue that the bjd@xmath0 is nearly the ideal time stamp , being as reliable as any time stamp can be without being unique to each target system . on the other hand , bjd@xmath5 and the hjd in any form should be avoided whenever possible .
most importantly , we emphasize that the time standard should always be explicitly stated .
any time stamp that is quoted without a time standard should be assumed to be uncertain to at least 1 minute . unless the time standards used in programs or algorithms have been independently confirmed
, one should avoid using ones that do not precisely specify the input and output time standard .
in addition , the arrival time at the observing site along with its time standard ( e.g. , jd@xmath5 ) should also be specified .
this will remove any ambiguity in the time stamp , allow others to apply improved corrections should more precise ephemerides become available in the future , and allow others to check that original conversion was done accurately enough for their purpose .
finally , we have written an idl program for general use that facilitates the use of bjd@xmath0 to an accuracy of 1 @xmath1s , provided that the inputs are sufficiently precise , and we provide a web - based interface to its most useful features .
we would like to thank craig markwardt for his fundamental routines that make ours possible , the help desks at the various space telescopes and at iers for answering our questions , the anonymous referee , steve allen , richard pogge , joseph harrington , roberto assef , andrew becker , mercedes lopez - morales , christopher campo , drake deming , ryan hardy , heather knutson , eric agol , and joshua winn for useful discussions , and wayne landsman and jrn wilms for managing the idl astronomy libraries .
we looked in detail at several readily available tools for the bjd / hjd calculation , and have been in contact with many people in the exoplanet community and the help desks for several major space telescopes .
we summarize our findings here to demonstrate how easily errors of up to 1 minute can be introduced and to stress the importance of specifying the time precisely .
we caution the reader not to trust our general findings for specific cases , but always to confirm what has been done in each case where 1 minute timing accuracy is required but the time standard has not been specified explicitly .
ground - based observers have typically used one of the following methods to calculate the bjd
. however , most fits image headers give the date - obs and time - obs keywords in utc .
we have found that most people , when starting with jd@xmath5 end up quoting hjd@xmath52 or bjd@xmath52 .
jpl s horizons ephemeris calculator , which is used by many to calculate the bjds from space telescopes , and can be used to calculate ground - based bjds , returns the time in jd@xmath62= t@xmath28= jd@xmath0when the ephemeris type is `` vector table '' .
any conversion that uses a horizons ephemeris in jd@xmath0 but indexes it with jd@xmath5 , as had been done by several people we spoke with , will calculate bjd@xmath52 , which can be offset from the true bjd@xmath5 by up to 13.4 ms ( as shown in fig .
[ fig : utcvutc ] ) , and offset from the uniform bjd@xmath0 by more than 1 minute ( as shown in fig .
[ fig : tdbvutc ] ) .
iraf s setjd calculates the hjd , but calls for ut , which is likely to be interpreted as jd@xmath5 . in this case
, it would calculate the drifting quantity hjd@xmath52 .
if tt were used instead , it would calculate the hjd@xmath14 , accurate to @xmath108 s. the idl routines helio_jd ( for hjd ) , from the idl astronomy library curated by wayne landsman , and barycen ( for bjd ) , from the institut fr astronomie und astrophysik idl library , maintained by jrn wilms , both call for the gjd , which , we remind the reader , can be specified in any time standard .
often , this is interpreted as jd@xmath5 , in which case they would calculate hjd@xmath52 or bjd@xmath52 , respectively . if tt were used , they would calculate the hjd@xmath14 , accurate to @xmath108 s , or bjd@xmath0 , accurate to the geocentric correction ( 21.3 ms ) .
nasa s high energy astrophysics science archive research center ( heasarc ) created the tools fxbary and later the improved version faxbary , both of which call axbary to calculate the bjd .
their documentation is precise and correct , but quite long and may be difficult for the uninitiated to follow .
therefore , it would not be surprising for users of axbary to input either utc or tt , in which case they could generate either the bjd@xmath52 to the accuracy of the leap seconds or the bjd@xmath0 to the accuracy of the geocentric correction ( 21.3 ms ) . currently , common google results turn up various applets , spreadsheets , programs , or algorithms to calculate the hjd that explicitly call for jd@xmath5 or jd@xmath63 as an input . unless explicitly mentioned otherwise , it is usually safe to assume the time standard used as input will be the time standard used throughout their calculation .
thus , these algorithms and applets will very likely calculate hjd@xmath52 .
however , they are perfectly capable of calculating hjd@xmath14 if given jd@xmath14 as an input .
epoxi has the mid - exposure time bjd@xmath0 in the header for the intended pointing under the fits header keyword kpkssbjt .
this can be used directly , as long as one is careful about the intended target , so it would be very surprising if a bjd from epoxi was not bjd@xmath0 .
we recalculated the bjd@xmath0 using the horizons ephemeris as described in
[ sec : calculating ] and an example fits header given to us by the epoxi help desk .
in their example fits header , they pointed at the moon , which is not infinitely far away , so we must use equation [ eq : exactroemer ] for @xmath6 . with this method , we agree with the kpkssbjt header value to @xmath64 ms , the limit of the precision of the keyword .
we also redid the bjd@xmath0 calculation of hat - p-4b as described in the _ report on the calibration of epoxi spacecraft timing and reduction to barycentric julian date _ of august 2009 by hewagama et al .. we find agreement with the quoted kpkssbjt fits header keyword to 47 ms . while this is much better than the 0.41 s difference calculated by hewagama et al .
( a difference they attribute to `` cumulative rounding limits '' in their method ) , we believe our method to be far more precise .
however , we could not obtain access to the original headers and were unable to determine the source of the discrepancy .
given the very good agreement with the calculation of the moon above , our best guess is that the target coordinates used by the epoxi pipeline differed from the published values for hat - p-4b .
the 47 ms difference could be explained by a @xmath65 discrepancy in r.a .
, a @xmath66 discrepancy in declination , or some combination thereof .
chandra stores their date - obs keyword in tt .
their more precise tstart and tstop keywords are expressed in seconds after 1998 jan 1 , 00:00:00 tt .
this departs from what is typically done , which may lead to confusion , but it makes the conversion to a uniform time stamp much more straightforward and less likely to drift by the leap seconds .
they provide extensive directions online to calculate the bjd@xmath0 using axbary , so it is likely that anyone using chandra who quotes a bjd is using bjd@xmath0 .
the fits headers of _ hst _ state that their date - obs and time - obs keywords are ut . we contacted the _ hst _
help desk for clarification , since ut is ambiguous .
help desk response stated that their clock reports utc accurate to @xmath1010 ms , but `` due to variabilities and quantization in the particular science instruments operations , the actual time light begins falling on the detector is not known to better than about @xmath101 second , @xmath26 50% ( rough estimate ) . ''
it is possible for the hst engineering team to calibrate these variations , but they have limited resources and have no plans to do so .
it is thought that this error is some combination of random and systematic errors , but the precise breakdown is unknown .
this potential @xmath101 s systematic error may have important implications for the reliability of the transit times quoted with hst observations , most importantly , the 3 s error of the transit time of hd189733b @xcite .
hst does not calculate the hjd or bjd at any point , leaving the calculation up to each individual observer .
our experience with ground based observers suggests that most people will end up quoting an hjd@xmath52 or bjd@xmath52 .
the kepler data release notes 2 describe how to calculate the bjd from utc , but do not include the correction to tdb .
they mention the horizons ephemeris , but neglect to mention that its output time is in ct , not utc ; thus it appears they calculate bjd@xmath52 , though we were unable to confirm this .
in addition , the time stamp uncertainty may be much larger than typical , so it is worth quoting from the kepler data release notes 5 ( released 2010 june 4 ) : the advice of the dawg [ data analysis working group ] is not to consider as scientifically significant relative timing variations less than the read time ( 0.5 s ) or absolute timing accuracy better than one frame time ( 6.5 s ) until such time as the stability and accuracy of time stamps can be documented to near the theoretical limit .
the _ spitzer _ pipeline calculates the hjd@xmath5 for the intended pointing ( presumably the target ) at the end of the exposure , subtracts the full exposure time , and records the result in the header as hjd .
depending on the exposure time , this will produce roughly a 10 ms effect similar to that shown in figure [ fig : utcvutc ] , and depending on how close the intended pointing was to the object of interest , may produce a @xmath67 s effect similar to figure [ fig : bjdra ] . however , this effect is negligible compared to both the @xmath108 s accuracy of the hjd ( fig .
[ fig : bjdvhjd ] ) and the number of leap seconds that may have elapsed between observations ( fig .
[ fig : tdbvutc ] ) .
one typically quotes the hjd at the midexposure time . since _
spitzer _ quotes the hjd@xmath5 at the beginning of the exposure , using the unmodified _ spitzer _ hjds would produce a systematic offset of half the exposure time , though experienced observers correct for this . also contributing to this confusion
, the fits header keyword utcs_obs is incorrectly documented . while the documentation states that it is seconds after j2000 et , it is actually seconds after january 1st , 2000 12:00 utc + @xmath25 - 32 .
therefore , trusting the documentation as is will unwittingly lead to a difference of @xmath68 s. | as the quality and quantity of astrophysical data continue to improve , the precision with which certain astrophysical events can be timed becomes limited not by the data themselves , but by the manner , standard , and uniformity with which time itself is referenced . while some areas of astronomy ( most notably pulsar studies ) have required absolute time stamps with precisions of considerably better than 1 minute for many decades , recently new areas have crossed into this regime . in particular , in the exoplanet community , we have found that the ( typically unspecified ) time standards adopted by various groups can differ by as much as a minute .
left uncorrected , this ambiguity may be mistaken for transit timing variations and bias eccentricity measurements .
we argue that , since the commonly - used julian date , as well as its heliocentric and barycentric counterparts , can be specified in several time standards , it is imperative that their time standards always be reported when accuracies of 1 minute are required .
we summarize the rationale behind our recommendation to quote the site arrival time , in addition to using bjd@xmath0 , the barycentric julian date in the barycentric dynamical time standard for any astrophysical event .
the bjd@xmath0is the most practical absolute time stamp for extra - terrestrial phenomena , and is ultimately limited by the properties of the target system . we compile a general summary of factors that must be considered in order to achieve timing precisions ranging from 15 minutes to 1 @xmath1s .
finally , we provide software tools that , in principal , allow one to calculate bjd@xmath0 to a precision of 1 @xmath1s for any target from anywhere on earth or from any spacecraft . |
imagine positive charge + q essentially fixed to the origin of coordinates of system @xmath0 by virtue of having a large rest mass , @xmath1 .
let a positive charge , q , of mass @xmath2 be placed distance @xmath3 from q in @xmath0 and experience the influence of a coulomb force , @xmath4 . the amounts of charge are chosen such that their repulsion just balances their gravitational attraction @xmath5 .
having set up this nice balance , you depart along the @xmath6-axis at speed @xmath7
. obviously , the masses will remain balanced from your new perspective in system @xmath8 .
but in your now moving frame it appears that both electric and magnetic forces act on @xmath9 such that a magnetic field @xmath10 is created by the motion of q relative to you , where @xmath11 is the ( foreshortened ) electric field at q , due to q. here @xmath12 . the sum of electric and magnetic forces as calculated in @xmath8 is @xmath13 , where @xmath14 is the velocity of q relative to s and , in this case @xmath15 .
this leads to combined electric and magnetic forces of @xmath16 clearly this is a weaker force than that observed in the original system @xmath0 and if gravity were not present it would lead to a slower separation of @xmath17 and @xmath9 as they repelled each other . from @xmath8 , you would see the separation occur in a time dilated way . but now consider the gravitational attraction between masses @xmath18 and @xmath19 .
the fact that you chose to go away and leave this balanced arrangement of electric and gravitational forces can surely have no bearing on its subsequent behavior . to remain motionless
as observed in system @xmath8 , the total force of gravitational origin between the objects in system s must surely be @xmath20 and as originally balanced in @xmath0 , @xmath21 . the part of the force in eq .
( 2 ) that is velocity dependent can be considered to be a gravitomagnetic force . in the s frame
it must exist and must , and does , balance the ordinary magnetic force .
one would expect that eq .
( 2 ) could be obtained by analogy with the electromagnetic case .
@xmath18 would be the source of a @xmath22 field as observed in @xmath8 to be @xmath23 and also a source of a gravitomagnetic field @xmath24 in which @xmath19 moves .
then if the sum of gravitostatic and gravitomagnetic forces would be @xmath25 we would obtain eq .
( 2 ) without further ado . but
the weak - field limit of general relativity produces a gravitomagnetic field that is @xmath26 ( see , e.g. ref [ 1 ] ) which would lead to gravitomagnetic forces four times stronger than required to balance the ordinary magnetic effect .
the force of gravitational origin in eq .
( 3 ) would become @xmath27 i leave the reader to ponder where this goes wrong .
to see what gravitomagnetic effects we should expect , let us dispense with the electric charges on m and m ( @xmath28 ) and just consider the free fall of @xmath19 toward @xmath18 down the @xmath3-axis starting from rest . in either frame , s or @xmath0 , the small test mass , @xmath19 , would follow a geodesic . in the weak field limit of @xmath0 and in isotropic
coordinates the metric is @xmath29 where @xmath30 .
a lorentz transform to frame @xmath8 [ @xmath31 yields @xmath32 where , at the instant depicted in figure 1 , @xmath33 .
the last terms with @xmath34 arise from the lorentz transformation .
the coefficient of @xmath35 is exactly 4x larger than expected from newtonian theory combined with special relativity .
it partially describes the `` gravitomagnetic '' effect .
of particular interest is that @xmath36 can be considered as arising from a vector potential @xmath37 where @xmath38 is the mass current density and the gravitomagnetic field is taken to be @xmath39 . but
this term with its extra factor of 4 can not completely describe all of the gravitomagnetic effects , for we have seen that that the assumption that it does leads to an unacceptable result for our gedanken experiment . in the weak - field limit , we have @xmath40 , neglecting terms involving @xmath41 .
we can use the inverse of the minkowski metric @xmath42 to raise indexes . the geodesic equation for the y - component of motion of @xmath19
is @xmath43 taking @xmath44 , there follows @xmath45 and @xmath46 .
then with a little algebra and using @xmath47 , we obtain @xmath48= ( 1-v^2/c^2)\frac{\partial \phi}{\partial y}\ ] ] which is exactly the same result that we get from eq .
( 2 ) if the inertial mass on which @xmath49 acts is @xmath50 .
( 9 ) is a geodesic equation that is correct to order @xmath51 .
what happens in eq .
( 9 ) is that the term @xmath52 manages to cancel @xmath53 of the effect of the factor of 4 in the gravitomagnetic field term .
it would have been very surprising to have obtained any different result , for that would have left the general relativistic approach in conflict with special relativity . from the standpoint of special relativity ,
the freely falling mass could serve as an interval timer , with the interval corresponding to the time to fall some particular small distance down the y-axis .
as observed in @xmath8 , the interval would be dilated and that is exactly what is described by eqs .
( 2 ) or ( 9 ) .
apparently magnetic and gravitomagnetic forces provide mechanism for enforcing time dilation .
the reason that the combination of eqs .
( 3 ) and ( 4 ) fails to yield eq . ( 9 ) and correctly describe the free - fall problem is that they are only correct to order @xmath54 . they have been combined to produce terms of order @xmath51 , but apparently not all of them . under these circumstances , it would also be expected that they would not correctly predict all gravitomagnetic effects to order @xmath51 .
the gedanken experiment can be made more interesting by allowing m to orbit m in a circular orbit of radius b as observed in @xmath0 .
as observed in @xmath8 , the orbital time period will be dilated and it can be shown from the kinematic equations of special relativity that the orbit diameter will be the usual 2b along y - axis , but be contracted to @xmath55 along the x - axis .
further , a special case in agreement with special relativity can be obtained from eq .
( 8) using @xmath56 , where @xmath57 would be the speed of m relative to m , as observed in s , when m is on the y - axis . neglecting terms in @xmath58 , the resulting acceleration of m for these positions
would be @xmath59 the last term of the equation accounts for the small difference of speeds of m and m relative to reference frame @xmath8 .
the foregoing ruminations have been motivated by recent discussions of laser lunar ranging .
murphy , nordtvedt and turyshev [ 2 ] have recently published new results that claim to have verified eq .
( 4 ) as a source of perturbing acceleration acting on the lunar orbit .
it is claimed that they have established this weak - field limit prediction of general relativity to an accuracy of about 0.1% , which is far better confirmation than expected from the gravity probe b experiment .
ciufolini [ 3 ] has pointed out that the result of murphy et al .
is merely a coordinate dependent effect .
their results were obtained relative to a solar system barycenter frame ( hereafter ssb ) and thus could be transformed away .
kopeiken , on the other hand , says [ 4 ] that there should be no observable effect at all relative to the ssb .
while the gravitomagnetic field due to the sun s rotation can not be transformed away , it is much too small to have a measurable effect on the lunar orbit .
if a gravitomagnetic effect exists relative to the ssb it must be associated with the motions of earth and moon relative to the ssb . for analysis of the lunar motions , the orbital motion of earth
would be regarded as the cause of a @xmath60 field that is some four orders of magnitude larger than that due to solar rotation at the position of the moon . at the time of a new moon
, the situation would correspond to the masses m and m positioned as shown in figure 1 , with the sun located a long way up the y - axis .
system @xmath0 would be comoving with the earth while system @xmath8 would remain at rest relative to the ssb .
thus whatever would be observed in s should be the same as what is measured relative to the ssb for this particular configuration .
considering the outcome of the previous gedanken experiment , the major effects that we should observe would be time dilation and length contraction . for a lunar orbit radius of @xmath61 and lunar velocity of @xmath62 , the contraction of the lunar orbit parallel to the nearly joint earth - moon orbital motion about the sun would be about 1.9 m. this
would be observed in frame s of fig . 1 , but not in the ssb frame .
in view of the gedanken experiment and eqs .
( 2 ) , ( 9 ) and ( 10 ) , we should not expect the combination of eqs .
( 3 ) and ( 4 ) to be confirmed in the ssb frame , notwithstanding the apparent confirmation by murphy , nordtvedt and turyshev [ 1 ] .
they transformed the llr to the ssb frame and then fitted the lunar trajectory using multiple adjustable parameters . in recent years the amplitude of the dominant gravitomagnetic oscillation has been reported as @xmath63 [ 1 ] , @xmath64 [ 5 ] and @xmath65 [ 6 ] .
although the residuals of these curve fits to the lunar range observations are typically less than one cm , the oscillation amplitudes are obviously much more variable . until the statistical errors of determination of the oscillation amplitude exclude the alternatives considered here , the claim that eq .
( 4 ) has been confirmed is unwarranted .
finally , we should note that calculating gravitomagnetic effects relative to a geocentric frame removes several opportunities for error . in this frame
the moon moves in a gravitomagnetic field due to the apparent motion of the sun .
shahid - saless has calculated the expected amplitude of gravitomagnetic lunar orbit oscillations to be about @xmath66 in the geocentric frame . | a gedanken experiment implies the existence of gravitomagnetism and raises a question about what we know about the weak - field limit of the gravitomagnetic field of general relativity . |
there has been great interest in using there darboux transformation @xcite for the analysis of physical systems @xcite and for finding new solvable systems @xcite .
it has been shown that the transformation method is useful in finding soliton solutions of the integrable systems @xcite and constructing supersymmetric quantum mechanical systems @xcite .
it is well known that susy qm is basically equivalent to the darboux transformation and the factorization properties of the schrdinger equation @xcite .
the darboux transformation of the one - dimensional stationary dirac equation is equivalent to the underlying quadratic supersymmetry and the factorization properties @xcite . despite of the growing number of papers in this field
many questions still remain open and require further study . in particular , the author are aware of only several papers @xcite devoted to the susy transformations at the level of the green functions . in @xcite was obtained the integral relation between the green functions for two susy partner hamiltonians of the one - dimensional schrdinger equation with discrete spectra . in @xcite the integral relation between the green functions corresponding two susy partner hamiltonians of the one - dimensional schrdinger equation
is generated to the case of continuous spectrum .
the exact green function of the time - dependent schrdinger equation was obtained in @xcite .
an interesting open problem is to find the analogous results for the dirac equation . in this paper
, we construct the darboux transformation of the green function for the regular boundary problem of the one - dimensional stationary dirac equation with a generalized form of the potential and obtain formulas for the unabridged trace of the difference of the modified and initial green functions .
the rest of the paper is organized as follows . in section 2
we construct a green function for the initial regular boundary problem of the one - dimensional dirac equation with the generalized form of the potential . in section 3
we consider the green function for the transformed regular boundary problem of the one - dimensional dirac equation .
we construct the darboux transformation of the green function . in section 4
we obtain formulas for the unabridged trace of the difference of the modified and the initial green functions and consider the spectral representation of the corresponding unabridged trace . in conclusion the summarize our results and speculate about some perspective .
the green function of the one - dimensional dirac equation @xmath0 with the dirac hamiltonian of the form @xmath1 @xcite is needed for obtaining the solution @xmath2of the following inhomogeneous equation @xmath3 where @xmath4 and @xmath5 are the endpoints of an closeg interval .
earlier this method allowed the authors of the paper @xcite to obtain the green function of the one - dimensional dirac equation with the potential @xmath6 in the present paper , we will obtain the green function of the dirac equation with the generalized form @xcite of the potential @xmath7 where @xmath8 , @xmath9 and @xmath10 are real functions of @xmath11 , @xmath12 is the mass of a particle , @xmath13 , @xmath14 are usual pauli matrices .
the generalized potential is the self - adjoint potential .
we consider the following boundary conditions for the components @xmath15 , @xmath16 of the solution ( [ thesolution ] ) to ( [ 2 ] ) : @xmath17 @xmath18 we suppose that the solutions @xmath19 , @xmath20 of the dirac equation with the generalized form of the potential obey the following conditions : @xmath21 let us construct the matrix @xmath22 where @xmath23 is the wronskian of the two functions @xmath19 and @xmath20 .
one can reality examine that the vector - function @xmath24 is the solution to the inhomogeneous equation ( [ 2 ] ) and therefore the matrix ( [ g ] ) is the green function of the regular boundary problem of the one - dimensional stationary dirac equation with the generalized form of the potential ( [ 3 ] ) .
let us construct the spinor @xmath25 @xmath26 in detail , eq .
( [ g ] ) becomes @xmath27 we write down a detailed form of the components of the spinor @xmath28 @xmath29 @xmath30 from these equations , taking into consideration ( [ gff ] ) and @xmath31 @xmath32 we find the components of the spinor @xmath33 , @xmath34 .
we consider the components of the spinor in the case @xmath35 , then @xmath36 $ ] , after that in the case @xmath37 , when @xmath38 $ ] , and as a result , we find the components of the spinor @xmath33 , @xmath34 in the general case : @xmath39 @xmath40 now we check that the vector - function @xmath41 is the solution of the inhomogeneous equation ( [ 2 ] ) @xmath42 @xmath43 where @xmath44 , @xmath45 are the components of the spinor @xmath46 .
let us show that equality ( [ 17 ] ) is true .
for that we differentiate @xmath33 : @xmath47 since the spinors @xmath48 and @xmath49 are the solutions to the dirac system of equations , the following equalities are correct : @xmath50 @xmath51 substituting equalities ( [ 12717 ] ) , ( [ 1272 ] ) into ( [ 127777 ] ) , we obtain @xmath52-f_2(x).\end{aligned}\ ] ] from ( [ 12 ] ) , ( [ 14 ] ) for @xmath33 ,
@xmath34 we get @xmath53 therefore , ( [ 17 ] ) is correct . similarly ,
equality ( [ 16 ] ) can be proven .
the spinor - function @xmath54 is the solution to ( [ 2 ] ) with the generalized form of the potential .
now we demonstrate realization of the regular boundary conditions ( [ s1 ] ) , ( [ s2 ] ) for spinor - function @xmath54 @xmath55 @xmath56 where @xmath33 , @xmath34 are the components of the spinor @xmath41 .
we compute ( [ 12 ] ) , ( [ 14 ] ) when @xmath57 , @xmath58 : @xmath59 @xmath60 @xmath61 @xmath62 since @xmath63 and @xmath64 , where @xmath65 is the arbitrary function of @xmath66 , the equalities ( [ 127 ] ) , ( [ 147 ] ) , ( [ 1277 ] ) , ( [ 1477 ] ) take the following forms : @xmath67 @xmath68 @xmath69 @xmath70 now , taking into account ( [ 1271 ] ) , ( [ 1471 ] ) , ( [ 12771 ] ) , ( [ 14771 ] ) , we check the conditions ( [ sy1 ] ) , ( [ sy2 ] ) : @xmath71 @xmath72 due to the boundary conditions ( [ 5 ] ) for the functions @xmath48 , @xmath49 the relations are ( [ sy1 ] ) , ( [ sy2 ] ) are valid .
hence , the matrix @xmath73 ( [ g ] ) , is the green function for the regular boundary problem of the one - dimensional stationary dirac equation with the generalized form of the potential .
the matrix ( [ g ] ) can be written in the following form : @xmath74 where @xmath75 , @xmath76 are the heaviside step functions .
thus , we have constructed the green function of the initial problem . + * example 1 * let us calculate the green function of the initial problem for the dirac ( [ 1 ] ) free particle equation with the @xmath77 , @xmath78 .
we first chose the two linearly independent solutions @xmath79 , @xmath80 for the free particle case : @xmath81@xmath82 next we apply the boundary conditions to functions @xmath80 , @xmath79 for the case @xmath83 , @xmath84 : @xmath85 from it follows that @xmath86 , @xmath87 .
also , we find the final expression for the green function of the initial regular boundary problem ( [ s1 ] ) , ( [ s2 ] ) with @xmath88 : @xmath89 @xmath90 @xmath91
before constructing darboux transformed green function , let us shortly consider the darboux transformation method for the dirac equation .
the so - called _ darbuox transformation operator _ @xmath92 and the potential @xmath93 of the transformed dirac hamiltonian @xmath94 have respectively the following form : @xmath95.\end{aligned}\ ] ] the so - called _ transformation function _ u is @xmath96 matrix consists from two solutions @xmath97 of the initial dirac equation with @xmath98 , where @xmath99 @xmath100 are the neighboring energy levels .
if the function @xmath80 is solution of initial equation and @xmath101 , then the function @xmath102 is solution of the transformed dirac equation @xmath103 the green function of the equation ( [ 111 ] ) with the regular boundary conditions for the components @xmath104 , @xmath105 @xmath106 @xmath107 of the spinor @xmath108 one can represent in the form : @xmath109 \(i ) if @xmath110 , @xmath111 , we have the green function of the dirac equation with the transformed potential and the initial regular boundary conditions .
\(ii ) if @xmath112 , @xmath113 , we have the green function of the dirac equation with the transformed potential and the modified regular boundary conditions .
further we consider only the case when @xmath110 , @xmath111 .
now let us consider the construction of the darboux transformed green function on the basis of the concrete example .
* example 2 * let us first chose the transformation functions as follows : @xmath114 next , obtain from the free particle case the transformed potential @xmath115 where @xmath116 it can be checked that only these two transformation matrices ( differing from each other only by sign of @xmath117 ) provide the conservation of the initial regular boundary condition .
, @xmath118,height=264 ] now the darboux transformed functions @xmath119 , @xmath120 one can represent in the form : @xmath121 finally , the green functions of the dirac equation with the transformed potential may be given by the formula : @xmath122 where @xmath123 @xmath124 thus , we have constructed the green function of the transformed dirac equation with the potential and the initial regular boundary conditions .
in this section we calculate the unabridged trace of the difference of the transformed and the initial green functions .
firstly , let us consider the unabridged trace of the transformed green function @xmath125 .
it can be represented in the form : @xmath126 it is known that the action of the transformation operator and the conjugate transformation operator on the spinors may be written in the following way @xcite : @xmath127 also , accounting these properties , we can rewrite ( [ 115 ] ) as : @xmath128 next let us consider the unabridged trace of the difference of the transformed and the initial green functions .
it is obvious @xcite that the spectral representation for the initial and the transformed green functions correspondingly look like as @xmath129 @xmath130 where the functions @xmath131 form an orthonormal set .
the completeness of will be shown in examples 3 and 4 .
the construction @xmath132 is called the _ unabridged trace of the difference of the transformed and the initial green functions_. the spectral representation of is as follows : @xmath133 where @xmath134 , @xmath135 are discrete eigenvalues of @xmath136 and @xmath94 respectively .
if the set is complete then the formula one can represent in the form : @xmath137 now let us show an another derivation of the trace formula for the green functions difference .
let us integrate ( [ 1153 ] ) by parts and apply the trace property @xmath138 to obtain the relation @xmath139 due to ( [ tildaw ] ) and the factorization property from @xcite we obtain : @xmath140 similarly , we would like to write @xmath141 finally , from ( [ 1155 ] ) , ( [ 1157 ] ) we find : @xmath142 here @xmath4 , @xmath5 are limits of the integration . as shown in appendix , right parts of , are equal . * example 3 * in this example we calculate with the help of .
more precisely , take into account , , , we calculate @xmath143 where @xmath83 , @xmath84 : @xmath144 using the lhospital s rule for evaluating of indeterminations , we obtain @xmath145 due to we get : @xmath146 where @xmath147 , @xmath148 .
this result indicate that the spectrum of the transformed dirac hamiltonian differs from the spectrum of the initial dirac operator .
the spectral lines @xmath149 disappear from the transformed spectrum and the set of functions @xmath150 is complete .
+ * example 4 * in this example we calculate in a direct way .
first , we consider @xmath151 next , we integrate the expression ( [ trg0 ] ) @xmath152 @xmath153 @xmath154 similarly , we calculate @xmath155 and obtain that @xmath156 thus , we can conclude that the set of functions @xmath150 is complete .
in this paper , we have studied the darboux transformation of the green functions of the regular boundary problem corresponding to the initial and the transformed potentials of the one - dimensional dirac equation for the case of the dirac hamiltonians with discrete spectrum .
the main results of the paper are the construction of the darboux transformed green function with initial regular boundary conditions and the formulae for an unabridged trace , . for the checking of formulae and we consider the unabridged trace of difference between transformed and initial green functions by usual possible and by formula , both results are equal .
the all results of this paper are studied only for discrete spectrum and regular boundary problem .
an interesting question for the future is generation of these results to the case of continuous spectrum of the dirac equation and the case of problem on the real line and a half - line .
we believe that these problem will investigate in a separate publication .
the author is grateful to dr .
d. antonov for helpful comments of the given work .
i would like also the express my thanks to prof .
b. g. bagrov for useful discussions .
this work was supported in part by the `` dynasty '' fund and moscow international center of fundamental physics .
in this appendix , taking into account , , , , we show the equivalence of expressions for right parts of , @xmath157 where @xmath83 , @xmath84 . | we consider the darboux transformation of the green functions of the regular boundary problem of the one - dimensional stationary dirac equation .
we obtained the green functions of the transformed dirac equation with the initial regular boundary conditions .
we also construct the formula for the unabridged trace of the difference of the transformed and the initial green functions of the regular boundary problem of the one - dimensional stationary dirac equation .
we illustrate our findings by the consideration of the darboux transformation for the green function of the free particle dirac equation on an interval .
+ keywords : dirac equation ; green function ; darboux transformation . |
modern _ real - time systems _ have incurred tremendous challenges to verification engineers .
the reason is that a model process running in a modern real - time system can be built with support from many server processes in the environment .
moreover , the model may also have to respond to requests from several user processes .
the fulfillment of a computation relies not only on the functional correctness of the model , but also on the reactions from the servers and the clients .
for example , a company may submit a task of dna sequencing to a server .
the server then develops a computing budget and decomposes the task into several subtasks ( e.g. , snp finding , alignments ) .
then the server may relegate the subtasks to several other servers .
the decompositions of subtasks may then go on and on .
if the task is to be completed , not only the server for the root task needs to function correctly , but also all the servers for the subtasks have to fulfill their assignments .
thus , to verify the function of the root server , it is only reasonable and practical to assume that all the other supporting servers work correctly . in many industrial projects
, the specification can be given in the concept of state - transition diagrams ( or tables ) .
in such a context , _ simulation - checking _ is an appropriate framework for verifying that a model conforms to the behavior of a specification @xcite .
intuitively , the specification simulates the model if every timed step of the model can be matched by the specification at the same time .
@xmath0 [ exmp.intro ] in figure [ fig.ms_ne ] , we have the state - transition diagrams of two _ timed automatas _ ( _ ta _ )
@xcite .
( 0,0 ) # 1#2#3#4#5 @font ( 5195,3861)(664,-3289 ) ( 976,-361)(0,0)[lb ] ( 1501 , 89)(0,0)[lb ] ( 1501,239)(0,0)[lb ] ( 1501,389)(0,0)[lb ] ( 2926,-811)(0,0)[lb ] ( 5176,-361)(0,0)[lb ] ( 4726,164)(0,0)[lb ] ( 4726,314)(0,0)[lb ] ( 2926,-961)(0,0)[lb ] ( 901,-1261)(0,0)[lb ] ( 2926,-2761)(0,0)[lb ] ( 1501,-1861)(0,0)[lb ] ( 1501,-1711)(0,0)[lb ] ( 2926,-2911)(0,0)[lb ] ( 901,-3211)(0,0)[lb ] ( 976,-2311)(0,0)[lb ] ( 3451,-436)(0,0)[lb ] ( 3376,-2161)(0,0)[lb ] ( 3451,-2386)(0,0)[lb ] ( 3376,-211)(0,0)[lb ] the one in figure [ fig.ms_ne](a ) is for a model @xmath1 while the one in figure [ fig.ms_ne](b ) is for a specification @xmath2 .
we use ovals for the _ control locations _ of the tas while arcs for the transition rules . in each oval , we label the invariance condition that must be satisfied in the location . for example , in location @xmath3 , @xmath1 can stay for at most 20 time units . by each transition rule , we stack its synchronization event , triggering condition ( guard ) , and actions . for convenience ,
tautology triggering conditions and nil actions are omitted .
an event starting with a ` ? ' represents a _ receiving event _ while one with a ` ! ' represents a _ sending event_. for example , for the transition from location @xmath4 to @xmath5 , @xmath1 must send out an event request , be in a state satisfying @xmath6 , and reset clock @xmath7 to zero .
the specification in figure [ fig.ms_ne](b ) does not simulate the model in figure [ fig.ms_ne](a ) since event !
end of @xmath1 can not be matched by any event of @xmath2 .
moreover @xmath2 can neither receive a ?
serve event 15 time units after issuing a ! request event while @xmath1 can
. however , the concept of simulation described in the last paragraph can be too restrictive in practice .
developers of a project usually can not make too much assumption on the environment .
the deadline constraints @xmath8 and @xmath9 can be too restrictive and hurt the extensibility of the model in the future .
another approach in this regard is using _ fairness assumptions _
for example , for the model and specification processes in figure [ fig.ms_ne ] , we may want to check whether @xmath2 simulates @xmath1 under the fairness assumption that the environment functions reasonably .
such an assumption can be captured with the fairness assumption that _ there will always be infinitely many occurrences of event serve_. under this assumption , the @xmath2 in figure [ fig.ms_ne](b ) actually simulates the @xmath1 in figure [ fig.ms_ne](a ) . in this work
, we propose the _ simulation _ with fairness assumptions for the processes in a dense - time setting .
in such a setting , the model and the specification are both
_ generalized bch timed automatas _ ( _ gbta _ ) @xcite with communication channels and dense - time behaviors .
we want to check whether the specification gbta can simulate the model gbta with multiple fairness assumptions . following the approach of @xcite
, we allow for the requirement and analysis of both strong and weak fairness assumptions .
a _ strong fairness _
assumption intuitively means something will happen infinite many times . a _ weak fairness _
assumption means something will hold true eventually forever . for convenience ,
we use two consecutive sets of formulas for fairness assumptions , the former for the strong fairness assumptions while the latter for the weak fairness assumptions .
@xmath0[exmp.intro.fstate ] for the system in figure [ fig.ms_ne ] , we may have the following fairness assumptions .
@xmath10 the fairness assumptions in the above say that a valid computation of the system must satisfy the following two conditions .
_ for the strong fairness assumption of _ @xmath11 : for every @xmath12 , there exists a @xmath13 with @xmath14 such that in the computation at time @xmath15 , the model process is in location @xmath5 .
this in fact says that the model must enter location @xmath5 infinitely many times along any valid computation .
_ for the weak fairness assumption of _ @xmath16 : there exists a @xmath12 such that for every @xmath13 with @xmath14 , the model process is in either locations @xmath4 or @xmath5 .
this in fact says that the model will stabilize in locations @xmath4 and @xmath5 .
the two types of fairness assumption complement with each other and could be handy in making reasonable assumptions .
furthermore , we also allow for both state formulas and event formulas @xcite in the description of fairness assumptions .
state formulas are boolean combinations of atomic statements of location names and state variables . for convenience ,
we use index @xmath17 for the model and index @xmath18 for the specification .
event formulas are then constructed with a precondition , a event name with a process index , and a post - condition in sequence .
@xmath0[exmp.intro.fevent ] for the system in figure [ fig.ms_ne ] , we may write the following strong event fairness assumption . @xmath19 the event specification of @xmath20 means there is an event serve received by process 1 .
the precondition for the event is @xmath5 while the post - condition is @xmath21 .
the strong fairness assumption says that there should be infinite many events serve received by process 1 in location @xmath5
. in general , an event specification can be either a receiving or a sending event .
such event formulas can be useful in making succinct specifications . without such event formulas
, we may have to use auxiliary state variables to distinguish those states immediately before ( or after ) an event from others .
such auxiliary variables usually unnecessarily exacerbate the state space explosion problem .
one goal of our work is to develop a simulation - checking algorithm based on symbolic model - checking technology for dense - time systems @xcite . to achieve this , we focus on a special class of simulations with the restriction of at most one fairness assumption for the specification . for convenience
, we call this class the _ usf _ ( _ unit - specification - fairness _ )
_ simulations_. then we propose a symbolic algorithm for this special class of simulations . to our knowledge , this is the first such algorithm for gbtas . also unlike the fair simulation @xcite checking algorithm based on ranking function in the literature ,
our algorithm is based on symbolic logic formulas manipulation , which has been proven useful in symbolic model checking @xcite .
thus , our algorithm style can be interesting in itself .
we also present a technique for the efficient simulation checking of concurrent systems by taking advantage of the common environment of a model and a specification . to apply the simulation checking algorithms mentioned in the above and in the literature @xcite , we need first construct a product automata of the environment @xmath22 and the model @xmath1 , in symbols @xmath23 .
then we construct a product of @xmath22 and the specification @xmath2 , in symbols @xmath24 .
then we check if @xmath24 simulates @xmath23 . as a result ,
such algorithms incur duplicate recording of the state information of @xmath22 while manipulating representations for the simulation of @xmath23 by @xmath24 .
moreover , different transitions in @xmath22 with the same observable events can also be matched in the simulation - checking .
such matching is not only counter - intuitive in simulation against the same environment , but also incur explosion in the enumeration of matched transitions between @xmath23 and @xmath24 .
our technique is embodied with the definition of a new simulation relation against a common environment .
we have implemented this technique and experimented with benchmarks with and without fairness assumptions .
we have the following presentation plan .
section [ sec.relwork ] is for related work .
section [ sec.prel ] reviews our system models @xcite .
sections [ sec.simf ] presents our simulation for dense - time systems with fairness assumptions .
section [ sec.usf.neg.char ] presents a characterization of the simulation when the specification is a bchi ta .
section [ sec.simf.alg ] presents our simulation checking algorithm based on the characterization derived in section [ sec.usf.neg.char ] .
section [ sec.sim.env ] presents the simulation against a common environment and techniques for performance verification in this context .
sections [ sec.imp ] and [ sec.exp ] respectively report our implementation and experiment .
section [ sec.conc ] is the conclusion .
cerans showed that the bisimulation - checking problem of timed processes is decidable @xcite .
tairan et al showed that the simulation - checking problem of dense - time automatas ( tas ) @xcite is in exptime @xcite .
weise and lenzkes reported an algorithm based on zones for timed bisimulation checking @xcite .
cassez et al presented an algorithm for the reachability games of tas with controllable and uncontrollable actions @xcite .
henzinger et al presented an algorithm that computes the time - abstract simulation that does not preserve timed properties @xcite .
nakata also discussed how to do symbolic bisimulation checking with integer - time labeled transition systems @xcite .
beyer has implemented a refinement - checking algorithm for tas with integer - time semantics @xcite .
lin and wang presented a sound proof system for the bisimulation equivalence of tas with dense - time semantics @xcite .
aceto et al discussed how to construct such a modal logic formula that completely characterizes a ta @xcite .
larsen presented a similar theoretical framework for bisimulation in an environment for untimed systems @xcite .
however no implementation that takes advantage of the common environment information for verification performance has been reported .
proposals for extending simulation with fair states have been discussed in @xcite .
our simulation game of gbtas stems from henzinger et al s framework of fair simulation @xcite .
techniques for simulation checking of gbas were also discussed in @xcite .
we have the following notations .
@xmath25 is the set of real numbers .
@xmath26 is the set of non - negative reals .
@xmath27 is the set of nonnegative integers .
also ` iff ' is if and only if
. " given a set @xmath28 of atomic propositions and a set @xmath29 of clocks , we use @xmath30 as the set of all boolean combinations of logic atoms of the forms @xmath31 and @xmath32 , where @xmath33 , @xmath34 , ` @xmath35'@xmath36 , and @xmath37 .
an element in @xmath30 is called a _ state - predicate_. a ta@xcite is structured as a directed graph whose nodes are _ modes ( control locations ) _ and whose arcs are _
transitions_. please see figure [ fig.ms_ne ] for examples .
a ta must always satisfy its _
invariance condition_. each transition is labeled with events , a _ triggering condition _ , and a set of clocks to be reset during the transitions .
at any moment , a ta can stay in only one _
mode_. if a ta executes a transition , then the triggering condition must be satisfied . in between transitions , all clocks in a
ta increase their readings at a uniform rate .
a ta @xmath38 is a tuple @xmath39 .
@xmath40 is a finite set of modes ( locations ) .
@xmath28 is a finite set of propositions .
@xmath29 is a finite set of clocks .
@xmath41 is the initial condition .
@xmath42 is the invariance condition for each mode .
@xmath43 is the set of process transitions .
@xmath44 is a finite set of events .
@xmath45 is a mapping that defines the events at each transition . @xmath46 and
@xmath47 respectively define the triggering condition and the clock set to reset of each transition . without loss of generality , we assume that for all @xmath48 with @xmath49 , @xmath50 is a contradiction .
we also assume that there is a null transition @xmath51 that does nothing at any location .
that is , the null transition transits from a location to the location itself .
moreover , @xmath52 , @xmath53 , and @xmath54 . given a ta @xmath55 , for convenience , we let @xmath56 , @xmath57 , @xmath58 , @xmath59 , @xmath60 , @xmath61 , @xmath62 , @xmath63 , @xmath64 , and @xmath65 . also , for convenience , we let @xmath66 be the _ invariance predicate _ of @xmath38 .
@xmath0 [ exmp.tas ] we have already seen examples of tas in figure [ fig.ms_ne ] . for the ta in figure [ fig.ms_ne](a ) ,
the attributes are listed in table [ tab.ms_ne.a.attr ] .
@xmath67 \\ e_{{\cal m } } & = & \{(\mbox{\tt idle}_1,\mbox{\tt wait}_1 ) , ( \mbox{\tt wait}_1,\mbox{\tt idle}_1 ) , ( \mbox{\tt wait}_1,\mbox{\tt stop}_1)\ } \\ \sigma_{{\cal m } } & = & \{\mbox{\tt request},\mbox{\tt serve},\mbox{\tt end}\ } \\
\epsilon_{{\cal m } } & = & [ ( \mbox{\tt idle}_1,\mbox{\tt wait}_1)\mapsto\{!\mbox{\tt request}\ } , ( \mbox{\tt wait}_1,\mbox{\tt idle}_1)\mapsto\{?\mbox{\tt serve}\ } , ( \mbox{\tt wait}_1,\mbox{\tt stop}_1)\mapsto\{!\mbox{\tt end}\ } ] \\
\tau_{{\cal m } } & = & [ ( \mbox{\tt idle}_1,\mbox{\tt wait}_1)\mapsto x_1>5 , ( \mbox{\tt wait}_1,\mbox{\tt idle}_1)\mapsto { \mbox{\em true } } , ( \mbox{\tt wait}_1,\mbox{\tt stop}_1)\mapsto x_1>10 ] \\ \pi_{{\cal m } } & = & [ ( \mbox{\tt idle}_1,\mbox{\tt wait}_1)\mapsto \{x_1\ } , ( \mbox{\tt wait}_1,\mbox{\tt idle}_1)\mapsto \{x_1\ } , ( \mbox{\tt wait}_1,\mbox{\tt stop}_1)\mapsto \emptyset ] \\ \end{array}$ ] + @xmath68 $ ] denotes a ( partial or total ) function @xmath69 with @xmath70
. a _ valuation _ of a set is a mapping from the set to another set .
given an @xmath71 and a valuation @xmath72 of @xmath73 , we say @xmath72 _ satisfies _ @xmath74 , in symbols @xmath75 , iff @xmath74 is evaluated @xmath21 when the variables in @xmath74 are interpreted according to @xmath72 .
suppose we are given a ta @xmath38 .
a _ state _
@xmath72 of @xmath38 is a valuation of @xmath76 with the following constraints . for each @xmath77 , @xmath78 .
there exists a @xmath79 such that @xmath80 and for all @xmath81 , @xmath82 .
given a @xmath79 , if @xmath83 , we denote @xmath31 as @xmath84 .
for each @xmath85 , @xmath86 .
in addition , we require that @xmath87 .
we let @xmath88 denote the set of states of @xmath38 .
note that we define a state as a mapping instead of as a pair of control locations and a real mapping as in @xcite .
this is for the convenience of presentation when latter we want to discuss the state - pairs in simulation relations . for any state @xmath72 and real number @xmath12
, @xmath89 is a state identical to @xmath72 except that for every clock @xmath85 , @xmath90 .
also given a process transition @xmath91 , we use @xmath92 to denote the destination state from @xmath72 through the execution of @xmath93 .
formally , if @xmath94 , then @xmath92 is a new state that is identical to @xmath72 except that the following constraints are true .
@xmath95 and @xmath96 . for every clock @xmath97 , @xmath98 . for every clock @xmath99 , @xmath100 . given a @xmath101 and a transition @xmath93
, we write @xmath102 iff @xmath103 , @xmath104 , @xmath105 , and for each @xmath106 $ ] , @xmath107 . for convenience
, we use @xmath108 $ ] to denote such a @xmath109 with @xmath102 . a _ run _ of a ta
@xmath38 is an infinite sequence of state - transition - time triples @xmath110 with the following restrictions .
* non - zeno requirement : * @xmath111 is a non - decreasing and divergent real - number sequence .
that is , @xmath112 and @xmath113 . for all @xmath114 , either @xmath115 or @xmath116{10mm}{0.5pt}\!\!\!\longrightarrow } } } \nu_{k+1}$ ] .
a _ run prefix _ is a finite prefix of a run .
a run prefix or a run @xmath117 of @xmath38 is _ initial _
iff @xmath118 .
suppose we are given a ta @xmath119 .
an _ event - predicate _ is of the form @xmath120 . here
@xmath121 and @xmath122 are two state - predicates in @xmath123 respectively for the precondition and the post - condition of the event .
@xmath124 is an event name .
event - predicate
@xmath120 " specifies the observation of event @xmath125 with precondition @xmath121 and post - condition @xmath122 .
in this work , we allow _ fairness assumptions _ either as state - predicates or as event - predicates .
a state fairness assumption is in @xmath123 .
an event fairness assumption is an event - predicate of @xmath119 .
given two sets @xmath126 and @xmath127 of fairness assumptions , @xmath128 denotes a _ multi - fairness assumption _ ( _ mf - assumption _ ) for @xmath119 .
all elements in @xmath126 are called _ strong fairness assumptions _ while all in @xmath127 are called _ weak fairness assumptions_. a run @xmath129 of @xmath119 satisfies @xmath128 iff the following constraints hold . for every state - predicate @xmath130 ,
there are infinitely many @xmath131 s such that for some @xmath132 $ ] , @xmath133 . for every event - predicate @xmath120 in @xmath126
, there are infinitely many @xmath131 s such that @xmath134 , @xmath135 , and @xmath136 . for every state - predicate @xmath137 ,
there is a @xmath131 such that for every @xmath138 and @xmath139 $ ] , @xmath140 . for every event - predicate @xmath120 in @xmath127 ,
there is a @xmath131 such that for every @xmath138 , if @xmath134 and @xmath135 , then @xmath136 . given a ta @xmath119 and a state @xmath141 , we let @xmath142 denote the set of runs of @xmath119 from @xmath72 satisfying @xmath128 .
the following definition shows how to formally model real - time systems with fairness assumptions .
[ def.gbcta ] a _ generalized bchi ta _
( _ gbta _ ) is a pair @xmath143 with a ta @xmath119 and an mf - assumption @xmath128 . if @xmath144 , the pair is also called a _ bchi ta _
( _ bta _ ) .
@xmath0 [ exmp.ctas ] for the model @xmath1 in figure [ fig.ms_ne](a ) , we may have a gbta @xmath145 that assumes @xmath1 should stay in location @xmath5 infinitely many times and event @xmath146 should be received by @xmath1 infinitely many times with post - condition @xmath4 .
we may also have the following gbta @xmath147 that assumes that @xmath1 should eventually stabilize in location @xmath148 .
suppose we are given two tas @xmath149 .
for any transitions @xmath150 and @xmath151 , @xmath93 and @xmath69 are _ compatible _ iff @xmath152 . that is , the observable events of the two automatas on the two transitions must be nontrivially identical .
for each @xmath150 with @xmath153 , we use @xmath154 to denote the subset of @xmath155 with elements compatible with @xmath93 . for each @xmath156 with @xmath157 , @xmath158 .
also , @xmath159 denotes the subset of @xmath155 with elements @xmath69 such that @xmath160 . in this section , from now on , we assume the context of two gbtas @xmath161 and @xmath162 respectively for the model and the specification . given a state @xmath163 of @xmath1 and a state @xmath72 of @xmath2 , we use @xmath164 to denote the state - pair of @xmath163 and @xmath72 .
operationally , @xmath164 can be viewed as @xmath165 , the functional composition of @xmath163 and @xmath72 .
a _ play _ between @xmath1 and @xmath2 is made of two matching runs , one of @xmath1 and the other of @xmath2 .
conceptually , it is a sequence @xmath166 of triples with the following restrictions .
@xmath167 is a run of @xmath1 . for convenience ,
we denote this run as @xmath168 .
@xmath169 is a run of @xmath2 . for convenience ,
we denote this run as @xmath170 . for each @xmath171 , @xmath172 .
the play is _ initial _
iff @xmath173 and @xmath174 .
play prefix _ is a finite prefix of a play .
given a play @xmath175 , we let @xmath176 be the prefix represented as the sequence of the first @xmath177 elements of @xmath175 . given a run ( prefix ) @xmath178 of @xmath1 and a play ( prefix ) @xmath179 between @xmath1 and @xmath2 , we say @xmath175 _ embeds _
@xmath180 iff there is a monotonically increasing integer function @xmath181 such that @xmath182 and for each @xmath114 , @xmath183 , @xmath184 , @xmath185 , and for each @xmath186 , @xmath187 .
notationally , we let @xmath188 denote the embedding relation between @xmath175 and @xmath180
. similarly we can define @xmath189 for the embedding relation between @xmath175 and a run @xmath190 of @xmath2 .
a _ strategy _ in a game tells a ta what to execute at a state - pair in a play that is developing .
specifically , a _ strategy _
@xmath191 for @xmath2 is a mapping from play prefixes of @xmath1 and @xmath2 to event sets of @xmath192 .
symmetrically , we can define strategies for @xmath1 . given a strategy @xmath191 for @xmath2 and a play @xmath193 between @xmath1 and @xmath2
, we say that @xmath175 _ complies _ to @xmath191 iff the following constraints are satisfied . for each @xmath171 and @xmath194 , @xmath195 . for each @xmath171 and @xmath196 with either @xmath197 or @xmath198 , @xmath199
similarly , we can also define the compliance of plays to strategies of @xmath1 . given a state - pair @xmath200 , a run @xmath180 of @xmath1 from @xmath163 , and a strategy @xmath191 of @xmath2 , we let @xmath201 be the play ( prefix ) from @xmath164 with the following restrictions .
@xmath175 complies to @xmath191 .
if @xmath175 is of infinite length , then it embeds @xmath180 .
if @xmath175 is of finite length , then there is a finite prefix @xmath202 of @xmath180 with the following restrictions .
@xmath175 embeds @xmath203 .
any prefix of @xmath180 that supersedes @xmath203 is not embedded by @xmath175 . note that it may happen that @xmath204 is of only finite length .
this can happen when at the end of the finite play , a player chooses a transition with an event set that the other player ( _ opponent _ ) can not choose a transition to match .
this can also happen when at the end of the finite play , a player can only execute matching transitions with post - condition falling outside the invariance predicate .
[ def.simf ] a _ simulation _ @xmath205 of @xmath161 by @xmath162 is a binary relation @xmath206 such that for every @xmath207 and every run @xmath180 of @xmath1 from @xmath163 that satisfies @xmath208 , there exists a play @xmath175 from @xmath164 such that @xmath175 embeds @xmath180 and @xmath170 satisfies @xmath209 .
we say that @xmath162 simulates @xmath161 , in symbols @xmath210 , if there exists a simulation @xmath205 of @xmath161 by @xmath162 such that for every @xmath211 , there exists a @xmath212 with @xmath207 .
@xmath0 [ exmp.simf ] for the tas in figure [ fig.ms_ne ] , we have that @xmath213 does not simulate @xmath214 .
also , @xmath215 does not simulate @xmath216 .
however , @xmath213 simulates @xmath217 .
if @xmath162 simulates @xmath161 , then for all initial states @xmath163 and runs @xmath180 of @xmath1 from @xmath163 satisfying @xmath208 , there exists a strategy @xmath191 such that @xmath218 satisfies @xmath209 .
we call such a @xmath191 a _ simulating strategy _ for @xmath180 by @xmath2 .
if @xmath162 does not simulate @xmath161 , then there exists an initial run @xmath180 of @xmath1 such that @xmath180 satisfies @xmath208 and for all initial states @xmath72 and all strategies @xmath191 of @xmath2 , all initial runs of @xmath2 embedded by @xmath218 do not satisfy @xmath209 .
we call such a run @xmath180 a _ refuting run _ of @xmath1 .
a strategy @xmath191 of a ta @xmath2 is _ memory - less _
iff for any two plays @xmath175 and @xmath219 that end at the same triple , @xmath220 .
it is well known that parity games and reachability games all have memory - less winning strategies for either player @xcite .
the following lemma shows that the simulation of gbas may need finite - memory refuting strategies .
[ lemma.simf.memory ] there is a simulation of gbas with a simulation strategy for the specification but without a memory - less simulation strategy for the specification .
+ in figure [ fig.simf.memory ] , we have the tas of two gbas @xmath221 and @xmath222 .
( 0,0 ) that need memory.,title="fig : " ] # 1#2#3#4#5 @font ( 2847,1550)(166,-863 ) ( 1863,-476)(0,0)[lb ] ( 2543,-476)(0,0)[lb ] ( 2471,-47)(0,0)[lb ] ( 2006,-47)(0,0)[lb ] ( 2901,275)(0,0)[lb ] ( 1577,311)(0,0)[lb ] ( 2221,167)(0,0)[lb ] ( 2185,-798)(0,0)[lb ] ( 789,132)(0,0)[lb ] ( 789,-512)(0,0)[lb ] ( 897,-190)(0,0)[lb ] ( 181 , 24)(0,0)[lb ] ( 575,-798)(0,0)[lb ] suppose we have a state - pair @xmath164 with @xmath223 and @xmath224 . as can be seen , for any memory - less strategy @xmath191 , either transition @xmath225
will always be chosen for any initial play prefix that ends at @xmath164 or transition @xmath226 will always be .
but such plays do not satisfy the strong fairness assumption of @xmath222 and can not be used to fulfill the strong fairness assumptions of @xmath2 .
thus we know there is no memory - less simulation strategy for @xmath222 . on the other hand
, we can devise a strategy for @xmath2 that chooses @xmath225 and @xmath226 alternately .
it is clear that such a strategy fulfills the strong fairness assumptions of @xmath227 .
in this work , we focus on characterization of the simulation of a model gbta by a specification bta . that is , we restrict that the specification @xmath228 is a bta with @xmath229 . for convenience , given an mf - assumption @xmath128 and a play @xmath193
, we may also define the satisfaction of @xmath128 by @xmath175 in a way similar to the satisfaction of @xmath128 by runs . according to definition [ def.simf ] ,
a state - pair @xmath164 is not in any simulation if there exists a run @xmath180 of @xmath1 from @xmath163 , satisfying @xmath208 , such that for every strategy @xmath191 for @xmath2 and play @xmath175 from @xmath164 complying to @xmath191 and embedding @xmath180 , @xmath175 does not satisfy @xmath209 .
put this description in a structural way , we have the following presentation .
@xmath230 according to the composition of @xmath209 , this can be broken down to cases described with the following four lemmas .
[ lemma.rsim.ss ] in case @xmath231 for a state - predicate @xmath74 , a state - pair @xmath164 is not in any simulation of @xmath232 by @xmath228 iff @xmath233 is true .
+ according to the argument in the beginning of the subsection , we only have to prove that the following two statements are equivalent in the context that @xmath175 embeds @xmath180 .
@xmath234 .
@xmath235 .
assume that @xmath193 .
we can prove this equivalence in two directions .
@xmath236 we assume that @xmath234 . according to the definition of strong fairness , we know that there are only finitely many @xmath131 s with a @xmath132 $ ] such that @xmath237 .
we let @xmath238 the maximum of such @xmath131 s .
then it is clear that for every @xmath239 and @xmath139 $ ] , @xmath240 .
this means that @xmath175 satisfies @xmath241 .
then the embedding of @xmath180 by @xmath175 implies that @xmath175 satisfies @xmath242 .
@xmath243 we assume that @xmath244 . then according to the definition of weak fairness , we know that there exists an @xmath238 such that for every @xmath239 and @xmath139 $ ] , @xmath245 . thus it is not true that there are infinitely many @xmath131 s with a @xmath246 $ ] such that @xmath237 .
according to the definition of strong fairness , @xmath234 . with the proof of the two directions , we know the lemma is proven . [ lemma.rsim.se ] in case @xmath247 for an event - predicate @xmath120 , a state - pair @xmath164 is not in any simulation of @xmath232 by @xmath228 iff @xmath248 is true .
+ suppose we are given @xmath193 .
the proof is similar to the one for lemma [ lemma.rsim.ss ] except that we need to show that for a @xmath249 , the equivalence between the following two statements .
it is not true that @xmath250 , @xmath251 , and @xmath252 .
if @xmath250 and @xmath251 , then @xmath253 .
this equivalence follows from the semantics of propositional logic . by treating the event - predicate as a state - predicate
, we can prove the lemma as we have proved lemma [ lemma.rsim.ss ] .
[ lemma.rsim.ws ] in case @xmath254 for a state predicate @xmath74 , a state - pair @xmath164 is not in any simulation of @xmath232 by @xmath228 iff @xmath255 is true .
+ by replacing @xmath74 with @xmath256 , we can use a proof similar to the one for lemma [ lemma.rsim.ss ] for this lemma .
[ lemma.rsim.we ] in case @xmath257 for a state predicate @xmath74 , a state - pair @xmath164 is not in any simulation of @xmath232 by @xmath228 iff @xmath258 is true .
+ by replacing @xmath259 with @xmath260 , we can use a proof similar to the one for lemma [ lemma.rsim.se ] for this lemma . for convenience , given two sets @xmath261 and @xmath262 of fairness assumptions ,
we let @xmath263 denote @xmath264 . according to lemmas [ lemma.rsim.ss ] , [ lemma.rsim.se ] , [ lemma.rsim.ws ] , and [ lemma.rsim.we ]
, we conclude with the following lemma .
[ lemma.rsim ] in case @xmath229 , a state - pair @xmath164 is not in any simulation of @xmath232 by @xmath228 iff @xmath265 is true . a procedure to construct a formula for states @xmath163 that starts a run of @xmath1 satisfying @xmath208 can be found in @xcite .
lemma [ lemma.rsim ] suggests that we still need to implement a procedure that constructs formulas for state - pairs that start all plays @xmath175 satisfying the following constraints .
@xmath266 such a play @xmath175 eventually stabilizes into a cycle of state - pairs along which each assumption in @xmath267 is satisfied once and all assumptions in @xmath268 are satisfied throughout the cycle .
the following definition characterizes state - pairs in such a cycle .
[ def.isr ] a state - pair @xmath164 is _ csr _ ( _ cyclically simulation - refuting _ ) with @xmath269 iff for every @xmath270 , there exists a run @xmath180 of @xmath1 with the following two constraints . for every strategy @xmath191 of @xmath2 with @xmath201 ,
if @xmath175 is of infinite length , then the following four constraints are satisfied .
all state - pairs along @xmath175 satisfy state - predicates in @xmath268 .
all transition - pairs along @xmath175 satisfy event - predicates in @xmath268 . for every state - predicate @xmath74 in @xmath267
, there is a csr state - pair in @xmath175 satisfying @xmath74 in more than 1 time units from the start of @xmath175 . for every event - predicate @xmath74 in @xmath267
, there is a transition - pair in @xmath175 satisfying @xmath74 in more than 1 time units from the start of @xmath175 .
there exists a strategy @xmath191 of @xmath2 with an infinitely long @xmath204 .
the 1-time - unit requirement at condition c1c is for making sure that the play is non - zeno .
a state - pair @xmath164 is _ inevitably sr _
( _ isr _ ) with @xmath271 iff there exists a run @xmath180 of @xmath1 from @xmath163 such that for all strategies @xmath191 of @xmath2 , if @xmath204 is infinite , then @xmath204 visits a csr state - pair .
the following lemma is important for our algorithm development .
[ lemma.simf.neg-evidence ] suppose we are given a gbta @xmath1 and a bta @xmath2 .
for any state - pair @xmath272 , the following two statements are equivalent .
@xmath163 starts a run @xmath180 of @xmath1 satisfying @xmath208 and for all plays @xmath175 from @xmath164 embedding @xmath180 , @xmath175 satisfies @xmath269 .
there exist an @xmath273 , a @xmath12 , and a @xmath274 with the following constraints
. @xmath275 .
@xmath276 starts a run satisfying @xmath208 .
for every @xmath277 and @xmath278 with @xmath279 , @xmath280 is an isr state - pair with @xmath281 .
we prove the lemma in two directions .
@xmath236 we assume that r1 is true .
conditions r2a and r2b are automatically true since @xmath180 must begin with a timed transition step @xmath282 for some @xmath283 , and @xmath274 . as for condition r2c
, we establish it in the following .
the truth of r1 means that for every strategy @xmath191 of @xmath2 , if @xmath201 embeds @xmath180 , then @xmath175 must satisfy @xmath269 .
this means that there exists a @xmath284 such that for every such infinite @xmath175 , after @xmath285 time units from the start of @xmath175 , all predicates in @xmath268 are satisfied and all predicates in @xmath267 are satisfied infinitely and divergently many times .
if such a @xmath285 does not exist , then we can construct a play that violates @xmath269 and the assumption of r1 .
we claim that all state - pairs @xmath286 happening after @xmath285 time units from the start in all infinite plays are csr state - pairs in definition [ def.isr ] .
the reasons are the following .
since @xmath286 happens @xmath285 time units after the start of the play , it must satisfy conditions c1a and c1b in definition [ def.isr ] .
moreover , along every infinite play from @xmath286 , for every predicate @xmath74 in @xmath267 , there are infinitely and divergently many state - pairs or transition - pairs that satisfies @xmath74 .
thus we can find the first state - pair @xmath287 in the tail with the following restrictions .
@xmath287 is at least one time unit from the start of the play .
either @xmath287 satisfies @xmath74 as a state - predicate or the transition - pair right before @xmath287 satisfies @xmath74 as an event - predicate .
this implies that conditions c1c and c1d in definition [ def.isr ] are satisfied at @xmath287 .
the assumption that leads to the satisfaction of @xmath269 by @xmath175 then implies that there exists such a play .
this implies that condition c2 in definition [ def.isr ] is satisfied .
the argument in the above establishes that @xmath286 is indeed a csr state - pair .
thus we know that along every infinite play from @xmath280 , we can reach such a @xmath286 .
this implies that @xmath280 is an isr state - pair and condition r2c is satisfied .
thus the lemma is proven in this direction .
@xmath243 we assume that r2 is true .
this implies that there exist an @xmath273 , a @xmath12 , and a @xmath274 with @xmath275 and @xmath276 starting a run satisfying @xmath208 .
there are two cases to analyze . by letting @xmath180
start with @xmath288 and followed by the tail from @xmath276 that satisfies @xmath208 , we deduce that @xmath163 also starts a run @xmath180 that satisfies @xmath208 .
then for all strategies @xmath191 of @xmath2 with @xmath201 , we can go to an isr state - pair @xmath280 with @xmath269 .
this implies that for all infinite plays from @xmath280 , we can visit a csr state - pair @xmath286 .
then according to the definition of csr state - pairs , for each predicate @xmath289 , we can go from @xmath286 along a play with all state - pairs and transition - pairs satisfying the predicates in @xmath268 .
moreover , the play visits a csr state - pair @xmath287 that either satisfies @xmath74 as a state - predicate or satisfies with the transition - pair immediately before @xmath287 as an event - predicate .
since @xmath287 is also csr , we can then repeat the same argument to fulfill another predicate assumption in @xmath267 . by repeating this procedure for all predicates in @xmath267 infinitely many times
, we can construct every infinite plays from @xmath164 that embeds @xmath180 mentioned in the last item .
this construction then leads to the conclusion that all plays from @xmath164 embedding @xmath180 satisfy @xmath269 .
this completes the proof of this direction .
since both directions of the proof are done , we know the lemma is true .
lemma [ lemma.simf.neg-evidence ] suggests the development of evaluation algorithm for csr state - pairs for the solution of usf - simulations of gbtas . in the following ,
we explain how to do this .
in this work , we focus on the simulation algorithm for a model gbta by a specification bta . our algorithm is based on the construction of formulas for csr and isr state - pairs . in the following , we assume the context of a model gbta @xmath232 and a specification bta @xmath228 . in subsection [ subsec.tck.mck ] , we present some symbolic procedures from model - checking technology of dense - time systems as our basic building blocks . in subsection
[ subsec.timed.inevit ] , we present algorithms for state - pairs that can be forced to a goal in one timed transition step . in subsection [ subsec.timed.minevit ]
, we use the procedures in subsection [ subsec.timed.inevit ] to construct a algorithms for state - pairs that can be forced to a goal in zero or more timed transition steps . in subsection
[ subsec.simcheck.alg ] , we present the algorithm for simulation - checking . in subsection
[ subsec.simf.complexity ] , we analyze the complexity of our algorithm . in this subsection , we adapt procedures for tctl model - checking @xcite for the evaluation of simulation - checking .
given a formula @xmath74 , a run prefix @xmath290 of @xmath1 is called an _ @xmath74-rprefix _ if for every @xmath291 and @xmath139 $ ] , @xmath292 .
similarly , a play prefix @xmath293 of @xmath1 is called an _ @xmath74-pprefix _ if for every @xmath291 and @xmath139 $ ] , @xmath294 .
given a state - pair set @xmath295 , we let @xmath296 . given a ta @xmath2 with @xmath297 and @xmath298 , we let @xmath299 be the following formula . @xmath300 .
also given a set @xmath301 and a set @xmath302 , we let @xmath303(\eta)$ ] be the following formula . @xmath304 .
standard procedures for constructing state - predicates of existentially quantified formulas can be found in @xcite .
given a transition - pair @xmath305 with @xmath306 and @xmath307 , we let @xmath308 be the formula of state - pairs that may go to state - pairs in @xmath74 through the simultaneous execution of @xmath93 and @xmath69 respectively . specifically , @xmath308 is defined as follows .
@xmath309\left ( \begin{array}{lll } \eta & \wedge & \lambda_{{\cal m}}(q'_1)\\ & \wedge & \lambda_{{\cal s}}(q'_2 ) \end{array}\right ) \end{array}\right)$ ] we also need the formulas for the precondition of time - progress to a state - pair satisfying @xmath122 through intermediate state - pairs satisfying @xmath121 . procedures for such formulas can be found in @xcite .
we present the formula , denoted @xmath310 , for the readers convenience in the following .
@xmath311 here @xmath312 represents a formula obtained from @xmath74 by replacing every clock variable @xmath313 in @xmath74 with @xmath314 .
we use adapted tctl formulas @xmath315 in our presentation of the algorithm .
specifically , @xmath315 characterizes those state - pairs @xmath164 with the following restrictions .
@xmath164 starts an @xmath121-pprefix @xmath175 that ends at a state - pair satisfying @xmath122 .
along the @xmath175 mentioned in the above , all the transitions are of the form @xmath316 with @xmath317 .
following the techniques in @xcite , we can construct a formula in @xmath318 that characterizes state - pairs satisfying @xmath315 . specifically , the formula is as follows .
@xmath319 here @xmath320 is the least fixpoint operator and @xmath321 represents the smallest solution to @xmath322 .
another type of formulas that we want to use is for states @xmath163 of @xmath1 that start runs satisfying @xmath208 .
we denote this formula as @xmath323 for convenience .
the construction of this formula can be found in @xcite .
given a set @xmath295 of states ( or state - pairs ) , we use @xmath324 to denote a formula that characterizes @xmath295 . given a formula @xmath74
, we use @xmath325\!]}$ ] to represent the set of states ( or state - pairs ) that satisfies @xmath74 . given an @xmath273 , a set @xmath127 of event weak fairness assumption , and a @xmath12
, we use @xmath326 to denote the set of state - pairs @xmath164 with the following restrictions .
there is a @xmath327-rprefix @xmath328,e , t+t_0)$ ] with the following two restrictions .
@xmath329 $ ] is in @xmath330 and satisfies @xmath323 . for every @xmath331-pprefix @xmath332 with @xmath333 , @xmath334 , @xmath335 , and @xmath336 , for every event weak fairness assumption
@xmath337 , if @xmath338 and @xmath339 , then @xmath329 \models\eta_4 $ ] .
@xmath340 is in @xmath341 .
note that in the just - mentioned @xmath331-pprefix , the strategy of @xmath2 can only use the internal transitions of @xmath2 .
we can use the following tctl formula to help us characterize @xmath326 . given two state - predicates @xmath342 , and a set @xmath127 of event formulas for weak fairness assumption , we let @xmath343 be defined as follows .
@xmath344 here @xmath345 is an auxiliary clock variable not used in @xmath346 .
the conjunction @xmath347 in the post - condition is used to make sure that no event weak fairness assumptions in @xmath127 is violated .
it is used to eliminate all state - pairs violating an event weak fairness assumption .
the following lemma shows how to use the above formulas to help us evaluating @xmath348 .
[ lemma.nxt.form.ext ] for every @xmath164 , @xmath349 , @xmath350 $ ] , formulas @xmath342 of state - pairs , and a set @xmath127 of event weak fairness assumptions , @xmath351\!]}{\bigcirc}^{e\psi}_t { [ \![}\eta_2{]\!]}$ ] iff @xmath352 .
we can rewrite condition @xmath353 of @xmath354\!]}{\bigcirc}^{e\psi}_t { [ \![}\eta_2{]\!]}$ ] as follows .
there is no @xmath121-pprefix @xmath332 with @xmath333 , @xmath334 , @xmath335 , @xmath336 , @xmath355 , and for every @xmath337 and @xmath356 , @xmath338 , and @xmath357 , it is not true that @xmath358\models \neg\eta_4 $ ] .
it is clear that a state - pair satisfies @xmath359 and @xmath353 if and only if it satisfies @xmath359 and @xmath360 . by renaming @xmath361 as a clock variable @xmath345 and @xmath362 as constant @xmath363
, we can use @xmath364 to represent @xmath365 .
this means that @xmath359 and @xmath360 can be rewritten as @xmath366 and the following two conditions .
there exists an @xmath367-rprefix @xmath368{8mm}{0.5pt}\!\!\!\longrightarrow}}],e,{{c^{{\cal m}}_{{\cal s}}}})$ ] with @xmath369{8mm}{0.5pt}\!\!\!\longrightarrow}}]\models \exists{{\cal s}}(\eta_2)\wedge\exists{\box}\phi_{{\cal m}}\psi_{{\cal m}}$ ] .
there is no @xmath121-pprefix @xmath293 with @xmath335 , @xmath370 , @xmath333 , @xmath371 , and for every @xmath337 and @xmath356 , @xmath338 , and @xmath357 , it is not true that @xmath372\models \neg\eta_4 $ ] .
@xmath373 means the following .
@xmath374\!]}\\ \wedge & \exists{\box}\phi_{{\cal m}}\psi_{{\cal m}}\end{array}\right)\right ) $ ] @xmath375 means the following .
@xmath376 combining these two formulas together and reduce them with the definition of @xmath343 , we find that @xmath164 must satisfy @xmath377 . thus the lemma is proven .
based on lemma [ lemma.nxt.form.ext ] , we can define the following notations for those state - pairs that can be forced into either certain destination or a transition of @xmath1 that @xmath2 can not match .
specifically , we let @xmath378 . correspondingly , given two formulas @xmath121 and
@xmath122 , we can construct @xmath379 , defined as follows .
@xmath380 . then according to lemma [ lemma.nxt.form.ext ]
, we can establish the following lemma .
[ lemma.until.form ] for every @xmath272 , formulas @xmath342 of state - pairs , and set @xmath127 of event weak fairness assumptions , @xmath381\!]}{\bigcirc}^\psi{[\![}\eta_2{]\!]}$ ] iff @xmath382 .
+ we have the following deduction .
@xmath383\!]}{\bigcirc}^\psi{[\![}\eta_2{]\ ! ] } } \\ \equiv & \mu\nu\in
\bigcup_{e\in e_{{\cal m}},t\in{{{\mathbb r}^{\geq 0 } } } } \langle { { \cal m}}\rangle d_1{\bigcirc}^{e\psi}_t d_2 \\ \equiv & \bigvee_{e\in e_{{\cal m}},t\in{{{\mathbb r}^{\geq 0 } } } } \mu\nu\in\langle { { \cal m}}\rangle d_1{\bigcirc}^{e\psi}_t d_2 \end{array}$ ] according to lemma [ lemma.nxt.form.ext ] , this implies the following .
@xmath384 since @xmath343 does not contain variable @xmath365 , the above formulas are equivalent to the following .
@xmath385 since @xmath386 is a tautology , we have the following .
@xmath387 the last step is from the definition of @xmath379 .
thus the lemma is proven .
note that before the fulfillment of @xmath122 , @xmath388 is satisfied with play prefixes with only transitions internal to @xmath2 . in general , we want to characterize state - pairs from which @xmath1 can force the fulfillment of @xmath122 through zero or more timed transition steps of @xmath1 that do not violate the weak fairness assumptions in @xmath127 .
we denote the set of such state - pairs as @xmath389\!]}{\mbox{$\cal u$}}^\psi { [ \![}\eta_2{]\!]}$ ] . for convenience ,
given two formulas @xmath342 for sets of state - pairs , we let @xmath390 here @xmath320 is the least fixpoint operator .
@xmath391 specifies a smallest solution to equation @xmath392 .
the procedure to construct formulas for such least fixpoints can be found in @xcite .
[ lemma.until-star.form ] for every state - pairs @xmath164 and formulas @xmath342 for state - pairs , @xmath351\!]}{\mbox{$\cal u$}}^\psi { [ \![}\eta_2{]\!]}$ ] iff @xmath393 . + we can prove this lemma in two directions .
@xmath236 we assume that @xmath394\!]}{\mbox{$\cal u$}}^\psi { [ \![}\eta_2{]\!]}$ ] is true
. we can prove this by induction on the maximum number @xmath395 of timed transition steps of @xmath1 to reach state - pairs in @xmath396\!]}$ ] through state - pairs in @xmath397\!]}$ ] .
in the base case , @xmath398 and @xmath399\!]}$ ] . then it is clear that @xmath164 also satisfies every formula of the form @xmath400 .
thus , @xmath393 in the base case and the lemma is proven .
now we assume that this direction of the lemma is true for every state - pairs with maximum number no greater than @xmath131 with @xmath249 .
now we have a state - pair @xmath164 with maximum number @xmath177 of timed transition steps to reach state - pairs in @xmath396\!]}$ ] through state - pairs in @xmath397\!]}$ ] .
this implies that there exist an @xmath273 and a @xmath101 with @xmath401\!]}{\bigcirc}^{e\psi}_t \left(\langle { { \cal m}}\rangle{[\![}\eta_1{]\!]}{\mbox{$\cal u$}}^{\psi } { [ \![}\eta_2{]\!]}\right)$ ] .
this means that in one timed transition step of @xmath93 and @xmath365 time units by @xmath1 , we end up in a state - pair @xmath280 such that within @xmath131 timed transition of @xmath1 steps through state - pairs in @xmath397\!]}$ ] , we can go from @xmath280 to state - pairs in @xmath396\!]}$ ] . according to the inductive hypothesis , we know that @xmath280 satisfies @xmath402 .
together , this implies the following deduction .
@xmath403 according to the definition of least fixpoint , the last step implies @xmath404 . by definition
, this implies that @xmath405 .
thus this direction of the lemma is proven by induction .
@xmath243 we assume that there exist @xmath406 such that @xmath407 , @xmath408 , and for every @xmath409 , @xmath410 .
we prove by induction on @xmath411 $ ] that @xmath412 implies @xmath394\!]}{\mbox{$\cal u$}}^\psi{[\![}\eta_2{]\!]}$ ] .
the base case is that @xmath413 and @xmath414 .
this implies that @xmath399\!]}$ ] and @xmath415\!]}{\mbox{$\cal u$}}^\psi{[\![}\eta_2{]\!]}$ ] .
thus the base case is proven .
now we assume that the lemma in this direction is true for all @xmath416 $ ] .
now we have a @xmath417 .
this means that @xmath418 .
there are two cases to analyze .
the first is @xmath419 and coincides with the base case .
thus the first case is already proven .
the second case is @xmath420 . according to lemma [ lemma.nxt.form.ext ] ,
this implies that we can force in one timed transition step through state - pairs in @xmath397\!]}$ ] to state - pairs @xmath280 in @xmath421\!]}$ ] .
moreover , the inductive hypothesis says that all such @xmath422 . according to the definition of @xmath423
, this implies that @xmath424 .
thus the lemma is proven in this direction . thus the lemma is proven .
our plan is first to use the procedures in subsections [ subsec.tck.mck ] , [ subsec.timed.inevit ] , and [ subsec.timed.minevit ] to construct a procedure for evaluating csr state - pairs .
then we use this procedure to evaluate isr state - pairs . for convenience , we denote @xmath425 . conceptually , @xmath426 denotes the state - predicates that a play satisfying @xmath427 must stabilize with .
also we let @xmath428 be the set of event - predicates in @xmath268 . for convenience ,
we also let @xmath429 and @xmath430 .
we present a greatest fixpoint characterization , denoted @xmath431 , of the csr state - pairs with an mf - assumption @xmath128 .
a state - pair @xmath164 satisfies @xmath431 if there is a fair run from @xmath163 such that all plays embedding the run from @xmath164 can not be fair for @xmath2 . the characterization follows .
@xmath432 .
here @xmath433 is the greatest fixpoint operator .
@xmath434 is a largest solution @xmath435 to @xmath436 .
the procedure to construct formulas for greatest fixpoints can be found in @xcite .
the following lemma establishes the correctness of the characterization .
[ lemma.mfsf ] a state - pair @xmath164 is csr with @xmath437 iff @xmath438 .
+ following definition [ def.isr ] , lemma [ lemma.until-star.form ] , the definition of @xmath426 , and the semantics of greatest fixpoint , @xmath439 is actually a rewriting of the csr definition with logic formulas , the greatest fixpoint procedure , and the @xmath440 procedure .
thus the lemma is proven .
now we use @xmath439 to evaluate isr state - pairs . given a fair run @xmath180 of @xmath1 , there are two classes of isr state - pairs .
the first class contains state - pairs that start no play embedding @xmath180 .
the second class contains state - pairs with a strategy @xmath2 to drive a play to stabilize to csr state - pairs .
the former can be evaluated with the traditional procedures for branching simulation @xcite .
specifically , state - pairs is in the first class can be characterized with the following lemma .
[ lemma.isr.1st ] a state - pair @xmath164 is a first - class isr state - pair iff @xmath441 .
+ @xmath164 is first class iff for all strategies @xmath191 of @xmath2 , @xmath204 is of finite length .
there can only be two causes for the termination of the plays . along a time progress operation , @xmath1 moves to a valid state while @xmath2 can not .
this is captured by formula @xmath442 . at a transition
@xmath93 by @xmath1 , no compatible @xmath277 can result in a valid state of @xmath2 .
this is captured by @xmath443 .
if and only if @xmath1 can drive all plays to state - pairs with these two causes , then it is clear all plays are finite in length . thus the lemma is proven .
the state - pairs in the second class can be forced into infinite plays that stabilize in csr state - pairs .
specifically , we have the following lemma . [ lemma.isr.2nd ]
a state - pair @xmath164 is a second - class isr state - pair iff @xmath444 .
+ this lemma follows from the definition of the second class state - pairs , lemma [ lemma.until-star.form ] , and lemma [ lemma.mfsf ] .
combining lemmas [ lemma.simf.neg-evidence ] , [ lemma.isr.1st ] , and [ lemma.isr.2nd ] , we present the following lemma for the characterization of state - pairs that is in no simulation of a gbta by a bta .
[ lemma.sim.refute ] a state - pair @xmath164 is in no simulation of a gbta @xmath445 by a bta @xmath446 iff @xmath164 satisfies either @xmath447 or @xmath448 .
the complexity of our algorithm relies on the implementation of the basic manipulation procedures of zones . like in @xcite
, we argue that we can implement the formulas as sets of pairs of proposition valuations and regions @xcite .
in such an implementation , basic operations like subsumption , intersection , union , complement , time progression , and variable quantification can all be done in exptime .
[ lemma.bisim.complexity ] proper implementations of the formulation in lemma [ lemma.sim.refute ] can be done in exptime .
+ according to @xcite , a zone can be implemented as a set of regions .
the number of regions is exponential to the input size of @xmath445 and @xmath446 .
all precondition calculations need at most polynomial numbers of region set operations and can all be done in exptime .
the numbers of iterations of the least and greatest fixpoint procedures are at most the number of regions .
thus , summing everything up , we conclude that our algorithm can be executed in exptime .
in real - world , we may usually want to check whether a system component satisfies its specification .
in such a context , the simulation - checking is carried out against the same behavior of the environment of the component .
such a context can usually make room for verification efficiency if we carefully represent the common environment state information . in this section ,
we extend the simulation defined in section [ sec.simf ] to simulation of a model by a specification against a common environment .
then we propose a technique to take advantage of the common environment information for simulation - checking efficiency . in figure
[ fig.mse ] , there are two tas for two environment processes .
( 0,0 ) # 1#2#3#4#5 @font ( 5195,3561)(664,-3139 ) ( 1501 , 89)(0,0)[lb ] ( 2926,-811)(0,0)[lb ] ( 5176,-361)(0,0)[lb ] ( 3226,-286)(0,0)[lb ] ( 4726 , 89)(0,0)[lb ] ( 1501,239)(0,0)[lb ] ( 901,-1111)(0,0)[lb ] ( 976,-361)(0,0)[lb ] ( 976,-2311)(0,0)[lb ] ( 1501,-1861)(0,0)[lb ] ( 2926,-2761)(0,0)[lb ] ( 5176,-2311)(0,0)[lb ] ( 3226,-2236)(0,0)[lb ] ( 3226,-2386)(0,0)[lb ] ( 4726,-1861)(0,0)[lb ] ( 1501,-1711)(0,0)[lb ] ( 901,-3061)(0,0)[lb ] note that location comp in figure [ fig.mse](b ) is labeled with a deadline @xmath449 .
this means that the environment process in figure [ fig.mse](b ) can only stay in location comp for at most 10 time units .
thus the environment process in figure [ fig.mse](a ) may deliver late service while the one in figure [ fig.mse](b ) always deliver service in 10 time units . against the environment described by figure [ fig.mse](a ) , the @xmath2 in figure [ fig.ms_ne](b ) does not simulate the @xmath1 in figure [ fig.ms_ne](a ) since the @xmath1 terminates the computation on late service while the @xmath2 never terminates the computation . in comparison , against figure [ fig.mse](b )
, the @xmath2 simulates the @xmath1 since the service is always in time .
we use _ ctas _ ( _ communicating timed automata _ ) to model the interaction between an environment and a model ( or a specification ) .
the formal definition is in the following .
[ def.cta ] a cta of two tas @xmath119 and @xmath450 , in symbols @xmath451 , is a ta with the following constraints .
@xmath452 .
@xmath453 .
@xmath454 .
@xmath455 .
@xmath456 . for each @xmath457 , @xmath458 . for simplicity , we assume that @xmath459 , @xmath460 , and @xmath461 . moreover , the transitions of a product ta needs to consider the synchronization between the two process tas . specifically , we let @xmath462
for each @xmath463 , one of the following constraints must hold .
@xmath464 represents the autonomous execution of a process ta with a transition without any events . formally speaking ,
this means at least one of @xmath93 and @xmath69 is @xmath51 , i.e. , no operation .
we have the following two cases to explain . if @xmath465 and @xmath466 , then @xmath150 , @xmath467 , @xmath468 , @xmath469 . if @xmath470 and @xmath471 , then @xmath151 , @xmath472 , @xmath473 , @xmath474 .
@xmath464 represents the synchronized execution of the two process tas respectively with a receiving event and a sending event of the same type . formally speaking
, this means that there is an @xmath475 with the following restrictions .
either of the following two is true .
@xmath476 , @xmath477 , and @xmath478 .
@xmath479 , @xmath480 , and @xmath481 .
note here we blend the process names and the operations into the name of the new events .
for example , @xmath482 and @xmath483 respectively represent the receiving and the sending of event @xmath125 by process @xmath119 .
@xmath484 .
@xmath485 .
@xmath0 [ exmp.cta ] for the specification @xmath2 in figure [ fig.ms_ne](b ) and the environment @xmath22 in figure [ fig.mse](a ) , we have @xmath486 with attributes in table [ tab.cta.attr ] .
@xmath487 \\ e_{{{{\cal e}}\times{{\cal s } } } } & = & \{((\mbox{\tt standby},\mbox{\tt idle}_2 ) , ( \mbox{\tt process},\mbox{\tt wait}_2 ) ) , ( ( \mbox{\tt process},\mbox{\tt wait}_2 ) , ( \mbox{\tt standby},\mbox{\tt idle}_2))\ } \\ \sigma_{{{{\cal e}}\times{{\cal s } } } } & = & \{\mbox{\tt request},\mbox{\tt serve},\mbox{\tt end}\ } \\
\epsilon_{{{{\cal e}}\times{{\cal s } } } } & = & \left [ \begin{array}{l } ( ( \mbox{\tt standby},\mbox{\tt idle}_2 ) , ( \mbox{\tt process},\mbox{\tt wait}_2 ) ) \mapsto\{!\mbox{\tt request}@({{\cal e } } ) , ?
\mbox{\tt request}@({{\cal s}})\ } , \\ ( ( \mbox{\tt process},\mbox{\tt wait}_2 ) , ( \mbox{\tt standby},\mbox{\tt idle}_2 ) ) \mapsto\{?\mbox{\tt serve}@({{\cal e } } ) , !
\mbox{\tt serve}@({{\cal s}})\ } \end{array}\right ] \\
\tau_{{{{\cal e}}\times{{\cal s } } } } & = & \left [ \begin{array}{l } ( ( \mbox{\tt standby},\mbox{\tt idle}_2 ) , ( \mbox{\tt process},\mbox{\tt wait}_2 ) ) \mapsto x_2>5 , \\ ( ( \mbox{\tt process},\mbox{\tt wait}_2 ) , ( \mbox{\tt standby},\mbox{\tt idle}_2 ) ) \mapsto{\mbox{\em true}}\end{array}\right ] \\
\pi_{{{{\cal e}}\times{{\cal s } } } } & = & \left [ \begin{array}{l } ( ( \mbox{\tt standby},\mbox{\tt idle}_2 ) , ( \mbox{\tt process},\mbox{\tt wait}_2 ) ) \mapsto\{x_3\ } , \\ ( ( \mbox{\tt process},\mbox{\tt wait}_2 ) , ( \mbox{\tt standby},\mbox{\tt idle}_2 ) ) \mapsto\{x_2\ } \end{array}\right ] \end{array}$ ] + @xmath68 $ ] denotes a ( partial or total ) function @xmath69 with @xmath70 .
since a cta is also a ta , we explain how to interpret the notations about tas for ctas . given a state @xmath488 of @xmath119 and a state @xmath163 of @xmath450 ,
@xmath489 is called a state of @xmath490 .
we say a state @xmath489 satisfies a state predicate @xmath491 , in symbols @xmath492 , with the following inductive rules . for any @xmath493 , @xmath494 iff @xmath495 . for any @xmath496 , @xmath494 iff @xmath497
for any @xmath498 , @xmath499 iff @xmath500 . for any @xmath501 , @xmath499 iff @xmath502 .
@xmath503 iff it is not the case that @xmath504 .
@xmath505 iff @xmath504 or @xmath506 .
the state after a transition @xmath464 from a state @xmath489 of cta , denoted @xmath507 , can also be interpreted as @xmath508 .
a timed transition of @xmath365 time units from a state @xmath489 , denoted @xmath509 , can be defined as @xmath510 . in this way
, we can also define the timed transition relation between two states @xmath511 through a transition @xmath464 in @xmath365 time units , denoted as @xmath512 , with the following restrictions . for all @xmath106 $ ] , @xmath513 .
@xmath514 .
@xmath515 .
then a run of @xmath490 can also be defined as a sequence @xmath516 with @xmath517{10mm}{0.5pt}\!\!\!\longrightarrow}}(\mu_{k+1},\nu_{k+1})$ ] for all @xmath249 .
given a cta @xmath518 and an mf - assumption @xmath128 of @xmath518 , @xmath519 is called a _ gcbta _ ( _ generalized communicating bta _ ) .
similarly , @xmath519 is a _ cbta _ ( _ communicating bta _ ) if @xmath144 .
[ def.simfe ] * simulation of gcbtas against an environment * a _ simulation _ @xmath520 of a model gcbta @xmath161 by a specification gcbta @xmath162 against an environment gcbta @xmath521 is a binary relation @xmath522 such that for every @xmath523 , the following restrictions are satisfied .
@xmath524 with @xmath525 .
@xmath526 .
@xmath527 . for every run @xmath180 of @xmath23 from @xmath489 that satisfies @xmath528
, there exists a play @xmath175 from @xmath529 with the following restrictions .
@xmath175 embeds @xmath180 and satisfies @xmath530 . for every transition
@xmath531 along @xmath175 , @xmath532 .
we say that @xmath162 simulates @xmath161 against environment @xmath521 , in symbols @xmath533 , if there exists a simulation @xmath520 of @xmath161 by @xmath162 against @xmath521 such that for every @xmath534 , there exists an @xmath535 with @xmath536 .
as can be seen , definition [ def.simfe ] is more restrictive than definition [ def.simf ] in their presentations .
however , we can prove that they are equivalent .
[ lemma.simfe.eq ] given an environment gcbta @xmath521 , a model gcbta @xmath161 , and a specification gcbta @xmath162 , @xmath537 iff @xmath538 .
+ the backward direction of the proof is straightforward since every simulation against an environment in definition [ def.simfe ] is also a simulation in definition [ def.simf ] .
thus we only have to focus on the forward direction of the proof .
we first assume that there is a simulation @xmath205 of @xmath539 by @xmath540 .
we can construct @xmath520 as follows .
@xmath541 . given an @xmath542 ,
it is apparent that @xmath543 satisfies conditions se1 , se2 , and se3 of definition [ def.simfe ] .
then for every runs @xmath180 of @xmath23 from @xmath489 satisfying @xmath528 , there exists a play @xmath175 from @xmath529 such that @xmath175 embeds @xmath180 and satisfies @xmath530 .
suppose @xmath544 this implies the following for all @xmath249 .
@xmath545 then we can construct a sequence @xmath546 as follows .
@xmath547 we have the following two claims to prove the lemma . @xmath546 is a play of @xmath548 by @xmath549 and embeds @xmath180 .
@xmath546 satisfies @xmath530 .
claim cl1 relies on the validity that for all @xmath249 , @xmath550{10mm}{0.5pt}\!\!\!\longrightarrow } } & ( \alpha_{k+1},\mu_{k+1})\\ ( \alpha_k,\nu_k ) & \stackrel{t_{k+1}-t_k,(e_k , g_k)}{{\rule[2pt]{10mm}{0.5pt}\!\!\!\longrightarrow } } & ( \alpha_{k+1},\nu_{k+1 } ) \end{array}$ ] these two statements rely on the following three statements .
@xmath551 the validity of the above three then follows from statements @xmath552 , and @xmath553 in the above .
thus we know that @xmath546 is indeed a play of @xmath554 .
furthermore , the validity of statements @xmath555 and @xmath556 implies that @xmath546 indeed embeds @xmath180 .
now we want to prove claim cl2 . for all assumptions in @xmath557 and @xmath558
, they are automatically satisfied since @xmath546 also embeds @xmath180 and @xmath180 satisfies @xmath559 . for a strong fairness assumption @xmath560
, we have the following two cases to analyze . _
@xmath561 is a state - predicate .
_ we claim that along @xmath546 , for every @xmath562 , there exists an @xmath138 and a @xmath563 $ ] with @xmath564 .
this is true since along @xmath175 , @xmath565 which implies that @xmath566 which in turn implies the claim . _
@xmath567 is an event - predicate . _
we claim that along @xmath546 , for every @xmath562 , there exists an @xmath138 with @xmath568 , @xmath569 , and @xmath570 .
this is true since along @xmath175 , @xmath571 , @xmath572 , and @xmath573 .
this further implies that @xmath574 , @xmath575 , and @xmath576 . in the end
, this implies the claim . for a weak fairness assumption @xmath577
, we have the following two cases to analyze . _
@xmath578 is a state - predicate .
_ we claim that there exists a @xmath562 such that for every @xmath138 and @xmath563 $ ] , @xmath579 .
this is true since along @xmath175 , @xmath580 which implies that @xmath581 which in turn implies the claim . _
@xmath582 is an event - predicate .
_ we claim that along @xmath546 , there exists a @xmath562 such that for all @xmath138 , if @xmath568 and @xmath583 , then @xmath570 .
this is true since along @xmath175 , if @xmath571 and @xmath584 , then @xmath573 .
this further implies that if @xmath574 and @xmath585 , then @xmath576 . in the end , this implies the claim .
with the proof of claims cl1 and cl2 , thus we conclude that the lemma is proven . according to lemma [ lemma.simfe.eq ]
, we can check the classic simulation in definition [ def.simf ] by checking the one in definition [ def.simfe ] .
this can be helpful in enhancing the verification performance when the common environment between the model and the specification is non - trivial .
lemma [ lemma.simfe.eq ] implies that we can use the following techniques to enhance the simulation algorithm against an environment .
based on condition se1 of definition [ def.simfe ] , we significantly reduce the sizes of the spaces of state - pairs by disregarding state - pairs of the form @xmath529 with @xmath586 .
since the number of different zones representing @xmath587 s can be exponential to the input size , the reduction can result in exponential speed - up . by mapping variables in @xmath587 in state - pairs @xmath529 , to those in @xmath488
, we actually only have to record one copy of values for each variables in @xmath488 .
since the size of bdd - like diagrams@xcite is exponential to the number of variables , this technique can also significantly reduce the memory usage in representations with bdd - like diagrams . in evaluating the precondition of state - pairs
, we need to enumerate all the transition pairs of the form @xmath531 with @xmath588 , @xmath589 , and @xmath590 .
if we use the classic simulation , the enumeration is of size @xmath591 .
but with the simulation against a common environment in definition [ def.simfe ] , the enumeration is of size @xmath592 .
thus significant reduction in time and space complexity can also be achieved with definition [ def.simfe ] .
we have implemented the techniques proposed in this manuscript in * red * 8 , a model / simulation - checker for ctas and parametric safety analysis for lhas based on crd ( clock - restriction diagram ) @xcite and hrd ( hybrid - restriction diagram ) technology @xcite .
the state - pair spaces are explored in a symbolic on - the - fly style . to our knowledge , there is no other tool that supports fully automatic simulation checking with gbtas .
we used parameterized networks of processes as our benchmarks . for a network of @xmath238 processes , we use integer @xmath17 through @xmath238 to index the processes .
users supply two index lists , the first for the indices of the model processes and the second for indices of the specification processes .
the process indices not in the two lists are treated as indices of the environment processes .
for example , we may have a system of 10 processes .
the following describes a simulation - checking task of process 1 ( the model ) by process 1 ( the specification ) . .... 1;2 ; .... here processes 3 through 10 are the environment processes . to support convenience in presenting fairness assumptions , we allow parameterized expressions .
for example , in table [ tab.sim.reqs](a ) , we have a simulation requirement with parameterized strong fairness assumptions .
\(a ) one simulation requirement + ' '' '' .... # ps-1 assume { strong event { execute@(#ps-1 ) } ; } ; # ps assume { strong true event { execute@(#ps ) } true ; } ; assume { |k:2 .. #ps-2 , strong true event { execute@(k ) } ; } .... ' '' '' \(b )
another simulation requirement + ' '' '' .... # ps-1 assume { strong event { execute@(#ps-1 ) } ; } ; # ps assume { weak idle@(#ps ) ; } ; assume { |k:2 ..
#ps-2 , strong true event { execute@(k ) } ; } .... ' '' '' here ` # ps ` is a parameter for the number of processes .
thus for a system of 10 processes , process 9 is the model , process 10 is the specification , while the others are the environment .
the last assume statement is for the fairness assumption of the environment .
the specification of event - predicates is in the following form .
type [ @xmath121 ] a [ @xmath122 ] here type is either ` strong ' or ` weak . '
[ @xmath121 ] and [ @xmath122 ] are respectively the optional precondition and the optional post - condition .
we may also use quantified expressions to present several fairness assumptions together .
for example , in the above , .... assume { |k:2 .. #ps-2 , strong true event { execute@(k ) } ; } .... presents the following strong fairness assumptions . ....
strong true event { execute@(2 ) } strong true event { execute@(3 ) } ... ... strong true event { execute@(8 ) } ....
to our knowledge , there is no other tool that supports fully automatic simulation checking with fairness assumptions for tas as ours .
so we only experimented with our algorithms .
we report two experiments .
the first is for timed branching simulation against a common environment without fairness assumptions in subsection [ subsec.exp.tsim ] . especially , we report the performance enhancement of the simulation in definition [ def.simfe ] ( without fairness assumption ) over the simulation in definition [ def.simf ] .
the second experiment is for simulation against a common environment with fairness assumptions in subsection [ subsec.exp.fsim ] . especially , we use liveness properties in the experiment .
we used the following three parameterized benchmarks from the literature .
_ fischer s timed mutual exclusion algorithm _
@xcite : the algorithm relies on a global lock and a local clock per process to control access to a critical section .
two timing constants used are 10 and 19 .
_ csma / cd_@xcite : this is the ethernet bus arbitration protocol with collision - and - retry . the timing constants used are 26 , 52 , and 808 .
_ timed consumer / producer_@xcite : there is a buffer , some producers , and some consumers .
the producers periodically write data to the buffer if it is empty .
the consumers periodically wipe out data , if any , in the buffer .
the timing constants used are 5 , 10 , 15 , and 20 . for each benchmark , we use one model process and one specification process .
all the other processes are environment .
also for each benchmark , two versions are used , one with a simulation and one without . for the versions with a simulation , @xmath1 and @xmath2 are identical . for the version without ,
@xmath1 and @xmath2 differ in only one process transition or invariance condition .
for example , for the fischer s benchmark , the difference is that the triggering condition of a transition to the critical section of @xmath2 is mistaken .
the performance data is reported in table [ tab.perf.tsim ] .
[ cols="<,<,^ , > , > , > , > " , ] + for each benchmarks , there are a model process , a specification process , and @xmath238 environment processes . `
n / a ' means not avaiable . "
+ data collected on a pentium 4 1.7ghz with 380 mb memory running linux ; + s : seconds ; k : kilobytes of memory in data - structure ; m : megabytes of total memory + as can be seen from the performance data , our techniques show promise for the verification of fulfillment of liveness properties in concurrent computing .
in this work , we investigate the simulation problem of tas with multiple strong and weak fairness assumptions . for the succinct presentation of fairness assumptions , we also allow for event fairness properties .
we then present an algorithm for the usf - simulation of gbtas .
the algorithm is based on symbolic model - checking and simulation - checking techniques and can be of interest by itself .
we then propose a new simulation against a common environment between the model and the specification
. we then present efficiency techniques for this new simulation .
our implementation and experiment shows the promise that our algorithm could be useful in practice in the future .
the work is partially supported by nsc , taiwan , roc under grants nsc 97 - 2221-e-002 - 129-my3 .
+ part of the work appears in the proceedings of formats 2007 , lncs 4763 , springer - verlag@xcite and the proceedings of hscc 2009 , lncs 5469 , springer - verlag @xcite .
f. wang .
symbolic parametric safety analysis of linear hybrid systems with bdd - like data - structures .
, 31(1):3851 , 2005 .
a preliminary version is in proceedings of 16th cav , 2004 , lncs 3114 , springer - verlag .
f. wang .
symbolic simulation checking of dense - time automata . in _
5th formats ( international conference on formal modelling and analysis of timed systems ) _ , volume lncs 4763 .
springer - verlag , october 2007 .
f. wang .
time - progress evaluation for dense - time automata with concave path conditions . in _
atva ( international symposium of automated technology on verification and analysis ) _ , volume lncs 5311 .
springer - verlag , 2008 .
f. wang , g .- d .
huang , and f. yu .
tctl inevitability analysis of dense - time systems : from theory to engineering . , 32(7 ) , 2006 . a preliminary version of the work appears in the proceedings of 8th ciaa ( conference on implementation and application of automata ) , july 2003 , santa barbara , ca , usa ; lncs 2759 , springer - verlag . | we investigate the simulation problem in of dense - time system .
a specification simulates a model if the specification can match every transition that the model can make at a time point .
we also adapt the approach of emerson and lei and allow for multiple strong and weak fairness assumptions in checking the simulation relation .
furthermore , we allow for fairness assumptions specified as either state - predicates or event - predicates .
we focus on a subclass of the problem with at most one fairness assumption for the specification .
we then present a simulation - checking algorithm for this subclass .
we propose simulation of a model by a specification against a common environment . we present efficient techniques for such simulations to take the common environment into consideration .
our experiment shows that such a consideration can dramatically improve the efficiency of checking simulation .
we also report the performance of our algorithm in checking the liveness properties with fairness assumptions .
* keywords : * branching simulation , fairness , verification , bchi automatas , concurrent computing , timed automata , algorithms , experiment |
the gravitational wave background from smbh coalescence during galaxy mergers is proving to be a diagnostic of galaxy formation models because of the useful constraints on the gwb from improved understanding of timing noise in pulsar data @xcite .
recent studies of the expected rate of merger - based smbh formation based on current observations ( corrections to bh - bulge relation , and new estimates of galaxy merger timescales ) @xcite suggest that the rates derived earlier from semi - analytic models for galaxy formation implemented in dark matter simulations @xcite may be seriously underestimated .
this has led to the prediction of imminent detection of the gwb by pulsar timing arrays @xcite . on the other hand ,
the gwb from coalescing supermassive binaries might be substantially diminished via environmental interactions @xcite and owing to the `` last parsec '' problem @xcite .
axis precession has long been invoked to explain certain inversion symmetric morphologies in radio galaxies where presence of close companion galaxy and binary smbh models have been used in some cases to explain the phenomenon @xcite .
@xcite suggested that the structures of radio galaxies , which are hosted by giant ellipticals , offer independent and direct clues to the prevalence of smbh binaries .
in particular , they pointed out that `` x - shaped '' radio galaxies ( xrg s ) may be signposts of smbh coalescence and hence their abundance amongst the extended radio source population might be an indicator of the true coalescence rate . using the estimate of the fraction of radio galaxies that show x morphology from @xcite and a relic lobe lifetime of @xmath0 they estimated the merger rate to be about @xmath1galaxy@xmath2 .
x - shaped radio sources in particular , and the directional stability of the jets in active galactic nuclei in general , may be probes of the coalescence rate and smbh binary formation ( e.g. , @xcite ) . however , the relevance of xrg s in this context depends on their formation mechanism whether their radio structures arise from reorientations of the bh spin axes ( spin flips ) , or slow precession or drifts in the axes , or whether the x - shaped structures are caused by the hydrodynamic deflection of backflows or jet - shell interactions or have nothing to do with changing directions of the central beams ( for references and a discussion of the various proposed mechanisms see @xcite .
these different mechanisms may be operating in different cases , generating the variety in radio structures ; therefore , it is crucial to determine the specific abundance of `` genuine xrg s '' that are candidates for reorientations of smbh spin axes as these are what are relevant for the gwb .
all of the above mechanisms for generating off - axis distortions lead to radio structures with low axial ratios , and to deduce the precise mechanism in individual cases might be difficult . for this reason , good quality radio , optical , and x - ray studies of populations of complete samples of extended radio sources that have small axial ratios are critical to the problem .
in @xcite we presented an initial step in this direction using available data in the very large array archives on the sample of 100 xrg candidates compiled by @xcite from the nrao first survey .
characterization of the extended structure in all of the sources showed that a large majority of the sample 33 of 52 sources , or 63% ) constitutes sources where the off - axis emission is traced to individual lobes .
these sources were further divided into those where the emission occurs near the inner ends of the lobes ( 25 of 52 sources , or 48% ) and those where the emission occurs near the outer ends ( 8 of 52 sources , or 15% ) .
an example of each of these classes is shown in figure [ fig : notx ] .
the sources with distortions near the inner ends of the on - axis lobes , and in which the off - axis emission on the two sides are not collinear , appear to be the best candidates for back flow deflections by thermal halos of the host ellipticals ( @xcite ; but see @xcite ) .
it is possible that such inversion symmetric structures ( which would classify as xrgs in the literature if the off - axis emission is long enough ) might arise from rapid swings or flips of the jet axis during a merger ; however , in explaining the variety in source structures this would require models where in some cases nuclei along with the host galaxy are displaced a considerable distance during mergers .
additionally , in most cases the timescales associated with the flip would require to be fine tuned so that on one hand it is short enough to produce distinct wings and lobes and long enough for the older lobes of edge - brightened type to have morphed into edge - darkened or fr - i type .
hydrodynamic forces associated with the thermal gas halo of the host elliptical would now have to be invoked to create the emission gap between the two lobe pairs .
most importantly , wings are almost always along the host minor axis and the active main lobes are towards the major axis @xcite , also for a sub - sample of the present set of sources , @xcite ) : this requires that all of these inversion symmetric xrgs would have to be formed as a result of mergers that flip the axis from the minor to major axis . on the other hand ,
the sources with distortions near the outer ends of the on - axis lobes appear to be candidates for cases of `` axis precession or drift , '' with radio structures that may signal the presence of binary black holes @xcite , and thus be precursors of genuine x systems . excluding these two source categories , eleven sources ( 21% )
have been identified as genuine x - shaped radio galaxy candidates.0830 , j1008 + 0030 , j1015 + 5944 , j1043 + 3131 , j1327@xmath30203 , j1345 + 5233 , j1406 + 0657 , j1408 + 0225 , j1606 + 0000 , j1614 + 2817 and j1625 + 2705 . ]
two examples of candidate `` genuine x '' sources are shown in figure [ fig : genuinex ] .
we point out that our source morphology classification into outer - end and inner - end deviation sources can have implications for the general population of sources presently classified as xrgs . in our classification scheme
several of the well - known xrgs will be classified as inner - end deviation sources and as argued above are more easily explained via the backflow deflection phenomenon rather than the spin - flip model . for
the sources in figure [ fig : genuinex ] the observations that we have do not indicate that the central transverse emission originates in either of the two lobes .
hence they are classified as x - shaped radio sources . in the present paper
we discuss the implications of these results for the gwb .
, title="fig : " ] + , title="fig : " ] , title="fig : " ] + , title="fig : " ]
our imaging of a sample of low - axial ratio radio sources has allowed us to identify examples where the central transverse emission may be genuinely independent radio emission unrelated to the current activity .
these are sources in which , within the limitations of sensitivity and resolution of the observations , the central transverse emission on the two sides can not be traced to either of the two lobes , and appears as a swathe of emission through the central core or galaxy but at a large angle with respect to the main radio axis .
we have identified 11 such sources ; these sources are out of the sample of 52 sources whose radio structures we have examined carefully .
these 52 are out of the sample of 100 sources that have been identified by @xcite to have low axial ratios and hence represent sources with substantial off axis emission .
the sample itself was selected from the first survey that was searched for extended sources by selecting those with fitted major axis exceeding 15 ( the first beam was about 5 fwhm ) ; this yielded a total of 1648 extended radio sources .
the search of the first images was limited to sources with peak flux density exceeding 5 mjy beam@xmath2 .
altogether this leads to an estimate that at most 1.3% of extended radio sources might be genuine x - shaped radio sources . if in these sources
( which we call genuine xrg s ) the central swathe of off - axis emission arises because of axis reorientation alone then our observations point to an occurrence rate of less than 1.3% for axis reorientations amongst the extended radio source population
. it may be noted that this rate of axis - reorientations in radio galaxies is much smaller than the occurrence rate ( of 7% ) estimated by @xcite for xrgs ( in the luminosity range of @xmath4@xmath5 w hz@xmath2 ) .
however , leahy & parma derived only the abundance fraction of radio galaxies with wings , without examining whether they were genuine x - shaped radio sources and without excluding sources where the off - axis emissions were obviously inversion - symmetric extensions of the lobes of a double lobed structure .
based on the first survey , the number of such candidate x - shaped sources was deemed to be 100 @xcite , which is indeed a 6% fraction of the 1648 extended radio sources that were selected for examination . on a related note , if the eight sources that appear to have off - axis emission connected to the outer lobe ends are representative of axis precession or drifts then our observations point to an occurrence rate of 0.9% for axis drifts ( occurring over a smaller angle of few tens of degrees ) .
it has been argued that sources with substantial off - axis emission that is traceable to the individual inner - lobes might form via a precession of the axis @xcite . while it is possible that precession may explain some source structures , it may be noted that @xcite showed that the radio structures in a more inclusive sample of x - shaped sources , which also included these sources with wings traceable to the inner - lobe ends as well , have a statistical relationship with the shape of the host elliptical in that the current radio axis is aligned closer to the major axis of the galaxy .
we see no obvious reason why axis reorientations ought to be almost always from an initial direction along the host minor axis to one along the major axis .
it is more reasonable therefore to suppose that the off - axis emission in most sources in such an inclusive sample arises from backflow deflections , as concluded by @xcite , and that the number of genuine x - shaped sources is a small subset of these candidates , as we suggest here .
prediction of the level of the gwb arising from the mergers of massive galaxies depends on estimation of the merger rate @xcite .
this has been done in a variety of ways including studies of close pairs of galaxies @xcite , simulations of structure formation @xcite .
for example , recent observations of a large number of close galaxy pairs @xcite show that a typical @xmath6 galaxy has undergone @xmath7 mergers since @xmath8 . taking the mean of these numbers implies a major merger rate somewhat less than @xmath9 .
estimation of the level of the gwb also depends on understanding the process of smbh coalescence as driven by dynamical friction @xcite and gravitational radiation @xcite , as well as knowing how the last parsec problem is overcome @xcite
. it would be useful to have an independent observational normalization for all of this , and the x - shaped radio galaxies offer one avenue in that direction . in this model , the coalescence rate for luminous ellipticals
might be estimated by examining the occurrence rate of genuine x - shaped radio sources in the population of extended radio sources .
if we use our finding above that at most 1.3% of extended sources might have genuine x - shapes , then the occurrence rate of coalescence in their hosts is given by 0.013/@xmath10 , where @xmath10 is the lifetime of relic lobes of radio galaxies .
if we adopt the value of @xmath11 yr for the lifetimes of the relics ( as assumed by @xcite ) then we arrive at an estimate of 0.13 gyr@xmath2galaxy@xmath2 as an upper limit to the event rate for radio galaxies , and presumably therefore for the parent population of massive ellipticals . on the other hand ,
if we include the eight sources with off - axis emission traced to outer ends of lobes as sources that have undergone precession due to the presence of binary black holes @xcite then we estimate a coalescence rate of at most twice this value , of @xmath12 0.2 gyr@xmath2galaxy@xmath2 . in either case
the event rate is substantially smaller than the event rate of about 1 gyr@xmath2galaxy@xmath2 usually assumed in calculations of the gwb from galaxy formation @xcite .
our revision of the event rate is owing to the morphology classification adopted .
if we consider the four examples of xrgs in @xcite in their figure 2 , only 3c223.1 would have been included as a potential `` genuine xrg '' as per our scheme .
for the other three xrgs ( 3c52 , 3c403 , & ngc326 ) given the available images we would note them as cases where the wings link with individual lobes rather than forming an independent transverse emission feature .
in @xcite we analyzed 1.4 and 5 ghz archival vla continuum data on a sample of 52 first radio sources selected on the basis of low - axial ratio radio structures . in the current paper
we discuss the impactions of the results for the gravitational wave background .
our principal results are : 1 .
we argue that a large fraction of the sample ( 63% ) consists of sources where the off - axis emission is traced to individual lobes , which are more likely to be due to back flows or axis drifts rather than axis flips .
2 . in this case
the number of genuine x - shaped radio sources is at most eleven ( 21% ) .
we then derive a five times lower occurrence rate for genuine x - shaped radio galaxies than the value estimated by @xcite .
4 . if x - shaped radio galaxies arise because of major mergers alone then our observations point to an occurrence rate of less than 1.3% for such mergers amongst the extended radio source population . 5 . the low occurrence rate for major mergers suggests a significantly smaller event rate for coalescences of supermassive black holes of less than 0.13 gyr@xmath2galaxy@xmath2 for hosts of radio galaxies .
inclusion of radio galaxies with off - axis emission traced to outer ends of lobes as sources that have undergone axis precession due to a close binary black hole only increases the coalescence rate by a factor of two
if the class of sources with inner deviations are also cases of axis flips , then the event rate would be substantially greater , and similar to that derived by merritt and ekers ( 2002 ) . indeed a measurement of the gwb might be a clue to the origins of the xrg structures
our estimate of the coalescence rate of giant ellipticals could be used to normalize calculations of the gwb from such events .
the elucidation of the structures in non - classical radio sources is important in that it has the power to yield event rates for smbh coalescence and occurrence rates for a variety of phenomena that perturb the stability of the bh spin axis and interrupt the accretion flow that powers the beams .
the work presented in @xcite and herein using primarily vla archival data has been impactful in its conclusions for the event rate that determines the gwb , and clearly the conclusions could be substantially firmer with better data .
improved imaging in radio and optical broad band continuum , followed by derivation of source orientations in 3d with respect to the host ellipticity using depolarization distributions , and deriving the abundance statistics for the different types of off - axis distortions , may reveal the formation mechanisms for small - axial - ratio radio sources and hence provide a useful constraint on gwb from bh coalescence that accompanies interactions that reorient the spins of radio loud agns . while the improved resolution ( compared to the first survey ) in the observations presented in @xcite has helped in unraveling basic structure in several of the sources ,
the sensitivity is too limited in tracing fully the larger lobe and wing structures , all of which are critical in isolating genuine x - shaped radio galaxies .
deeper observations are needed to determine the structures more accurately than what have been imaged with the limited sensitivity archival observations presented here .
radio observations at multiple wavelengths would be useful in detecting gradients in spectral index within the lobes and hence indicate the time evolution in the structure .
we anticipate that with improved data the number of genuine candidates for axis reorientations would be further constrained , thereby reducing the event rates further and the discordance with formation simulations would be on firmer ground .
critical to the estimation of the smbh binary formation rate is the assumed lifetime of the relic radiation in an xrg .
the estimate of @xmath13 @xcite was based primarily on modeling of the spectral aging of the off - axis wings . with improved observations of the @xcite sample
we will be able to make similar estimates for a larger sample of xrgs and thus firm up this critical parameter in the computation of the coalescence rate . thus new jvla observations are needed to test different formation models , estimate occurrence rate of axis reorientations , and hence also to place more accurate limits on the expected gwb signal . if models for axis flips are the explanations for inner lobe distortions , then we may explore the consequence of the inference that flips almost always occur from minor to major axis .
indeed , giant radio sources do have their axes along the minor axis of the hosts @xcite .
this leads to the suggestion that the minor axis is the stable state for smbh spins .
explorations of the torques and tidal couplings between the large - scale structure of the host elliptical and the inner accretions disk and hence bh spin axis may suggest an explanation for this phenomenology .
the national radio astronomy observatory is a facility of the national science foundation , operated under cooperative agreement by associated universities , inc .
d. h. r. gratefully acknowledges the support of the william r. kenan , jr
. charitable trust .
we thank the referee , prof .
david merritt , for his useful suggestions . | coalescence of super massive black holes ( smbh s ) in galaxy mergers is potentially the dominant contributor to the low frequency gravitational wave background ( gwb ) .
it was proposed by @xcite that x - shaped radio galaxies are signposts of such coalescences , and that their abundance might be used to predict the magnitude of the gravitational wave background . in @xcite we present radio images of all 52 x - shaped radio source candidates out of the sample of 100 selected by @xcite for which archival vla data were available .
these images indicate that at most 21% of the candidates might be genuine x - shaped radio sources that were formed by a restarting of beams in a new direction following a major merger .
this suggests that fewer than 1.3% of extended radio sources appear to be candidates for genuine axis reorientations ( `` spin flips '' ) , much smaller than the 7% suggested by @xcite .
thus the associated gravitational wave background may be substantially smaller than previous estimates .
these results can be used to normalize detailed calculations of the smbh coalescence rate and the gwb . |
one of the most promising candidates for quantum computation is spin systems @xcite due to their long decoherence and relaxation time .
combined with nanostructure technology , they have the potential advantage to scale up to large systems . just as other quantum systems ,
the spin systems are inevitably influenced by their environment , especially the spin environment . as a result ,
decoherence due to the presence of the environment will cause the transition of a system from pure quantum states to mixed ones .
the decoherent behavior of a single spin or several spins interacting with a spin bath has attracted much attention in recent years @xcite .
the interaction in such a spin bath system often leads to strong non - markovian behavior .
the usual markovian quantum master equations which are widely used in the area of atomic physics and quantum optics may fail for many spin bath models .
therefore it becomes more and more important to develop methods that are capable of going beyond the markovian approximation @xcite .
entanglement has been recognized as one of the most amazing aspects of quantum mechanics .
it has been considered as important resources for applications in quantum communication and information processing , such as quantum teleportation @xcite , quantum cryptography @xcite , quantum dense coding @xcite , and telecloning @xcite .
it is also believed to be one of the features that make quantum computers more powerful than classical ones . for spin systems ,
much attention has been dedicated to the problem of thermal entanglement @xcite , i.e. , to quantify entanglement arising in spin chains at thermal equilibrium with an environment or a reservoir . in this approach ,
a thermal distribution of the system energy levels is determined by the environment temperature , but the detailed interaction between the system and environment , and the evolution of the system toward the thermal equilibrium are explicitly ignored .
a quantum system exposed to environmental modes is described by the reduced density matrix when the environment modes are traced over .
the time evolution of the reduced density matrix is usually very difficulty to obtain in the case of non - markovian process .
recently , the dynamics of the reduced density matrix for one- , two- and three - spin - qubit systems in a spin bath described by the transverse ising model has been analyzed without making the markovian approximation , but using a perturbative expansion method @xcite or a mean - field approximation @xcite .
the interaction between the system and the spin bath for these cases was assumed to be of a ising type .
it has also been reported recently that the exact reduced dynamics and for one- and two - spin - qubit systems in a spin - star environment @xcite has been derived and analyzed @xcite . there
, the interaction between the system and environment was assumed to be of a heisenberg @xmath0 interaction . in this paper
, we study a two - spin - qubit system in a spin star configuration , similar to the case studied in ref . .
there are however several important differences between our model and that of ref . .
in ref . , the interaction between the two - spin qubits and the interaction between the constituents of the spin environment are neglected , i.e. , no internal dynamics for both the spin qubit system and the spin environment is considered .
in addition , the spin environment is assumed to be initially in an unpolarized infinite temperature state . as a result ,
no dependence of the environment temperature on the dynamics and entanglement is present .
it is under these conditions that the `` exact '' dynamics is reported in ref . .
neglecting the direct interaction among the constituents in the environment and considering only the infinite temperature initial state may not be proper in dealing with spin baths . in this paper
, we investigate a more general case . using a novel operator technique , we present an exact calculation of the dynamics of the reduced density matrix of two coupled spins interacting with a thermal spin bath at finite temperatures in the thermodynamic limit . in our model ,
the interaction between the constituents of the spin environment , the interaction between the two spin qubits , and the interaction between the the spin qubit system and the spin environment are all of the heisenberg @xmath0 type and can all be taken into account simultaneously .
in addition , we include also the zeeman coupling between the spin qubits and a locally applied external magnetic field . to quantify quantum entanglement dynamics of the two coupled spins under the influence of the spin bath at arbitrarily finite temperatures in the thermodynamics limit , we calculate the exact time evolution of their concurrence .
our model involves the heisenberg @xmath0 coupling which has extensive applications for various quantum information processing proposals @xcite .
in addition , the transformed hamiltonian of the total system in our approach becomes effectively jaynes - cumming like and thus our analysis is also very relevant to cavity quantum electrodynamics @xcite .
the paper is organized as follows . in sec .
ii the model hamiltonian is introduced and the operator technique is employed to obtain the reduced density matrix , taking into account the memory effect of the environment . from the reduced density matrix , the entanglement measure of concurrence of the coupled spin system
is calculated in sec .
conclusions are given in sec .
we consider a two - spin - qubit system interacting with bath spins via a heisenberg _ xy _ interaction . the system and bath are composed of spin-@xmath1 atoms . we restrict ourselves to a star - like configuration with coupling of equal strength , similar to the cases considered in refs . .
the interactions between bath spins are also of _ xy _ type . in refs . , a similar but somewhat different type of ising interactions between the constituents of the spin bath was considered .
the hamiltonian for the total system is @xmath2 here , @xmath3 and @xmath4 are the hamiltonians of the system and bath respectively , and @xmath5 is the interaction between them @xcite . they can be written as @xmath6,\\ h_b&=&\frac{g}{n}\sum_{i\neq j}^n\left(s_i^+s_j^-+s_i^-s_j^+ \right),\end{aligned}\ ] ] where @xmath7 represents the coupling constant between a locally applied external magnetic field in the @xmath8 direction and the spin qubit system .
@xmath9 is the the coupling constant between two qubit spins . @xmath10 and @xmath11 ( @xmath12=1,2 ) are the spin - flip operators of the qubit system spins , respectively . @xmath13 and @xmath14 are the corresponding operators of the @xmath12th atom spin in the bath .
the indices of the sums for the spin bath run from @xmath15 to @xmath16 , where @xmath16 is the number of the bath atoms .
@xmath17 is the coupling constant between the qubit system spins and bath spins , whereas @xmath18 is that between the bath spins .
both coupling strengths are rescaled such that the free energy is extensive and a nontrivial finite limit of @xmath19 exists @xcite . by using collective angular momentum operators
@xmath20 , we rewrite the hamiltonians , eqs .
( 3 ) and ( 4 ) , as @xmath21,\\ h_b&=&\frac{g}{2j}\left(j_+j_-+j_-j_+ \right)-g,\end{aligned}\ ] ] where @xmath22 is the length of the pseudo - spin . after the holstein - primakoff transformation @xcite , @xmath23 with @xmath24=1 $ ] , the hamiltonian , eqs . ( 5 ) and ( 6 ) , can be written as @xmath25,\\ h_b&=&g\left [ b^+\left(1-\frac{b^+b}{n}\right ) b+\sqrt{1-\frac{b^+b}{n}}bb^+\sqrt{1-\frac{b^+b}{n}}\right]-g.\end{aligned}\ ] ] in the thermodynamic limit ( i.e. @xmath26 ) at finite temperatures , we then have @xmath27,\\ h_b&=&2 g b^+b.\end{aligned}\ ] ] equations ( 2 ) , ( 10 ) and ( 11 ) are then effectively equivalent to the hamiltonian of a jaynes - cumming type .
they describe two coupled qubits interacting with a single - mode thermal bosonic bath field , so the analysis of the problem is also relevant to cavity quantum electrodynamics quantum information processing proposals @xcite .
we note here that due to the high symmetry of our model , the coupling to the environment is actually represented by a coupling to a single collective environment spin .
after the holstein - primakoff transformation and in the thermodynamic limit , this collective environment spin is transformed into a single - mode bosonic thermal field .
the effect of this single - mode environment on the dynamics of the two coupled qubits is extremely non - markovian .
this reflects onto , for example , the revival behavior of the reduced density matrix or entanglement evolution of the two coupled spins , which will be shown later .
this is different from the usual environment models which consist of very large degrees of freedom ( e.g. many bosonic modes ) and often cause the reduced dynamics of the system of interest displaying an exponential decay in time behavior .
so the markovian approximation usually used in quantum optics master equation will not work in our model .
one may perform perturbation theory for weak - coupling case , but the single - mode environment in our model will not remain in thermal equilibrium state as usually assumed for an environment with very large degrees of freedom in the weak - coupling master equation approach . using a special operator technique ,
we can obtain the exact reduced density matrix for the two coupled qubits by tracing over the degrees of freedom of the bosonic bath at arbitrarily finite temperatures .
reference reported the theoretical results of entanglement dynamics of a coupled two - level atoms interacting with a cavity mode embedded in an effective atomic environment . however the influence of the environmental temperature was not considered . in ref .
, a decoupled two - qubit system interacting with a single - mode thermal field at resonance ( i.e. , zero detuning ) in the context of cavity electrodynamics was studied . there
, the dynamics of the reduced density matrix for the two - qubit system is obtained using the method of the kraus operator representation . in this paper
, we use a different approach of operator technique to obtain the exact non - markovian dynamics of the reduced density matrix for the two - qubit system for our model of eqs .
( 2)-(4 ) with the bath spins in the thermodynamics limit , or equivalently eqs .
( 2 ) , ( 10 ) and ( 11 ) . different from the case considered in ref .
, our model furthermore includes the coupling between the two qubits and investigates the effect of detuning ( i.e. , the single bosonic bath mode is not necessarily resonant with the qubit transition frequency ) .
the detuning in our model is represented by @xmath28 and it could be controlled by the strength of a locally applied magnetic field , i.e. , the @xmath7 term in eq .
we find that the detuning has a significant effect on the entanglement generation between the two qubits .
we assume the initial density matrix of the total system to be separable , i.e. , @xmath29 .
the density matrix of the spin bath satisfies the boltzmann distribution , that is @xmath30 , where @xmath31 is the partition function , and the boltzmann constant has been set to one . at absolute zero temperature ,
no excitation will exist .
the bath is in a thoroughly polarized state with all spins down . with the increase of temperatures , the number of spin up atoms increases .
note that in ref . , the non - interacting bath spins are assumed to be initially in the unpolarized infinite temperature state .
the most general form of an initial pure state of the two - qubit system is @xmath32 we might proceed the calculation with this general initial state , but the final analytical solution would , however , be somewhat complicated . for analytical simplicity , in the following we set @xmath33 .
we note that the general initial qubit state case can be calculated in a similar way presented below . by taking the initial state of the two qubit system to be @xmath34
, the reduced density matrix can be written as @xmath35\nonumber\\ & & + \frac{1}{z}\alpha\beta^*\textrm{tr}_b\left[e^{-iht}|00\rangle e^{-h_b / t}\langle11| e^{iht}\right]\nonumber\\ & & + \frac{1}{z}\alpha^*\beta\textrm{tr}_b\left[e^{-iht}|11\rangle e^{-h_b / t}\langle00| e^{iht}\right]\nonumber\\ & & + \frac{1}{z}|\beta|^2\textrm{tr}_b\left[e^{-iht}|11\rangle e^{-h_b / t}\langle11| e^{iht}\right],\end{aligned}\ ] ] where @xmath36 the matrix @xmath37 is a @xmath38 matrix in the standard basis @xmath39 , @xmath40 , @xmath41 , @xmath42 . in order to obtain the exact reduced density matrix elements
, we have to evaluate eq .
( 14 ) exactly .
however , it is difficult to do so using usual methods , because the treated system and environment could all be strongly coupled mutually and internally . in the following
, we will present a special operator technique to obtain the exact density matrix elements . as shown below
, our treatment is mathematically simple and physically clear , and may be easily extended to more complicated systems with strong coupling .
note also that our method also applies to the case that the two - qubit system is initially in a mixed state .
for example , if the initial state for the qubits is @xmath43 , the corresponding reduced density matrix is eq .
( 14 ) provided that the second and third terms on its right hand side are removed .
the basic idea of our operator technique is as follows . before tracing over the environmental degrees of freedom
, we will first convert the time evolution equation of the qubit system under the action of the total hamiltonian into a set of coupled non - commuting operator variable equations . then by introducing a new set of transformation on the operator variables
, we turn the coupled non - commuting operator variable equations into commuting ones . as a result
, they can be solved exactly by using the general method of solving coupled first - order ordinary differential equations for ordinary variables .
after that , the trace over the environmental degrees of freedom can be performed and the exact reduced dynamics of the qubit system can be obtained . from the total hamiltonian @xmath44 , we can see that it consists of operators @xmath45 , @xmath46 , @xmath11 and @xmath10 , where @xmath11 , and @xmath10 change the @xmath12th ( @xmath47 ) qubit spin from state @xmath48 to @xmath49 , and _ vice versa_. it is then obvious that we can write in a most general form that @xmath50 where @xmath51 , @xmath52 , @xmath53 , and @xmath54 are functions of operators @xmath45 , @xmath46 , and time @xmath55 . using the schrdinger equation identity @xmath56 and eq .
( 16 ) , we obtain @xmath57,\\ \frac{d}{dt}b&=&-i\left[g_0ba+2gb^+bb+\omega c+g_0b^+d\right],\\ \frac{d}{dt}c&=&-i\left[g_0ba+\omega b+2gb^+bc+g_0b^+d\right],\\ \frac{d}{dt}d&=&-i\left[g_0bb+g_0bc+\mu_0d+2gb^+bd\right],\end{aligned}\ ] ] with initial conditions from eq .
( 16 ) being @xmath58 , @xmath59 , @xmath60 , and @xmath61 . as @xmath51 , @xmath52 , @xmath53 ,
@xmath54 are functions of @xmath46 and @xmath45 , they are operators and do not commute with each other . equations ( 18)-(21 ) are thus coupled differential equations of non - commuting operator variables , which can not be solved by using conventional methods for ordinary number variables .
the crucial observation to solve the problem is that the hamiltonian , eqs .
( 2 ) , ( 10 ) and ( 11 ) , is of an effective jaynes - cumming type and it can be block - diagonalized in the dressed state subspace of @xmath62 , with @xmath63 . here
@xmath64 represent the qubit states and @xmath65 are the bosonic field number states . as a result ,
we may rewrite eqs .
( 18)-(21 ) in such a subspace .
by introducing the following transformation @xmath66 equations ( 18)-(21 ) then become @xmath67,\\ \frac{d}{dt}c_1&=&-i\left[g_0(2+\hat{n})a_1+\omega b_1+g_0d_1\right],\\ \frac{d}{dt}d_1&=&-ig_0(1+\hat{n})\left(b_1+c_1\right)-i(\mu_0 - 2g)d_1,\end{aligned}\ ] ] where @xmath68 .
note that for initial qubit state in @xmath42 on the left hand side of eq .
( 16 ) , the transformation , eqs .
( 22)-(25 ) , is chosen in such a way that the bosonic field operator(s ) in front of the exponential term in eqs .
( 22)-(25 ) together with its corresponding qubit state on the right hand side of eq .
( 16 ) make @xmath69 a constant value and the initial condition @xmath70 .
that is , the operator coefficient @xmath51 of @xmath39 state on the right hand side of eq .
( 16 ) requires two field creation operators in eq .
( 22 ) , @xmath52 and @xmath53 of @xmath40 and @xmath41 respectively require only one field creation operator in eqs .
( 23 ) and ( 24 ) , and @xmath54 of @xmath42 does not need any field operator in eq .
the exponential term in eqs .
( 22)-(25 ) is introduced to make the resultant equations more concise . as a consequence , the coefficients of eqs .
( 26)-(29 ) after the transformation ( 22)-(25 ) involve only the operator @xmath71 .
therefore @xmath72 , @xmath73 , @xmath74 , and @xmath75 are functions of @xmath71 and @xmath55 , and commute with each other .
we can then treat eqs .
( 26)-(29 ) as coupled complex - number differential equations and solve them in a usual way .
this novel operators approach thus allows us to solve eq .
( 16 ) and then consequently the non - markovian dynamics of the reduced density matrix of the qubit system .
we note again that the crucial point of the method used here is to find proper transformations to change the coupled differential equations of non - commuting operator variables to the coupled differential equations of complex - number variables .
this can be done when the effective hamiltonian can be block - diagonalized .
this is the case of jaynes - cumming model and other models which contain interaction hamiltonian of the forms of , for example , @xmath76 , @xmath77 , etc .
, regardless how strong their interaction strengths are .
if the total effective hamiltonian can not be block - diagonalized , for example , for the effective spin - boson model in ref .
, the operator method used here will then not apply to solve the problem exactly . as we are working in the schrdinger picture ,
the basic operator @xmath68 is time independent ( sometimes the operators could have time dependence explicitly ; however this is not the case here ) . from eq .
( 16 ) and eqs .
( 22)-(25 ) , the initial conditions at @xmath78 are given by @xmath79 in general , we can solve eqs .
( 26)-(29 ) exactly via the initial conditions eqs .
( 30)-(33 ) . as we aim to obtain analytical expressions for the reduced qubit dynamics , for the sake of analytical simplicity we consider the on - resonant case , i.e. , @xmath80 .
we can easily tune the locally applied external magnetic field to satisfy this condition
. we will give the numerical results for the off - resonant case in fig .
we then obtain for the on - resonant case @xmath81,\\ b_1&=&c_1=-\frac{g_0}{\sqrt{\omega^2 + 8(3 + 2\hat{n})g_0 ^ 2}}\left[e^{i\lambda_1t}-e^{i\lambda_2t}\right],\\ d_1&=&\frac{2+\hat{n}}{3 + 2\hat{n}}+\frac{2g_0 ^ 2(1+\hat{n})}{\sqrt{\omega^2 + 8(3 + 2\hat{n})g_0 ^ 2}}\left[\frac{e^{i\lambda_1t}}{\lambda_1}-\frac{e^{i\lambda_2t}}{\lambda_2}\right],\end{aligned}\ ] ] where @xmath82 following the similar calculations above , we can evaluate the time evolution for the initial two - qubit spin state of @xmath39 . let @xmath83 in a similar way , we have @xmath84 and then obtain @xmath85,\\ f_1&=&g_1=-\frac{g_0}{\sqrt{\omega^2 + 8(2\hat{n}-1)g_0 ^ 2}}\left[e^{i\lambda_1't}-e^{i\lambda_2't}\right],\\ k_1&=&\frac{-1}{2\hat{n}-1}+\frac{2g_0 ^ 2}{\sqrt{\omega^2 + 8(2\hat{n}-1)g_0 ^ 2}}\left[\frac{e^{i\lambda_1't}}{\lambda_1'}-\frac{e^{i\lambda_2't}}{\lambda_2'}\right],\end{aligned}\ ] ] where @xmath86 from eq .
( 14 ) and all the results that we obtained , the reduced density matrix can be written as @xmath87 where @xmath88 in eqs .
( 48)-(51 ) , the trace over the environmental degrees of freedom has been performed and the operator @xmath71 has been replaced by its eigenvalue @xmath89 . in a similar way , the solutions for the reduced dynamics of the two - coupled spins under the influence of the quantum heisenberg @xmath0 spin star bath in the thermal dynamics limit at arbitrarily finite temperatures for arbitrary initial states of @xmath90 can be obtained .
we use the concurrence @xcite to measure the entanglement between the two coupled qubit spins .
it is defined as @xcite @xmath91 where the quantities @xmath92 are the square roots of the eigenvalues of the operator @xmath93 we find @xmath94 are values in decreasing order of @xmath95 , @xmath96 , @xmath97 , and @xmath98 . for the system with an initial state of @xmath99 ,
i.e. , the both spins in the ground state , we plot the time evolution of the concurrence in fig . 1 .
although there is no initial entanglement and no coupling between the two spins , it is interesting to notice that the entanglement between the two spins after some time is present as shown in fig .
this confirms that the environment which usually causes the decoherence of the system can nevertheless entangle qubits that are initially prepared in a separable state @xcite .
this is mainly due to the fact that the two spins are coupled to the same , common environment which then in turn generates some effective interaction between the two spins even if they were originally decoupled .
the result , however , depends on the environmental temperature .
further numerical calculations show that no entanglement is generated , for example , for @xmath100 . as the coupling between the two qubit spins
is switched on even though the value is small , the `` collapse '' and `` revival '' of the entanglement as a function of time are demonstrated in fig .
this is in analogy to the collapse and revival of atomic population inversion of a single two - level atom interacting with a single mode field initially in a coherent state @xcite , a fock state @xcite or a squeezed state @xcite in quantum optics . here from the hamiltonian , eqs .
( 2 ) , ( 10 ) and ( 11 ) , this novel phenomenon of entanglement arises from two - coupled qubits interacting with a single mode field initially however in a thermal state . the reasons for causing the collapse and revival behaviors in these cases could be similar , that is the rabi ( or time evolution ) oscillations associated with different excitations have different frequencies .
consequently , as the time increases , these rabi ( or time evolution ) oscillations become uncorrelated leading to a collapse behavior . as time
is further increased , the correlation is restored and the revival occurs . for different values of @xmath9 ; ( a ) @xmath101 and ( b ) @xmath102 .
other parameters are @xmath80 , @xmath103 , @xmath104.,title="fig:",width=309 ] for different values of @xmath9 ; ( a ) @xmath101 and ( b ) @xmath102 .
other parameters are @xmath80 , @xmath103 , @xmath104.,title="fig:",width=309 ] figure 2 shows the time evolution of the concurrence for the system in the initial state of @xmath105 , i.e. , an maximally entangled state . at high temperature
, the state loses its entanglement completely for a short period of time , and then it is partially entangled again some times later .
this is in agreement with the results of ref .
where an initially unpolarized infinitely temperature states of the spin bath is assumed . however , our results are temperature dependent . at a very low temperature , the concurrence exhibits the behavior of oscillation between 1 and 0.35 . with increasing temperatures , the concurrence decreases more quickly and oscillates disorderly in the lower value region . at a fixed time @xmath55 ,
the concurrence decreases with temperatures and a critical temperature @xmath106 exists , above which the entanglement vanishes . however , this critical temperature @xmath106 is time - dependent and sensitive to the initial state of the system .
figure 3 illustrates the time evolutions of the concurrence for different values of the coupling constant @xmath17 . as expected , increasing the value of the coupling constant has similar effects as increasing the value of the environmental temperature , i.e. , the decay rate of the concurrence increases .
figure 4 presents the time evolution of the concurrence for different inner - bath - spin coupling constants @xmath18 .
we see that the concurrence increases with the increase of @xmath18 .
this confirms that strong quantum correlations within the environment suppress decoherence @xcite and thus perhaps also disentanglement .
as shown in the inset , the concurrence is regained sometime later .
however , it appears disorderly without a particular pattern .
similar behaviors arise for fig . 2 , fig . 3 , and fig . 5 in the long time scales , which reflect the non - markovian dynamics of the system
. in fig . 5
, we show the effect of the coupling between the two qubit spins on the concurrence .
it is obvious that the coupling benefits the entanglement .
note that the initial state of the system in this case is different from that in fig . 1 .
so , if the system is initially prepared in a maximally entangled state , the larger the coupling constant @xmath9 is , the more slowly the entanglement decays . for different temperatures ; @xmath107 ( solid curve ) , @xmath108 ( dashed curve ) and @xmath109 ( dot dashed curve ) .
other parameters are @xmath80 , @xmath103 , @xmath101.,width=309 ] for different values of @xmath17 ; @xmath103 ( solid curve ) , @xmath110 ( dashed curve ) and @xmath111 ( dot dashed curve ) .
other parameters are @xmath80 , @xmath108 , @xmath101.,width=309 ] for different values of @xmath18 ; @xmath112 ( solid curve ) , @xmath113 ( dashed curve ) and @xmath114 ( dot dashed curve ) .
other parameters are @xmath80 , @xmath115 , @xmath101 .
the inset shows the long time behavior of concurrence for @xmath112.,width=309 ] for different values of @xmath9 ; @xmath101 ( solid curve ) , @xmath116 ( dashed curve ) , and @xmath117 ( dot dashed curve ) .
other parameters are @xmath80 , @xmath103 , @xmath108.,width=309 ] for different values of detuning ; @xmath118 ( solid curve ) , and @xmath119 ( dashed curve ) .
other parameters are @xmath103 , @xmath104 , @xmath101.,width=309 ] when the two spins are initially prepared in their excited state , i.e. @xmath120 , the result is quiet different from that in fig .
1 . if the detuning @xmath80 , i.e. , at the on - resonant case , no entanglement between the two qubit spins exists at any temperatures , even with a strong interaction @xmath9 between them .
this is consistent with the result for two qubit atoms obtained by kim et al .
, @xcite where the coupling between the two qubit atoms is not considered and the detuning between the two atoms and the single - mode field is zero . if the detuning @xmath121 , i.e. , in the off - resonant case , the two qubit spins will entangle via the environment again as shown in fig .
so the entanglement generation of the two spins in this case is very sensitive to the detuning which can be controlled in our model by the locally applied external magnetic field .
we have studied the exact entanglement evolution of two coupled qubit spins in a model of a quantum heisenberg @xmath0 spin star environment in the thermodynamic limit .
the dynamics of the reduced density matrix of the two coupled spins is analytically obtained in terms of a novel operator technique which is mathematically simple and physically clear . in our analysis ,
the transformed hamiltonian becomes effectively jaynes - cumming like and thus the results are also relevant to cavity quantum electrodynamics .
the time evolutions of the concurrence of the two coupled spin qubits for different initial conditions are evaluated exactly .
the results show that the dynamics of the entanglement strongly depends on the initial state of the system , the coupling between the two spin qubits , the interaction between the qubit system and the environment , the interactions between the constituents of the spin environment , the environmental temperatures , as well as the detuning controlled by a locally applied external magnetic field .
we have also found that if the two coupled spin qubits are initially prepared in the ground state , the entanglement between them will exhibit the `` collapse '' and `` revival '' behavior with time due to the interaction between the two spin qubits and the environment , in analogy to the collapse and revival of the atomic population inversion in quantum optics .
x.z.y . and h.s.g .
would like to acknowledge support from the national science council , taiwan , under grant numbers nsc94 - 2112-m-002 - 028 , nsc95 - 2112-m-002 - 018 and nsc95 - 2112-m-002 - 054 .
also thanks support from the national taiwan university under grant number 95r0034 - 02 and is grateful to the national center for high - performance computing , taiwan , for computer time and facilities . d. loss and d. p. divincenzo , phys .
a * 57 * , 120 ( 1998 ) .
b. e. kane , nature * 393 * , 133 ( 1998 ) .
g. burkard , d. loss , and d. p. divincenzo , phys .
b * 59 * , 2070 ( 1999 ) .
r. vrijen , e. yablonovitch , k. wang , h. w. jiang , a. balandin , v. roychowdhury , t. mor , and d. divincenzo , phys .
a * 62 * , 012306 ( 2000 ) .
r. raussendorf and h.j .
briegel , phys .
lett . * 86 * , 5188 ( 2001 ) .
a. j. skinner , m. e. davenport , and b. e. kane , phys .
lett . * 90 * , 087901 ( 2003 ) .
t. d. ladd , j. r. goldman , f. yamaguchi , y. yamamoto , e. abe , and k. m. itoh , phys .
lett . * 89 * , 017901 ( 2002 ) .
r. de sousa , j. d. delgado , and s. das sarma , phys .
a * 70 * , 052304 ( 2004 ) .
m. friesen , p. rugheimer , d. e. savage , m. g. lagally , d. w. van der weide , r. joynt , and m. a. eriksson , phys .
b * 67 * , 121301(r ) ( 2003 )
. w. h. zurek , rev .
phys . * 75 * , 715 ( 2003 ) .
a. melikidze , v. v. dobrovitski , h. a. de raedt , m. i. katsnelson , and b. n. harmon , phys .
b * 70 * , 014435 ( 2004 ) . n. v. prokofer and p. c. e. stamp , rep .
. phys . * 63 * , 669 ( 2000 ) .
j. schliemann , a. khaetskii , and d. loss , j. phys .
: condens matter * 15 * , r1809 ( 2003 ) .
h. p. breuer , phys .
a * 69 * , 022115 ( 2004 ) .
h. p. breuer , d. burgarth , and f. petruccione , phys .
b * 70 * , 045323 ( 2004 ) .
y. hamdouni , m. fannes , and f. petruccione , phys .
b * 73 * , 245323 ( 2006 ) . c. h. bennett , g. brassard , c. crpeau , r. jozsa , a. peres , and w. k. wootters , phys . rev .
lett . * 70 * , 1895 ( 1993 ) . c. h. bennett , g. brassard , and a. k. ekert , sci .
( int . ed . )
* 267 * , 50 ( 1992 ) . c. h. bennett and s. j. wiesner , phys .
* 69 * , 2881 ( 1992 ) .
m. murao , d. jonathan , m.b .
plenio , and v. vedral , phys .
rev . a * 59 * , 156 ( 1999 ) .
d. gunlycke , v. m. kendon , v. vedral , and s. bose , phys . rev .
a * 64 * , 042302 ( 2001 ) .
t. j. osborne and m. a. nielsen , phys .
a * 66 * , 032110 ( 2002 )
. n. canosa and r. rossignoli , phys .
a * 73 * , 022347 ( 2006 ) .
s. paganelli , f. de pasquale , and s. m. giampaolo , phys .
a * 66 * , 052317 ( 2002 ) .
m. lucamarini , s. paganelli , and s. mancini , phys .
a * 69 * , 062308 ( 2004 ) .
x. san ma , a. min wang , x. dong yang , and h. you , j. phys .
a 38 , 2761 ( 2005 ) .
a. hutton and s. bose , phys .
a * 69 * , 042312 ( 2004 ) .
a. imamoglu , d. d. awschalom , g. burkard , d. p. divincenzo , d. loss , m. sherwin , and a. small , phys . rev
. lett . * 83 * , 4204 ( 1999 ) .
s. b. zheng and g. c. guo , phys .
. phys . * 85 * , 2392 ( 2000 ) .
d. a. lidar and l. a. wu , phys .
lett . * 88 * , 017905 ( 2002 )
. l. a. wu , d. a. lidar , and m. friesen , phys .
lett . * 93 * , 030501 ( 2004 ) .
x. wang , phys .
a * 64 * , 012313 ( 2001 ) . c. d. hill and h .- s .
goan , phys .
a * 68 * , 012321 ( 2003 ) .
a. blais , r .- s .
huang , a. wallraff , s. m. girvin , and r. j. schoelkopf , phys .
a * 69 * , 062320 ( 2004 ) .
m. sarovar , h. -s .
goan , t. p. spiller , and g. j. milburn , phys .
a * 72 * , 062327 ( 2005 ) .
m. frasca , ann . phys . * 313 * , 26 ( 2004 ) .
t. holstein and h. primakoff , phys . rev . * 58 * , 1098 ( 1949 ) .
i. sainz , a. b. klimov , and luis roa , phys .
a * 73 * , 032303 ( 2006 ) .
m. s. kim , j. lee , d. ahn , and p. l. knight , phys .
a * 65 * , 040101(r ) ( 2002 ) .
e. k. irish , j. gea - banacloche , i. martin , and k. c. schwab , phys .
b * 72 * , 195410 ( 2005 ) .
s. hill and w. k. wootters , phys .
lett . * 78 * , 5022 ( 1997 ) ; w.k .
wootters , _ ibid _ * 80 * , 2245 ( 1998 ) .
s. bose , i. fuentes - guridi , p. l. knight , and v. vedral , phys .
lett . * 87 * , 050401 ( 2001 ) .
f. benatti , r. floreanini , and m. piani , phys .
91 * , 070402 ( 2003 ) . m. b. plenio and s. f. huelga , phys .
* 88 * , 197901 ( 2002 ) .
a. napoli , f. palumbo , and a. messina , journal of physics : conference series * 36 * , 154 ( 2006 ) .
m. o. scully and m. s. zubairy , _ quantum optics _ , ( cambridge university press , cambridge , 1997 ) . m. fleischhauer and w. p. schleich , phys .
a * 47 * , 4258 ( 1993 ) .
l. tessieri and j. wilkie , j. phys .
a * 36 * , 12305 ( 2003 ) . c. m. dawson , a. p. hines , r. h. mckenzie , and g. j. milburn , phys .
a * 71 * , 052321 ( 2005 ) . | the exact quantum dynamics of the reduced density matrix of two coupled spin qubits in a quantum heisenberg @xmath0 spin star environment in the thermodynamic limit at arbitrarily finite temperatures is obtained using a novel operator technique . in this approach ,
the transformed hamiltonian becomes effectively jaynes - cumming like and thus the analysis is also relevant to cavity quantum electrodynamics .
this special operator technique is mathematically simple and physically clear , and allows us to treat systems and environments that could all be strongly coupled mutually and internally . to study their entanglement evolution ,
the concurrence of the reduced density matrix of the two coupled central spins is also obtained exactly .
it is shown that the dynamics of the entanglement depends on the initial state of the system and the coupling strength between the two coupled central spins , the thermal temperature of the spin environment and the interaction between the constituents of the spin environment .
we also investigate the effect of detuning which in our model can be controlled by the strength of a locally applied external magnetic field .
it is found that the detuning has a significant effect on the entanglement generation between the two spin qubits . |
many complex systems display a very heterogeneous degree distribution @xcite characterized by a power law decay of the form @xmath0 .
this form implies the absence of a characteristic scale hence the name of `` scale - free network '' ( sfn ) @xcite . among these networks ,
a certain number are of a great interest to epidemiology @xcite and it is thus very important to understand the effect of their topology on the spreading dynamics of a disease .
one of the most relevant results is that disease spreading does not show an endemic threshold in sfn when the population size is infinite and @xmath1 @xcite .
this result means that a disease propagates very easily on a large sfn whatever the value of its transmission probability .
in addition , recent studies showed that the presence of hubs in sfn not only facilitates the spread of a disease but also accelerates dramatically its outbreak @xcite .
the long - tailed degree distribution of sfn is the signature of the presence of a non - negligible number of highly connected nodes .
these hubs were already identified in the epidemiological literature as superspreaders @xcite .
consequently , from a public health point of view , studying the spreading of epidemics on sfn is all the more appropriate .
superspreading events affect the basic reproductive number @xmath2a widely used epidemiological parameter @xcite making its estimate from real - world data difficult @xcite . as a matter of fact
, it seems that superspreading events appeared in the onset of the recent sars outbreak @xcite and could be crucial for the new emergent diseases and bioterrorist threats .
their potential threat justifies detailed studies of the incidence of the degree distribution at the initial stage of epidemics .
the variability plays an important role in the accuracy and the forecasting capabilities of numerical models and has thus to be quantified in order to assess the meaningfulness of simulations with respect to real outbreaks @xcite . using a numerical approach
, we analyze the evolution of epidemics generated by different sets of initial parameters , both for sfn and homogeneous random networks ( rn ) .
we use the barabsi - albert model ( ba ) @xcite for generating a sfn and the erds - renyi network ( er ) @xcite as a prototype for rn . concerning the epidemic modeling ,
a simple and classical approach is to consider that individuals are only in two distinct states , infected ( i ) or susceptible ( s ) .
there is initially a number of @xmath3 infected individuals and any infected node can pass the disease to his neighbors @xcite . the probability per unit time to transmit the disease
the spreading rate is denoted by @xmath4 and once a susceptible node is infected it remains in this state . in more elaborated models ,
an infected individual can change its state to another category , for example , coming back to susceptible ( sis ) , or going to immunized or dead ( sir ) @xcite .
this s @xmath5 i approach ( si ) , in spite of its simplicity , is a good approximation at short times to more refined models such as the sis or sir models .
the si model on both sfn and rn is thus well adapted to the characterization of the variability of the initial stages of epidemic outbreaks spreading in complex networks , which is the focus of this article .
the outline of the paper is the following . in section
ii , we study the fluctuations of the prevalence and we identify different parameters controlling them .
in particular , we highlight the effects due to different realizations of the network as well as different initial conditions .
we also investigate the influence of the nodes degree on the prevalence variability . in section iii , we present results on the infection time and its variation with the degree and with the distance from the origin of infections .
we also discuss the effect of the number of paths between two nodes on the infection time .
finally , we discuss our results and conclude in section iv.=-1
we analyze in this section the effect of the underlying network topology on the variability of outbreaks .
it is indeed important to understand whether the local fluctuations of the structure of the network can have a large impact on the development of epidemics . in order to analyze this effect
, we measure the variability of outbreaks as the relative variation of the prevalence ( density of infected individuals @xmath6 ) given by @xmath7= \frac{\sqrt{\langle i(t)^2\rangle -\langle i(t ) \rangle^2}}{\langle
i(t ) \rangle}.\ ] ] in order to evaluate this quantity we run simulations for different `` model sets '' : first , for a given number of outbreaks on a single network , second for a single outbreak on different networks , and finally several outbreaks on different networks .
we show in fig .
[ fig : cv_1vs1000 ] the curves @xmath8 $ ] computed for both the rn ( thin lines ) and the sfn ( bold lines ) and for two of these model sets : @xmath9 outbreaks spreading on the same network ( dashed curves ) , and a single outbreak per network for @xmath9 different networks ( plain curves ) .
the curves representing these two model sets are nearly superimposed for both network topologies .
the curves obtained from model sets made of @xmath10 outbreaks on @xmath11 networks and @xmath11 outbreaks on @xmath10 networks coincide with the other model sets ( not shown in the figure ) .
these results indicate that the contribution to the variability of @xmath6 given by a particular network realization is essentially the same as the one generated by different outbreaks on the same network .
this confirms the intuitive idea that sampling different parts of a large network is equivalent to average over different networks .
consequently , studying variability of epidemics simulated on one large enough network ( intra - network ) will lead practically to the same conclusions as studying variability on several instances of that network ( inter - network ) .
furthermore , it means that the results described in the next sections for one network can be generalized to any instances of ba and er networks .
[ b]@xmath12 $ ] $ ] ) in ba networks ( bold ) and er network ( thin ) for outbreaks simulated on the same network ( dashed curve ) , or on different networks ( plain curve).the results are obtained for @xmath13 and on networks of size @xmath14 nodes , and average degree @xmath15 .
, title="fig : " ] fig .
[ fig : cv_1vs1000 ] also reveals interesting facts about the time behavior of @xmath8 $ ] on complex networks . since the initial prevalence is fixed and is the same for all instances , @xmath16 is initially equal to zero and can only increase . at very large times , almost all nodes are infected implying that @xmath17 .
this argument implies the existence of a peak which as shown in fig .
[ fig : cv_1vs1000]is located for ba networks at the beginning of the outbreak , with a maximum value larger than the one obtained for er networks . in order to characterize the relation between the variability peak and the network heterogeneity , we define @xmath18 as the time at which the maximum of @xmath8 $ ] is reached .
we also use the fact that the heterogeneity of the network degree often quantified by @xmath19is related to the typical outbreak timescale @xmath20 given by @xcite @xmath21
a discussion of the validity of this equation is provided in ref .
@xcite . in order to understand to
which regime corresponds @xmath18 , we plot in fig .
[ fig : tau_v_vs_tau ] @xmath18 and @xmath20 for ba networks with different values of @xmath22 .
we use networks with different sizes ( from @xmath23 to @xmath24 nodes ) and with different values of @xmath25 ( @xmath26 ) in order to obtain a broad range of @xmath20 values .
we see in fig .
[ fig : tau_v_vs_tau ] that @xmath18 is increasing linearly with @xmath20 ( with a pre - factor of order @xmath27 ) .
this implies that @xmath18 is of the same order of the typical time @xmath28 where the diversity of degree classes of infected nodes is the largest ( @xmath29 ) @xcite .
the result @xmath30 confirms the intuitive idea that the variability is maximal when the diversity of different classes of infected nodes is the largest , which happens at the beginning of the spread . versus @xmath20 for several ba networks with @xmath31 ranging from @xmath32 to @xmath33 , and different sizes ( @xmath34 : @xmath35 , @xmath36 : @xmath14 , @xmath37 : @xmath38 , @xmath39 : @xmath40 nodes ; @xmath41 ) .
the line is a linear fit with slope of order @xmath27 . ]
[ ] @xmath12 $ ] [ ] @xmath6 $ ] ) in ba networks for outbreaks seeded with infected nodes of different degrees @xmath42 ( from top to bottom , @xmath43 ) .
inset : initial evolution of the prevalence i(t ) .
the order of the curves is reversed between both plots ( results are averaged over @xmath44 epidemics on one network , with @xmath13 , @xmath14 nodes , @xmath45).,title="fig : " ] in this si model , the parameter @xmath4 simply fixes the time unit .
in contrast , we expect that other parameters such as the degree of the seed may have a more interesting effect on the outbreak and its variability .
[ fig : cv_graine ] displays the evolution of @xmath8 $ ] for outbreaks starting from initial infected nodes with a given degree @xmath42 ( from @xmath46 up to @xmath47 ) .
this figure shows that the variability peak decreases when @xmath42 is increased . in other words , when an outbreak begins from a highly connected node , the early stages of the spreading tend to be less variable .
one might think that the number of paths available on a highly connected node leads to a higher overall variability , it is however not the case . as shown in the inset of fig .
[ fig : cv_graine ] , the prevalence increases with the seed degree , which may explain the variability for different @xmath42 .
indeed , when the seed is a hub , the number of infected becomes rapidly very large and thus leads to smaller relative variations of the prevalence .
this result leads us to investigate more thoroughly the degree of infected nodes and analyze the differences between ba and er networks . in this section
, we study in detail the degree properties of the infected nodes during the outbreak of the disease . [
cols="<,^ " , ] the reason for this behavior lies in the difference of the numbers of shortest paths in these networks . indeed ,
if we enumerate these paths , we observe that their numbers relatively differ between both ba and er topologies .
we have computed the size and the number of shortest paths between a randomly selected node , i.e. a potential seed of infection and the rest of the network and we present in table [ tab : nb_sp ] the average number of shortest path at distance @xmath48 .
results are computed over @xmath9 random selection of the potential seed in order to get an accurate picture of the network .
the table exhibits a difference in the number of path for @xmath49 ( @xmath50 difference for @xmath51 , @xmath52 for @xmath53 , @xmath54 for @xmath55 ) which confirms the fact that on ba networks , nodes have more paths to go from one to another in a small number of hops . for two different configurations . in the first case infection
occurs in one step and in the second case another path is added .
the dotted curve represents the average time of infection for the first case , @xmath56 and the plain curve represents @xmath57 for the second case and is given by eq .
( [ eq : mintdti ] ) .
the result of a numerical simulation are shown by @xmath58 symbols ) . ]
table [ tab : nb_sp ] describes the statistics of shortest paths but longer paths also contributes to the spreading of the disease .
their role can be highlighted by studying the following simple cases . in the first case
an infected node @xmath59 is in contact with a susceptible node @xmath60 . in the second case
, there is an additional path from @xmath59 to @xmath60 going through a susceptible node @xmath61 ( see fig . [
fig : avg_t ] ) . in the first `` direct '' case , the average time of infection @xmath62 of @xmath60
is given by @xmath63 the addition of a longer path in the second case ( fig .
[ fig : avg_t ] ) changes the behavior of @xmath64 and eq .
( [ eq : td ] ) no longer holds for this case .
in fact , the time of infection of the susceptible node @xmath60 is given by @xmath65 , \label{eq : mintdti}\ ] ] where @xmath66 is the time of a direct infection @xmath67 and @xmath68 of an indirect 2-steps infection process : @xmath69 .
the statistics of @xmath70 can be easily computed and its first moment reads @xmath71 eq .
( [ eq : avg_t ] ) predicts values always smaller than @xmath72 ( see fig .
[ fig : avg_t ] ) .
this result could appear as paradoxical since adding a _ longer _
path actually _ reduces _ the average infection time .
in fact , the probability that the disease is not transmitted on both paths is very small and the existence of another path cuts off large direct infection time and thus reduces the average infection time of @xmath60 . since ba networks have a clustering coefficient larger than er networks @xcite this result explains the small difference of infection times for @xmath73 seen in fig .
[ fig : density_seq_dtopo_sfn_rn ] . from the seed .
top panel : spreading on a ba network ; bottom panel : er network .
both panels show @xmath74 , for every nodes of a single network , @xmath75 nodes , @xmath76 , and computed over @xmath9 outbreaks , @xmath41 , originating from exactly the same seed of degree @xmath77 . ] concerning the relationships between the relative dispersion of infection time @xmath74 and @xmath48 , their behavior on both topologies are reported on fig .
[ fig : ell_cv ] .
this figure shows that the nodes in both networks exhibit higher values of @xmath74 when they are closer to the seed , i.e. for @xmath78 . for larger distances
, @xmath74 is practically constant in both cases .
we have analyzed in detail the variability of a simple epidemic process on sfn .
first , we have shown that different realizations of ba networks do not display significant statistical differences in outbreak variability .
consequently , it is statistically reliable to consider a single realization of the network , provided it is large enough .
we have also shown that the prevalence fluctuations are maximal during the time regime for which the diversity of the degrees of the infected node is the largest . in order to analyze in detail this variability , we examined the temporal degree pattern of infected nodes .
in particular , we demonstrated the high variability of superspreaders prevalence .
we found that for the hubs the infection time is usually small but with fluctuations which can be large .
even if the hubs are good candidates for being chosen as surveillance stations given their short average infection time , they present non - negligible fluctuations which limit their reliability . in this respect ,
the ideal detection stations should be nodes with the best trade - off between a short average infection time and a high reliability as given by small infection time fluctuations .
the topological distance to the seed is also an important parameter in epidemic spreading pattern .
nodes at a short distance from the seed are infected at small time in the high variability regime and thus have a large infection time variability .
maybe more surprising is the importance of the number of paths not only the shortest one going from the seed to another node .
the larger this number and the smaller the average infection time .
this is an important conclusion for containment strategies since the reduction of epidemic channels will increase the delay of the infection arrival and will thus allow for a better preparation against the disease ( for example vaccination )
. these results could be helpful in designing early detection and containment strategies in more involved models which go beyond topology and which include additional features such as passenger traffic in airlines or city populations @xcite .
the authors thank a .- j .
valleron for his support during this work .
p.c . acknowledges financial support from a.c.i .
systmes complexes en sciences humaines et sociales , and f.a . from frm
( fondation pour la recherche mdicale ) .
we also thank m. loecher and j. kadtke for sharing with us their manuscript prior to publication . | we study numerically the variability of the outbreak of diseases on complex networks . we use a si model to simulate the disease spreading at short times in homogeneous and in scale - free networks . in both cases , we study the effect of initial conditions on the epidemic dynamics and its variability .
the results display a time regime during which the prevalence exhibits a large sensitivity to noise .
we also investigate the dependence of the infection time of a node on its degree and its distance to the seed .
in particular , we show that the infection time of hubs have non - negligible fluctuations which limit their reliability as early - detection stations .
finally , we discuss the effect of the multiplicity of paths between two nodes on the infection time . in particular , we demonstrate that the existence of even long paths reduces the average infection time .
these different results could be of use for the design of time - dependent containment strategies . |
the detection of supernova signatures in the late afterglow of grbs ( galama et al . 1998 ; stanek et al .
2003 ; hjorth et al . 2003 ) has established , beyond reasonable doubt , that at least a fraction of classical long duration grbs are associated to the final evolutionary stages of massive stars .
such discovery has triggered speculation on the mechanisms that allow a relativistic jet to form , propagate in the core of a massive star , reach the surface of the star , and expand in space without being choked by baryon entrainment .
these questions have been addressed mainly numerically , with increasingly sophisticated codes and powerful computers .
pioneering work ( macfadyen & woosley 1999 ) showed that a light jet can propagate unpolluted through a massive star core by pushing the cold dense stellar material aside . the head of the jet propagates sub - relativistically inside the star , allowing for the shocked material to flow in a cocoon that surrounds and hydrodynamically collimates the outflow .
these results were subsequently confirmed by more refined special - relativistic computations ( aloy et al . 2000 ; zhang et al .
lazzati & begelman ( 2005 ) showed with an analytical model that the interaction between the jet and the star must create time evolution in the jet properties as it reach the stellar surface , even if the energy release at the core of the star is constant . in this paper
, we show the results of high resolution two - dimensional simulations , performed with the special relativistic adaptive mesh refinement ( amr ) code flash .
simulations were targeted to investigate the role of the jet - star interaction in shaping the structural and temporal properties of the jet outside the star surface .
we show that the jet propagation takes place in four phases , three of which are radiative and can be observed in the grb light curve .
these phases can be separated by moments of quiescence , possibly explaining long dead - times observed between precursors and prompt emission in grb light curves ( lazzati 2005 ) .
in contrast to previous lower resolution simulations , we find that the highly relativistic part of the jet can be well described with a sharp cutoff or with a top - hat toy model configuration . for more details about the code , the simulations , and some of the results reported here ,
see morsony et al .
with the aim of exploring the general properties of a confined light jet and of understanding the features introduced by a specific stellar progenitor we ran two different set of simulation . in one set , the star is described as a polytropic sphere with a power - law density profile @xmath0 . in a second set of simulations the stellar properties
are taken from model 16ti of woosley & heger ( 2006 ) , which is considered a possible grb progenitor .
the jet injection is modeled as a boundary condition on the lower edge of the grid at @xmath1 cm .
the opening angle and lorentz factor of the incoming jet are varied between simulations ( we explore jets with injection opening angles between 5 and @xmath2 ) , but are constant at all times in each run . the terminal lorentz factor , or lorentz factor at infinity , @xmath3 ,
is defined as the lorentz factor that the material would achieve if all its internal energy were converted to kinetic energy .
it is calculated as @xmath4 , where @xmath5 is the local bulk lorentz factor .
@xmath3 for the jet material is set to 400 for all simulations and the luminosity of the central engine is set to @xmath6 erg s@xmath7 .
the injection lorentz factor is either 2 or 5 .
each simulation is run for 50 seconds , giving an energy input of @xmath8 ergs per jet , or a total energy of @xmath9 ergs assuming symmetric jets .
these total energies are comparable to those assumed in previous works ( see zhang et al .
2003 for a discussion ) .
a visual result of simulations are shown in fig . [ fig : phases ] .
the figure shows stills of the simulation for a @xmath2 jet propagating in a polytropic star with @xmath10 .
the figure ( see also http://rocinante.colorado.edu/@xmath11morsony/grb/index.html for animations ) shows how the light relativistic jet propagates through the star and is affected by the propagation .
the jet breaks on the stellar surface at approximately 10 seconds after the injection , releases a broad mildly relativistic component and eventually accelerate almost to its potential lorentz factor of @xmath12 .
we find that the jet propagation takes place in four phases , three of which are of great interest since are potentially radiative and can therefore be identified in the grb light curve .
* the confined phase .
* initially the jet is confined inside the star .
it immediately becomes supersonic and develops a shock structure at its head .
the jet head propagates subrelativistically in the star , powered by the relativistic material behind that goes through the reverse shock .
since the head is sub - relativistic , and well in causal contact , the jet and stellar material do not accumulate but rather move backward to the sides of the jet , forming a high pressure cocoon .
this cocoon pressure hydrodynamically collimates the jet ( begelman & ciotti 1989 ) .
the jet propagation results therefore from the feedback between the head working surface , that increases the pressure of the cocoon , and the recollimation , that decreases the head surface and weakens the cocoon ( lazzati & begelman 2005 ) .
analytical considerations for a simple jet model ( morsony et al .
2006 ) suggest that the speed of the jet head is fairly independent on the stellar properties ( see also fig.[fig : betah ] ) . *
the breakout .
* as the head of the jet reaches the surface of the star , it opens a channel through which the hot material stored in the cocoon can escape .
the cocoon material is dominated by the radiation energy and can in principle accelerate to relativistic speed .
it s asymptotic lorentz factor is , however , lower than that of the jet due to mixing with the stellar baryons . unlike the jet material
, the cocoon material is released through a nozzle on the surface of the star . as a consequence the cocoon material forms a quasi - spherical fireball .
taking advantage of the constancy of the propagation speed of the jet , we can evaluate the cocoon energy to be : @xmath13 the energy stored in the cocoon can give rise to a precursor ( ramirez - ruiz et al .
a measurement of the precursor energy can therefore give us valuable informations on the stellar dimension , assuming a radiative efficiency .
* the shocked jet . * as mentioned above , as the jet propagates in the confined phase , it is recollimated with tangential shocks . the first phase of jet propagation outside the star is therefore characterized by a jet whose structure is severely affected by the presence of multiple shocks .
figure [ fig : shj ] shows a cross section of the jet properties during the shocked phase .
the cross section is taken just outside the stellar surface , and the star model is again the polytropic one .
the jet in the figure is not clearly defined , and the vertical dashed line is somewhat arbitrary .
there are shocks inside the jet boundary and the properties of the material are highly variable .
the opening angle during this phase is somewhat constant , with small variations overlaid ( see morsony et al . 2006 ) .
* the un - shocked jet . *
the simulations show that several tens of seconds after the jet breaks out of the star surface , its configuration relaxes to a much more stable one . as can be seen in fig .
[ fig : usj ] , the core of the jet is made by a free streaming outflow , accelerating according to the adiabatic expansion @xmath14 .
it is unclear whether the feature in the very core of the jet is due to the dimensionality of the computations or is physics .
this free streaming part of the jet is clearly bounded by a strong shock .
the pressure of the material outside the shock is balanced by the jet ram pressure .
the shocked area constitutes the boundary layer between the unperturbed core and the non - relativistic material outside . inside the boundary layer ,
the density and lorentz factor of the material join smoothly with the outside .
the width of the layer is loosely associated to its highest lorentz factor : @xmath15 . the inner free streaming part of the jet is very interesting since its propagation is unperturbed from the inner engine out to the radiative phase . as a consequence , it would be very interesting to isolate this component in observations and study its variability in order to gain information on the processes that take place at the very core of the star , where the jet is released and energized . * the angular structure *
the three phases of jet propagation outside the star are also characterized by a different angular structure , or energy distribution @xmath16 . during the prompt phase
any given observer receives radiation produced only from the material moving along its line of sight . as a consequence , different observers will see different contribution from the three phases .
figure [ fig : dedo ] shows the energy distribution in the three phases as a function of the off - axis angle .
let us first consider an on - axis observer .
he will see a very dim precursor , if any .
this is due to the lack of energy emission in the isotropic precursor along the jet axis .
it is not entirely clear whether this is an artifact of the 2d simulations or it is intimately linked to the process of vortex shedding that shapes the cocoon ( morsony et al .
2006 ) . immediately after the precursor ,
if any , the on - axis observer will enter a very bright phase , during which the flow is dominated by the shocked jet .
this is very well collimated material , concentrated at small angles .
it is also characterized by strong variability ( see fig . [
fig : shj ] ) imprinted by the recollimation shocks .
it should therefore be a favorable phase for the production of @xmath5-rays through internal shocks and/or other dissipation mechanisms .
finally , the on - axis observer will see emission from the un - shocked jet .
this emission is weak since most of the energy in the un - shocked phase is stored in the boundary layer .
even though the opening angle of the jet increases in time ( lazzati & begelman 2005 ) , the luminosity of the unshocked part of the jet is constant because the freely expanding material is not in causal contact with the jet boundary . during this phase no variability
is imprinted in the flow by the interaction with the star . on the other hand , this is the phase in which the intrinsic variability of the inner engine , if any , is preserved .
observers located at intermediate angles , with @xmath17 in fig . [ fig : dedo ] , will experience a completely different emission pattern .
they will see a relatively bright precursor , followed by a dark phase . during the shocked jet phase the jet is too collimated and no emission
is observed at those angles .
eventually , as the jet enters the un - shocked phase , its opening angle spreads and the observer can see a bright grb . the dead time will be in this case longer than any timescale associated with the jet or the central engine , and this effect
could explain the early precursors observed with batse ( lazzati 2005 ) and swift .
finally , observers located outside the maximum opening angle of the jet ( about @xmath18 in the case of fig . [ fig : dedo ] ) , will only see the precursor .
whether this weak emission could explain the observation of faint events like grb 980425 and grb 031203 is however a matter of open debate ( e.g. , waxman , 2004 ) there are two caveats to this discussion .
first , the relative normalization of the curve in fig . [ fig : dedo ] can be changed significantly .
especially , the un - shocked jet phase can last an arbitrarily long time , and therefore can dominate most of the emission .
it is however unlikely that the inner engine will be fueled for much longer than several hundred seconds and therefore the figure should give a fairly correct representation of the energy budgets .
also , the normalization of the first two phases is somewhat related to the stellar radius and can therefore vary by approximately a factor 10 .
more importantly , the energy distributions plotted in fig .
[ fig : dedo ] are computed just outside the stellar radius , at several@xmath19 cm .
the material has to travel out to the transparency radius , located at approximately @xmath20 cm , before the grb photons are produced .
to what extent the energy distribution will be modified by the expansion over two orders of magnitude in distance is unclear . to estimate whether there will be sizable modifications we show in fig .
[ fig : evol ] the changes in the energy distribution over a factor 2 in radius .
the left panel shows the case of the shocked jet , while the right panel shows the un - shocked jet . in both panels we select a thin shell in the outflow and we follow it through the expansion over a factor of two in radius , from @xmath21 cm to @xmath22 cm .
the shocked jet is only marginally modified during the expansion , the main effect being the filling of an energy minimum in the center .
the un - shocked jet , instead , changes significantly . in particular , the strong shock that separates the freely expanding flow from the boundary layer moves inward ( in angular coordinates ) , reducing the size of the freely flowing outflow .
it seems unlikely that this shrinking can proceed to completely choke the free jet , but it is clear that energy distributions in the un - shocked jet phase should be taken with some caution .
we presented high resolution numerical simulations of the hydrodynamic propagation of a light relativistic jet through a massive star , performed with the amr code flash .
we inject the jets as boundary conditions at the base of the grid and follow their propagation through the star , leaving the engine active for a timescale much longer than the time it takes the jet to break out on the stellar surface .
we identify four phases of the jet propagation , three of which are radiative and can possibly contribute to the grb light curve .
we describe the temporal and angular properties of the outflow in these three phases , showing how the outflow is initially very wide ( the precursor phase , caused by the release of the cocoon on the stellar surface ) . following this initial release , the jet emerges from the star hydrodynamically re - collimated by tangential shocks , producing a very bright highly collimated and variable jet . eventually , as the stellar influence disappears with the cocoon release , the jet set into a stable condition whereby the core is freely expanding and is surrounded by a boundary layer that connects the highly relativistic core to the static dense stellar material .
the jet opening angle grows steadily in this phase .
such a scenario allows us to explain the presence of long dead - time between the precursor emission and the subsequent main grb discovered by lazzati ( 2005 ) in approximately @xmath23 per cent of bright batse light curves .
the software used in this work was in part developed by the doe - supported asc / alliance center for astrophysical thermonuclear flashes at the university of chicago .
this work was supported by nsf grant ast-0307502 , nasa astrophysical theory grant nng06gi06 g , and swift guest investigator program nnx06ab69 g .
this work was partially supported by the national center for supercomputing applications under grant number ast050038 and utilized the ncsa xeon cluster . | gamma rays : bursts hydrodynamics shock waves supernovae : general the association of at least some long gamma - ray bursts with type ic supernova explosions has been established beyond reasonable doubt . theoretically , the challenge is to explain the presence of a light hyper - relativistic flow propagating through a massive stellar core without losing those properties .
we discuss the role of the jet - star interaction in shaping the properties of the outflow emerging on the surface of the star .
we show that the nature of the inner engine is hidden from the observer for most of the evolution , well beyond the time of the jet breakout on the stellar surface .
the discussion is based on analytical considerations as well as high resolution numerical simulations .
finally , the observational consequences of the scenario are addressed in light of the present capabilities .
[ firstpage ] |
numerical integration of the einstein equations is the only way to investigate highly dynamical and nonlinear gravitational space - time .
the detection of gravitational wave requires templates of waveform , among them mergers of compact objects are the most plausible astrophysical sources .
numerical relativity has been developed with this purpose over decades . for neutron star ( ns ) binaries ,
a number of scientific numerical simulations have been done so far , and we are now at the level of discussing the actual physics of the phenomena , including the effects of the equations of state , hydrodynamics , and general relativity by evolving various initial data @xcite .
mergers of black holes ( bhs ) are also available after the breakthrough by pretorius @xcite in 2004 .
pretorius s implementation had many novel features in his code ; among them he discretizes the four - dimensional einstein equations directly , which is not a conventional approach so far . however , after the announcements of successful binary bh mergers by campanelli et al .
@xcite and baker et al .
@xcite based on the standard 3 + 1 decomposition of the einstein equations , many groups began producing interesting results @xcite .
the merger of ns - bh binary simulations has also been reported recently , e.g. @xcite .
almost all the groups which apply the above conventional approach use the so - called bssn variables together with `` @xmath0''-type slicing conditions for the lapse function and `` @xmath1-driver '' type slicing conditions for the shift function .
bssn stands for baumgarte - shapiro @xcite and shibata - nakamura @xcite , the modified arnowitt - deser - misner formulation initially proposed by nakamura @xcite .
( the details are described in [ subsec : bssn ] . )
there have already been several efforts to explain why the combination of this recipe works from the point of view of the well - posedness of the partial differential equations ( e.g. @xcite ) .
however , the question remains whether there exists an alternative evolution system that enables more long - term stable and accurate simulations .
the search for a better set of equations for numerical integrations is called the formulation problem for numerical relativity , of which earlier stages are reviewed by one of the authors @xcite . in this article , we report our numerical tests of modified versions of the bssn system , the _ adjusted bssn systems _ , proposed by yoneda and shinkai @xcite .
the idea of their modifications is to add constraints to the evolution equations like lagrange multipliers and to construct a robust evolution system which evolves to the constraint surface as the attractor .
their proposals are based on the eigenvalue analysis of the constraint propagation equations ( the evolution equations of the constraints ) on the perturbed metric . for the adm formulation ,
they explain why the standard adm does not work for long - term simulations by showing the existence of the constraint violating mode in perturbed schwarzschild space - time @xcite . for the bssn formulation
, they analyzed the eigenvalues of the constraint propagation equations only on flat space - time @xcite , but one of their proposed adjustments was immediately tested by yo et al .
@xcite for the numerical evolution of kerr - schild space - time and confirmed to work as expected .
( the details are described in [ subsec : adjbssn ] . )
our numerical examples are taken from the proposed problems for testing the formulations of the mexico numerical relativity workshop 2001 participants @xcite , which are sometimes called the apples - with - apples test . to concentrate the comparisons on the formulation problem ,
the templated problems are settled so as not to require technical complications ; e.g. , periodic boundary conditions are used and the slicing conditions do not require solving elliptical equations .
several groups already reported their code tests using these apples tests ( e.g. @xcite ) , and we are also able to compare our results with theirs .
this article is organized as follows .
we describe the bssn equations and the _ adjusted _ bssn equations in sec .
[ subsec : bssn ] and [ subsec : adjbssn ] .
we give our three numerical test problems in sec .
[ sec : setup ] .
comments on our coding stuff are in sec .
[ sec : code ] .
[ sec : num ] is devoted to showing numerical results for each testbeds , and we summarize the results in sec . [ sec : summary ] .
we start by presenting the standard bssn formulation , where we follow the notations of @xcite , which are widely used among numerical relativists .
the idea of the bssn formulation is to introduce auxiliary variables to those of the arnowitt - deser - misner ( adm ) formulation for obtaining longer stable numerical simulations .
the basic variables of the bssn formulation are @xmath2 , which are defined by @xmath3,\label{eq : ext - trldess}\\ \tilde{\gamma}^i & = & \tilde{\gamma}^{jk}{\tilde{\gamma}^i}_{jk } , \label{eq : conf - con}\end{aligned}\ ] ] where @xmath4 are the intrinsic and extrinsic adm 3-metric .
the conformal factor @xmath5 is introduced so as to set @xmath6 $ ] as unity , @xmath7 is supposed to be traceless , and @xmath8 is treated independently in evolution equations .
therefore these three requirements turn into the new constraints [ below ( [ eq : cons - calg])-([eq : cons - cals ] ) ] .
the set of the bssn evolution equations are @xmath9^{\text{tf } } + \alpha k \tilde{a}_{ij } - 2 \alpha \tilde{a}_{ik } { \tilde{a}^k}_{~j } \nonumber \\ & & + \partial_i \beta^k \tilde{a}_{kj } + \partial_j \beta^k \tilde{a}_{ki } - \frac{2}{3 } \partial_k \beta^k \tilde{a}_{ij } + \beta^k \partial_k \tilde{a}_{ij},\label{eq : ev - ext - trless}\\ \partial_t \tilde{\gamma}^i & = & - 2 \partial_j \alpha \tilde{a}^{ij } + 2 \alpha \left [ { \tilde{\gamma}^i}_{jk } \tilde{a}^{jk } - \frac{2}{3}\tilde{\gamma}^{ij}\partial_j k + 6 \tilde{a}^{ij } \partial_j \phi \right ] \nonumber \\ & & + \tilde{\gamma}^{jk}\partial_j\partial_k\beta^i + \frac{1}{3}\tilde{\gamma}^{ij}\partial_j\partial_k \beta^k + \beta^j \partial_j \tilde{\gamma}^i - \tilde{\gamma}^j \partial_j \beta^i + \frac{2}{3}\tilde{\gamma}^i \partial_j \beta^j \label{eq : ev - conf - con},\end{aligned}\ ] ] where @xmath10 is the covariant derivative with respect to the 3-metric @xmath11 and tf means trace - free operation , i.e. , @xmath12 .
the ricci tensor is computed with the conformal connection @xmath8 as @xmath13 where @xmath14 is a covariant derivative associated with @xmath15 .
similarly to the adm formulation , this system has constraint equations .
the two kinematic " constraints , the hamiltonian and momentum constraint equations , are expressed in terms of the bssn basic variables and are written as @xmath16 additionally , the bssn formulation requires three algebraic " constraint relations ; @xmath17 where ( [ eq : cons - calg ] ) and ( [ eq : cons - cala ] ) are from the definitions of ( [ eq : conf - con ] ) and ( [ eq : ext - trldess ] ) , respectively . equation ( [ eq : cons - cals ] ) is from the requirement on @xmath18 . these five constraints are , theoretically , supposed to be zero at all times ; therefore they can be used to modify the dynamical equations .
for example , alcubierre et al .
@xcite announced that the replacement of the terms in ( [ eq : ev - conf - con ] ) using the momentum constraint drastically changes the stability feature .
actually , such replacements of terms using constraints are applied ( with / without intentions ) in many terms in ( [ eq : ev - conf])-([eq : ev - conf - con ] ) , which are expressed as eqs .
( 2.27)-(2.31 ) in @xcite .
alcubierre et al .
@xcite also pointed out that the re - definition of @xmath7 by @xmath19 during the time evolutions improves the numerical stability .
this technique again can be understood as the trace - out of the @xmath20-constraint ( [ eq : cons - cala ] ) from the evolution equations . in our numerical code
, we do not apply this technique because we recognize the trace - free property as the new constraint @xmath20 in the bssn system , and our purpose is to construct a system preventing the violation of constraints .
recently , several groups applied artificial dissipation ( e.g. @xcite ) to obtain stable evolutions ( see , e.g. @xcite ) .
we , however , do not introduce such dissipations in our code , since we try to clarify the difference of stability from the viewpoint of _ formulations _ of the einstein equations . to understand the stability property of the bssn system , yoneda and shinkai @xcite studied the structure of the evolution equations , ( [ eq : ev - conf])-([eq : ev - conf - con ] ) , in detail , especially how the modifications using the constraints , ( [ eq : cal h])-([eq : cons - cals ] ) , affect to the stability .
they investigated the signature of the eigenvalues of the constraint propagation equations ( dynamical equations of constraints ) , and explained that the standard bssn dynamical equations are balanced from the viewpoints of constrained propagations , including a clarification of the effect of the replacement using the momentum constraint equation .
moreover , they predicted that several combinations of modifications have a constraint - damping nature , and named them _ adjusted _ bssn systems .
( their predictions are based on the signature of eigenvalues of the constraint propagations , and the negative signature implies a dynamical system which evolves toward the constraint surface as the attractor . ) among them , in this work , we test the following three adjustments : 1 .
an adjustment of the @xmath21-equation with the momentum constraint : @xmath22 where @xmath23 is predicted ( from the eigenvalue analysis ) to be positive in order to damp the constraint violations .
an adjustment of the @xmath18-equation with @xmath24 constraint : @xmath25 with @xmath26 .
an adjustment of the @xmath27-equation with @xmath24 constraint : @xmath28 with @xmath29 .
these three adjustments are chosen as samples of best candidates " , eq .
( 4.9)-(4.11 ) in @xcite .
best " comes from their conjecture on the eigenvalue analysis of the constraint propagation matrix ; that is , ( a ) all the resultant eigenvalues from above adjustments can be less than or at most equal to zero , which indicates the decay of constraint errors , and ( b ) the resultant constraint propagation matrix is diagonalizable , which guarantees the predictions of above eigenvalue analysis ( see table ii in @xcite ) .
however , since above eigenvalues include zero elements and also above analysis assumes a linearly perturbed metric about the flat space - time , the effects of the adjustments ( [ b1-adj])-([b2b - adj ] ) need to be demonstrated via numerical experiments .
following the proposals of the mexico numerical relativity workshop @xcite , we perform three kinds of tests . in this section
, we explicitly give some details of the models .
the first test is the trivial minkowski space - time , but sliced with the time - dependent 3-metric , which is called the gauge - wave test .
the 4-metric is obtained by coordinate transformation from the minkowski metric as @xmath30 where @xmath31 which describes a sinusoidal gauge wave of amplitude @xmath32 propagating along the @xmath33-axis .
the non - trivial extrinsic curvature is @xmath34 following @xcite , we chose numerical domain and parameters as follows : * gauge - wave parameters : @xmath35 and @xmath36 * simulation domain : @xmath37 $ ] , @xmath38 * grid : @xmath39 with @xmath40 , where @xmath41 with @xmath42 * time step : @xmath43 * boundary conditions : periodic boundary condition in @xmath33 direction and planar symmetry in @xmath44 and @xmath45 directions * gauge conditions : @xmath46 the 1d simulation is carried out for a @xmath47 crossing - time or until the code crashes , where one crossing - time is defined by the length of the simulation domain . the second test is to check the ability of handling a travelling gravitational wave .
the initial 3-metric and extrinsic curvature @xmath48 are given by a diagonal perturbation with component @xmath49 where @xmath50 for a linearized plane wave traveling in the @xmath33-direction .
here @xmath51 is the linear size of the propagation domain and @xmath32 is the amplitude of the wave
. the non - trivial components of extrinsic curvature are then @xmath52 following @xcite , we chose the following parameters : * linear wave parameters : @xmath35 and @xmath53 * simulation domain : @xmath37 $ ] , @xmath54 , @xmath55 * grid : @xmath39 with @xmath40 , where @xmath41 with @xmath42 * time step : @xmath43 * boundary conditions : periodic boundary condition in @xmath33 direction and planar symmetry in @xmath44 and @xmath45 directions * gauge conditions : @xmath56 and @xmath57 the 1d simulation is carried out for a @xmath47 crossing - time or until the code crashes . the third test is to check the formulation in a strong field context using the polarized gowdy metric , which is written as @xmath58 here time coordinate @xmath59 is chosen such that time increases as the universe expands .
simple forms of the solutions , @xmath60 and @xmath61 , are given by @xmath62 \nonumber\\ & & - \frac{1}{2}[(2\pi)^2 [ j_0 ^ 2(2\pi ) + j_1 ^ 2(2\pi ) ] \nonumber \\ & & - 2\pi j_0(2\pi ) j_1(2\pi)],\label{eq : gowdy - funcl}\end{aligned}\ ] ] where @xmath63 is the bessel function .
the non - trivial extrinsic curvatures are then @xmath64 according to @xcite , the new time coordinate @xmath65 , which satisfies harmonic condition , is obtained by coordinate transformation as @xmath66 where @xmath67 and @xmath68 are arbitrary constants . using this freedom ,
we can set the lapse function in the new time coordinate to be unity at the initial time .
concretely , we set @xmath69 where @xmath70 is the initial time . following @xcite
, we perform our evolution in the collapsing ( i.e. backward in time ) direction .
parameters are chosen as follows : * simulation domain : @xmath45 @xmath71 @xmath72 $ ] , @xmath73 * grid : @xmath74 with @xmath40 , where @xmath75 with @xmath42 * time step : @xmath76 * boundary conditions : periodic boundary condition in @xmath45-direction and plane symmetry in @xmath33- and @xmath44-directions * gauge conditions : the harmonic slicing ( [ eq : gwave - gauge ] ) and @xmath57 the 1d simulation is carried out for a @xmath47 crossing - time or until the code crashes .
we have developed a new numerical code based on the adjusted bssn systems .
the variables are @xmath77 , and the evolution equations are ( [ eq : ev - conf])-([eq : ev - conf - con ] ) with / without adjustment ( [ b1-adj ] ) , ( [ b2a - adj ] ) , and/or ( [ b2b - adj ] ) .
the time - integration is under the free - evolution scheme , and we monitor five constraints , ( [ eq : cal h])-([eq : cons - cals ] ) , to check the accuracy and stability of the evolutions . our time - integration scheme is the three - step iterative crank - nicholson method with centered finite difference in space @xcite .
this scheme should have second - order convergence both in space and time , and we checked its convergence in all the testbeds .
as we have already mentioned in the end of ii a , we do not apply the trace - out technique of @xmath7 , ( [ redef_traceout ] ) in our code .
we also remark on our treatment of the conformal connection variable @xmath8 . as was pointed out in @xcite , it is better not to use @xmath8 in all the evolution equations .
we surmise this is because the amplification of the error due to the discrepancy of the definition ( [ eq : conf - con ] ) , i.e. , the accumulations of the violations of @xmath78-constraint ( [ eq : cons - calg ] ) .
therefore , we used the evolved @xmath8 only for the terms in ( [ eq : ev - conf - con ] ) and ( [ eq : conm - ricci ] ) , and not for other terms , so as not to implicitly apply the @xmath78-constraint in time evolutions .
it is crucial that our code can produce accurate results , because the adjustment methods are based on the assumption that the code represents the bssn system ( [ eq : ev - conf])-([eq : ev - conf - con ] ) accurately .
we verified our code by comparing our numerical data with analytic solutions from the gauge - wave and gowdy - wave testbeds in sec .
[ sec : setup ] .
the actual procedures are as follows : 1 . evolve only one component , e.g. @xmath79 , numerically , and express all the other components with those of the analytic solution .
in this situation , the origin of the error is from the finite differencing of the analytic solution in the spatial direction and from that of the numerically evolved component ( @xmath79 ) both in spatial and time directions .
we checked the code by monitoring the difference between the numerically evolved component ( @xmath79 ) and its analytic expression .
this procedure was applied to all the components one by one .
2 . evolve only several components , e.g. , @xmath79 and @xmath80 , numerically , and express the other components by the analytic solution .
the error can be checked by a procedure similar to the one above .
3 . evolve all the components numerically , and check the error with the analytic solution .
we repeated these procedure three times by switching the propagation directions ( @xmath33 , @xmath44 , and @xmath45-directions ) of gauge - wave and gowdy - wave solutions .
we also applied these procedures in a 2d test @xcite , and checked the off - diagonal component .
it should be emphasized that the adjustment effect has two meanings , improvement of stability and of accuracy . even if a simulation is stable
, it does not imply that the result is accurate .
we judge the stability of the evolution by monitoring the l2 norm of each constraint , @xmath81 where @xmath82 is the total number of grid points , while we judge the accuracy by the difference of the metric components @xmath83 from the exact solution @xmath84 , @xmath85 adjusted systems , ( [ b1-adj])-([b2b - adj ] ) , require to specify the parameter @xmath86 . from the analytical prediction in @xcite we know the signature of @xmath86 , but not for its magnitude . by definition of the adjustment terms in eq .
( [ b1-adj])-([b2b - adj ] ) , applying small @xmath86 should produce the close results with those of the plain system . on the contrary
, the large @xmath86 system will violate the courant - friedrich - lewy condition @xcite .
hence , there exists a suitable region in the adjustment parameters . at this moment
, we have to chose @xmath86 experimentally , by observing the life - time of simulations .
the value of @xmath86 , used in our demonstrations , is one of the choices of which the adjustment works effectively in all the resolutions .
as the first test , we show the plain bssn evolution ( that is , no adjustments ) in fig .
[ 1d - gauge - plain ] for the gauge - wave test . in fig .
[ 1d - gauge - plain ] , the l2 norms of the hamiltonian and momentum constraints ( [ eq : errorc ] ) are plotted as a function of the crossing - time .
the second - order convergent nature is lost at an early time , the 20 crossing - time , and the simulation crashes at about the 100 crossing - time .
the poor performance of the plain bssn system for the gauge wave test has been reported in @xcite ( see their fig .
this drawback , on the other hand , can be overcome if one uses the fourth - order finite differencing scheme , an example of which can be seen in @xcite ( see their fig . 2 ) .
[ cols="^,^ " , ] in order to check the _ accuracy _ of the simulations , we prepare fig . [ 1d - gowdy - gzz ] to show the error of the @xmath87 component of the metric . unlike the gauge - wave or the linear wave test , in this gowdy - wave test the amplitude of the metric functions
damps with time .
therefore we use the criterion that the error normalized by @xmath87 be under @xmath88 for an _ accurate evolution_. this criterion is the same as the one used in zlochower et al .
@xcite .
figure [ 1d - gowdy - gzz ] shows the normalized error in @xmath87 versus time for the plain bssn , adjusted bssn with @xmath89-equation , and adjusted bssn with @xmath18-equation systems .
we find that these three systems produce accurate results up to @xmath90 , @xmath91 , and @xmath92 , respectively .
this proves that the adjustments work effectively , i.e , they make possible a stable and accurate simulation , especially the @xmath20-adjusted bssn system .
in this article , we presented our numerical comparisons of the bssn formulation and its adjusted versions using constraints .
we performed three testbeds : gauge - wave , linear wave , and collapsing polarized gowdy - wave tests with their evolutions by three kinds of adjustments , which were previously proposed by yoneda and shinkai @xcite based on their constraint propagation analysis .
the idea of the adjusted systems is to construct a system robust against constraint violations by modifying the evolution equations using the constraint equations .
we can summarize our tests as follows : * when the plain ( original ) bssn evolutions already show satisfactory good evolutions ( e.g. , the linear wave test ) , the constraint violations ( i.e. , adjusted terms ) are also small or ignorable . therefore the adjusted bssn equations become quite similar to the plain bssn equations , and their results coincide with the plain bssn results . * among the adjustments we tried , we observed that the adjusted bssn system with the @xmath21-eq .
( [ b1-adj ] ) is the most robust for all the testbeds examined in this study .
it gives us an accurate and stable evolution compared to the plain bssn system .
quantitatively , the life - time of the simulation becomes 10 times longer for the gauge - wave testbed and 5 times longer for the gowdy - wave testbed than the life - time of the plain bssn system .
however , it should be noted that for the gauge - wave testbed , the convergence feature is lost at a comparatively early time , the 200 crossing - time in the hamiltonian constraint and the 50 crossing - time in the momentum constraint .
recently , it has been claimed that the set up of the gauge wave problem in apples - with - apples has a problematic point @xcite , which arises from the harmonic gauge condition . in @xcite , it is argued that this gauge has a residual freedom in the form @xmath93 , where @xmath61 is an arbitrary and @xmath94 is a function in eq .
( [ eq : gwave - metric ] ) . of course
, our set up corresponds to the @xmath95 case , but numerical error easily excites modes that result in either exponentially increasing or decaying metric amplitude .
actually , we find the amplitude of the error decays with time in this testbed .
so , we conclude that due to the adjustment , the growing rate of the gauge mode is suppressed and the life - time of the simulation is extended as a result .
* the other type of adjustments ( [ b2a - adj ] and [ b2b - adj ] ) show their apparent effects while depending on a problem .
the @xmath27-adjustment for the gauge - wave testbed makes the life - time longer slightly .
the @xmath18-adjustment for the gowdy - wave testbed makes possible a simulation twice as long as the plain bssn system .
we can understand the effect of the adjustments in terms of adding dissipative terms . by virtue of the definition of the constraints
, we can recognize that the adjusted equation corresponds to the diffusion equation ( see , for example , eq .
( [ b1-adj ] ) ) and the signature of @xmath86 determines whether the diffusion is positive or negative . in the adjusted @xmath21-eq .
system , ( [ b1-adj ] ) , the adjustment term corresponds to the positive diffusive term , due to the definition of @xmath96 and the positiveness of @xmath97 ( see eq .
( [ eq : cal mi ] ) and ( [ b1-adj ] ) ) .
this fact might explain why the adjusted @xmath21-eq . system works effectively for all the testbeds .
in contrast , why are not all the adjustments effective in all testbeds ? as we mentioned in sec .
iib , the eigenvalue analysis was made on the linearly perturbed violation of constraints on the minkowski space - time .
since the constraint violation grows non - linearly as seen in the appendix of @xcite , the candidates may not be the best in their later evolution phase .
we remark upon two more interesting aspects arising from our study .
the first is the mechanism of the constraint violations . as was shown in the appendix of @xcite , each constraint propagation ( behavior of their growth or decrease ) depends on the other constraint terms together with itself .
that is , we can guess @xmath20 and @xmath98 constraints ( [ eq : cons - cala ] and [ eq : cons - cals ] ) in this article , propagate independently of the other constraints , while the violation of the @xmath24-constraint , ( [ eq : cons - calg ] ) is triggered by the violation of the momentum constraint , and both the hamiltonian and the momentum constraints are affected by all the other constraints .
such an order of the constraint violation can be guessed in fig .
[ fig : c - vio - gauge ] ( earlier time ) , where we plot the rate of constraint violation normalized with its initial value , @xmath99 , as a function of time , for the gauge - wave testbeds with the plain bssn evolution .
( note that the constraints at the initial time , @xmath100 , are not zero due to the numerical truncation error . )
the parameters are the same as those shown in sec .
[ sec : gaugewave ] , and the lowest resolution run is used . from this investigation
, we might conclude that to monitor the momentum constraint violation is the key to checking the stability of the evolution .
the second remark is on the lagrange multipliers , @xmath86 , used in the adjusted systems . as discussed in sec .
[ subsec : adjbssn ] , the signatures of the @xmath86s are determined _ a priori _ , and we confirmed that all the predicted signatures of @xmath86s in @xcite are right to produce positive effects for controlling constraint violations .
however , we have to search for a suitable magnitude of @xmath86s for each problem .
therefore we are now trying to develop a more sophisticated version , such as an auto - controlling @xmath86 system , which will be reported upon in the future elsewhere .
although the testbeds used in this work are simple , it might be rather surprising to observe the expected effects of adjustments with such a slight change in the evolution equations .
we therefore think that our demonstrations imply a potential to construct a robust system against constraint violations even in highly dynamical situations , such as black hole formation via gravitational collapse , or binary merger problems .
we are now preparing our strong - field tests of the adjusted bssn systems using large amplitude gravitational waves , black hole space - time , or non - vacuum space - time , which will be reported on in the near future .
k.k . thanks k. i. maeda and s. yamada for continuing encouragement . k.k .
also thanks y. sekiguchi and m. shibata for their useful comments on making numerical code .
this work was supported in part by the japan society for promotion of science ( jsps ) research fellowships and by a grant - in - aid for scientific programs .
was partially supported by the special research fund ( project no .
4244 ) of the osaka institute of technology .
a part of the numerical calculations was carried out on the altix3700 bx2 at yitp at kyoto university .
99 m. shibata , k. taniguchi and k. uryu , phys .
d * 68 * , 084020 ( 2003 ) , phys .
d * 71 * , 084021 ( 2005 ) m. shibata and k. taniguchi , phys .
d * 73 * , 064027 ( 2006 ) p. marronetti and s. l. shapiro , phys .
d * 68 * , 104024 ( 2003 ) p. marronetti , m. d. duez , s. l. shapiro and t. w. baumgarte , phys .
lett . * 92 * , 141101 ( 2004 ) j. a. faber , t. w. baumgarte , s. l. shapiro and k. taniguchi , astrophys . j. * 641 * , l93 ( 2006 ) f. pretorius , phys . rev . lett . * 95 * , 121101 ( 2005 ) m. campanelli , c. o. lousto , p. marronetti and y. zlochower , phys . rev .
lett . * 96 * , 111101 ( 2006 ) j. g. baker , j. centrella , d - i .
choi , m. koppitz , j. van meter , phys .
* 96 * , 111102 ( 2006 ) .
p. diener , f. herrmann , d. pollney , e. schnetter , e. seidel , r. takahashi , j. thornburg , j. ventrella , phys .
* 96 * , 121101 ( 2006 ) f. herrmann , i. hinder , d. shoemaker , p. laguna , gr - qc/0601026v2 z. b. etienne , j. a. faber , y. t. liu , s. l. shapiro and t. w. baumgarte , arxiv:0707.2083 [ gr - qc ] .
w. tichy and p. marronetti , phys . rev .
d * 76 * , 061502 ( 2007 ) m. campanelli , c. o. lousto , y. zlochower and d. merritt , phys .
* 98 * , 231102 ( 2007 ) j. a. gonzalez , m. d. hannam , u. sperhake , b. brugmann and s. husa , phys .
lett . * 98 * , 231101 ( 2007 ) m. campanelli , c. o. lousto , y. zlochower and d. merritt , astrophys .
j. * 659 * , l5 ( 2007 ) j. thornburg , p. diener , d. pollney , l. rezzolla , e. schnetter , e. seidel and r. takahashi , class . quant .
* 24 * , 3911 ( 2007 ) j. a. gonzalez , u. sperhake , b. bruegmann , m. hannam and s. husa , phys .
lett . * 98 * , 091101 ( 2007 ) j. g. baker , j. centrella , d. i. choi , m. koppitz and j. van meter , phys .
d * 73 * , 104002 ( 2006 ) m. shibata and k. uryu , phys .
d * 74 * , 121503 ( 2006 ) , class . quant .
* 24 * , s125 ( 2007 ) t.w .
baumgarte and s.l .
shapiro , phys .
d * 59 * , 024007 ( 1999 ) .
t. nakamura , k. oohara and y. kojima , prog .
. suppl . * 90 * , 1 ( 1987 ) .
t. nakamura and k. oohara , in _ frontiers in numerical relativity _ edited by c.r .
evans , l.s .
finn , and d.w .
hobill ( cambridge univ .
press , cambridge , england , 1989 ) .
n. jansen , b. bruegmann and w. tichy , phys .
d * 74 * , 084022 ( 2006 ) y. zlochower , j. g. baker , m. campanelli and c. o. lousto , phys .
d * 72 * , 024021 ( 2005 ) b. bruegmann , j. a. gonzalez , m. hannam , s. husa , u. sperhake and w. tichy , arxiv : gr - qc/0610128 . m. alcubierre , g. allen b. brgmann , e. seidel and w .- m .
suen phys .
d * 62 * , 124011 ( 2000 ) m. c. babiuc _ et al . _ [ apples with apples collaboration ] , arxiv:0709.3559 [ gr - qc ] .
m. alcubierre , b. brugmann , p. diener , m. koppitz , d. pollney , e. seidel and r. takahashi , phys .
d * 67 * , 084023 ( 2003 ) s. a. teukolsky , phys .
d * 61 * , 087501 ( 2000 ) g. yoneda and h. shinkai , class .
* 18 * , 441 ( 2001 ) .
m. babiuc , b. szilagyi and j. winicour , in analytical and numerical approaches to mathematical relativity eds . by j. frauendiener , d. giulini , and v. perlick , ( springer , heidelberg , 2006 ) .
[ arxiv : gr - qc/0404092 ] . w. press , b. p. flannery , s. teukolosky , and w. t. vetterling , numerical recipes in c ( cambridge university press , cambridge , england , 1986 ) . | we present our numerical comparisons between the bssn formulation widely used in numerical relativity today and its adjusted versions using constraints .
we performed three testbeds : gauge - wave , linear wave , and gowdy - wave tests , proposed by the mexico workshop on the formulation problem of the einstein equations .
we tried three kinds of adjustments , which were previously proposed from the analysis of the constraint propagation equations , and investigated how they improve the accuracy and stability of evolutions .
we observed that the signature of the proposed lagrange multipliers are always right and the adjustments improve the convergence and stability of the simulations .
when the original bssn system already shows satisfactory good evolutions ( e.g. , linear wave test ) , the adjusted versions also coincide with those evolutions ; while in some cases ( e.g. , gauge - wave or gowdy - wave tests ) the simulations using the adjusted systems last 10 times as long as those using the original bssn equations .
our demonstrations imply a potential to construct a robust evolution system against constraint violations even in highly dynamical situations . |
the ground state(s ) of a spin system are important since they determine its behavior and properties at low temperatures .
although at low temperatures quantum fluctuations are prominent yet the classical ground state(s ) may contain valuable information .
these classical ground states can be calculated by numerical or analytical methods for a number of special cases but a general theory does not exist despite a few attempts towards general results as , e. g. , @xcite or @xcite . in this paper
we consider a special class of anisotropic systems , namely classical magnetic dipoles located at the vertices of a regular @xmath5polygon interacting via their magnetic fields , in short : dipole rings .
there exists an overwhelming numerical evidence that the ground states of dipole rings are given by the two clockwise or anti - clockwise spin orientations tangent to the circle encompassing the dipole ring , see figure [ figgs ] .
these states , denoted by @xmath6 , are consequently assumed to be the ground states in a couple of publications , e. g. , @xcite @xcite .
nevertheless , a rigorous proof of this fact seems not to exist . the motivation to publish such a proof is not to dispel remaining doubts about the ground state property of @xmath6 but rather to illustrate the problems that may occur even in analytically solvable cases and to provide methods to cope with these problems . in section [ sec : r ] we shortly provide the basic definitions for magnetic dipole rings . for more details the reader
is referred to @xcite and @xcite .
the proof of the ground state property of the two tangential states @xmath6 requires a large number of mostly elementary steps that are , however , intricately intertwined . in order to help the reader to keep track of the structure of the proof it will be in order to sketch the main ideas without going into details .
the hamiltonian of the dipole ring is a bilinear function of the @xmath7 spin components @xmath8 and hence can be represented by a real , symmetric @xmath1 matrix @xmath2 such that the energy @xmath9 has the form of an expectation value " @xmath10 .
it turns out that in our case @xmath2 can be completely diagonalized due to the @xmath11 symmetry of the dipole ring .
in particular , there exists an eigenvector @xmath12 of @xmath2 that can be identified with the conjectured tangential ground state of the dipole ring .
it only " remains to show that the corresponding eigenvalue @xmath13 is the lowest eigenvalue @xmath14 of @xmath2 for all @xmath15 . in this way we have reduced the ground state problem to a matrix problem in close analogy to the luttinger - tisza approach @xcite .
the matrix in question is even diagonalized in closed form and hence the problem should be tractable .
actually , the ground state problem could be solved along these lines of thought for any given @xmath0 provided that it could be treated by hand or by computer - algebraic software .
we have done this in section [ sec : a ] for @xmath3 .
the challenging problem is rather to prove @xmath16 for _ all _ @xmath0 .
it can be further reduced to the problem whether the determinant of a @xmath17 matrix @xmath18 is strictly positive , where @xmath19 is a certain discrete parameter ( number of the finite fourier coefficient or wave number ) . however , the matrix entries are not explicitly given numbers but sums of approximately @xmath0 trigonometric functions and hence the positivity of the determinant is not obvious .
as for many other problems it seems to be a good strategy to look at special cases . as noted above , the special case of small @xmath0 is well understood . what about the special case of @xmath20 ?
this leads to the strategy of evaluating the mentioned sums asymptotically , i. e. , in the leading order w. r. t. @xmath0 .
the next step , see section [ sec : p ] , would then be to replace the asymptotic reasoning by strict inequalities and thus to obtain a proof of the ground state property of @xmath6 that is valid for , say , @xmath21 .
here we encounter the next complication : already the asymptotic reasoning , and the more the formulation of rigorous estimates heavily depends on case distinctions .
to explain this we remark that the above - mentioned determinant can be written as a double sum of the form @xmath22 .
an obvious attempt to control the sign of the determinant is to split the terms @xmath23 into two parts , say , @xmath24 : the first part @xmath25 will be strictly positive whereas the second part @xmath26 may be positive or negative , depending on the parameters .
one is then looking for a lower bound @xmath27 of the sum of the positive parts and an upper bound @xmath28 of the absolute value of the sum of the possibly negative parts and tries to show @xmath29 which is sufficient to complete the proof .
unfortunately , the form of the splitting depends on @xmath30 , more precisely , whether @xmath31 or @xmath32 ( the other domains of @xmath30 being reduced to the former ones by means of symmetry arguments ) .
moreover , the form of the bounds @xmath27 and @xmath28 , restricted to partial sums , depends on @xmath33 and @xmath34 and we are led to a variety of case distinctions , see table [ tab2 ] . especially , the domain of the parameter @xmath34 has to be divided into three parts .
one reason is that , in the case of @xmath35 , the lower bound of the positive terms @xmath27 is of order @xmath36 . without any restriction of
@xmath34 we could only show that @xmath37 which would not be sufficient even for an asymptotic proof . only with the restriction @xmath38
we could achieve the result @xmath39 which entails @xmath29 for @xmath21 . in a similar way we found it necessary to introduce the further sub - division @xmath40 and @xmath41 in order to obtain reasonable bounds . here
the three real parameters @xmath42 and @xmath43 are first considered as variables and only after all bounds have been established , have to be chosen in an optimal way to obtain an @xmath43 as small as possible .
some estimates crucially depend on the assumption @xmath4 ; hence we introduce the variable @xmath44 that is used in order to make some equations more transparent but has nevertheless the constant value @xmath45 .
our final aim is to show @xmath46 but we must not assume this from the outset .
one intermediate attempt gave an @xmath47 which is slightly above the desirable value of @xmath48 .
hence we had to improve the bounds @xmath28 by using the explicit result @xmath49 that can be found , e. g. , in @xcite , thereby partly abolishing the former case distinction . after this move
the choice @xmath50 was happily consistent with @xmath48 and i stopped improving estimates .
however , the latter result was at first only obtained by numerical means . here
we face a subtle problem connected with the question to what extent a mathematical proof is allowed to be based on computer - algebraic means .
it is clear that the present proof practically would not be possible without such means .
but the crucial question is which means can be tolerated without vitiating the standard of a mathematical proof .
i will adopt the position that computer - algebraic means are admissible as long as they could be replaced by paper - and - pencil operations albeit long and arduous ones .
this needs some clarification .
any computer program can , in principle , be simulated by a suitable turing machine and hence by paper - and - pencil operations .
thus the above definition of admissible computer - algebraic means " only makes sense if it is _ not _ understood to be applicable in principle " but rather in practice " , even though this introduces some vagueness into the definition .
for example , the simplifications leading from ( [ a5a ] ) to the results of table [ tabev ] have been performed using the computer - algebra software mathematica and not been checked by hand . but
this appears to be a harmless use of computer software since it is completely clear what the equivalent paper - and - pencil operations would be and that they could be performed in a reasonable time . on the other hand , the claim that @xmath29 for @xmath4 is a statement about infinitely many real numbers and can not be justified by an inspection of a numerically produced graph .
the latter can not be considered as a proof but at most as a strong numerical evidence .
otherwise already the statement @xmath51 for all @xmath15 and all @xmath52 could be proven " by inspection of a graph and the whole proof presented in this paper would be pointless . given this restriction of the use of computer - algebraic means in a mathematical proof , how should we then complete the present proof ?
i will explain the chosen strategy for the situation given in figure [ figib ] .
we replaced the condition @xmath29 by the equivalent one @xmath53 and plotted the two functions of @xmath0 for @xmath54 .
it is evident that there is an intersection at @xmath55 and that @xmath53 for @xmath56 , but this will not suffice for the proof .
what we can prove is that @xmath57 is an increasing ( linear ) function of @xmath0 and that @xmath58 is decreasing with the limits @xmath59 for @xmath20 and @xmath60 for @xmath61 .
( by an _ increasing _ function i always mean a _
strictly monotonically increasing _
one throughout this paper , analogously for _ decreasing _ functions . )
the graphs of both functions hence intersect at a unique point with @xmath62 and @xmath53 for @xmath56 .
for our purposes we need the stronger result that @xmath63 and this will be obtained by means of mathematica .
this use of computer - algebraic software is now legitimate since it only involves the approximate evaluation of two elementary functions for two arguments , say , @xmath64 and @xmath65 .
the rationale for allowing computer - algebraic means here is that the approximate evaluation of elementary functions could be done by hand and would yield the same results and only require more time .
one might object that the possibility of errors in applying the computer - algebraic software or even in the software itself can not be excluded but this is beside the point since even an alleged traditional mathematical proof may contain errors .
the remaining part of the proof follows these guidelines .
we consider systems of @xmath0 classical point - like dipoles .
the normalized dipole moments are described by unit vectors @xmath66 .
each dipole moment performs a precession about the momentary magnetic field vector that results as a sum over all magnetic fields produced by the other dipoles .
the @xmath0 dipoles are fixed at the positions of the vertices of a regular @xmath5 polygon @xmath67 for the sake of simplicity , the length of the vectors @xmath68 is chosen as @xmath69 , but it can be scaled arbitrarily .
the dimensionless energy of the dipole system is @xmath70 see @xcite ( 6.35 ) and @xcite . here
@xmath71 denotes the unit vector pointing from the @xmath34-th dipole to the @xmath72-th one : @xmath73 by definition , the _ ground states _ of the system are spin configurations that minimize the energy ( [ r4 ] ) .
numerical studies suggest that there are exactly two ground states , namely @xmath74 and @xmath75 , see figure [ figgs ] for an illustration . the present paper is devoted to the proof of this fact . of the @xmath76 dipole ring .
[ figgs ] , width=453 ]
obviously , the hamiltonian ( [ r4 ] ) is bilinear in the components @xmath77 of the moment vectors @xmath78 and hence can be written in the form @xmath79 where we have introduced multi - indices @xmath80 that run through a finite set of size @xmath7 .
let @xmath81 be the lowest eigenvalue of the symmetric matrix @xmath2 .
then , by the rayleigh - ritz variation principle , @xmath82 , but the minimal energy @xmath83 need not be equal to @xmath84 in general .
we will prove that there exists a certain eigenvalue @xmath85 of @xmath2 such that the corresponding eigenvector can be identified with the state @xmath12 , see ( [ r5 ] ) .
hence @xmath12 is a ground state if @xmath86 since in this case the lower bound @xmath87 of the energy is assumed by the spin configuration @xmath12 . + to detail the above remarks it is convenient to introduce new cartesian coordinates @xmath88 for the moment vectors @xmath78 that are better adapted to the @xmath89-symmetry of the problem : @xmath90 where @xmath91 . in the following
we will express the energy @xmath92 in terms of the new coordinates ( [ a2 ] ) , where the transformed matrix will again be denoted by @xmath2 without danger of confusion .
to this end we consider the part of the energy that is linear in @xmath93 : @xmath94 for the intermediate steps of the calculation we set @xmath95 and @xmath96 .
after elementary transformations we obtain @xmath97\;,\\ \nonumber { \mathbf s}_0\cdot{\mathbf s}_\mu -3\,{\mathbf s}_0\cdot { \mathbf e}_{0\mu}\;{\mathbf s}_\mu\cdot { \mathbf e}_{0\mu } & = & \frac{1}{2}\left [ \xi_0\xi_\mu\,(3-c)+\xi_0\eta_\mu\,s-\eta_0\xi_\mu\ , s-\eta_0\eta_\mu ( 3 + c)+\zeta_0\zeta_\mu \right]\;. \\&&\label{a3f}\end{aligned}\ ] ] from these equations one can read off the first , the @xmath98-th and the @xmath99-th row ( and the analogous columns ) of the matrix @xmath2 .
the other rows can be obtained by cyclic permutations of @xmath100 .
more precisely , the matrix @xmath2 assumes the form @xmath101 where @xmath102 denote @xmath103 sub - matrices that are so - called circulants , see @xcite .
circulant _ is an @xmath103-matrix that commutes with the cyclic permutation matrix of @xmath100 . as an example we display the sub - matrix @xmath104 for @xmath105 : @xmath106 one notes that @xmath104 has constant secondary diagonals even if these are periodically extended .
the eigenvectors @xmath107 of a circulant form the fourier basis , i. e. , are of the form @xmath108 and the eigenvalues are the fourier transform ( times @xmath109 ) of the circulant s first row , see @xcite .
+ @xmath110 and @xmath111 are symmetric , whereas @xmath112 is anti - symmetric .
the matrices @xmath102 pairwise commute since they have the fourier basis ( [ l3a ] ) as a common system of eigenvectors .
since they are circulants it suffices to give the entries of the first row of the respective matrices .
these values can be read off from ( [ a3c ] ) and ( [ a3f ] ) : @xmath113 we note that , except a vanishing diagonal , @xmath104 and @xmath111 have only positive entries and @xmath114 has only negative ones .
the row sum of @xmath112 vanishes since @xmath112 is an anti - symmetric circulant .
moreover , the eigenvalues of @xmath112 are purely imaginary since it is also anti - hermitian .
( from this it follows again that the row sum , which is a real eigenvalue of @xmath112 , must vanish . )
let @xmath115 be the @xmath116 matrix where @xmath117 are the eigenvalues of the corresponding sub - matrices @xmath102 of @xmath2 and @xmath118 .
further let @xmath119 be the eigenvalues of @xmath120 with eigenvectors @xmath121 .
then the general eigenvector of @xmath122 has the form @xmath123 corresponding to the eigenvalue @xmath124 . in this way we have , in principle , diagonalized the matrix @xmath2 .
in particular , its eigenvalues corresponding to @xmath125 can be determined explicitely .
the fourier basis vector @xmath126 is the vector with constant entries @xmath127 .
the eigenvalues @xmath128 considered above are the constant row sums of @xmath102 , where @xmath129 , since @xmath112 has vanishing row sums .
it follows that @xmath130 . obviously , @xmath131 is the lowest eigenvalue of @xmath132 since @xmath133 but @xmath134 and @xmath135 .
the corresponding eigenvector of @xmath2 is @xmath136 .
it is , up to normalization , identical with the conjectured ground state @xmath12 according to ( [ r5 ] ) . to prove that @xmath12 is actually a ground state it would suffice to show that @xmath131 is the lowest eigenvalue of @xmath2 since then the equality sign in @xmath137
would be assumed .
+ , see ( [ a3 ] ) and ( [ a4a])([a4d ] ) , for @xmath3 .
the lowest eigenvalue @xmath138 corresponding to the ground state energy @xmath139 can be analytically calculated , see table [ tabev ] .
[ figev ] , width=566 ] recall that @xmath140 for small @xmath0 there exists an even simpler analytical form of @xmath131 , see table [ tabev ] , except for @xmath141 , where we could only simplify ( [ a5b ] ) to @xmath142 obviously , not only @xmath131 but all eigenvalues of @xmath2 can be calculated in closed form for , say , @xmath3 .
hence the claim that @xmath138 can be confirmed for these cases by numerical evaluation of given expressions involving only elementary functions .
.[tabev]table of the analytical form of @xmath138 for @xmath3 .
+ [ cols="^,^,^",options="header " , ] now we consider the positive terms of the double sum ( [ adouble ] ) in the case c defined by @xmath143 and .
+ + the calculations are similar to the case a1@xmath144p except that we only consider one term of the double sum ( [ adouble ] ) as a lower bound .
again we utilize the result of lemma [ lmono ] ( iv ) in order to find a lower bound of @xmath146 : @xmath147 hence we conclude @xmath148 ( blue graph ) and @xmath149 ( red graph ) at @xmath150 .
[ figib ] , width=566 ] the complete results are displayed in table [ tab3 ] as far as the leading order w. r. t. @xmath0 is concerned .
we note that for each case , a , b or c , the leading order ( of the lower bound ) of the positive terms is larger than the leading order ( of the upper bound of the absolute value ) of the possibly negative terms .
this implies that @xmath151 for all @xmath152 and sufficiently large @xmath0 .
hence the following holds : [ tgs ] there exists an @xmath43 such that for all @xmath21 the state @xmath153 , see ( [ r4 ] ) , is a ground state of the dipole ring of length @xmath0 . in the remainder of this section
we will show that the number @xmath43 in theorem [ tgs ] can be chosen as @xmath48 which is compatible with the assumption @xmath154 we have made from the outset , see ( [ ngn0 ] ) .
we begin with the cases b and c that turn out to be simpler than the case a. + + after inserting ( [ epsilon ] ) , ( [ delta ] ) and @xmath45 we divide the lower bound of the positive terms ( [ b1p6 ] ) by @xmath155 and obtain @xmath156 this is an increasing linear function of @xmath0 . on the other hand ,
the upper bound of the absolute value of the possibly negative terms is evaluated as follows : @xmath157 this is a decreasing function of @xmath158 with the limits @xmath159 for @xmath20 and @xmath60 for @xmath61 .
the graphs of both functions ( [ bound1a ] ) and ( [ bound2b ] ) hence intersect at a unique point with @xmath150 , see figure [ figib ] .
hence for @xmath160 we have @xmath161 and the state @xmath153 , see ( [ r4 ] ) , is a ground state of the dipole ring of length @xmath0 .
+ + ( blue graph ) and @xmath162 ( red graph ) at @xmath163 .
[ figic ] , width=566 ] after inserting ( [ epsilon ] ) , ( [ delta ] ) and @xmath45 we divide the lower bound ( [ c1p3 ] ) of the positive terms by @xmath164 and obtain @xmath165 this is an increasing linear function of @xmath0 . on the other hand ,
the upper bound of the absolute value of the possibly negative terms is evaluated as follows : @xmath166 this is a decreasing function of @xmath158 with the limits @xmath167 for @xmath20 and @xmath60 for @xmath61 .
the graphs of both functions ( [ bound3a ] ) and ( [ bound4b ] ) hence intersect at a unique point with @xmath163 , see figure [ figic ] .
hence for @xmath168 we have @xmath169 and the state @xmath153 , see ( [ r4 ] ) , is a ground state of the dipole ring of length @xmath0
. + + ( red graph ) and @xmath170 ( blue graph ) at @xmath171 .
[ figia ] , width=566 ] after inserting ( [ epsilon ] ) , ( [ delta ] ) and @xmath45 we divide the lower bound ( [ lb3 ] ) of the positive terms by @xmath155 and obtain the constant function @xmath172 on the other hand , the scaled upper bound of the absolute value of the sum of the possibly negative terms is evaluated as follows : @xmath173 consider the real function @xmath174 defined for @xmath158 .
we want to show that [ lna ] @xmath175 is a decreasing function for @xmath4 . * proof * : it is obvious that all terms @xmath176 are positive for @xmath177 and vanish for @xmath20 .
hence for @xmath178 the largest zero @xmath179 of the derivative @xmath180 represents a local maximum ( or a saddle point ) and @xmath181 is decreasing for @xmath182 . for @xmath183
the decrease is obvious .
it is a straightforward task to calculate the zeroes @xmath179 and we will only give the results : @xmath184 this completes the proof since all @xmath185 , and @xmath186 is a sum of four decreasing functions for @xmath187 .
@xmath188 + we calculate the value @xmath189 , see ( [ bound5b ] ) . hence , according to lemma [ lna ] , @xmath190 for @xmath191 .
numerical calculations show that there is a zero of @xmath192 at @xmath171 , see figure [ figia ] , and hence even @xmath190 for @xmath193 , but this result will not be used in the proof .
this completes the proof that @xmath6 is a ground state for @xmath4 .
together with the analytical results for @xmath3 , see section [ sec : a ] , the main result of this paper is hence proven . by the present method we can not exclude the existence of other ground states beside @xmath6 assuming the same ground state energy @xmath83 but this seems to be extremely unlikely .
i thank my co - autors of @xcite , christian schrder and marshall luban , for the permission to use parts of our joint publication for the present paper and for their continuous support .
moreover , i am indebted to thomas brcker for valuable discussions about the subject of this article . | we investigate the two well - known ground states of rings of @xmath0 classical magnetic dipoles that are given by clockwise or anti - clockwise spin orientations tangent to the circle encompassing the dipole ring .
in particular , we formulate a rigorous proof of the ground state property of the states in question .
the problem can be reduced to the determination of the lowest eigenvalue of a @xmath1 matrix @xmath2 .
we show that all eigenvalues of @xmath2 can be analytically calculated and , at least for @xmath3 , the lowest one can be directly determined .
the main part of the paper is devoted to the completion of the proof for @xmath4 based on various estimates and case distinctions .
we also discuss the question to what extent computer - algebraic results should be allowed to contribute to a mathematical proof .
_ keywords _ : ground states , magnetic dipole - dipole interaction + |
the subject is not new .
the problem is known since 1955 when gell - mann and pais [ 1 ] predicted the existence of two neutral kaons .
earlier , in 1953 in a scheme for classifying the various newly - found particles , gell - mann represented the neutral kaon @xmath0and its antiparticle @xmath1as two distinct particles .
the decay of both particles into @xmath2 was observed .
if so , how do we know which particle has originated it : the @xmath0or the @xmath3 the problem has been solved by realizing that what we observe is the mixture of two states , @xmath0and @xmath4 @xmath5 ^{1/2}}\left [ \left ( 1+\varepsilon \right )
\left| k^0\right\rangle + \left ( 1-\varepsilon \right ) \left|
\overline{k}% ^0\right\rangle \right ] , \ ] ] @xmath6 ^{1/2}}\left [ \left ( 1+\varepsilon \right ) \left|
k^0\right\rangle -\left ( 1-\varepsilon \right ) \left|
\overline{k}% ^0\right\rangle \right ] , \ ] ] where @xmath7 is a small , complex , later measured parameter responsible for cp symmetry breaking [ 2 ] . in this way first time
the interference between states of slightly different masses has appeared in quantum mechanics .
inspired by the work of gell - mann and pais , bruno pontecorvo turned to consider the possibility of quantum mechanical mixing in another neutral particle - the neutrino . in 1957
he first suggested that a neutrino may oscillate into its antipartner [ 3 ] .
oscillation among the different kinds of neutrinos was then proposed by maki , nakagawa and sakata in 1962 [ 4 ] and later by many others [ 5 ] .
the neutral @xmath8 @xmath9boson system is not the only one where the quantum mechanical mass mixing can be considered .
we can expect to observe the same phenomena in other neutral boson systems : @xmath10 @xmath11 and @xmath12 @xmath13 generally , flavour oscillations of particles can occur when states produced and detected in a given experiment , are superpositions of two or more eigenstates with different masses .
the oscillation of k and b meson has been observed experimentally in 1961 [ 6 ] and later [ 7 ] and has been used to place stringent constraints on physics beyond the standard model . if neutrinos are massive and oscillate it is possible to resolve the well - known solar neutrino problem [ 8 ] .
there are also first experiments in which the neutrino oscillations are observed [ 9 ] .
the flavour oscillation of particles is a very fascinating demonstration of quantum mechanics in the macroscopic world .
it has served as a model for many interesting systems and problems .
various aspects of quantum mechanics , as for example coherence , decoherence , wave packets , measurements , similarity and differences between pure and mixed states , wave function collapse , epr paradox are in action . on the other hand particle mixing
is the place where fundamental symmetries and properties of fundamental interactions are studied .
discovering of the cp symmetry violation and the measurement of differences between neutral mesons masses are connected with k , b bosons mixing .
neutrino oscillations have a chance to be the first place where problem of neutrino masses can be resolved . in this review
we will concentrate only on the quantum mechanical description of particle oscillations .
problems connected with testing of the fundamental interactions will not be discussed .
first of all we should mention that interference between states with different masses is not allowed in non - relativistic quantum mechanics .
the galilean invariance forbids a coherent superposition of such states ( the so called bargman superselection rule [ 10 ] ) . beyond the non - relativistic limit such restrictions
do not hold ( which clearly follows from experiment ) .
it means that all of our considerations should be done in relativistic quantum mechanics ( nevertheless non - relativistic approximations are possible ) .
first , we would like to describe briefly the traditional approach to the particle oscillation problem .
this treatment is simple and elegant but immediately raises a number of conceptual questions .
we specify more of them ( chapter 2 ) .
next we show the wave packet treatment , where some of the problems disappear ( chapter 3 ) .
the current density approach which is closely connected with the experimental setting , is described in chapter 4 .
the problem of constructing the probability current density for a particle with undetermined mass is also considered there .
next , in chapter 5 , we give some remarks on the field theoretical approach to particle oscillations .
usually ( as in the case of neutrino oscillations ) the oscillating particle is not directly observed . only particles accompanying neutrinos , hadrons and charged leptons created in the decay
are observed .
the proper approach should take all these circumstances into account .
the creation of the neutrino in the source , its propagation to the detector and the detection process are treated in the framework of quantum field theory as one large feynman diagram . in chapter 6
we discuss the controversial problem of the oscillation of particles recoiling against kaons or neutrinos from the production process .
a detailed approach using wave packets explains the problem of four - momentum nonconservation raised in the literature .
in chapter 7 we discuss the modern example of an einstein - podolsky - rosen correlation in k@xmath14and @xmath15 systems .
the amplitude approach does not entail the somewhat mysterious
collapse of the wave function which is usually invoked to describe the epr effects .
finally in chapter 8 we summarize our main conclusions .
the usual description of kaon mixing phenomena can be found in many textbooks [ 11 ] .
suppose , that we produce @xmath0at @xmath16 by the reaction @xmath17 from ( 1 ) and ( 2 ) the @xmath0state at @xmath16 is @xmath18 after time t , as @xmath19and @xmath20 states are definite mass eigenstates , we have @xmath21 where @xmath22 and @xmath23 are masses and inverse mean lifetimes respectively of the long ( short)-lived component of k. the @xmath24 fraction of the beam after time t is just @xmath25 \end{aligned}\ ] ] where @xmath26 from eq.(6 ) we can see that the fraction of @xmath27 becomes smaller ( because of decay ) and changes with time with frequency @xmath28 neutrino oscillations are described in a very similar way [ 12 ] .
let us assume that at @xmath16 neutrino with flavour @xmath29 was born with momentum @xmath30 perfectly defined ( as for example neutrino @xmath31 in the pion decay @xmath32 ) . at this time
the neutrino state is described by @xmath33 where states @xmath34 are energy - momentum eigenstates for neutrinos with mass @xmath35 and @xmath36 are elements of a flavour - mass mixing matrix .
+ then @xmath37 where @xmath38 with the same momentum @xmath39 for each neutrino .
after time t the state will evolve into @xmath40 then the probability that a neutrino born at @xmath16 with flavour @xmath29 at time t has flavour @xmath41 is given by @xmath42 where @xmath43 is the number of interfering light neutrinos this means that we must introduce some normalization volume and momentum and energy are quantized . ] .
now usually relativistic approximations are made .
as for real , light neutrinos @xmath44 , we have \(i ) @xmath45 and \(ii ) a neutrino born in @xmath46 , at time t will be approximately at position @xmath47 then from ( 10 ) we can find that the probability for our neutrino , born with flavour @xmath29 , to have new flavour @xmath48 after traveling a distance @xmath49 is @xmath50 where @xmath51 known as * oscillation length * between @xmath52 and @xmath53is defined by @xmath54 and @xmath55 are phases responsible for cp violation . from ( 11 )
it follows that the oscillation will disappear ( the @xmath56 does not depend on @xmath57 ) if ( i ) all neutrino masses are equal @xmath58 and/or ( ii ) only diagonal elements of the mixing matrix @xmath59 do not vanish .
the presented arguments seem to be clear and elegant but they are wrong .
many conceptual questions arise when we look at the presentation shown above .
a complete treatment of particle oscillation must address the following additional issues .
+ ( 1 ) a necessary condition for particle oscillation to occur is that particle source and detector are localized within the region @xmath60 much smaller than the oscillation length @xmath61@xmath62 + ( 2 ) from eqs .
( 4 ) and ( 7 ) we see that different mass eigenstates are produced and detected coherently .
this is possible only if the momentum @xmath63 and energy @xmath64 of the oscillating particle are spread in such a way that the error in @xmath65 measurements given by @xmath66 ^{1/2}\ ] ] is larger than @xmath67@xmath68 if this condition is not satisfied and @xmath69 then also @xmath70 but from the uncertainty relation @xmath71 and eq.(17 ) gives @xmath72 and the wave packet approach must be constructed .
\(3 ) the energy and momentum conservation in processes in which oscillating particles are created ( e.g. @xmath73or @xmath74 implies that different mass eigenstate components have different energy and momentum [ 14 ] .
approaches where all oscillating particles have the same momentum and different energies [ 11,12 ] , or the same energies and different momenta [ 15 , 16 , 17 ] are conceptually not correct .
\(4 ) in the traditional approach to find the oscillation probability we calculate the wave functions@xmath75 overlap ( compare eqs .
( 6 ) and ( 10 ) ) .
this procedure gives the probability which depends on time . in the real experiment the distance between the source and the detector is known ( not the moment in which the measurement is done ) . to transform @xmath76 into @xmath77 the classical formula @xmath78 is invoked .
however to find the probability that the beam of particles produced at @xmath79 will reach a physical detector at a distance @xmath80 the current density @xmath81 should be integrated over the surface of the detector and over the time of observation [ 18 ] @xmath82 in such an approach there is no problem of how to change @xmath83 into @xmath57 .
\(5 ) in case of neutrinos additional conceptual problems arise because neutrinos are not seen directly .
the only things which can be seen are hadrons or / and charged leptons in points where neutrinos are produced and detected .
so in a realistic description , the external ( initial and final ) particles should be described by wave packets , and the mass - eigenstate neutrinos should propagate from the production region to a detector [ 19 , 20 ] .
all points which have been mentioned above are not only purely academic . we do not try to derive in a more precise way something which is known from the beginning .
we will see that the more precise approach to the particles@xmath84 oscillation phenomenon gives us new predictions and elucidates mysteries in many points .
on the other hand we will see quantum mechanics in action on macroscopic distances .
the wave packet approach to neutrino oscillation was first proposed by b. kayser [ 13 ] and later considered in more detail in [ 21 , 22 ] . nowadays
the neutral boson oscillation is also treated in the same way .
we will present the formalism for neutrinos , but everything can be repeated also for bosons . in eq.(7 ) states @xmath34 have definite energy and momentum ( also the spin direction of neutrinos is defined ) so for the sake of precision we should write @xmath85 we can easily construct a state with momentum distributed around a mean value @xmath86 . let us assume that , instead of a plane wave , our new state @xmath34 ( eq.(20 ) ) has a gaussian form , which in the momentum representation , is given by @xmath87 ^{3/2}}\exp \left [ -\frac{\left ( \stackrel{\rightarrow } { p}-\stackrel{\rightarrow } { p}_a% \right ) ^2}{4\sigma _ { pp}^2}\right ] , \ ] ] where the width @xmath88 is the same for each massive neutrino in the production @xmath89 process and the same along all three directions .
the average momenta @xmath86 of the different mass eigenstates are determined by the kinematics of the production process .
in the wave packet approach , the flavour states @xmath90 after time t ( given by eq.(9 ) in the plane wave formalism ) are now @xmath91 where @xmath92 the same states in the position representation @xmath93 are given by @xmath94 where now the function @xmath95 is defined by @xmath96 since the gaussian wave packet in momentum space is picked around the average momentum @xmath86 we can neglect the spreading of the wave packet and approximate @xmath97 where @xmath98 then @xmath99 ^{3/2}}\exp \left [ i\left ( \stackrel{\rightarrow } { p_a}\stackrel{\rightarrow } { x}-e_at\right ) -\frac{\left ( \stackrel{\rightarrow } { x}-\stackrel{% \rightarrow } { v_a}t\right ) ^2}{4\sigma _ { xp}^2}\right ] , \ ] ] with the width @xmath100 in coordinate space given by @xmath101 as earlier in eq.(10 ) , to find the amplitude of the flavour changing process , we project the states @xmath102 on the flavour states @xmath103@xmath104
if @xmath103 is the same as before ( eq.(7 ) ) this means that the momentum of each neutrino @xmath105 is measured precisely .
but this is not realistic , so let us assume that also detection process is characterized by the spatial coherence width @xmath106 connected with the uncertainties in momentum and energy measurements @xmath107 the average values of the momentum @xmath108 are the same as in the incoming wave packets eq.(22 ) . to calculate the spatial decomposition of the detecting flavour state we have to take into account that the detector is placed at a distance @xmath109 from the origin of the coordinates , so we have @xmath110 and after the same approximation as before ( eq.(24 ) ) we obtain @xmath111 @xmath112 is given exactly by eq.(25 ) after replacements @xmath113 the amplitude of the flavour changing process is given by the overlap @xmath114 we have to remember that the origin of coordinates @xmath115 and @xmath116 are not the same , so @xmath117 and we have @xmath118 , \end{aligned}\ ] ] @xmath119 where the total production and detection width is now @xmath120 everything we have done up to now can be applied either to neutrino or neutral boson oscillations .
the only assumption about the narrow wave packets in the momentum representation ( eq.(24 ) ) can be used in both cases .
next we calculate the oscillation probability for neutrinos which are relativistic @xmath121 and following ref.[22 ] we approximate @xmath122 and @xmath123 e is the energy determined by kinematics of the production process for a massless neutrino and @xmath124 is a dimensionless quantity of order unity .
we will see ( next chapter ) how the relativistic approximation for neutrinos causes that the production and detection processes can be factorized out and the standard quantum mechanical approach describes the oscillation phenomenon properly . in all realistic experiments the distance @xmath109 is a fixed and known quantity , whereas time t is not measured .
the quantity which we measure is the time integral of the probability . now
the time integral can be done , and it is possible to avoid the not properly legitimated ( in quantum mechanics ) replacement @xmath125 the time integral for @xmath126 ( eq.(34 ) ) can be done [ 22 ] and after normalization @xmath127 , instead of eq.(11 ) we have @xmath128 the oscillation lengths are the same as before ( eq.(12 ) ) , namely @xmath129 and @xmath130 known as * coherence lengths * [ 23 ] are given by @xmath131 comparing eq.(37 ) to the usual expression for the neutrino oscillation probability we can see that two additional terms appear .
the second factor @xmath132 is equal to unity if @xmath133 this inequality must be satisfied to observe any oscillation .
the presence of the term ( 40 ) which goes to zero for @xmath134 reflects the requirement which we qualitatively discussed in the previous chapter : to see the oscillations , the localization of the source and the detector must be much better determined than the oscillation length .
the first factor @xmath135 was predicted long ago [ 23 ] .
it is connected with the fact that two wave packets each with different momentum and energy , have slightly different group velocities .
it means that after some time the mass eigenstate wave packets no longer overlap and can not interfere to produce oscillations .
it is very easy to predict the value of coherence length .
if both wave packets have width @xmath100 along the direction of propagation and the difference between group velocities is @xmath136 then we can expect that after traveling a distance l @xmath137 both wave packets cease to overlap each other .
this l is just the coherence length . for relativistic neutrinos , using approximations given by eq.(36 )
, we reproduce @xmath130 from eq.(39 ) , to a factor of @xmath138 we can expect that the coherence length becomes longer if the spreading of the wave packets is taken into account .
it is indeed the case as it was proved in ref.[21 ] . from eq.(39 )
we see that the coherence length @xmath130 is proportional to @xmath139 and not only to @xmath100 as in eq.(42 ) .
it means that precise measurements of momenta of all particles appearing in the detection process ( which implies small @xmath106 , thus large @xmath140 ) can increase the coherence length [ 22 , 24 ] .
this is a wonderful example of quantum mechanics in action .
a measurement can restore the coherence .
two wave packets having negligible overlap in the detector ( thus without detector influence they can not interfere , and the oscillation disappears , @xmath141 ) , because of precise measurements @xmath142 may still interfere to give rise to oscillations @xmath143 .
this feature of quantum mechanics disagrees with causality .
however , it is not the first time when quantum mechanics is at variance with common sense .
but eq.(37 ) also restores some common sense .
measurements of momenta and energies of detected particles can not be too precise if we want to maintain the particle oscillation . as a matter of fact
, we have a longer and longer coherence length , but on the other hand increasing @xmath144 makes the position of the detector to be more and more undefined . if @xmath145 the wave packets of neutrinos @xmath52 and @xmath146 lose coherence ( eq.(40 ) ) , the oscillation between them is wash away .
we see , particularly , that the plane wave approach and oscillations are incompatible . for a plane wave
@xmath147 @xmath148 and the factor ( 40 ) teaches us that oscillations disappear .
there are also approaches where the neutrino oscillation is treated in a manifestly lorentz invariant way [ 25 ] .
the final answer is exactly the same as was presented up to now , but for one difference . in the fully covariant treatment , besides the spatial width @xmath149 also temporal width @xmath150 should appear .
it causes only one change in the oscillation probability formula eq.(37 ) . instead of the spatial width
@xmath149 a new effective one , @xmath151 appears in the coherence length ( 39 ) and in the factor ( 40 ) @xmath152 where @xmath105 and @xmath153 are group velocities of wave packets .
if @xmath154 is given by the lifetime of particles which produce neutrino , @xmath154 = @xmath155 @xmath156 $ ] ( e.g. pions , kaons or muons ) then the second term in ( 43 ) is usually much bigger than the first one and the role of factor ( 40 ) becomes more important .
we should stress however that the independent widths of spatial and temporal characteristics of wave packets cause that freely propagating particles are not necessary on mass - shell . if we insist to have our particle on mass shell ( all the time , not for mean values exclusively , which is equivalent with the requirement that our particle s state satisfies the equation of motion ) only the momentum ( or only the energy ) distribution should be applied .
then energy ( or momentum ) is distributed also but in agreement with the on - shell relation @xmath157 such an approach , which we also used in our presentation above is the standard one .
a typical experiment which tries to observe particle oscillations measures the flux of flavour @xmath41 type particles in the detector localized at some distance l from the source which produces particles with flavour @xmath158 the time of measurements is not known .
usually typical measurements last hours , days or even years ( like the observation of solar neutrinos ) .
so the most appropriate way to find the probability ( or number of particles ) to cross the surface @xmath159 of the detector ( see fig . ( 1 ) ) is to integrate the probability current density over the surface and integrate the result once more over the duration of measurements @xmath160 this procedure seems to be so easy and natural , that we can ask why people try to use other , more complicated methods .
the answer is very simple , there is a problem with the correct definition of the current @xmath161 this current should be defined for particles which we measure , that is @xmath162 or @xmath31 .
but these particles have undefined masses , and we do not know how to define the probability current for such particles .
the problem is more general : how to define properly the creation and annihilation operators for undefined mass states [ 26,27 ] ? here we will not discuss all trials [ 26 ] to resolve this problem in quantum field theory .
we will concentrate only on a simple example of definition of the current @xmath163 for particles with flavour @xmath164 for relativistic currents the problem has not been solved . for kaons
it was done in ref.[18 ] . in the non - relativistic case the free schrdinger equation for particles with mass @xmath35
can be written in the form @xmath165 which is the appropriate non - relativistic limit of either klein - gordon or dirac equations @xmath166 .
for the schrdinger equation ( eq.(45 ) ) we know how to define the probability current : @xmath167 for which the usual continuity equation is satisfied @xmath168 the problem arises while we try to define the current for the states @xmath169 which are an orthogonal mixture of two states with different masses @xmath35 and @xmath170 .
we expect that , because of mass mixing , the currents for @xmath171 will not be conserved [ 18 ] , so let us propose a modified continuity equation for @xmath172 states in the form @xmath173 with the following requirements , concerning the new current @xmath174[18]@xmath175 that : ( 1 ) only velocity terms with one gradient are included , ( 2 ) for @xmath35 @xmath176 @xmath177 the diffusion terms @xmath178 should vanish @xmath179 , and ( 3 ) the sum of both flavour currents @xmath180 is conserved @xmath181 the currents @xmath182 and the diffusion terms @xmath178 can be found .
@xmath183 , \ ] ] and @xmath184 .\ ] ] for @xmath185 and @xmath186 we have @xmath187 calculations for @xmath188mixing probability , using the definition ( eq.(44 ) ) have been done in ref.[18 ] . the standard formula ( eq.(6 ) ) , with the same oscillation frequency @xmath189 was recovered , supporting previous results .
there is , however , one objection concerning this approach .
the flavour currents which we use are not conserved .
this nonconservation is given by the diffusion term ( eq.(52 ) ) which is proportional to @xmath190 .
so , we can expect that all our calculations have been done with the some precision . then
, if the result is proportional to @xmath191 our probability flux calculations give the correct answer .
the diffusion terms will change the result in higher powers of @xmath192
there are several papers where authors investigate the neutrino oscillation problem in the framework of quantum field theory [ 19,20,24,28,29,30 ] .
what are the main reasons for those studies ?
first until now , to describe particle oscillations , we have used states with undefined masses ( eq.(4 ) for kaons and eq.(7 ) for neutrinos ) .
but , as we have seen in the previous chapter , there is a problem of proper definition of such states [ 26 ] . only in the extremely relativistic limit
the flavour states are defined correctly [ 27 ] .
as we will see , in quantum field theory the particle oscillation can be treated without resort to weak eigenstates .
secondly , we have completely neglected the effect of the production and detection processes .
it has been shown [ 27 ] , that the neutrino oscillation probability is independent from the details of the production and detection processes only in the case of extremely relativistic neutrinos . and finally , in real neutrino oscillation experiments only associated particles , hadrons and charged leptons are observed .
neutrinos are not prepared and not observed directly .
one can measure the energy and momentum distributions of other particles which appear in the production and detection processes . as we will see , only the quantum field theoretical approach
gives the opportunity to express the neutrino oscillation probability in terms of measured quantities .
let us now describe briefly how the particle oscillation is treated in field theory .
as in ref.[20,30 ] we will describe the process of neutrino production ( p ) and neutrino detection ( d ) as one feynman diagram with a virtual neutrino propagating itself on macroscopic distances between the source and the detector .
let us consider the process [ 20,30 ] @xmath193 where @xmath194 and @xmath195 ( @xmath196 and @xmath197 ) are the particles in the production ( detection ) processes .
the production and detection processes are localized in coordinates @xmath198 and times @xmath199 ( see fig.2 ) .
all initial @xmath200 and final @xmath201 particles are described by wave packets .
their shapes depend on the measurement precision in the production and detection processes .
the amplitudes for the full process can be written in the form @xmath202 we can see that there are no neutrinos in the initial and final states . only particles which really appear in the production and detection
equipments are observed .
neutrinos with mass @xmath203propagate virtually between the source and the detector and are described by the feynman propagators @xmath204 we will not present the details of all calculations , which are straightforward but tedious .
a clear presentation can be found in refs .
[ 20 , 28 , 30 ] .
we will concentrate only on the discussion of the final results .
first of all , neutrinos are not directly present in eq.(55 ) , but this is not necessary .
the amplitude @xmath205 depends on points @xmath206 and @xmath207 where neutrinos were born and detected , and this is enough to study oscillations .
next , the amplitude @xmath205 depends also on amplitudes of the production and detection processes and the full structure of @xmath208 is the following @xmath209 where @xmath210 describes the process of neutrino creation and annihilation .
the standard oscillation formula is recovered only if @xmath210 can be factorized .
this happens , when amplitudes @xmath210 become independent of neutrino masses , @xmath211 .
if all neutrinos are relativistic then @xmath212 and the oscillation probability can be defined . in case of relativistic neutrinos the time integrated neutrino flavour changing probability is given by the similar formula to eq.(37 ) with two changes .
first , the dumping factor ( eq.(40 ) ) is slightly modified and now equals @xmath213 where @xmath124 is the same quantity as before ( eq.(36 ) ) , but @xmath214 is the new factor which depends on the production and detection dynamics and can be large ( e.g. @xmath215 is possible [ 30 ] ) .
the second modification is a little different definition of the coherence length . instead of ( eq.(39 ) )
we now have @xmath216 with the same factor @xmath214 as in eq.(58 ) .
the oscillation length @xmath217 and the spatial width @xmath144 are given by the same formulae as before ( eqs.(38 ) and ( 35 ) ) .
there is also an additional very important difference between the present and the former wave packet approaches .
before , @xmath100 and @xmath218 ( eq.(35 ) were two spatial widths of neutrinos specified in some way by the production and detection processes respectively .
now , these quantities are defined by spatial widths of hadrons and leptons which are measured .
it turns out ( for detail see ref.[30 ] ) @xmath219 and @xmath220 the widths of observed particles in the production and detection processes define the width of the neutrino
, even if there is no place in this approach for physical neutrinos ( only virtual ones appear ) . in the configuration space ( eq.(60 ) )
the sum of the inverse squares of widths for the observed particles gives the inverse square of the resultant width .
then the smallest ingredient width dominates the values of @xmath100 or @xmath221 .
it is the opposite for the resultant width @xmath149 where @xmath222 @xmath223 from the definition of the momentum width @xmath224 , it follows that it is just opposite in momentum space , so then @xmath225 but @xmath226 and the same for particles in the detection process .
we can see explicitly , that precise measurements of momenta of all particles involved in the neutrino detection process ( small @xmath227 and @xmath228 ) give a small resultant width @xmath106 , thus large @xmath140 and large @xmath144 . the same subtle thing , which we have discussed before in chapter 3 , that the final measurement is able to recover
the interference , in the present interpretation has found a much stronger background .
for many years oscillations of particles like kaons or neutrinos were treated in isolation .
the circumstances in which oscillating particles were produced have not been considered . recently
a series of papers has appeared [ 31 , 32 ] , in which the kinematics of the production process have been taken into account in detail .
authors claim that , because the produced oscillating particles have neither momentum nor energy defined in explicit way , this fact should have consequences not only for them but also for the recoiling particles .
let us consider the @xmath229 production in the reaction @xmath230 or the neutrino production in the @xmath231 decay @xmath232 if the invariant mass of the initial system is denoted by m @xmath233 or @xmath234 then energies @xmath235 and momenta @xmath236 of outgoing particles depend on masses @xmath237 of @xmath238 or @xmath31 . in the cm system
there is @xmath239 ^{1/2}}{2m},\ ] ] and @xmath240 where @xmath241 is the mass of the recoiling particle @xmath242 or @xmath243 . from the four - momentum conservation in the production processes ( 63 ) or ( 64 ) , the energy and momentum of @xmath242 or @xmath243
are also defined by ( 65 ) .
authors of ref .
[ 31 ] claim that if @xmath229 or @xmath244 oscillate , also the recoiling particles @xmath242 or @xmath243 do the same .
if it is true , the existence of such phenomena could give a chance for indirect observation of neutrino oscillations which are very difficult to observe in a direct way .
however other papers have immediately appeared [ 33 , 34 , 35 ] where authors have been strongly against the oscillation of particles produced in association with kaons or neutrinos .
we will present here our approach to the problem [ 36 ] which also supports the opinion against a visible oscillation of the associated particles . to fix notations
, everything will be described for the process ( 63 ) , but equally well we can show the lack of visible muon oscillations in the pion decay ( 64 ) .
first of all we would like to specify the kind of oscillation , we can consider for @xmath242 ( or @xmath243 ) .
two @xmath242 s with different masses do not exist .
but even without mass differences the @xmath242 s are produced in association with the long - live @xmath245 and the short - live @xmath246 as @xmath245 and @xmath247 have different masses @xmath242 s will be produced in two orthogonal states with different energy and momentum . @xmath248 and @xmath249 where @xmath250 and @xmath251 are the momentum and the energy of the @xmath252 in total cm frame of the process ( 63 ) .
we do not know in which state @xmath253 or @xmath254 the @xmath242 particles are produced , so let us assume that at @xmath16 they are produced in some state which is a linear combination of both states ( 66 ) @xmath255 as both ingredient states have different energy they evolve with time in a different way , so there is some chance that after some period of time the state @xmath256 will oscillate to the orthogonal one @xmath257 do we have a chance to recognize both states @xmath258 and @xmath259 ? in the neutral bosons system , because of the strangeness conservation , @xmath229 and @xmath260interact strongly in a completely different way and are easily distinguishable .
here we have the same particle @xmath242 with only one decay width @xmath261 . in spite of that ,
in principle we can distinguish @xmath258 from @xmath259 but in a much more sophisticated way .
@xmath242 s in both states will decay in the same way ( mostly to @xmath262 but because two states @xmath263 have slightly different momenta ( in cm frame ) also the angular distribution ( e.g. for protons ) will be slightly different . as two states
@xmath264 and @xmath259 are various mixtures of @xmath265 the angular distribution of the protons in cm frame which come from @xmath266 or from @xmath267 will be different .
we can see that in principle both states ( 67 ) and ( 68 ) could be distinguished . but do we have anything which may be distinguished ?
in other words , if we have flux of @xmath242 s produced with kaons in the reaction ( 63 ) , will their number in the state @xmath258 or @xmath259 change with distance from the reaction point ?
for simplicity we will present the answer to this question in the plane wave language .
we know that it is not precise , but the value of oscillation length obtained in this way is correct .
the full wave packet approach , together with particle correlations ( the epr effect ) will be presented in the next section .
let us assume that at @xmath268 @xmath242 s are produced in the pure state @xmath269 ( in the reaction ( 63 ) , the coefficients @xmath270 , but it is more transparent to leave them undefined ) .
we will consider the production of @xmath242 s and the decay @xmath271 together .
the amplitude for @xmath242 production and decay after a period of time t can be written in a form , where two indistinguishable ways of reaching the final state are added coherently [ 37 ] @xmath272 where @xmath273 are amplitudes for @xmath242 production in the states @xmath274 , @xmath275 are decay amplitudes from both states , @xmath276 and @xmath277 are mass and decay width of the @xmath242 . the different proper times which elapse in the @xmath278 s rest frames during the propagation are the crucial points in our discussion .
if we denote @xmath279 the probability for @xmath242 production and decay from the initial state @xmath266 can be written @xmath280 \ } \nonumber\end{aligned}\ ] ] the oscillation can possibly arise from the term @xmath281 .
if instead of @xmath242 we consider the production and decay of the initial kaon a similar formula would be obtained but with one , as we will see , crucial difference . as masses of @xmath282bosons are different , the oscillation factor is equal to @xmath283 .
how to calculate the proper times ?
they are measured in different lorentz frames , in the rest frames of @xmath284 and @xmath285 .
the basic principle of quantum mechanics - the superposition principle tells , that we can add two states at the same time @xmath286 in the position representation we add wave functions @xmath287 in the same position @xmath116 and at the same time t. it means that the proper times are not suitable variables . we have to add wave function at the same point @xmath288 of the same lorentz system .
it is necessary to transform @xmath289 and @xmath290 to the same common frame .
the cm frame for the whole @xmath291 process is the most convenient in this place . for convenience ,
we consider only the one dimensional
. then lorentz transformations between the rest frames for @xmath293 and the cm frame are given by @xmath294 and @xmath295 where @xmath296 at the beginning @xmath297 and two ingredients of the @xmath242 particle , @xmath284 and @xmath285 are created at @xmath298 and @xmath299 but particles in two states @xmath300and @xmath301have different speeds and after time t they are in different points in the cm frame ( see fig.3 ) .
in classical mechanics , for point particles , it is impossible to have a situation that two particles which were born in the same point and at the same time but moving with different speeds would be still in the same , common points at the same time later ( fig.3 ) .
accordingly to our previous statement ( eq.(73 ) ) such particles will not interfere for any time @xmath302 but in qm particles are described by wave packets ( in the limiting case - plane waves ) .
we do not know at what place the particle was born inside the wave packet and what was the speed of it ( see fig.4 ) .
it is not strange that different parts of two wave packets still interfere . inside wave packets , energy and momentum
are distributed in agreement with qm prescriptions and it is not a surprise that they can be not conserved [ 32 ] . as we remember , in the wave packet approach , to find the probability as a function of position it is not necessary to assume any relation between t and x. we simply integrate over t [ 36 ] .
it is possible , however , to find such a relation between t and x that the oscillation length , which we obtain in this frame , will be ( to the first order ) the same as in a proper wave packet approach .
such a frame was found [ 35 ] .
it is the cm frame for @xmath303 and @xmath304 where their momenta are opposite @xmath305 @xmath306 ( see fig.4 ) .
the velocity @xmath41 of the origin of @xmath307 center of mass frame in the laboratory system can be easily found from the relation @xmath308 so @xmath309 then the movement of the origin of the @xmath310 cm frame in our laboratory system is described by the obvious relation @xmath311 with @xmath41 given by eq.(76 ) . using this classical relation eq.(77 ) , the proper times @xmath312and @xmath313from eq.(74 ) ) are given by @xmath314 and taking eqs .
( 74 ) and ( 77 ) we have [ 35 ] @xmath315 = 0\ ] ] and the oscillation length ( eq.(71 ) ) is infinitely large .
if we calculate , in a analogous frame , the oscillation factor for kaons we obtain [ 35 ] @xmath316 @xmath119 which reproduces the well known result for the oscillation frequency ( eq.(6 ) ) @xmath317 the relations ( 79 ) and ( 80 ) which characterize the oscillation length are obtained in this special lorentz frame .
in other frames , these results are correct only to first order in @xmath318 . instead of checking the dependence of the oscillation length on the lorentz frame , in the next section we will present the more complete wave packet approach .
here we have found that the oscillation length of particles recoiling against mixed states is very large .
it means that even if we consider the oscillation of such particles @xmath319 or @xmath320 separately , without connection to kaons or neutrinos it is impossible to observe such oscillations on any acceptable terrestrial distance .
now we consider the oscillation of both particles @xmath319 and @xmath321 or @xmath322 and @xmath323 together .
up to now we have considered the oscillation of one particle without taking into account possible correlations which may appear for two or more particles in the final states from which at least one oscillates in the traditional way .
there are many such cases .
some of them , with one oscillating particle , have been discussed above @xmath324 there are also interesting processes with two oscillating bosons e.g. @xmath325 or @xmath326 let us describe the last of them . at @xmath16 the state of two bosons is a combination of states with definite masses @xmath327 where @xmath328 and @xmath329 are elements of unitary matrix which describes the mixing .
the momentum conservation in the production process gives altogether @xmath330 independent momenta for all @xmath331 pairs @xmath332 .
sometimes there are additional correlations between various mass states @xmath333 and @xmath334 in eq.(82 ) .
if , for example , the @xmath335pairs are produced by the @xmath336 decay then the state @xmath337 must be totally antisymmetric [ 38,39 ] ( since @xmath336 has intrinsic spin @xmath338 but b mesons are spinless , the b pair is in a p wave ) .
the factors @xmath339 in eq.(71 ) are responsible for such correlations ( see for details ref.[36 ] ) .
each state @xmath340 is described by a wave packet which in the momentum representation is given by @xmath341 where @xmath342 for simplicity is taken as the gauss function @xmath343 ^{3/2}}e^{-\frac{% \left ( \stackrel{\rightarrow } { p}-\stackrel{\rightarrow } { p}_a\right ) ^2}{% \scriptstyle 4\sigma _ p^2}},\ ] ] with @xmath86 - the average momentum and @xmath344 - the width of the distribution .
after time t ( taking into account the particle decay ) the state @xmath345 will evolve into @xmath346 where @xmath347 is the lifetime of the @xmath348 particle in a chosen lorentz frame .
the states @xmath349 in the position representation will be given by @xmath350 where @xmath351 is the fourier transform of the momentum distribution ( eq.(84 ) ) and is given by eq.(25 ) .
let us assume that two detectors are placed at points @xmath352 and @xmath353 .
the detectors will measure the particles with beauty 1 and 2 ( @xmath354 , respectively ) .
the states of the b mesons measured by the two detectors are defined by @xmath355 and @xmath356 notations in eqs .
( 87 ) and ( 88 ) are similar to previously presented in eq.(30 ) .
we can find the amplitude for two oscillating particles in the same way as before ( chapter 3 ) .
then the amplitude of the process where two particles @xmath357 and @xmath358 produced at @xmath16 at point @xmath359 are detected as particles with beauty 1 ( 2 ) at point @xmath360 at @xmath361 , is the following @xmath362 @xmath119 with the normalization factor @xmath363 ^{1/2},\ ] ] and the effective total widths @xmath364 the amplitude ( 89 ) can be used in various situations .
if we apply the formula ( 89 ) to the description of the epr effect in the @xmath365 decay ( refs [ 38,39 ] ) , the collapse of the bb wave function is included in a natural way .
our approach is an alternative to the amplitude description ( refs [ 38,39 ] ) and , in case of particle mixing , takes into account the epr correlations in a much more transparent way ( see ref.[36 ] for detail ) .
the formula ( 89 ) can also be used for the @xmath366 pair oscillation from the @xmath367 decay . in this case only one particle , the neutrino , mixes .
then , in the application of eq.(89 ) to our present purpose , we have to take one diagonal mixing matrix ( e.g. @xmath368 ) and @xmath369 .
usually neutrinos are considered as stable or very long living particles .
let us also assume that the oscillation of the muons is measured on a distance much shorter than their decay length
. then both factors in eq.(89 ) , which are responsible for particle decay , may be neglected . in such circumstances , the probability that neutrinos produced as @xmath370 type together with muons at @xmath359
are observed as a @xmath41-type neutrino at distance @xmath371 and the muons at distance @xmath372 after integrating over times is given by ( for details see ref.[36 ] ) , @xmath373 ^{-1}\sum\limits_{ab}\sqrt{\frac 4{\left ( v_{\mu a}^2+v_{\mu b}^2\right ) \left ( v_{\nu a}^2+v_{\nu
b}^2\right ) } } \times u_{\beta b}u_{\alpha b}^{*}u_{\beta a}^{*}u_{\alpha a}\times \\ \\ e^{-2\pi i\left ( \frac{l_\mu } { l_{ab}^{\mu osc}}\right ) } e^{-\left ( \frac{% l_\mu } { l_{ab}^{\mu coh}}\right ) ^2}e^{-\left ( \frac{\sigma _ { \mu x}}{% l_{ab}^{\mu osc}}\right ) ^2n_{ab}^\mu } \times e^{-2\pi i\left ( \frac{l_\nu } { l_{ab}^{\nu osc}}\right ) } e^{-\left ( \frac{l_\nu } { l_{ab}^{\nu coh}}% \right ) ^2}e^{-\left ( \frac{\sigma _ { \nu x}}{l_{ab}^{\nu osc}}\right ) ^2n_{ab}^\nu } . \end{array}\ ] ] the oscillation @xmath374and coherence @xmath375lengths , and the factor @xmath376 for muons are ( for neutrinos the appropriate expressions are similar ) @xmath377 ^{-1},\ ] ] @xmath378 and @xmath379 ^2},\ ] ] where @xmath380 and @xmath381 are energy , momentum and velocity of the muon associated with the neutrino a(b ) .
first of all , we can see from eq.(92 ) that the muon oscillation disappears if we do not measure separately the neutrinos with flavour @xmath41 .
in such case the probability given by amplitude ( 92 ) is constant in @xmath382 and @xmath383 and can be normalized @xmath384 but even if we measure the @xmath41-type neutrino , the muon oscillation will not be seen
. we can prove this statement because we know precisely the muon oscillation length ( eq.(93 ) ) .
if we denote the difference between the masses of two neutrinos a and b , as @xmath385 the inverse of the oscillation length may be decomposed in powers of @xmath386 .
for muons , the term proportional to the first power of @xmath386 vanishes , and @xmath387 where @xmath388 is the factor in the decomposition @xmath389 while for neutrinos there is @xmath390 and we reconstruct the previous formula for mixing particle oscillation length eqs.(12 ) , ( 38 ) . from eq.(98 )
it follows that @xmath391 is very large and for the acceptable neutrino mass difference the muon oscillation length is much bigger than its decay length @xmath392 we can see that even if the neutrino and the muon are both measured , the oscillation of muon will not be observed
. taking into account opinions presented in the latest exchange of views [ 31 - 35 ] we agree with the statement that , in practice , oscillations of @xmath322 or @xmath393particle are impossible to observe .
let us briefly summarize the main results of this review paper * ( i ) * many conceptual difficulties arise in the plane wave approach to the particle oscillation problem .
this approach gives us the shortest way to get the correct expression for the oscillation length , but it fails if we try to describe other aspects of the particle oscillation . - gives a possibility to understand in a simple way the phenomenon , that a precise measurement of the detected particles momenta may restore the coherence between various eigenmass states and , as a consequence , the oscillation between particles , - temporal and spatial distributions in the wave packet are correlated by the requirement , that particles are on mass shell .
independent distributions for time and space give a wave packet which does not satisfy the free particle wave equation . *
( iv ) * a problem arises with the proper definition of the fock space for flavour states . as a consequence ,
the defined probability currents for such states are not conserved .
then the calculated flavour changing probability is correct only to the first power of the mass difference @xmath396 * ( v ) * the most adequate approach to particle oscillation is given by quantum field theory .
it can be seen especially for neutrinos , which are neither prepared nor observed , and only propagate between sources and detectors . in this approach * ( vi ) * we have specified the meaning of oscillation of particles which recoil against mixed states ( as @xmath242 in the process @xmath397 or @xmath322 in the decay @xmath398 ) . even if we consider oscillations of such particles separately , without connection to kaons or neutrinos , it is impossible to observe these oscillations on acceptable terrestrial distances . - description of the epr correlations in decays like @xmath399 or @xmath400 , including the mysterious collapse of the wave function in a natural way and giving the possibility to discuss the relativistic epr correlations on distances longer than coherence lengths
. - description of two particles oscillation from which only one has indeterminate mass like @xmath401 or @xmath366 .
oscillations of particles with known mass ( eq . @xmath242 or @xmath322 ) can be defined only if , in the same time , flavours of the unknown mass particles are measured ( @xmath229 or @xmath31 ) . in this case , however , the oscillation length of particles with determinate mass is very large , much larger than the particle decay length , which makes it impossible to observe their oscillation in practice .
99 m. gell mann , a. pais , phys . rev . 97 ( 1995 ) 1387 .
j. h. christensen , j.w .
cronin , v.l .
fitch , r. turlay , phys .
( 1964 ) 138 .
b. pontecorvo , zh .
33 ( 1957 ) 549 . z. maki , m. nakagawa , s. sakata , prog .
28 ( 1962 ) 870 .
v. gribov , b. pontecorvo , phys .
b28 ( 1969 ) 493 ; j.n .
bahcall , s. frautschi , phys .
b29 ( 1969 ) 623 ; s. eliezer , d.a .
ross , phys .
d10 ( 1974 ) 3088 ; s. eliezer , a.r .
swift , nucl .
phys . b105 ( 1976 ) 45 ; s.m .
bilenky , b. pontecorvo , phys .
lett . b61 ( 1976 ) 248 ; h. fritzsch , p. minkowski , phys .
lett . b62 ( 1976 ) 72 ; a.k .
mann , h. primakoff , phys .
d15 ( 1977 ) 655 .
good et al . , phys .
124 ( 1961 ) 1223 . c. geweniger et al . ,
b48 ( 1974 ) 487 ; for the recent test see , for example r. adler et al .
( cplear collaboration ) , phys .
b363 ( 1995 ) 237 ; phys .
b363 ( 1995 ) 243 ; d. buskulic et al .
( aleph collaboration ) , z. phys .
c75 ( 1997 ) 397 .
b. pontecorvo , sov .
jetp 26 ( 1968 ) 984 ; l. wolfenstein , phys . rev .
d17 ( 1978 ) 2369 ; phys .
d20 ( 1979 ) 2634 ; s.p .
mikheyev , a. yu .
smirnov , yad .
42 ( 1985 ) 1441 ; il nuovo cimento c9 ( 1986 ) 17 ; see also : j.n .
bahcall , neutrino astrophysics ed .
cambridge university press , cambridge , 1989 ; for the last data see : superkamiokande collaboration , hep - ex/9805021 . c. athanassopoulos et al .
( lsnd ) , phys .
( 1995 ) 2650 ; ibid . 77 ( 1996 ) 3082 ; nucl - ex/9706006 ; for oscillation of atmospheric neutrinos see : superkamiokande collaboration , hep - ex/9803006,hep - ex/9805006 ; hep - ex/9807003 .
v. bargmann , annals of math.59(1954)1 ; see also : a. galindo , p .
pascual , quantum mechanics , springer verlag , 1990 , p. 288 .
see for example : e.d .
commins , p.h .
bucksbaum , weak interaction of leptons and quarks , cambridge university press , cambridge , 1983 , p. 247 ; w.e .
burcham and m. jobes , nuclear and particle physics , longmans , harlow , uk , 1995 .
see for example : s.m .
bilenky , s.t .
petcov , rev .
mod . phys .
59 ( 1987 ) 671 ; b. kayser , f. gibrat - debu , f. perrier , the physics of massive neutrinos,world scientific , singapore , 1988 , p.10 ; r.n .
mohapatra , p.b .
pal , massive neutrinos in physics and astrophysics,world scientific , 1991 , p. 156 ; t.p .
cheng and l.f .
li , gauge theory of elementary particle physics
, clarendon press , oxford , 1984 , p. 410 .
b. kayser , phys .
d24 ( 1981 ) 110 .
winter , lett .
nuovo cim .
30 ( 1981 ) 101 ; f.boehm , p. vogel , physics of massive neutrinos , cambridge univ .
press,1992 , p. 92 ; t. goldman , la - ur-96 - 1349,hep - ph/9604357 .
k. grotz , h.v .
klapdor , the weak interaction in nuclear , particle and astrophysics , adam higler , bristol , 1990 , p. 296 .
h. j. lipkin , phys .
b348 ( 1995 ) 604 . y. grossman and h.j .
lipkin , phys .
d55 ( 1997 ) 2760 .
b. ancochea , a. bramon , r. munoz - tapia , m. nowakowski , phys .
b389 ( 1996 ) 149 .
j. rich , phys .
d48 ( 1993 ) 4318 . c. giunti , c.w .
kim , j.a .
lee , u.w .
lee , phys .
d48 ( 1993 ) 4310 . c. giunti , c.w .
kim , u.w .
lee , phys . rev .
d44 ( 1991 ) 3635 ; c.w .
kim and a. pevsner , neutrinos in physics and astrophysics , contemporary concepts in physics , vol
. 8 ed . by h. feshbach ( harwood academic chur ) , switzerland , 1993 .
c. giunti , c.w .
kim , phys .
d58 ( 1998 ) 017301 ; hep - ph/9711363 .
s. nussinov , phys .
lett . b63 ( 1976 ) 201 .
k. kiers , s. nussinov , weiss , phys .
d53 ( 1996 ) 537 ; k.kiers , n. weiss , phys .
d57 ( 1998 ) 3091 ; hep - ph/9710289 .
s. mohanty , hep - ph/9702428 ; hep - ph/9706328 ; hep - ph/9710284 . m. blasone , g. vitiello , ann.phys .
( n.y . ) 244 ( 1995 ) 283 ; e. alfinito , m. blasone , a. iorio , g. vitiello , acta phys . pol . 27b ( 1996 ) 1493 ; hep - ph/9510213 ; m. blasone , g. vitiello , annals phys . 244 ( 1995 ) 283 , erratum-
( 1996 ) 249 ; hep - ph/9501263 ; m. blasone , p.a .
henning , g. vitiello , in proceedings of results and perspectives in particle physics
, la thuile , aosta valley , march 1996 ; e. sassaroli , hep - ph/9609476 ; hep - ph/9805480 ; m. blasone , hep - ph/9810329 ; f. fujii , ch .
haba , t. yabuki , hep - ph/9807266 . c. giunti , c.w .
kim , u.w .
lee , phys .
d45 ( 1992 ) 2414 .
w. grimus , p. stockinger , phys .
d54 ( 1996 ) 3414 . yu .
v. shtanov , phys .
d57 ( 1998 ) 4418 ; hep - ph/9706378 .
c. giunti , c.w .
kim , u.w .
lee , phys .
lett . b421
( 1998 ) 237 ; hep - ph/9709494 .
srivastava , a. widom , e. sassaroli , phys .
b344 ( 1995 ) 436 ; hep - ph/9509261 ; z. phys .
c66 ( 1995 ) 601 .
srivastava , a. widom , hep - ph/9511294 ; hep - ph/9605399 ; hep - ph/9612290 ; hep - ph/9707268 .
j. lowe , b. bassalleck , h. burkhardt , a. rusek , g.j .
stephenson jr , t. goldman , phys .
b384 ( 1996 ) 288 .
dolgov , a. yu . morozov , l.b .
okun , m.g .
schepkin , nucl .
b502 ( 1997 ) 3 .
h. burkhardt , j. lowe , g.j .
stephenson jr , t. goldman , hep - ph/9803365 .
m. marganska , m. zralek , in preparation .
b. kayser , l. stodolsky , phys .
b359 ( 1995 ) 359 . b. kayser , in proceedings of the mariond workshop on electroweak interaction and unified theories , les ares , france , march 1995 ; hep - ph/9509386 .
b. kayser , in proceedings of the 28th conference on high energy physics , warsaw , july 1996 ; hep - ph/9702327 . | the problem of particle oscillation is considered in a pedagogical and comprehensive way .
examples from k , b and neutrino physics are given .
conceptual difficulties of the traditional approach to particle oscillation are discussed .
it is shown how the probability current density and the wave packet treatments of particle oscillations resolve some problems .
it is also shown that only full field theoretical approach is free from conceptual difficulties .
the possibility of oscillation of particles produced together with kaons or neutrinos is considered in full wave packet quantum mechanics language .
precise definition of the oscillation of particles which recoil against mixed states is given .
the general amplitude which describes the oscillation of two particles in the final states is found . using this epr - type amplitude the problem of oscillation of particles recoiling against kaons or neutrinos
is resolved .
the relativistic epr correlations on distances of the order of coherence lengths are considered . |
recently , a statistical model for the structure of the nucleon was proposed @xcite .
the nucleon was described as a non - interacting gas of valence quarks , sea quarks and gluons in equilibrium .
this picture was then improved by adding finite - size corrections ( fsc ) to the expression for the parton density function ( pdf ) .
these finite size corrections take into account the fact that the gas is enclosed in a finite volume .
the model was successful in reproducing a large body of polarized and unpolarized nucleon structure function data .
since the ideas from statistical mechanics worked well inside one hadron , namely the nucleon , it is quite motivating to test these ideas in other hadrons as well .
however , compared to the vast amount of precise data available for the nucleon , much less is available for others . among hadrons other than the nucleon ,
the pion is the most widely explored .
there exist data on @xmath0 drell - yan production @xcite and prompt photon production in @xmath5 @xcite which have been used by some groups @xcite to obtain parameterizations for pdfs of the pion .
the drell - yan data constrains the shape of the pion valence densities and the prompt photon data constrains the pionic gluon distribution in the large - x region .
these data are , however , not sufficient to fix the gluon and sea in the pion uniquely .
the number of parameters appearing in the expressions for the pdfs obtained from such global fits are very large and with no physical meaning .
hence it is worth investigating the scope of a statistical model which has fewer parameters based on a physical interpretation . in this letter
, we present a statistical model for the parton distributions in the pion .
the expressions for the pion pdfs are written assuming the pion to be a gas of massless partons .
these are then improved to include fsc as in the case of the nucleon . in principle
the two parameters associated with the fsc can be determined theoretically @xcite .
however , these parameters are sensitive to the equation of motion employed , the boundary conditions on the wave function , shape of the enclosure and details such as whether the particles are strictly massless . in ref.@xcite , in the absence of a complete understanding of the nucleon as a qcd bound state , a practical approach was taken and the two parameters for the nucleon were determined by fitting unpolarized structure function @xmath6 data at one value of @xmath7 . in the case of the pion , due to scarcity of precise experimental data giving direct information on the pion pdfs , we prefer to use as a first guess , the same parameters for the pion as in the case of the proton @xcite .
interestingly , with these parameters we get good quantitative agreement with the available data on pion valence quark distribution and @xmath0 drell - yan production .
the agreement with data is nearly as good as that obtained by the existing parameterizations @xcite .
we visualise the pion to be made up of a gas of massless partons in equlibrium at temperature @xmath8 in a spherical volume @xmath9 with radius @xmath10 .
the parton number density @xmath11 in the infinite momentum frame ( imf ) and the density @xmath12 in the pion rest frame are related to each other by , @xmath13 where the superscript @xmath14 refers to the imf , @xmath15 is the pion mass and @xmath16 is the parton energy in the pion rest frame @xcite . using the standard procedure in statistical mechanics to introduce the effects of the finite size of an enclosure , in the expression for the density of states @xcite we write @xmath12 for the pion as , @xmath17 where @xmath18 is the spin - color degeneracy factor , @xmath19 is the usual fermi or bose distribution function @xmath20\pm 1\}^{-1}$ ] , @xmath9 is the pion volume and @xmath10 is the radius of a sphere with volume @xmath9 .
we take @xmath21 , where @xmath22 is the root - mean - square radius of the pion @xcite .
the three terms in ( 2 ) are the volume , surface and curvature terms , respectively .
the coefficients @xmath23 and @xmath24 are , in principle , the free parameters of the model . for reasons mentioned in the introduction , though @xmath23 and @xmath24 can be determined theoretically , we prefer to choose these values to be the same as those determined phenomenologically for the proton in ref.@xcite . theoretically this makes sense , as the values @xmath25 and @xmath26 obtained from the fitting procedure in ref.@xcite are close to the values ( @xmath27 and @xmath28 ) determined theoretically by morse and ingard @xcite .
we will come back to this point at the end of the paper . to determine the values of the temperature and chemical potentials appearing in the distribution function @xmath19
, we consider the fact that any model of the pdfs for the pion has to obey the number and momentum constraints , the former being motivated by the constituent quark picture for mesons at a low scale ( this scale , being our input scale , should not be to too small as compared with @xmath29 to allow for the evolution of the partons densities , hence we choose it to be @xmath30 where @xmath31 is the mass of the nucleon ) .
if @xmath32 denotes the number of quarks ( antiquarks ) of flavour @xmath33 , then in the case of a @xmath34 for example , the constraints can be given as , @xmath35 the numbers @xmath32 in eqs.(3 - 5 ) are obtained from eqs.(1 ) and ( 2 ) by integrating the appropriate @xmath11 over @xmath36 .
the momentum fractions are obtained by integrating @xmath37 over @xmath36 .
the temperature @xmath8 and chemical potentials ( @xmath38 ) appearing in eqs.(3 - 6 ) are then not free parameters . for a given set of parameters @xmath23 and @xmath24 ,
they are determined uniquely by solving the four coupled non - linear eqs.(3 - 6 ) . using the values of @xmath23 and @xmath24
mentioned above , we obtain @xmath39 mev , @xmath40 mev and @xmath41 in the case of a negatively charged pion .
the left and right hand sides of eqs.(3 - 6 ) with these values of t and @xmath38 , agree with each other up to an accuracy of one part in @xmath42 .
the chemical potentials for antiquarks are determined by using the relation @xmath43 .
the model , described above , fixes the parton distributions in a pion uniquely at an input scale @xmath44 .
it remains to put it to test and compare the predictions with available experimental data . in order to compare with data , the parton distributions
are evolved to different @xmath7 values using the standard dglap evolution equations at next - to - leading order , starting from the input scale @xmath45 .
to start with , we compare the @xmath34 valence quark distribution calculated within the statistical model with the available data at @xmath46 gev@xmath47 . in fig .
1a we show the valence structure function extracted from @xmath48 drell - yan data by the e615 collaboration .
the solid curve corresponding to our calculation of @xmath49 done within the statistical model , shows reasonably good agreement with the e615 data .
our values of @xmath50 are , however , higher than those due to the smrs ( dashed line ) and grs ( dash - dotted line ) parameterizations .
both these parameterizations have determined their valence densities by making a global fit to the @xmath51 drell - yan data . to further compare our valence densities with other calculations , we calculate the first two moments of the pion valence quark distributions given by , @xmath52 in fig .
1b we plot the @xmath7 dependence of the pion valence moments .
the solid curve corresponds to our calculation and the smrs result is shown by the hashed region .
the points with error bars are the results from a lattice qcd calculation @xcite performed at @xmath53 gev@xmath47 .
next , we put the statistical model to test with the @xmath54 drell - yan data .
we calculate the double - differential cross - section @xmath55 where @xmath56 and @xmath57 . at leading order ,
@xmath58 and @xmath59 are the bjorken @xmath36 variables of the pion and target nucleon respectively .
m is the invariant mass of the muon pair and @xmath60 is the centre of mass energy .
we calculate the drell - yan cross - sections using the full next - to - leading order expressions @xcite and with the choice @xmath61 .
the proton pdfs required for the calculation of the @xmath62 drell - yan cross - sections are also calculated within the statistical model .
both the pion and nucleon pdfs are calculated using the same values of the fsc parameters @xmath23 and @xmath24 .
a detailed description of the evaluation of the nucleon pdfs can be found in ref.@xcite . to take into account uncertainties due to normalization and higher order qcd corrections ,
we multiply our cross - sections by the standard @xmath63-factor .
we analyse two different sets of data obtained from a @xmath34 beam incident on a tungsten target , @xmath64 , by the na10 and e615 collaboration .
to make a correct comparison of our calculated cross - sections for @xmath0 with the @xmath65 data , we multiply our cross - sections with a factor @xmath66 .
this form of the correction factor r is consistent with the observed values @xcite of r as shown in ref.@xcite . in fig.2
we compare our calculated results with the na10 data taken at @xmath34 beam momenta of 194 and 286 gev / c . with @xmath63 factors of 1.02 and 1.06 at beam momenta 194 and 286 gev / c respectively , we find reasonably good agreement with data at different values of @xmath67 .
our @xmath63 values are close to those in ref.@xcite which range between 1.1 and 1.4 depending on the type of fit .
a discussion on the @xmath63 factor can be found in ref.@xcite where it was also noted that it should be fairly close to unity . in fig.2
we also show the dependence of the cross sections on the choice of the parameters @xmath23 and @xmath24 .
the solid curves correspond to @xmath68 and @xmath69 which are the values used throughout this work . varying the parameters arbitrarily we found that the agreement with data reduces as we move away from the values of @xmath23 and @xmath24 quoted above .
to demonstrate this fact , we plot the cross sections at beam momentum 286 gev / c for two different sets of @xmath23 and @xmath24 in fig .
the dashed curves correspond to the calculations with @xmath70 , @xmath71 and the dot - dashed curves correspond to those with @xmath72 , @xmath73 . in fig.3
we show our calculations in comparison with the e615 data at @xmath74 gev / c .
the @xmath63 factor in this case is 0.85 .
it differs from the @xmath63 obtained by us for the na10 data due to normalization errors in the data .
since the ratio @xmath75 in this work is very similar to that obtained in refs.@xcite , we do not worry about the fact that @xmath76 .
as mentioned above it could be due to errors in normalization .
the agreement of our results with data is again quite good .
we are , however , not able to reproduce data at very high or low values of @xmath67 very well . although the reason for this disagreement is not clear to us , it certainly is not a short - coming of the statistical model .
we say so because this kind of disagreement was also found in ref.@xcite .
they included the data only in a certain range of @xmath67 for fitting as they were unable to reproduce the data at very high and low values of @xmath67 .
finally in fig .
4 we compare the valence , sea and gluon distributions of the pion within the statistical model with those obtained by the smrs and grs parameterizations . since the pion sea distributions are not well determined by data , the smrs parameterization made different fits by varying the fraction of the pion momentum carried by the sea between 10% and 20% .
interestingly , we find that the sea calculated within the statistical model carries a fraction roughly equal to 15% of the pion momentum which is like the smrs best fit value . to summarize
, we can say that we have presented a simple physical model for the pion , using ideas from statistical mechanics .
the only parameters in the description of the pionic parton densities are the two ` finite - size correction ' parameters .
considering the fact that the only two parameters appearing in the model for the pion _ have not been determined by any fit to the pionic data _ , the success of the model in reproducing data on the pion is surprising .
since we use the same parameters for the proton and pion ( for reasons explained in the introduction ) , we have in a sense , a unified model for the proton and pion .
we tried different sets of the values @xmath23 and @xmath24 for the pionic partons and having the same values as that for the proton is an outcome of our work rather than an assumption . at this stage , however , it would be too early to speculate about the significance of this result .
it could be that given a wealth of precise enough data on pionic parton distributions , the ` true ' values of @xmath23 and @xmath24 determined by a fit , will differ from the ones presently used .
however , it is clear from our results that the difference will not be drastic .
it seems therefore that the values for @xmath23 and @xmath24 are always close to the theoretical values , @xmath27 and @xmath28 determined in ref .
this could partly be the reason why the statistical model works so well for the proton as well as for the pion using the same @xmath23 and @xmath24 .
further efforts to understand why the ideas from statistical mechanics work inside hadrons could be worthwhile and could improve our understanding of hadron structure . in future
we intend to understand this point in more detail , by using for comparison a larger set of data ( i.e. prompt photon production ) , by extending the model to other hadrons like kaons and hyperons , and relating their structure functions to fragmentation functions .
this could also settle the issue about the parameters @xmath23 and @xmath24 .
13 r. s. bhalerao , n. g. kelkar , b. ram , phys .
b 476 ( 2000 ) 285 .
na10 collaboration , b. betev et al .
, z. phys .
c 28 ( 1985 ) 9 .
the cross sections in this paper have been revised with a better estimate of the fermi motion effects .
e615 collaboration , j. s. conway et al .
, phys . rev .
d 39 ( 1989 ) 92 .
wa70 collaboration , m. bonesini et al . , z. phys .
c 37 ( 1988 ) 535 .
p. j. sutton , a. d. martin , w. j. stirling , r. g. roberts , phys .
d 45 ( 1992 ) 2349 . m. glck , e. reya , i. schienbein , eur . phys .
j. c 10 ( 1999 ) 313 .
r. balian , c. bloch , ann . of phys .
60 ( 1970 ) 401 ; r. k. bhaduri , j. dey , m. k. srivastava , phys .
d 31 ( 1985 ) 1765 ; p.
m. morse , k. u. ingard , theoretical acoustics ( mcgraw - hill , new york , 1968 ) p. 587 .
r. s. bhalerao , phys .
b 380 ( 1996 ) 1 ; b 387 ( 1996 ) 881 ( e ) .
d.l . hill and j.a .
wheeler , phys .
89 ( 1953 ) 1102 ; h.r .
jaqaman et al . , phys .
c 29 ( 1984 ) 2067 .
v. bernard , n. kaiser , ulf - g .
mei@xmath77ner , phys .
c 62 ( 2000 ) 028201 .
g. martinelli , c. t. sachrajda , nucl .
b 306 ( 1988 ) 865 .
na10 collaboration , p. bordalo et al .
b 193 ( 1987 ) 368 .
g. altarelli , s. petrarca , f. rapuano , phys .
b 373 ( 1996 ) 200 . | we present a model for the structure of the pion . based on ideas of a recently developed statistical model of the nucleon
, we assume the pion to be a gas of partons .
the finite - size corrections ( fsc ) are incorporated through two parameters . using the same two fsc parameters for the proton and pion we reproduce quantitatively the data on @xmath0 drell - yan production and valence quark distribution of the pion .
* statistical model for pionic partons * * n. g. kelkar@xmath1 and m. nowakowski@xmath2 * 0.25 cm @xmath3 _ nuclear physics division , bhabha atomic research centre , _ _ mumbai 400 085 , india .
_ @xmath4 _ departamento de fisica , universidad de los andes _ _
a.a.4976 , bogota d.c . ,
colombia _
pacs numbers : 12.40.ee , 14.40.aq , 14.65.-q _ keywords _ : pion structure function , parton distribution , statistical model , phenomenological models of the pion , finite - size effects |
following the launches of several high energy astrophysical satellites in the last decade , for example , exosat , ttm and hete aboard mir space station , rosat , _ ginga _ sigma / granat , cgro , asca , sax and rxte , our knowledge of high energy spectra from xb sources has increased enormously . in this review paper , we shall focus on the high energy spectra of these binary systems whose primary electromagnetic energy output is above 1 kev .
we will , however , not discuss narrow or broad spectral line features .
in a binary system involving a compact object ( a ns or a bh ) and a more ordinary star ( called companion star ) , matter can be transferred , either through roche lobe overflow or via stellar wind , from the companion to the compact object . in order for
the transferred or accreted matter to reach the compact object , angular momentum loss has to occur , presumably via the viscous mechanisms .
thus an accretion disk may be formed and the gravitational energy loss via the viscous loss will be radiated in the form of electromagnetic waves . the energy of the radiated photons will become higher when the accreted material gets closer to the central object . in the standard shakura - sunyaev ( 1973 ) model , a geometrically thin and optically thick accretion disk is formed . the local radiation can be approximated by a blackbody spectrum .
the overall spectrum is thus the integration of the local ( effective ) blackbody emission from the inner disk region to the outer disk edge ( see e.g. , mitsuda _ et al . _
1984 , 1989 ) .
such a spectrum model is called the mcb model .
the majority of the total luminosity is radiated between the inner disk edge ( a few @xmath1 , the schwarzschild radius ) to a radius of the order of 100 @xmath1 .
the inner edge temperature for a stellar mass bh is usually around 1 kev .
a mcb spectrum is shown in figure 1 together with other spectral models to be discussed in this section .
when the central compact object is a ns , the remaining gravitational energy will be released when the accreted matter finally reaches the ns . in the case of a weakly magnetized ns ( b@xmath2@xmath3 gauss ) , usually found in atoll sources ,
the matter can reach the ns surface directly and the radiated spectrum can be approximated by the scb model ( 1 - 2 kev ) ( mitsuda _ et al . _ 1984 ) , although the boundary layer of the ns is quite complicated .
the measured temperature is thus the effective temperature of the ns surface . in the 1 - 10 kev region ,
the mcb spectrum is considerably softer than the scb spectrum .
this is why the mcb spectrum is usually called an ultra - soft ( us ) spectrum . in case of a strongly magnetized ns ( b@xmath4@xmath5 gauss ) , usually found in x - ray pulsars ,
the accreted matter can not reach the ns surface directly and is funneled onto its polar cap region by the strong magnetic field .
a hot plasma ( 5 - 20 kev ) is thus formed and the continuum emission can be described by the tb , characterised by a hard power law ( photon index -2.0 to -1.0 ) with an exponential cutoff below 20 kev ( white , swank & holt 1983 ) . for a mildly magnetized ns ( @xmath6@xmath7b@xmath7@xmath5 gauss ) , usually found in z - sources , both the scb and the tb components may be present ,
so that the overall spectrum becomes quite soft compared with x - ray pulsars . in some systems
roche lobe overflow can not occur .
significant wind accretion will take place if the companion can produce strong winds .
such systems exist with either a ns ( e.g. , vela x-1 ) or a bh ( e.g. , cyg x-1 ) as the compact object .
as well as forming an accretion disk , the wind accretion in a bh system may cause some interesting consequences . as suggested by shapiro and lightman ( 1976 ) ,
an asymmetric shock front ( bow shock ) will be formed with a temperature of the order of 10@xmath8 k , when the bh accretes directly from the stellar wind of its high mass ( hm ) companion of cyg x-1 .
it is reasonable to expect that the thermal radiation from the accretion disk , if formed , will be inverse compton scattered to higher energy photons ( zhang _ et al . _
1997a ) by the hot electrons in the bow shock ( the electrons are heated up by collisions with protons ) .
as calculated by sunyaev and titarchuk ( 1980 ) the spectrum will be characterized by a power law with a high energy cutoff , whose energy is proportional to the mean electron temperature in the shock .
this mechanism may naturally explain the observed hard state spectrum of cyg x-1 ( photon index -1.5 to -2.0 with cutoff between 50 - 200 kev ) .
this is the so - called tc process . in figure 2
, we show an illustrative model of cygnus x-1 ( zhang _ et al . _
1997a ) .
even in low mass ( lm ) xbs , some accretion disk models also suggest the existence of a hot cloud surrounding the central compact object .
thus the tc process may also work here .
several models are invoked to explain the formation of this cloud .
the most recent ones are the two - phase model of haardt and marashi ( 1993 ) , advection - dominated model of narayan ( 1996 ) and the post - shock model of chakrabarti and titarchuk ( 1995 ) .
although this issue is still controversial , this hot cloud seems to exist in both bh and ns xbs and seyfert galaxies as well , as inferred from the characteristic tc spectra .
when the supply of the low energy photons is over - abundant , this tc process can cool down the cloud and eventually destroy it . in the case of a ns xb ,
the power law component will decrease and finally disappear . for a bh xb ,
the situation is more interesting . because of the strong gravitational field and the absorptive horizon of the bh
, the low temperature electrons will move towards the bh relativistically to form a converging flow ( chakrabarti and titarchuk 1995 ) .
the relativistic electrons will up - scatter the low energy photons , either trapped in the flow when it sweeps through the surface of the accretion disk or simply intercepted by the flow , to form a power law .
we call this radiation mechanism bulk motion comptonization ( bmc ) process .
calculations made by chakrabarti and titarchuk ( 1995 ) have shown that the bmc power law is steeper ( photon index around -2.5 ) than the tc power law formed from the hot comptonization cloud ( photon index -1.5 to -2.0 ) .
as a matter of fact , up to the present time , the only reliable way of determining if a binary system contains a bh is by measuring the system parameters .
if one can estimate the mass of the compact object , it will be considered to be bh if its mass is @xmath0 3 m@xmath9 , currently the best known theoretical and observational upper mass limit of a ns .
several systems have been identified in this way . however
, this method is quite restrictive in principally two aspects .
one is that the masses of the compact objects of the majority of the bh candidates ( see below ) can not be reliably measured due to various observational constraints .
the other is that this method in principle remains ambiguous for those with mass limits @xmath73 m@xmath9 .
therefore we have to look for alternative methods of selecting bhbs . by comparing the common characteristics of the high energy emission from these dynamically determined bh systems
, it has been hoped that some signatures " of bhs may be found when accurate dynamical estimates are not available .
such signatures " include ultra - soft / ultra - hard energy spectra , rapid variability , qpos etc .
all such features have , however , been observed from ns systems , especially those with a weak magnetic field .
this is , perhaps , not surprising since most of these phenomena are associated with properties of the accretion disk common to both types of systems .
it has been recently suggested that perhaps luminous hard x - ray emission or a combination of an ultrasoft component and hard x - ray power - law are more indicative of the existence of a black hole ( van der klis 1994 ; van der klis and van paradijs 1993 ; white 1993 ; tanaka & lewin 1995 ; zhang _ et al . _
1996a ; 1996b ; barret , mcclintoch & grindlay 1996 ) . in this section
we try to understand how to distinguish between ns and bh xbs , based on the similarities and differences in the formation process of high energy spectra of these systems we discussed in the previous section .
we will first identify some signatures of nsxbs and then discuss how a bhxb might be identified in the absence of any signature of nsxbs .
the high energy spectra of magnetized nsxbs can be relatively easily distinguished from that of bhxbs .
the tb spectral shape with a cutoff energy below 20 kev is a strong indication of the existence of the ns magenetosphere and , in fact , has never been detected from a bhxb .
the 1 - 2 kev scb component , if unambiguously identified , is a strong indication of the detection of the neutron star surface radiation .
this component has also never been detected from a bhxb .
thus both characteristic spectral components , together with pulsations and type i x - ray bursts , can be used as signatures of nsxbs . in the absence of any one of the above signatures ,
a source may be considered a bhc .
we now discuss how to further identify those bhxbs from the list of bhcs , such as that reviewed by tanaka and lewin ( 1995 ) .
one common high energy radiating component between a bh and a weakly magnetized ns xb ( or simply ` nsxb ' for the rest of this section since the strongly magnetized nsxbs can be easily identified as discussed above ) , is the inner disk region very close to the compact object , which produces the mcb spectrum . due to the ns scb emission
, the overall spectrum of such a nsxb may be harder than that of a bhxb .
however , when the mass accretion rate is low so that the ns surface temperature is also low , the nsxb spectrum may appear indistinguishable from that of a bhxb .
_ thus a luminous ( @xmath010@xmath10 erg / s ) mcb component is a strong indication of the existence of a bh in a xb . _ using the mcb model , in principal one can determine the temperature and radius of the inner disk edge .
for a nsxb , the inner disk edge should be very close to the ns surface , i.e. , about 10 km . for a bhxb
, the inner disk radius can be much larger .
_ therefore an inner disk radius significantly larger than 10 km , for example , greater than 30 km , is another indication of the existence of a bh .
_ this is why dynamically determined bhxbs all appear to have much larger inner disk radii than that of nsxbs ( ebisawa _ et al . _ 1991
however , for a maximally rotating kerr bh with a mass comparable to the ns mass , its inner disk radius will not be very different from that of a nsxb .
so a small inner disk radius is not always associated with a nsxb .
another common high energy radiation component between a bhxb and a weakly magetised nsxb is a hot electron cloud , producing the observed tc power law with cutoff between 50 and 200 kev .
this cloud is different from the extended corona in many the so - called ` dipper ' sources ( the coronae in the ` dipper ' sources have much lower temperatures and extend well above the plane of the disk , as inferred from the significant residual fluxes of the high inclination systems during the accretion disk eclipse by the companion . ) .
although such tc spectra are quite similar between ns and bh xbs , there are , however , still some differences .
one is that the spectral break energy in a bhxb is frequently observed to be @xmath0 100 kev , while for nsxbs the highest observed spectral break energy is only 65 kev ( zhang _ et al . _
perhaps this is because there are two soft photon sources in a nsxb ( one is the disk emission and the other is the ns surface emission ) , in contrast to the single soft photon source in a bhxb ( the disk emission ) .
therefore the electron temperature in a nsxb is usually lower than that in a bhxb .
the same argument can be applied to the fact that the hard x - ray luminosity can be significantly higher in a bhxb .
_ therefore the detection of a combination of a high energy cutoff ( @xmath0100 kev ) and a bright hard x - ray luminosity ( @xmath010@xmath10 erg / s above 10 kev ) indicates strongly the existence of a bh in the system .
_ the hard x - ray power law emission will disappear in a nsxb when the disk accretion rate is so high that the hot electron cloud is cooled down and consequently the tc mechanism can no longer effectively produce hard x - ray photons . in a bhxb , a steeper power law ( spl )
should be formed in the converging flow by the bmc mechanism ( chakrabarti & titarchuk 1995 ) .
this emission mechanism is related to the lack of a solid surface of a bh , and therefore may provide a signature of the bh .
the question is whether a similar high energy spectrum can be also formed from other emission mechanisms not related to the horizon of a bh .
although hypothetically many high energy radiation mechanisms can form such a power law energy spectrum in this energy range ( 10 - 500 kev ) , the conditions when such a power law is detected restricts the possibilities severely . almost in all cases the spl ( without cutoff up to 300 - 500 kev ) is accompanied with a bright mcb component observed from a bhxb .
synchrotron radiation is an unlikely mechanism , since the bright mcb component tends to cool down the assumed high energy electrons by inverse compton scattering , thus preventing non - thermal radiation up to several hundred kev .
it has been widely considered that the tc process with a cooler electron cloud as the result of the more effective cooling by the higher soft x - ray flux , may be a good candidate for producing the spl .
there are , however , several problems with this picture .
first , the lower electron temperature should result in a lower spectral break energy when the spl is observed , contrary to the detected unbroken power law up to at least 300 - 500 kev .
the other problem is the near constancy of the spl index when both the mcb and the spl luminosity change significantly as observed from gs1124 - 68 in the soft state ( ebisawa _ et al . _
other radiation mechanisms , involving relativistic outflows or jets , are also unlikely to work , since not all bhxbs are observed with jets or signficant radio emission , or such spl detection is not always associated with jet ejection events ( zhang _ et al._1996b ) .
moreover , all the above mentioned mechanisms should also operate in a nsxb ; but such spl / mcb have never been detected from any confirmed nsxb .
_ we therefore conclude that such unbroken spl ( up to 300 - 500 kev ) associated with a luminous mcb component ( @xmath1110@xmath10 erg / s ) is mostly likely produced in the converging flow near the bh horizon and may be used as a distinctive signature of bhxbs .
_ in summary , there appears to be several strong suggestive indicators of the existence of a bh in a xb .
these are : a ) a luminous ( @xmath1110@xmath10 erg / s ) mcb spectrum without the contamination of a scb component ; b ) a bright ( @xmath1110@xmath10 erg / s ) hard x - ray power law ( photon index : -1.5 to -2.0 ) with high energy cutoff ( @xmath0100 kev ) ; and c ) a spl ( photon index @xmath11 -2.5 ) without cutoff up to 300 - 500 kev associated with a luminous mcb component ( @xmath1110@xmath10 erg / s ) .
the first two indicators are related to the lack of surface radiation from the bh , in contrast to a ns , while last one seems to be due to the horizon of the bh and thus is a more secured signature of a bh .
we do not include the large inner disk radius as a bh indicator here , since this parameter depends upon the details of the rather complicated spectral model which is not well understood yet . as a matter of fact , none of the known nsxbs , including all known pulsars , z - sources and type i x - ray bursters , have been observed with one of the three high energy spectral types , with the only possible exception of terzan-2 , a type i x - ray burster . during one observation with sigma / granat ,
a hard power law up to about 200 kev was detected with a luminosity of exceeding @xmath115@xmath1210@xmath10 erg / s ( barret _ et al . _
it is , however , not clear if the emission originated from terzan-2 , or another high energy transient from that direction or within the same globular cluster .
at all other occasions when it is detectable in the hard x - ray band above 20 kev , a much lower luminosity was seen .
all the dynamically confirmd bhxbs , i.e. , with the compact object mass greater than 3 m@xmath9 , have displayed at least one of three indicators .
these systems are cygnus x-1 ( indicator : b and c ) , 0620 - 00 ( a , c ) , lmc x-3 ( a , c ) , lmc x-1 ( a , c ) , gs 1124 - 683 ( a , b , c ) , gs 2000 + 25 ( a , c ) , gs 2023 + 33 ( b ) , gro j1655 - 40 ( a , c )
. here we also further examine all those remaining bhcs listed by tanaka and lewin ( 1995 ) .
these systems should also contain a bh according to the three indicators : gx 339 - 4 ( a , b , c ) , 1354 - 645 ( c ) , 1543 - 475 ( a , b ) , 1630 - 472 ( a , b , c ) , 1705 - 250 ( a , c ) , 1740.7 - 2942 ( b ) , 1741 - 322 ( c ) , grs 1915 + 105 ( a ) and 1758 - 258 ( b ) . the most recently discovered transient grs 1739 - 278 also belongs to this class since the indicator ( c ) was detected from it ( vargas _ et al . _
we briefly summarise the spectral state transitions in bh and ns xbs , aiming at understanding the dynamics in xbs . here
we call these states ` hard ' or ` soft ' , instead of ` low ' or ` high ' used by some others , since we will show below that the bolometric luminosity , presumably proportional to the overall mass transfer rate , actually changes insignificantly or at least the hard state bolometric luminosity is not necessarily always lower than that in the soft state . the most recent soft state of cyg x-1 has been extensively observed by several high energy instruments from aboard cgro , rxte , asca and sax satellites .
here we only show some results from the broad band monitoring achieved with asm / rxte ( 1.3 - 12 kev ) and batse / cgro ( 20 - 200 kev ) ( zhang _ et al . _
the main results are : a ) during the whole period the soft x - ray fluxes are anti - correlated with the hard x - ray fluxes ; b ) the 1.3 - 200 kev luminosity ( and the bolometric luminosity after the color correction , see zhang _
et al . _
1997c for details ) does not change significantly , and is not correlated with the state transition ; c ) the transition time of the lowest energy band ( 1.3 - 3.0 ) , whose flux is dominated by the mcb component , is much shorter than that of the hard x - ray flux ; d ) flares with durations of around 10 - 30 days are seen in all energy bands , superposed on top of the state transition ; and e ) the soft x - ray flares lag behind the hard x - ray flares by 10 - 20 days .
the near constant luminosity during the state transitions defies most of the current models of high energy radiation in bhxbs .
all these models require a significant luminosity increase in the soft state , although some fine tuning of some of these models may be able to remove this requirement .
for example , in the chakarabarti and titarchuk ( 1995 ) model , at a certain mass accretion rate the re - distribution of the mass flow rates between the keplerian and sub - keplerian mass accretion components may also produce a constant luminosity during a state transition ( titarchuk 1996 ) , instead of requiring a total mass acrretion rate increase .
however , the other results , especially the longer hard x - ray flux transition time and the lag between the soft and hard flares can not be naturally accounted for within these current models .
there appears to be several spectral state transitions observed from the lm bhxb gs 1124 - 683 during its 1991 outburst ( see ebisawa _ et al .
_ for details of this outburst observed with _ ginga _ ) .
here we just briefly mention two behaviors .
the first one is during the initial outbursting phase , when the power law component reached its peak flux about 3 days earlier than the mcb component .
this behavior is not really a spectral state transition , but instead reflected a lag of the mcb component increase behind the power law component increase , in responding perhaps to the mass accretion rate increase .
it is interesting to note that when the soft component reached its peak , the hard component also experienced a dip , without significant changes of the spectral shape of each component .
the second one is a real spectral state transition between april and july , when the hard component reached a secondary maximum and the soft component continued to decrease .
the significant features of this state transition are the low luminosity level ( @xmath111% of the peak total luminosity ) when the transition happened and the much harder power law ( photon index changed from between -2.2 and -2.7 to @xmath11 -1.6 ) during this secondary outburst . in figure 3
we show the spectral transition of the lm nsxb 4u 1608 - 522 when the energy spectrum changed from a typical soft state spectrum ( mcb plus a scb ) to a hard state ( power - law dominated ) spectrum when the total luminosity decreased from about 5@xmath1210@xmath13 to about 7@xmath1210@xmath14 erg / sec .
a transition happened at 10@xmath10 erg / sec .
at this point the mcb component ( f@xmath15 ) decreased by nearly a factor of 3 , the hard component ( f@xmath16 ) ( believed due to tc in an optically thin region ) increased by about a factor of 2 and the scb ( f@xmath17 ) remained almost constant or increased slightly .
the total x - ray luminosity tracks the total mass accretion in the system since the majority of the gravitational energy is released in the x - ray energy band .
therefore during this transition the change in total mass accretion rate is very small .
this is also indicated by the near constant scb component during the spectral state transition .
the sudden decrease of the mcb component indicates the mass accretion rate through the inner edge of the disk decreased by about a factor of 3 .
these material must be channeled to the region producing the observered hard x - ray component which increased simultaneously .
a successful model has to account for several major phenomena .
these are : _ a ) _ the soft x - ray component lags behind the hard component during the flares or onsets of outbursts ( hard precursors are common among soft x - rat transients or x - ray novae ) ; b ) no significant luminosity changes occur during state transitions ( the spectral state transitions are usually shown as soft / hard component anti - correlations . after removing the trend of long term light curve exponential decay during the gs 1124 - 68 spectral state transition , the total luminosity changes are also not significant . ) ; and c ) the state transition time of the soft component is much shorter than that of the hard component ( probably only observed from cyg x-1 , since no other sources have been monitored continuously over the whole transition period in a broad energy band ) . since all three phenomena are observed from cyg x-1 , we will examine this system first . in the model
we show in figure 2 for cyg x-1 , phenomenon-_a _ can be explained as due to the time difference between the disk mass transfer and the direct wind capture by the bh .
when the mass flow rate carried by the stellar wind increases suddenly , the flow will reach the bow shock promptly , resulting in an increased hard x - ray flux , due to the increased comptonization factor , although the supply of the seed photons remains the same .
the mass accretion component transported through the optically thick disk will arrive later at the inner disk region , thus producing the delayed soft x - ray flux .
the delay time of around 15 days constrains the optically thick keplerian disk to fill only 1 - 10% of the roche lobe of the system ( zhang _ et al . _
phenomenon-_b _ suggests that the the total mass accretion rate remains nearly a constant while both soft and hard x - ray fluxes change significantly .
it is therefore reasonable to assume what really happened was the re - distribution of mass flow between the direct wind capture component and the optically thick disk component .
phenomenon - c is explained as due to the mass flow exchange between the two mass accretion components when an instability is transported between the outer and inner disk regions ( see zhang _ et al . _ 1997a for details ) .
although no significant wind accretion exists in lmxbs , many models will require the existence of an optically thin , sub - keplerian region , responsible for the hard x - ray production , although the origin and the nature of this region distinguishes these models . in this picture
, the optically thick disk nearly fills the roche lobe , since it is formed out of the roche lobe overflow , and the optically thin region surrounds the inner disk region nearly spherically .
the time lag is then the same time difference , but starting from the boundary of the optically thin region since its radius is expected to be smaller than the outer radius of the optically thick disk .
_ thus the time lag measures the size of the optically thin region in a lmxb . _ in gs 1124 - 68 , the lag time of @xmath113 days suggests the size of this region is between 10@xmath8 to 10@xmath18 cm , i.e , about 10 - 100 times larger than the inner disk radius for a stellar mass bh . during the cen x-4 outburst in 1979 ( bouchacourt _ et al . _ 1984 ) , a similar lag was also found , indicating a similar geometry in both systems .
a significant difference between cyg x-1 and these lmxbs , in terms of spectral state transitions , is that the hard state is the persistent and dominating state and occurs at a much higher luminosity level in cyg x-1 than the lmxbs .
this is , perhaps due to the relatively higher mass flow rates in the optically thin region originated from the direct stellar wind capture in our model of cyg x-1 , than that in lmxbs , where the optically thin region is probably formed out of the optically thick disk and is much less effective when the disk mass accretion rate is very high .
it appears that a characteristic high energy spectral component is formed from each component of a ns or bh xb .
based on the similarities and differences of these spectral formation process in ns and bh xbs , we discussed possible ways to distinguish between ns and bh xbs .
spectral state transitions in xbs are discussed for three systems , hmbh xb cyg x-1 , lmbh xb gs 1124 - 68 and lmns xb 4u 1608 - 522 .
geometries of these systems can be constrained from the observed spectral transition characteristics .
we benefitted greatly from the many stimulating discussions between the batse team members , especially alan harmon , bill paciesas and craig robinson at marshall space flight center and with drs .
ken ebisawa , lev titarchuk , sandip chakrabarti , wei cui , wan chen , jan van paradijs , didier barret , ron remillard , bob hjellming and ramesh narayan .
barret , d._et al . _ 1991 , 379 , 21 barret , d. , mcclintock , j.e . ,
grindlay , j.e . , 1996 , 372 , in press bouchacourt , p. _ et al . _ 1984 ,
apjl 285 , 67 - 70 chakrabarti , s.k . & titarchuk , l. g. 1995 , , 455 , 623 ebisawa , k. , titarchuk , l. & chakrabarti , s. k. , 1996 , , 48 , 1 ebisawa , k. _ et al . _ 1994 , 46 , 375 haardt , f. , & marashi , l. 1993 , 413 , 507 mitsuda , k. _ et al
1984 , 36 , 741 mitsuda , k. _ et al . _ 1989 , 41 , p. 97
narayan , r. 1996 , 462(1 ) , 136 - 41 shakura , n. i. & sunyaev , r. a. 1973 , , 24 , 337 shapiro , s.l . & lightman , a.p .
1976 204 , 555 sunyaev , r.a .
, & titarchuk , l. , 1980 , a&a 86 , p. 121
tanaka , y. , & lewin , w. h. g. 1995 , in `` x - ray binaries '' , eds .
w. h. g. lewin , j. van paradijs , & e. p. j. van den heuvel titarchuk , l. 1996 , integral workshop , st .
malo , france van der klis , m. & van paradijs , j. 1994 , 281(1 ) , l17 - 20 van der klis , m. 1994 , 92 , 511 - 519 vargas , m. _ et al . _
1996 , in press white , n.e .
1993 , aip conf .
308 , 53 - 60 white , n.e . ,
swank , j.h . ,
holt , s.s .
1983 , 270 , 711 zhang , s.n .
_ et al . _ 1996a ,
in press zhang , s.n .
_ et al . _ 1996b , in press zhang , s.n
_ et al . _ 1997a , submitted zhang , s.n
_ et al . _ 1997b , in press zhang , s.n
_ et al . _ 1997c , | a variety of high energy ( @xmath01 kev ) spectra have been observed in recent years from black hole ( bh ) and neutron star ( ns ) x - ray binaries ( xb ) .
some common physical components exist between bhxbs and nsxbs , resulting in some high energy spectral features .
a common component between a bhxb and a weakly magnetized nsxb is the inner accretion disk region extending very close to the surface ( for a ns ) or the horizon ( for a bh ) .
the inner disk radiation can be described by a multi - color blackbody ( mcb ) spectral model .
the surface radiation of the ns can be approximated by a single color blackbody ( scb ) spectrum .
for a strongly magnetized nsxb , the high energy emission is from its magnetosphere , characterised by a thermal bremsstrahlung ( tb ) spectrum . in both bhxbs and weakly magnetized nsxbs ,
a hot electron cloud may exist , producing the hard x - ray power law ( photon index -1.5 to -2.0 ) with thermal cutoff ( 50 - 200 kev ) .
it has been recently proposed that a converging flow may be formed near the horizon of a bh , producing a softer power law ( photon index about -2.5 ) without cutoff up to several hundred kev .
based on these concepts we also discuss possible ways to distinguish between bh and ns xbs .
finally we discuss briefly spectral state transitions in both bh and ns xbs .
s.n .
zhang ( [email protected] ) _ usra / marshall space flight center , es-84 , huntsville , al 35812 _ |
a basic problem in non - equilibrium statistical mechanics is the following @xcite : take a microscopic statistical mechanics model at sufficiently low temperature so that there are , say , two pure thermodynamics phases .
assume that such system evolves according to a microscopic dynamics defined via local evolution rules .
then , the goal is to derive macroscopic , deterministic equations which describes , on large space - time scales , the evolution of spatial boundaries separating the two coexisting thermodynamic phases . an example to keep in mind is the nearest - neighbor ising model on @xmath11 , @xmath12 . below the critical temperature , in the absence of an external magnetic field
, there are two translation invariant equilibrium gibbs measures ( the `` @xmath13 '' and the `` @xmath10 '' thermodynamic phase ) .
one can then easily define a markov dynamics , the so - called glauber dynamics , where individual spins are flipped with rates chosen so that the gibbs measures are invariant and reversible . if at time zero a region of the space is occupied by the `` @xmath13 '' phase and the rest by the `` @xmath10 '' phase , we are interested in how the shape of these region will evolve with time .
if the dynamics does not conserve the order parameter ( e.g. the total magnetization for the ising model ) , it is well understood phenomenologically @xcite that a droplet of one phase immersed in the opposite phase will shrink in order to decrease its surface tension until it disappears in finite time ; also ( roughly speaking ) the normal velocity at a point of its boundary will be proportional to the local mean curvature .
based on this idea , one expects ( `` lifschitz law '' ) that , if the initial droplet has diameter @xmath14 , it will `` evaporate '' within a time of order @xmath8 ( as would be the case for a sphere evolving via mean curvature motion ) .
moreover , the droplet evolution should become deterministic and follow some version of a mean curvature flow in the `` diffusive limit '' where @xmath9 , space is rescaled by @xmath14 ( the initial droplet is then of size @xmath15 ) and time is accelerated by @xmath8
. the resulting large - scale deterministic evolution equation will in general be anisotropic ( i.e. the normal velocity will depend also on the local orientation of the droplet boundary ) when the microscopic model is defined on a lattice , as it is for the ising model .
the main difficulty in implementing this program is that there is no obvious way how to separate the `` fast modes '' related to relaxation inside the bulk of the pure phases from the `` slow modes '' , responsible for the @xmath8 time scaling , related to the interface motion .
such problem is absent in so - called `` effective interface models '' of ginzburg - landau @xmath16 type : for these models , under an assumption of strict convexity of the interaction , funaki and spohn @xcite derived the full mean - curvature motion in the diffusive scaling .
another case @xcite where mean - curvature motion is known to appear in the scaling limit are spin models with kac - type interactions ( the interaction range tends to infinity with the droplet size ) : in this case , however , the system is close to mean - field , the deterministic flow is isotropic and there is no sharp interface separating the phases .
results are much more incomplete for genuine lattice models : for instance , for the two - dimensional ising model at low but non - zero temperature @xmath17 , it is only known that a droplet of `` @xmath10 phase '' immersed in the `` @xmath13 phase '' will disappear in a time of order at most @xmath18 @xcite , to be compared with the expected @xmath8 scaling . in the present work ,
we study the two - dimensional nearest - neighbor ising model on the square lattice , at zero temperature . each spin variable @xmath19 is updated on average once per time unit : after the update , the spin takes the same sign as the majority of its @xmath20 neighbors , or the value @xmath21 with equal probabilities in case of a tie .
assume that the initial `` @xmath10 '' droplet , when the lattice spacing tends to zero , converges to a smooth enough domain @xmath0 of @xmath1 .
then , in the diffusive scaling limit the droplet boundary is expected to be given by a deterministic evolving curve @xmath22 .
such curve should move according to the following `` ( anisotropic ) curve - shortening flow '' : the normal velocity equals the local ( signed ) curvature , times a function @xmath23 with @xmath24 the local tangent .
the function @xmath25 is explicitly given , cf . .
this result was conjectured in @xcite by spohn , who gave some very reasonable supporting arguments , based on the local analysis of the dropled boundary evolution in terms of interacting particle systems . in @xcite ,
cerf and louhichi computed the `` drift at time @xmath26 '' of the droplet ( for the non - modified dynamics ) , but their result does not allow to get any information on the evolution for positive time @xmath27 . the full convergence to the curve shortening flow for initial _ convex _
droplets @xmath0 was recently obtained in @xcite .
the main result of the present work , theorem [ mainres ] , is a proof of spohn s conjecture , for smooth enough initial droplets @xmath0 , without any convexity assumption .
smoothness of the initial condition is required essentially so that the limit flow is unambiguously defined . as it was the case
also in @xcite for the convex initial condition , a preliminary but essential step before proving convergence of the stochastic evolution to the limit deterministic one is to show that the anisotropic curve - shortening flow does admit a global ( in time ) solution , and that it does not develop singularities before it shrinks to a point . for the isotropic case ( @xmath28 )
this was proven by grayson in a celebrated work @xcite . in the anisotropic case ,
grayson s result has been extended ( e.g. @xcite ) under the assumption that @xmath25 is at least @xmath29 .
the reason is very simple : the first step in the procedure is to write down the evolution equation satisfied by the curvature of @xmath22 , and in such equation a second derivative of @xmath25 appears . in our case ,
the anisotropy function @xmath25 is not even @xmath30 ( which reflects singularities of the zero - temperature surface tension of the ising model ) . to prove existence , uniqueness and regularity of the ( classical ) solutions of the curve - shortening flow ( cf .
theorem [ th : existence ] ) , we first regularize the function @xmath25 and then analyze the regularized equation following the ideas of @xcite .
of course , it is crucial to check that all the estimates we need are uniform in the regularization parameter , which is sent to zero in the end .
let us emphasize that the regularity estimates of @xcite are far from being quantitative in terms of the smoothness of the anisotropy function @xmath25 .
comparing our present result with that of @xcite , it is important to realize that dropping the convexity assumption is not at all a technical point .
first of all , various monotonicity arguments that were used in @xcite do not work here .
the basic reason is that such ideas crucially relied on the fact that , in the convex case , the normal velocity is always directed inward ( which is clearly false for non - convex droplets , at points where the curvature is negative ) .
secondly , proving existence and regularity of solution requires very different analytic and geometric arguments in the non - convex case with respect to the convex one ( there , we were able to use ideas from @xcite ) . at any rate ,
our previous result @xcite is important in section [ sec : mainres ] , where the evolution of the droplet boundary is controlled by locally comparing it with that of a suitable convex droplet .
let us mention some recent related works by one of the authors . in @xcite the issue of the evolution of a convex planar `` minus droplet '' in the presence of a positive magnetic field has been investigated . in this case the right time scaling is @xmath14 instead of @xmath8 for a droplet of size @xmath31 and the scaling limit is given by the anisotropic eikonal equation : the drift of the interface loses its dependence on the curvature .
generalizing such a result in higher dimensions is a very challenging problem , see e.g. @xcite for a non - rigorous attempt in this direction . in @xcite , the dynamical evolution of a half - droplet on a substrate that attracts the interface ( a situation that also corresponds to `` dynamical polymer pinning '' ) was studied , and the scaling limit was shown to be the solution of stefan - type equation where the motion of the point of contact between the droplet and the substrate depends on the local curvature .
we close this introduction by mentioning a couple of intriguing open problems .
first of all , one would like to know what are the finite-@xmath14 fluctuations of the droplet boundary around its limit shape @xmath22 , along the evolution .
secondly , it is natural to wonder what happens for the zero - temperature dynamics of the _ three- ( or higher- ) dimensional _ ising model .
recently , a weak version of the lifshitz law was proven for the three - dimensional ising model at zero temperature : the evaporation time of a `` @xmath10 '' droplet is of order @xmath8 , up to multiplicative logarithmic corrections @xcite .
an analogous upper bound was proven in higher dimensions @xcite .
however , it is still not clear ( even at a heuristic level ) what should be the precise macroscopic equation , analogous to , describing the droplet evolution in the diffusive limit .
given @xmath32 we consider the zero - temperature stochastic ising model on @xmath33 ( the square lattice with lattice spacing @xmath4 ) .
the state space is the set @xmath34 of spin configurations @xmath35 with @xmath19 .
the dynamics is a markov process @xmath36 , with @xmath37 .
each spin @xmath38 is updated with unit rate : when the update occurs , @xmath38 takes the value of the majority of its four neighbors , or takes values @xmath39 with equal probabilities if exactly two neighbors are @xmath40 and two neighbors are @xmath41 .
we consider a compact , simply connected subset @xmath42 ^ 2 $ ] whose boundary @xmath43 is a jordan curve of finite length .
the initial condition of the stochastic dynamics will be set to be `` @xmath10 '' inside @xmath44 and `` @xmath13 '' outside : @xmath45 we want to compute the scaling limit of the set of `` @xmath10 '' spins at positive times , when @xmath46 . in order to identify a set of `` @xmath10 '' spins as a subset of @xmath1 ,
let for @xmath47@xmath48 ^ 2\ ] ] be the square of side @xmath4 centered at @xmath49 and define @xmath50 which is the `` @xmath10 droplet '' at time @xmath51 for the dynamics .
our goal is to prove that , as @xmath9 , @xmath52 converges to the compact set @xmath53 whose boundary @xmath54 is the solution of the anisotropic curve shortening flow @xmath55 with initial condition @xmath56 .
this equation has to be read as follows .
the normal velocity at a point @xmath57 is given by the curvature @xmath58 at point @xmath59 times @xmath60 , with @xmath61 and @xmath62 the tangent angle to @xmath22 at @xmath59 .
the normal vector @xmath63 at point @xmath59 points inward and the curvature is positive ( resp .
negative ) at points of local convexity ( resp .
concavity ) of @xmath22 . since @xmath23 is not differentiable for @xmath24 multiple of @xmath64 , the existence of a solution for does not follow from the standard literature , that assumes @xmath25 to be at least @xmath29 ( see @xcite ) .
our first result is an existence , uniqueness and regularity theorem for the solution of .
define @xmath65 [ th : existence ] consider a domain @xmath66 ^ 2 $ ] whose boundary is a jordan curve @xmath67 of finite length , with curvature everywhere defined and @xmath68 as a function of the arc - length coordinate .
suppose moreover that @xmath67 has a finite number of inflection points .
+ there exists a unique solution @xmath69 of that is a jordan curve for @xmath70 and : 1 .
the area enclosed by @xmath22 is @xmath71 and @xmath22 shrinks to a point @xmath72 when @xmath73 .
2 . for every @xmath74 ,
the curvature function is equicontinuous on @xmath75 $ ] in the following sense : for every @xmath76 there exists @xmath77 such that if @xmath78 and @xmath79 with @xmath80 then @xmath81 .
our main result gives convergence of the stochastic droplet @xmath82 to the deterministic flow @xmath83 , that is the compact domain enclosed by @xmath22 ( for @xmath84 , we set by convention @xmath85 ) .
we introduce some notations . for @xmath86 let @xmath87 denote the ball of radius @xmath88 centered at @xmath89 and for any compact set @xmath90 , we define @xmath91 finally , we will say that an event holds with high probability ( w.h.p . ) if its probability tends to @xmath92 as @xmath14 tends to infinity .
[ mainres ] consider @xmath93 such that @xmath94 satisfies the assumptions of theorem [ th : existence ] .
let us consider the zero temperature stochastic ising model with initial condition .
then for any @xmath86 the following holds w.h.p . :
1 . for all @xmath95 ,
@xmath96 2 . for all @xmath97
, @xmath98 is empty .
we emphasize that the regularity estimates stated in theorem [ th : existence ] are not given just for the sake of completeness , but on the contrary are crucial in the proof of theorem [ mainres ] .
let us mention a few immediate generalizations of our result , and an interesting open problem .
a. _ more general initial condition_. instead of , let us assume only that the ( possibly random ) initial droplet @xmath99 converges w.h.p . in hausdorff distance to @xmath93 as @xmath9 .
then , still holds . just note that , for any given @xmath76 , w.h.p .
@xmath100 : then the claim follows from theorem [ mainres ] plus monotonicity of the dynamics , cf . section [ monoton ] . b. _ non - connected initial droplet_. if @xmath93 is not connected but each of its connected components @xmath101 verifies the assumptions of theorem [ th : esistenza ] , it is easy to see that theorem [ mainres ] applies to each of the @xmath58 components of the stochastic droplet @xmath102 .
essentially , the various components evolve independently .
c. _ non - simply connected initial droplet_. suppose that @xmath93 is compact , connected but non - simply connected ( say , an annulus ) . if each connected component @xmath103 of @xmath94 verifies the assumptions of theorem [ th : esistenza ] , define @xmath53 to be the domain with boundary @xmath104 .
then , it is not hard to see that theorem [ mainres ] still holds .
again , roughly speaking the @xmath58 macroscopic components of the boundary of @xmath105 evolve essentially independently and approach the deterministic evolution of the @xmath58 components of @xmath106 .
d. _ @xmath107 is not a simple curve_. if @xmath108 has self - intersections the situation is definitely more subtle . to fix ideas ,
consider the case of figure [ fig:8_1 ] .
note that @xmath93 can be seen either as the @xmath109 limit of a domain @xmath110 with jordan boundary and an @xmath111-narrow pinch or as the limit of two @xmath111-close simple domains @xmath112 . in this case
, we expect that the evolution of @xmath113 remains random in the @xmath9 limit .
more precisely , we expect that the ising droplet will follow with some probability @xmath59 the _ deterministic _ evolution @xmath114 and with probability @xmath115 the deterministic evolution @xmath116 $ ] . here , @xmath117 is the domain enclosed by the solution of with initial condition @xmath118 , and similarly for @xmath119 .
more importantly , we expect that the law of the limit evolution ( i.e. the probability @xmath59 ) depends crucially on the way the initial droplet @xmath99 microscopically approximates @xmath93 .
this will be considered in future work .
+ = 10 cm + = 10 cm
let us pretend for a moment that equation is replaced by @xmath120 where @xmath121 is a smooth ( at least @xmath29 ) function , that is strictly positive and @xmath122-periodic .
then , it is known from @xcite that does admit a solution @xmath22 that is a jordan curve . furthermore from @xcite the solution exists until a maximal time where it shrinks to a point .
the function @xmath25 in is instead only lipschitz , which is why we can not apply known results to our case .
call @xmath123 the arc - length coordinate on @xmath22 .
it is not important to establish what point of @xmath22 is assigned the coordinate @xmath124 , since we will always evaluate only derivatives or differences with respect to @xmath123 . in the following equations ( to ) ,
@xmath125 denotes derivation w.r.t .
arc - length at fixed time , while @xmath126 means derivation along the flux lines described by points @xmath127 who move with velocity at time @xmath51 given by @xmath128 times the normal vector @xmath63 to @xmath22 at point @xmath129 .
it is well known ( see ( * ? ? ?
* lemma @xmath130 ) in the isotropic case ) that the derivatives with respect to @xmath51 and @xmath123 do not commute , as motion affects the arc - length : in fact , we have @xmath131 we collect here a few useful formulas .
first of all , define for @xmath57 @xmath132 with @xmath62 the tangent angle at the point @xmath59 .
then , one has ( see for instance ( * ? ? ?
1 ) ; for the isotropic case @xmath133 , see @xcite ) @xmath134 with @xmath135 the length of @xmath136 and the area enclosed by it . remark also that @xmath137 ( this is a simple geometric fact that has nothing to do with the curve shortening flow ) .
let us call @xmath138 a family of @xmath68 regularizations of @xmath25 that are uniformly lipschitz ( this is possible because @xmath25 itself is @xmath92-lipschitz ) , have the same @xmath64-periodicity as @xmath25 and converge uniformly to @xmath25 when @xmath139 . from the previous section ,
we know that has a solution if @xmath140 is replaced by @xmath141 .
let @xmath142 be this solution and @xmath143 denote its curvature function .
the time when the curve shrinks to a point is given by ( cf . ) @xmath144 and the maximal curvature is bounded for times smaller than @xmath145 . moreover , up to time @xmath145 the curve is @xmath68 @xcite .
set for every @xmath146 @xmath147 note that @xmath148 for all @xmath149 .
indeed , if @xmath150 ( cf . ) , one has from , using @xmath151 and dropping the argument @xmath152 @xmath153 from , recalling that @xmath154 is uniformly bounded and calling @xmath155 the maximal value of @xmath156 along @xmath22 , we obtain @xmath157 and since @xmath158 is bounded away from zero uniformly in @xmath152 , we find that @xmath156 ( and therefore the curvature ) can not explode instantaneously .
[ th : esistenza ] let the initial condition @xmath67 be a jordan curve of finite length , whose curvature is everywhere defined and @xmath68 as a function of the arc - length coordinate .
suppose moreover that @xmath67 has a finite number of inflection points .
fix @xmath159 .
there exists a unique solution @xmath160 of that is a jordan curve for @xmath161 .
the area enclosed by @xmath22 is @xmath162 .
the curvature function is equicontinuous in the following sense : for every @xmath76 there exists @xmath163 such that if @xmath164 and @xmath79 with @xmath80 then @xmath81 .
further regularity properties of the limit flow and precise the connection with the one - dimensional heat equation are given in lemma [ analytic ] below .
note that , while @xmath145 in converges to @xmath17 ( cf . ) for @xmath165 , it is not guaranteed that @xmath166 : in principle , the curvature of the regularized solution could take larger and larger values ( when @xmath165 ) at some time @xmath167 .
the following result rules out this pathological behavior : [ th : gray ] one has @xmath168 and the curve @xmath22 shrinks to a point when @xmath169 .
clearly , theorems [ th : esistenza ] and [ th : gray ] imply theorem [ th : existence ] .
in this section we give some additional properties of the anisotropic curve shortening flow , in addition to those stated in theorem [ th : esistenza ] . these will be crucial for the proof of proposition [ th:35 ] .
let @xmath22 satisfy equation and consider a portion of @xmath22 that is the graph @xmath170 of a function in the coordinate frame @xmath171 obtained by rotating by @xmath172 anti - clockwise the usual cartesian frame .
the curvature is given by @xmath173 the evolution of @xmath174 , of the angle @xmath175 and of the `` curvature '' @xmath176 are given by the parabolic equations @xmath177
@xmath178 @xmath179 @xmath180 equation is obtained just by projecting equation in the chosen frame of coordinates . for and , one just needs to compute the derivatives with some patience ( is given also in ( * ? ? ?
* equation ( 1.2 ) ) with a different notation ) .
equation can be derived from by noticing that @xmath181 hence @xmath182 and a line of computation leads to .
[ unpointneuf ]
let @xmath22 be a solution of with @xmath121 smooth .
* for a given choice of cartesian coordinates @xmath171 , local minima for @xmath174 and @xmath24 strictly increase with time and local maxima strictly decrease with time . *
the total curvature of an arc connecting two isolated inflection points is decreasing with time , and the @xmath24-intervals of tangent directions of such arc ( angle span ) strictly nest with time . *
local minima of @xmath156 or @xmath183 that are not inflection points are strictly increasing with time .
in particular , isolated points with curvature zero that are not inflection points disappear instantaneously .
such points can arise only when two or more inflection points merge . *
if @xmath27 , @xmath22 does not contain flat pieces and the number of inflection points of the curve is non - increasing with time .
[ lemma : lourd1 ] point @xmath184 is a consequence of the maximum principle ( * ? ? ?
7 ) applied to equation and .
these equations are non - linear , but since we know a priori that @xmath22 is @xmath68 , one can treat e.g. the factor @xmath185 in as a smooth positive coefficient .
point @xmath186 is a consequence of @xmath184 , since inflection points are local maxima or minima of @xmath24 ( see also lemma 1.6 in @xcite ) .
point @xmath187 is obtained from and .
say for instance that one has a local minimum of @xmath188 that is non - negative .
then , omit the term @xmath189 that is locally positive and apply the maximum principle to the remaining equation .
point @xmath190 is proven in lemma 1.4 of @xcite and is based on a theorem by angenent @xcite that says that the set of zeros of a parabolic equation is finite at all positive times and its cardinality is non - increasing with time .
when the regularization parameter @xmath152 tends to zero , theorem [ th : esistenza ] says that holds , with @xmath121 replaced by @xmath25 , but the analog of equations to is not guaranteed to hold .
however , the following result gives extra information with respect to theorem [ th : esistenza ] : [ analytic ] let @xmath22 solve .
* in the cartesian frame @xmath171 with @xmath191 , along the portions of @xmath22 that are locally the graph of a @xmath92-lipshitz function @xmath170 , the evolution of @xmath174 is given by the one - dimensional heat equation @xmath192 ( which explains the title of the article ) .
* for any positive time , @xmath22 consists of a finite number of analytic arcs separated by points where the tangent angle belongs to @xmath193 .
* the total curvature of an arc connecting two isolated inflection points is decreasing with time , and the @xmath24-intervals of tangent directions of such arc strictly nest with time .
the number of inflection points is decreasing with time ( they can merge ) .
* flat portions of @xmath22 with zero curvature disappear instantaneously and can in principle be present only at a finite number of times @xmath194 .
* inflection points evolve continuously in time ( except possibly at the times @xmath194 of point @xmath190 ) .
* isolated points where @xmath195 , that are not inflection points , can be present only at a finite set of times and disappear instantaneously .
we emphasize that this proof assumes the existence and regularity results of theorem [ th : existence ] . for point @xmath184 ,
just look at and and observe that , for @xmath196 and @xmath197 , one has @xmath198
fix @xmath199 , consider any point @xmath200 where the tangent angle is not a multiple of @xmath64 and rotate the cartesian coordinate frame by a suitable multiple of @xmath201 so that locally the curve is the graph of a @xmath92-lipschitz function .
call @xmath202 the horizontal projection of @xmath59 in this frame .
there exists an open interval @xmath203 and @xmath204 such that the evolution of the curve is described by the solution @xmath170 of the heat equation , for @xmath205 and @xmath206 , with time - dependent boundary conditions at the endpoints of @xmath207 .
then ( * ? ?
* theorem 10.5.1 ) tells us that this solution is analytic in @xmath49 in the interior of the domain @xmath207 , for @xmath208 .
this guarantees that the curve is composed of analytic arcs whose evolution in a certain coordinate frame is given by the heat equation .
they are separated either by points or by flat segments where the tangent angle is a multiple of @xmath64 .
point @xmath186 is proven .
for point @xmath187 it is sufficient to prove that local minima of @xmath24 are increasing , and local maxima are decreasing ( this also implies that the number of inflection points can not increase ) .
we remark that in the neighborhood of an extremum of @xmath24 ( i.e. of an inflection point ) the curve can always be represented as the graph of a @xmath92-lipshitz function @xmath209 in the cartesian frame @xmath171 with @xmath191 ( this is the case even when the tangent angle is a multiple of @xmath64 ) .
the inflection point corresponds to a local extremum of @xmath210 .
as @xmath210 is the solution of the heat equation , the result follows .
if @xmath67 contains flat segments of slope not multiple of @xmath64 , then by points @xmath184-@xmath186 they disappear instantaneously and can not be created at later times .
flat segments with slope multiple of @xmath64 require a different argument .
if such a segment ( say , with slope @xmath26 ) is present at some time @xmath199 , then * either it is locally a maximum / minimum for the curve in the usual cartesian frame ( see fig . [
fig : partipiatte ] ( a ) ) and in that case it disappears immediately since the equation is strictly parabolic ; * or it is not ( see fig .
[ fig : partipiatte ] ( b ) ) , in which case its evolution in the cartesian frame rotated by @xmath211 is locally described by , so for @xmath51 just after @xmath199 the curve is analytic and can not contain flat segments . the only way to create a flat segment ( say with slope zero ) is to have an analytic arc , connecting two isolated points of zero slope ( `` poles '' ) that degenerates at some time @xmath194 to a horizontal segment , see fig .
[ fig : partipiatte ] ( c ) . at time @xmath194 , the number of poles in @xmath22 therefore decreases by @xmath212 .
however , poles can not be created ( again because is strictly parabolic ) , so this scenario can happen only a finite number of times ( @xmath67 has a finite number of poles because we assumed it has a finite number of inflection points ) .
this proves point @xmath190 .
= 10 cm for point @xmath213 , continuity of the curvature in space and time obtained in theorem [ th : esistenza ] makes it impossible for the inflection points to jump , except possibly at times @xmath194 ( at such times , the position of the inflection point on the flat segment is not well defined anyway ) . for point @xmath214 ,
let us call for simplicity `` zero - curvature points '' the points in question .
it is immediate ( from the heat equation ) that zero - curvature points with slope not multiple of @xmath64 that are present in @xmath67 disappear immediately and can not be created at later times .
a zero - curvature point with ( say ) slope zero is a pole .
if such a point existed for a time interval @xmath215 , the pole would not move in @xmath215 ( its velocity would be zero according to the curve - shortening flow ) , which is not possible since the equation is strictly parabolic . at later times , zero - curvature points of zero slope can appear only when an odd number of poles merge ( see fig . [
fig : k0 ] ) : this can happen only a finite number of times .
isolated poles can not become zero - curvature points , as can be easily obtained from the fact that , for the regularized evolution , local minima of @xmath58 where @xmath216 are non - decreasing , cf .
point @xmath187 of lemma [ lemma : lourd1 ] .
= 10 cm the scheme of the proof is the following .
first we prove that @xmath142 has a limit for every fixed @xmath51 , when @xmath152 tends to zero ( this is done in section [ sec : limflow ] ) .
then in sections [ sec : conclu1 ] and [ sec : conclu ] we prove that @xmath22 satisfies the regularity properties stated in theorem [ th : esistenza ] , and that it does solve equation . given a smooth simple curve @xmath136 we write @xmath217 for the curve obtained by shifting each point of @xmath136 by an amount @xmath218 in the ( inward ) normal direction at @xmath49 . set @xmath219 if @xmath220 , then the function @xmath221 defined by @xmath222 is a bijection that preserves the normal direction @xmath223 as a consequence , if @xmath44 is the interior of @xmath136 , then the interior of @xmath224 is @xmath225 .
identifying the points that are in bijection , we have that the curvatures of @xmath136 and @xmath224 are related by @xmath226 [ rem : bij ] [ ouille ] for the evolution @xmath142 set @xmath227 and @xmath228 to be the minimal value of the curvature radius @xmath229 for @xmath230 .
then @xmath231 furthermore @xmath232 is an increasing function .
what is meant precisely in the definition of @xmath233 is that @xmath234 are such that for all sufficiently small @xmath235 , the minimum @xmath236 with @xmath237 is positive and realized for @xmath238 . if the set in is empty as in the case where @xmath136 is convex then then we set @xmath239 to be infinite .
we drop all superscripts @xmath152 in this proof .
let us consider the evolution of @xmath240 for fixed @xmath51 when @xmath241 increases .
there are two ways for the curve to create self - intersections : either the maximal curvature explodes when @xmath242 and a loop is created when @xmath243 , or two distant arcs of the curves kiss when @xmath244 .
the first can occur when @xmath245 if @xmath246 is attained at a point of positive curvature and the second for @xmath247 if the chord linking the two points where @xmath248 is attained is enclosed in @xmath136 .
a similar argument holds for the evolution of @xmath240 for fixed @xmath51 when @xmath249 decreases ; hence our result . to see that @xmath248 is increasing , fix @xmath51 and consider two points @xmath250 realizing the minimum @xmath248 .
assume to fix ideas that the segment connecting them is vertical , enclosed in @xmath136 and that @xmath202 is above @xmath251 .
if @xmath252 are the curvatures at @xmath253 , then the definition of @xmath248 ( the distance being locally minimized by @xmath253 ) implies that @xmath254 let @xmath255 ( cf .
remark [ rem : bij ] , see fig/ [ fig : uff ] for a graphical construction ) and consider the points @xmath256 on @xmath257 with same horizontal coordinate as @xmath253 , that tend to @xmath253 when @xmath258 . for short times @xmath123 , locally around @xmath259 and @xmath260 , the curve @xmath257 is the graph of functions @xmath261 and @xmath262 . for @xmath124 one
has @xmath263 by definition of @xmath248 . to prove that @xmath248 is increasing , we just need to show that this inequality remains valid ( locally around @xmath264 ) for small positive @xmath123 .
= 13.5 cm from remark [ rem : bij ] one has for fixed @xmath51 @xmath265 with @xmath266 the curvature of @xmath267 .
the reason is that , if a point of @xmath257 has slope @xmath24 and curvature @xmath266 , the point in @xmath268 that is in bijection with it has the same slope @xmath24 and curvature @xmath269 , see with @xmath247 .
see also lemma [ lemma : inclu ] and in particular ( where one has to take @xmath270 ) for the more detailed proof of a similar result . projecting this equation vertically ( like what we did to get ) we obtain differential equations for @xmath271 and @xmath272 and ( keep in mind that the normal vector @xmath63 is pointing downward around @xmath259 and upward around @xmath260 ) we deduce that @xmath273 satisfies the parabolic equation @xmath274 where @xmath275 and @xmath276 are two smooth functions and @xmath277.\ ] ] here , @xmath278 denotes the curvature along the portions of the curve @xmath267 whose graph is @xmath279 .
if we can prove that @xmath280 , then from the ( weak ) maximum principle it follows that locally @xmath281 for positive time and thus the result is proved . to see that @xmath282 is positive , note that ( again thanks to ) , @xmath283 with @xmath284 the curvature on @xmath268 of the point in bijection with the point of @xmath257 with curvature @xmath285 . for @xmath124 and looking at point @xmath202 , the r.h.s .
of is larger than @xmath286 thanks to , since @xmath287 reduces to @xmath288 .
thanks to continuity of curvature in space and time ( the regularized evolution is @xmath68 ) , the positivity continues to hold for @xmath123 small , close to @xmath259 . a similar argument shows that the second term in @xmath282 is positive . we first show that , for @xmath289 ( cf .
definition ) , @xmath142 converges to a limit curve as @xmath290 : [ prop : cauchy ] take any sequence @xmath291 .
for any fixed @xmath292 , @xmath293 is a cauchy sequence for the hausdorff distance @xmath294 , uniformly for @xmath161 .
let us consider @xmath152 and @xmath295 close enough to zero so that the curvature of @xmath296 is bounded by @xmath292 uniformly on @xmath297 $ ] .
[ lemma : inclu ] let @xmath298 , @xmath299 be the compact sets enclosed by the curves @xmath300 . consider
@xmath301 and set @xmath302 for all @xmath161 , for all @xmath303 sufficiently small , the curves @xmath304 are simple curves and are the boundaries of @xmath305 ( cf .
remark [ rem : bij ] ) .
furthermore @xmath306 we will see that it is sufficient to choose @xmath307 sufficiently small so that they satisfy @xmath308 with @xmath309 the regularized version of @xmath25 . as a consequence of and of @xmath310 , @xmath311 and proposition [ prop : cauchy ] follows . to prove that @xmath304 are simple curves for all @xmath161 , it is sufficient to remark that by lemma [ ouille ] @xmath312 we prove the inclusions . set @xmath313 and observe that @xmath314 is the boundary of @xmath315 , see remark [ rem : bij ] .
let @xmath316 be the respective curvature functions of @xmath317 , that are related to @xmath318 via @xmath319 , cf . .
the normal velocity of a point on @xmath320 is given by @xmath321 with @xmath322 the @xmath51-derivative of @xmath323 .
similar considerations concerning @xmath324 implies that the curves @xmath317 are solution of the following equations : @xmath325 the curve @xmath326 is clearly enclosed by @xmath327 at initial time , since @xmath328 . to show that this inclusion holds for all times ,
observe that the normal velocity associated to the evolution of @xmath329 can be written @xmath330 if @xmath331 were zero , @xmath329 would just evolve with anisotropy function @xmath332 ( like @xmath327 ) , so it would be a standard fact that @xmath329 would be included in @xmath327 for all times .
this clearly remains true if we can show that @xmath333 , since the term @xmath334 gives a further inward push to @xmath327 .
as @xmath335 for all @xmath161 and @xmath336 , from we have @xmath337 . hence @xmath338\\
\ge 2k^2 \eta \exp(-10 k^2t_k)-2k\|a^{{\omega}}-a^{{\omega}'}\|_{\infty}\ge 0,\end{gathered}\ ] ] where in the second inequality we used that @xmath339 and that @xmath340 , and in the last one we used condition . with a similar reasoning , the proof that @xmath327 is enclosed by @xmath341 can be reduced to check that @xmath342 whose proof is analogous to that of @xmath333 . here
we prove that the curvature function is an equicontinous function of arc - length , uniformly in @xmath343 , up to time @xmath344 .
this is an essential step in establishing the regularity properties stated in theorem [ th : esistenza ] , point @xmath186 .
recall the definition of @xmath345 given just before .
[ prop : holder ] given @xmath346 there exists a constant @xmath347 such that , for every @xmath161 and @xmath348 , @xmath349 this @xmath350-hlder regularity is not optimal ( one can improve it to @xmath351-hlder regularity ) but is sufficient for our purposes . it is sufficient to show that ( omitting for simplicity the argument @xmath343 everywhere ) @xmath352 ( we will see that the dependence of @xmath353 on @xmath67 is only through the value @xmath354 ) . indeed , and cauchy - schwarz imply @xmath355 to show , write @xmath356 where the second term comes from the fact that the length of @xmath136 decreases with time according to ( see also the proof of proposition 2.7 in @xcite for the formula in the case @xmath357 ) and @xmath140 should be seen as @xmath358 , with @xmath359 the angle at the point of @xmath22 with arc - length coordinate @xmath123 . applying and with @xmath121 replaced by @xmath360 ,
the right - hand side of can be rewritten as @xmath361 the first term is bounded by @xmath362 . for the second one
, we observe that , as @xmath363 , @xmath364 altogether , recalling that @xmath141 is lipschitz uniformly in @xmath152 , the second term is bounded by @xmath365 $ ] .
the third term in looks more problematic .
however , by integration by parts it equals @xmath366 now we use that @xmath140 is bounded away from zero and that @xmath154 is bounded away from infinity ( uniformly in @xmath152 ) : is then upper bounded by @xmath367 ^ 2+c''(k)\psi(t)\le c''(k)\psi(t).\end{gathered}\ ] ] altogether we have obtained @xmath368\end{aligned}\ ] ] and , since @xmath369 ( @xmath67 was assumed to have a @xmath68 curvature ) we are done .
uniqueness is trivial , so we concentrate on existence .
proposition [ prop : cauchy ] implies that we can define @xmath370 and actually that the limit does not depend on the chosen sequence @xmath291 .
we have to prove that @xmath22 does solve equation and that it has the desired regularity properties .
the first step is : [ th : continua ] fix @xmath159 .
a. given a sequence of points @xmath371 , for @xmath372 , that converges to a point @xmath57 , the curvature and the tangent angle of @xmath142 at @xmath373 converge to the curvature and tangent angle of @xmath22 at @xmath59 .
b. the curvature function @xmath58 and the angle function @xmath24 of the family @xmath160 are equicontinuous in the sense that for @xmath76 there exists @xmath374 such that if @xmath375 with @xmath376 , @xmath377 then @xmath81 and @xmath378 . _
( i ) _ the angle @xmath379 at @xmath373 converges when @xmath165 to a limit @xmath24 , as a simple consequence of the convergence in hausdorff distance of @xmath142 to @xmath22 , plus the fact that the curvature of @xmath142 is bounded by @xmath292 .
assume to fix ideas ( and without loss of generality ) that @xmath380 $ ] ( if this is not the case , the cartesian coordinate frame below has to be rotated by a multiple of @xmath64 ) .
then , using proposition [ prop : cauchy ] : 1 .
there exists @xmath381 such that for @xmath152 small enough @xmath142 is locally the graph of a function @xmath382 in the usual cartesian coordinate frame , for @xmath49 in an interval @xmath383 $ ] , where @xmath202 is the horizontal coordinate of @xmath59 ; 2 .
the same holds for the limit curve @xmath22 and the function @xmath384 converges to @xmath385 uniformly on @xmath207 ; more than that , @xmath386 where the right - hand side does not depend on @xmath51 as long as @xmath387 , where @xmath388 tends to zero when @xmath165 ; 3 .
the @xmath49-derivative of @xmath389 is bounded by @xmath212 on @xmath207 , for @xmath152 small enough , where the value @xmath212 is chosen simply because @xmath390 . from ,
the curvature of @xmath142 at the point with horizontal coordinate @xmath205 is @xmath391 ^ 2)^{3/2}}\end{aligned}\ ] ] from which we infer that the second derivative of @xmath392 is uniformly bounded .
then , the ascoli - arzel theorem implies that @xmath393 converges to @xmath394 uniformly on @xmath207 , and in particular that @xmath24 is the tangent angle of @xmath22 at @xmath59 .
similarly , since @xmath318 is uniformly continuous w.r.t .
arc - length ( proposition [ prop : holder ] ) , @xmath318 converges uniformly and the limit ( that is a continuous function ) is the curvature function of @xmath22 .
let us emphasize that @xmath395 i.e. the estimates are uniform in @xmath161 , otherwise one would easily find a contradiction with .
_ ( ii ) _ take @xmath161 and @xmath57 . as in point
( i ) , assume without loss of generality that @xmath396 $ ] , so that the curve is locally the graph of a function @xmath170 with @xmath49 in some interval @xmath207 of width depending only on @xmath292 .
as long as @xmath397 we know that the evolution is smooth , in particular for @xmath398 very close to @xmath51 the curve @xmath399 is still the graph of a function in @xmath207 and ( see below for a bit more of detail ) @xmath400 for some @xmath401 that may diverge as @xmath165 .
then , for @xmath398 very close to @xmath51 @xmath402 the right - hand side clearly tends to zero with @xmath403 ( choose a sequence @xmath404 tending to zero . if @xmath405 does not diverge we are done .
otherwise , compute the right - hand side for @xmath406 with the largest value of @xmath407 such that @xmath408 ) . a similar argument
gives that @xmath409 uniformly for @xmath164 . to conclude the proof of point ( ii ) , take @xmath410 close to each other , with @xmath411 .
just write @xmath412 with @xmath413 the point on @xmath414 close to @xmath59 with the same horizontal coordinate as @xmath59 .
we have just proven that the first term in the right - hand side is @xmath415 as @xmath416 goes to @xmath26 ; as for the second one , it vanishes when @xmath417 , since we have shown in the proof of point ( i ) ( using proposition [ prop : holder ] ) that the curvature function of @xmath414 is uniformly continuous w.r.t .
arc - length .
similarly one proves the continuity statement for the angle function @xmath24 .
it remains to prove . from (
* theorem 3.1 ) , we know that , since the regularized anisotropy function @xmath309 is @xmath68 , the curve @xmath142 is also @xmath68 at all times @xmath418 ( cf . ) and thus follows from and . back to the proof of theorem [ th : esistenza ] , we still have to prove that the limit flow solves . as in the proof of proposition [ th : continua ] , let us concentrate on a portion of the curve that is locally described by the graph of a function @xmath419 .
we have from ( with @xmath420 ) , for @xmath421 , @xmath422 with @xmath318 as in and @xmath423 . from the ( uniform ) convergence of @xmath424 shown above , plus
the uniform bound @xmath335 that allows to apply dominated convergence , we have for the limit curve @xmath22 @xmath425 continuity of @xmath58 and @xmath24 with respect to time ( proposition [ th : continua ] , point _
( ii ) _ ) allows to deduce that @xmath426 hence @xmath22 does solve .
finally , from and the uniform convergence of @xmath309 to @xmath25 , we see that the area of @xmath22 is @xmath427
. the basic ingredients of the proof are the following two results , that say that when the curvature diverges , it necessarily does so on arcs of total curvature at least @xmath122 .
this says that the curve can not develop `` corners '' .
[ th:21 ] let @xmath428 be as in .
for any given @xmath159 there exists @xmath429 and an arc of @xmath22 of total curvature at least @xmath122 on which the curvature has a constant sign and is larger than @xmath292 in absolute value .
[ th:35 ] there exists a constant @xmath430 ( which depends on @xmath67 ) such that , for @xmath431 , if the curvature at @xmath432 is larger than @xmath430 in absolute value then @xmath59 belongs to an arc on which the curvature has constant sign , and whose total curvature is at least @xmath122 .
propositions [ th:21 ] and [ th:35 ] are the analog of theorem 2.1 and lemma 3.5 of @xcite ; however , due to anisotropy ( @xmath433 in our case ) and to the need to regularize @xmath25 to @xmath309 while obtaining bounds that are uniform in @xmath152 , the proofs require many non - trivial modifications .
this is done in detail in sections [ sub:21 ] and [ sub:35 ] below .
given these two propositions , the proof of theorem [ th : gray ] becomes essentially identical to the proof of the main theorem of @xcite ( that says that under isotropic curve shortening flow ( @xmath434 ) jordan curves do not develop singularities before shrinking to a point ) .
let us just remind the general strategy ( giving details would amount to repeating the argument of @xcite ) .
let @xmath435 be the supremum of the angles @xmath24 such that there exist sequences @xmath436 , @xmath437 , @xmath438 and arcs @xmath439 satisfying : * the diameter of @xmath440 is at most @xmath441 ; * the curvature @xmath58 has constant sign on @xmath440 and is pointwise larger than @xmath430 in absolute value , with @xmath430 the constant of proposition [ th:35 ] ; * the total curvature @xmath442 is @xmath443 .
proposition [ th:21 ] implies directly that @xmath444 ( observe that a non - self - intersecting arc of curvature pointwise larger than @xmath292 has diameter @xmath445 ) .
one then excludes the following two possibilities : 1 .
@xmath446 and @xmath22 does not shrink to a point as @xmath447 ; 2 .
@xmath448 .
case 1 can be ruled out following the argument of ( * ? ? ?
* theorem 4.1 ) and case 2 is dealt with exactly like in ( * ? ? ?
* section 5 ) . more precisely : * whenever the author of @xcite invokes his theorem 2.1 ( resp .
lemma 3.5 ) , one should apply our proposition [ th:21 ] ( resp .
proposition [ th:35 ] ) instead ; * lemma 3.2 of @xcite ( the `` @xmath449-whisker lemma '' ) holds also in our case ( with the same proof ) , since it is based only on the maximum principle and on the fact that locally @xmath22 evolves according to a strictly parabolic equation .
the only remaining possibility is that @xmath22 _ does shrink to a point _
when @xmath450 .
since the area of @xmath53 is @xmath162 up to @xmath428 ( theorem [ th : esistenza ] ) , we obtain that @xmath168 , that is the claim of theorem [ th : gray ] . as a side remark , for the _ isotropic _
curve shortening flow grayson @xcite proves also that @xmath22 becomes convex at a time strictly smaller than the time when it shrinks to a point ( so that @xmath451 ) , and becomes asymptotically a circle ( of radius going to zero ) .
recall the definition @xmath150 ; in this section , for lightness of notation , we drop the argument @xmath343 in @xmath452 , etc .
we let @xmath453 ( resp .
@xmath454 ) denote the maximum ( resp .
minimum ) of @xmath188 along @xmath22 .
let us first prove a weaker result , that is , that the curvature explodes on an arc of total curvature larger than @xmath64 .
[ ptit21 ] suppose that @xmath156 is uniformly bounded by @xmath292 at time zero and that @xmath455 ( respectively , @xmath456 ) for some @xmath27 .
we recall that @xmath457 is the length of @xmath136
. then there exists @xmath458 and a sub - arc of @xmath459 with total curvature at least @xmath64 on which @xmath460 ( respectively , @xmath461 ) .
we will consider only the case @xmath462 , the proof of the other statement being essentially identical .
we are going to rely on the following lemma , whose proof is given at the end of this section : [ zenob ] let us call @xmath463 the subset of @xmath22 on which @xmath460 : it is a union of arcs of positive curvature and we assume that each @xmath464 has an angle span @xmath465 $ ] ( the angle span can be larger than @xmath466 ) .
the two following statements hold : * if at time @xmath51 one has @xmath467 then @xmath468 * if at time @xmath51 one has @xmath469 then @xmath470 suppose that @xmath471 and , by contradiction , that for all @xmath458 there are no sub - arcs of length at least @xmath472 on which @xmath460 .
integration of @xmath184 of lemma [ zenob ] on @xmath473 $ ] gives @xmath474 here we have used the fact that @xmath475 since at time zero @xmath188 is dominated by @xmath292 by assumption .
moreover from the assumption @xmath462 and from one has @xmath476^{-1/2}\ ] ] for any @xmath458 .
hence when @xmath477 , we have @xmath478 ( i.e. @xmath479 ) and thus @xmath480^{-1}{\mathrm{d}}u\\ = \frac{{a_{\mathrm{min}}}}{8}\log\left(1+\frac{a^2}{8k^2}\right)>\frac{{a_{\mathrm{min}}}}{4}\log ( a/(3k))=2k \mathcal l(\gamma(0))\end{gathered}\ ] ] which contradicts . [ gro21 ] suppose that @xmath156 is uniformly bounded by @xmath292 at time zero and that @xmath481 ( resp .
@xmath482 ) , where @xmath483 and @xmath121 is as in proposition [ ptit21 ] . then there exists @xmath484 and a sub - arc of @xmath459 with curvature at least @xmath122 on which @xmath460 ( resp .
@xmath461 ) . assume to fix ideas that @xmath485 . from proposition [ ptit21 ] ,
there exists a time @xmath486 such that @xmath487 if one assumes that until time @xmath488 there is no arc with curvature at least @xmath489 on which @xmath460 , from lemma [ zenob ] @xmath186 , we have ( using also ) @xmath490 which contradicts . when @xmath51 approaches @xmath428 , the maximum of @xmath183 and therefore the maximum of @xmath156 diverges ( by definition of @xmath428 ) .
therefore , for any @xmath292 there is @xmath289 such that either @xmath491 or @xmath482 and proposition [ gro21 ] allows to conclude .
one has @xmath492 as on the ( moving ) boundary of @xmath464 one has @xmath493 . we know from ( * ? ? ?
* lemma 2.1 ) that , on arcs where curvature has a constant sign , @xmath494 then the left - hand side of is equal to @xmath495.\ ] ] the last term is negative as @xmath496 , @xmath497 by assumption .
we have to control the integral . for this
we use wirtinger s inequality ( which can be obtained by a fourier decomposition of @xmath353 on a base of eigenfunction of the laplacian ) for @xmath498 for each arc : [ lem : wirt ] let @xmath353 be a @xmath30 function on an interval @xmath499 $ ] . if @xmath500 with @xmath501 , then @xmath502 under assumption @xmath186 we thus have for all @xmath503 @xmath504 therefore , @xmath505 using @xmath506 and @xmath151 in the second inequality and in the last step . under assumption
@xmath184 we do the same thing , except for the arc where the maximal curvature is attained if it is larger than @xmath292 . on this arc
( call it @xmath507 ) one has , using the fact that its total curvature is smaller than @xmath64 and wirtinger s inequality , @xmath508 the last step just uses the fact that if @xmath509 , the angle span of @xmath507 , satisfies @xmath510 , then @xmath511 ^ 2\ ] ] ( apply cauchy - schwartz for the first inequality ) . to fix ideas
we will assume that the curvature at the point @xmath59 mentioned in proposition [ th:35 ] is positive .
decompose @xmath22 into _ minimal arcs _ , i.e. arcs where the curvature has constant sign , and which are delimited by inflection points that evolve continuously ( see again lemma [ analytic ] ) .
we want to find @xmath430 such that the curvature on minimal arcs of total curvature smaller than @xmath122 is necessarily bounded by @xmath430 . by choosing @xmath430 large
, we can restrict ourselves to an arbitrarily small time neighborhood of @xmath428 .
recall from lemma [ analytic ] that the number of inflection points of @xmath22 is decreasing with time and that points of zero curvature that are not inflection points can be present only for a finite set of times .
hence we can , without loss of generality , consider a time interval @xmath512 where the number of inflection points is constant while points of the latter type are absent .
note that on the interval @xmath512 , by point @xmath187 of lemma [ analytic ] , the total curvature of minimal arcs is strictly decreasing .
as the number of such arcs is finite , we can suppose ( one might need to take @xmath111 smaller still ) that after time @xmath513 , on all such arcs , the total curvature is smaller than @xmath514 . we emphasize that @xmath515 depends only ( but in a very implicitly way ) on the initial condition @xmath67 . in the rest of the proof
one shows that the curvature remains bounded on each one of these arcs .
we look at one of them , that we call @xmath516 .
let us call @xmath517 , @xmath518 the two minimal arcs of negative curvature that are connected to the endpoints of @xmath516 .
it may happen that @xmath519 , when @xmath22 has only two inflection points . by definition of @xmath428 ,
there exists a @xmath292 such that @xmath520 .
hence we can assume that at time @xmath513 the curvature of @xmath136 and of all the curves @xmath521 for @xmath152 sufficiently small are bounded above by @xmath292 . for @xmath522 fixed , for @xmath523 $ ] ,
when @xmath152 is small enough , we want to approximate @xmath516 by an arc @xmath524 of @xmath142 .
what we do first is finding minimal arcs @xmath525 and @xmath526 of @xmath142 , of negative curvature , approximating @xmath517 and @xmath518 . at the midpoint @xmath527 of @xmath528 ,
the curvature is negative and bounded away from zero in @xmath529 $ ] .
hence , from proposition [ th : continua ] , if @xmath152 is sufficiently small , all the points of @xmath142 that lie in the vicinity of @xmath527 have negative curvature , hence they belong to a common minimal arc @xmath525 which has negative curvature .
a similar construction gives @xmath530 .
there are two arcs that connect @xmath525 and @xmath530 : we call @xmath524 the one whose distance from @xmath516 as @xmath165 tends to zero .
note that this is not necessarily a minimal arc .
we claim however that @xmath524 converges in hausdorff distance to @xmath516 as @xmath165 , uniformly for @xmath523 $ ] .
indeed by proposition [ th : continua ] its endpoints @xmath531 , on which curvature is zero , must converge to points of zero curvature and the only possible options are the inflection points @xmath532 that delimit @xmath516 .
= 6 cm fix @xmath522 . * for @xmath152 small enough , the integral of @xmath533 along @xmath534 is bounded by @xmath535 for @xmath523 $ ] .
* for all @xmath536 the curvature on @xmath524 is uniformly smaller than @xmath537 where @xmath538 is the number of changes of sign of the curvature along @xmath67 .
[ lem : dpr ] for point @xmath184 , remark that we have @xmath539 where @xmath540 .
we first show that the second integral in the r.h.s . , i.e. the total curvature @xmath524 restricted to sub - arcs where the curvature is negative , is smaller than @xmath541 . as @xmath524 converges to @xmath516 and
since curvature also converges , the total length of arcs with negative curvature in @xmath524 is vanishing . the curvature is bounded ( by @xmath542 ) and this is sufficient to conclude .
the first integral is the angle between the tangents at the endpoints of @xmath524 .
because the curvature is bounded ( by @xmath542 ) the convergence of @xmath524 to @xmath516 implies that the tangent directions at the extremities also converge .
hence @xmath543 converges to the total curvature of @xmath516 which is smaller than @xmath514 .
therefore , the first integral is smaller than @xmath544 for @xmath152 small enough . for point @xmath186
, we use the ideas from ( * ? ? ?
* lemma 3.7 ) .
we observe that @xmath524 is in general composed of adjacent minimal arcs @xmath545 for some @xmath546 , with curvatures of alternating sign and @xmath547 having positive curvature .
the number @xmath548 is decreasing in time , as arcs disappear when two inflection points merge , and is in any case upper bounded by @xmath549 , the number of positive curvature arcs on @xmath67 .
[ lem37 ] let us consider an arc @xmath550 where the curvature is non - negative and whose endpoints are inflection points .
assume that , at some time @xmath123 , * @xmath188 is bounded by @xmath551 on @xmath552 ; * the total curvature of @xmath552 is bounded by @xmath535 .
then until the first time @xmath488 when an inflection point at one of the extremities of @xmath553 disappears , @xmath554 on @xmath553 .
we apply lemma [ lem37 ] for each arc @xmath555 of positive curvature , starting at time @xmath556 , and iterate it when two such arcs merge .
since the number of such mergings can not exceed @xmath549 , we obtain that @xmath345 is bounded above by @xmath557 .
recall that @xmath111 is positive and depends only on the initial condition @xmath67 .
we conclude the proof of proposition [ th:35 ] by noting that , when @xmath152 tends to zero , the curvature on @xmath558 is well approximated by the one of @xmath524 ( proposition [ th : continua ] ) , and that @xmath542 in lemma [ lem : dpr ] is arbitrary .
therefore , the curvature on @xmath516 is pointwise bounded by say @xmath559 for @xmath560 and we are done ; we recall that @xmath515 is a positive constant that depends ( very implicitly ) on @xmath67 .
[ proof of lemma [ lem37 ] ] recall that an arc of positive curvature can be parametrized by the tangent direction @xmath24 and ( cf . ) that @xmath561 satisfies ( cf .
* lemma 2.1 ) ) @xmath562 with no loss of generality we can consider that the initial angle span @xmath207 of @xmath563 ( at time @xmath123 ) is included in @xmath564 $ ] and from lemma [ unpointneuf ] ( ii ) we know that this remains true up to @xmath488 . then we remark that initially for all @xmath24 in @xmath207 @xmath565 and that the r.h.s .
is a stationary solution of .
this , via the maximum principle , implies that @xmath566 for all @xmath567 .
we will use at various places ( often implicitly ) the well - known monotonicity or attractiveness of the dynamics , that we can formulate as follows . consider the dynamics in a domain @xmath568 , with boundary conditions @xmath569 on @xmath570 . in general , @xmath569 can be time - dependent , @xmath571 .
let @xmath572 denote the configuration at time @xmath51 , when the initial condition is @xmath88 and the boundary conditions are @xmath569 . also introduce in the space of spin configurations the partial order @xmath573 where @xmath574 if @xmath575 for every @xmath49 .
then , it is possible to define a global coupling @xmath576 such that , if @xmath577 and @xmath578 one has @xmath579 with @xmath576-probability @xmath92 . here ,
@xmath578 means @xmath580 for every @xmath581 .
the way to prove the result is slightly indirect .
we first show that if the initial droplet includes a strict neighborhood of @xmath44 then @xmath53 is w.h.p . a lower bound for the droplet @xmath105 ( we also prove an analogous upper bound ) . [ controlinterieurexterieur ] let @xmath93 be a compact set such that @xmath94 satisfies the assumptions of theorem [ th : existence ] . 1 .
if for some @xmath86 the initial condition of the stochastic dynamics satisfies @xmath582 then , w.h.p . ,
@xmath583 2 .
if instead @xmath584 then , w.h.p . , @xmath585 _ lower bound_. we prove that @xmath586 for @xmath587 , where @xmath588 is the time when @xmath589 becomes empty .
take @xmath76 and let @xmath590 be the deterministic flow started from initial condition @xmath591 and @xmath592 .
we assume @xmath388 is small enough ( cf . remark [ rem : bij ] ) so that the boundary of @xmath591 is a simple curve and satisfies the assumptions of theorem [ th : existence ] .
proposition [ controlinterieurexterieur ] gives @xmath593 for all times up to @xmath594 , which is larger than @xmath588 for @xmath388 small enough .
the domain @xmath595 is included in @xmath53 for all times .
we want to show that @xmath596 .
let us call @xmath597 with @xmath598 the distance from @xmath59 to @xmath136 .
assume for definiteness that there is a unique pair of points ( @xmath599 ) that is exactly at distance @xmath600 ( the general case is analogous ) : then necessarily the slopes @xmath601 of @xmath602 at these points are equal and the segment @xmath603 $ ] is normal to both @xmath329 and @xmath22 .
moreover , since @xmath604 are smooth curves , we can locally approximate them by arcs of circles of radii @xmath605 , @xmath606 respectively , with @xmath607 the curvature of @xmath22 at @xmath608 and @xmath609 the curvature of @xmath329 at @xmath610 . the fact that @xmath608 realizes the infimum of @xmath611 for @xmath612 ranging on @xmath613 implies that @xmath614 . on the other hand ,
the fact that @xmath610 maximizes @xmath615 for @xmath59 ranging on @xmath329 implies @xmath616 both inequalities are easy to check if @xmath617 are replaced by arcs of circles . a look at then shows that , as long as @xmath618 , @xmath619 where @xmath620 is the maximal curvature of @xmath22 up to time @xmath588 .
observe that @xmath621 .
then , choosing @xmath388 sufficiently small ( as a function of @xmath88 ) , ensures that @xmath600 remains smaller than @xmath622 up to @xmath588 . as a consequence , @xmath623 .
the inclusion @xmath624 is therefore proven up to time @xmath588 , as we wished .
_ upper bound_. let @xmath625 be the deterministic flow started from initial condition @xmath626 and fix @xmath627 small .
it follows from proposition [ controlinterieurexterieur ] that , until time @xmath628 , @xmath629 by the same argument that showed that @xmath630 we have that @xmath631 until time @xmath632 , provided @xmath388 is small enough .
this proves the upper bound up to time @xmath633 .
then we notice that , if @xmath627 has been chosen small enough , @xmath634 is included in @xmath635 , a ball of radius @xmath636 centered at @xmath637 , the point to which @xmath22 shrinks as @xmath169 . from (
* theorem 2.2 ) we know that , w.h.p .
, an initial droplet contained in @xmath635 disappears within time @xmath638 , and is included at all times in @xmath639 . on the other hand , always for @xmath627 small , @xmath640 for all @xmath641 .
this concludes the proof of both claims . to prove proposition [ controlinterieurexterieur ]
one needs two ingredients .
the first says essentially that in the diffusive scaling the speed of evolution of the random droplet boundary is finite : in a time @xmath642 , it moves at most a distance @xmath643 .
[ chips ] suppose that @xmath93 satisfies the assumptions of theorem [ th : existence ] , so that in particular @xmath644 ( cf
. definition ) , and consider the dynamics starting from initial condition .
there exists @xmath645 and @xmath646 such that for any @xmath647 , with high probability @xmath648.\ ] ] the constants @xmath649 and @xmath650 can be chosen to be decreasing in @xmath549 .
the second ingredient is a control of the droplet after a time small on the diffusive scale @xmath8 .
[ infinit ] let @xmath93 satisfy the assumptions of theorem [ th : existence ] and fix @xmath651 .
there exists @xmath652 and , for @xmath653 , there exists @xmath654 such that for all for @xmath655 , for all integer @xmath656 with @xmath657 , 1 . if the initial condition ( which might be random ) satisfies w.h.p .
@xmath658then w.h.p . @xmath659 2 .
if the initial condition satisfies w.h.p .
@xmath660 then w.h.p .
@xmath661 recall that @xmath662 is the deterministic domain @xmath93 at time @xmath663 , expanded by @xmath664 .
remark that we can assume that @xmath88 is small : indeed , if the claim of proposition [ controlinterieurexterieur ] holds for small values of @xmath88 , then it clearly holds also for larger values .
_ claim ( 1)_. choose @xmath665 , @xmath666 and @xmath111 smaller than @xmath667 , with @xmath668 and @xmath669 from proposition [ infinit ] . iterating the first claim of proposition [ infinit ] , we obtain w.h.p . for all @xmath656 such that @xmath670 @xmath671 this is the desired statement , but only at the discrete set of times @xmath672 . then , in order to also have an inclusion bound in the time intervals @xmath673 $ ] , we use proposition [ chips ] .
first of all , define @xmath674 ( recall ) .
one has @xmath675 .
if @xmath676 were zero , this would follow immediately from lemma [ ouille ] , since the inverse maximal curvature @xmath246 is bounded away from zero up to time @xmath677 .
for @xmath678 small ( i.e. @xmath88 small ) , instead , @xmath675 follows from remark [ rem : bij ] and lemma [ ouille ] , which relate explicitly the curvature function of @xmath679 with that of @xmath680 for small @xmath49 .
assume then that @xmath111 was chosen smaller than @xmath681 , with @xmath682 as in proposition [ chips ] .
starting at time @xmath663 with an initial condition `` @xmath10 '' in @xmath683 and `` @xmath13 '' outside , we get that w.h.p .
@xmath684 for all @xmath685 $ ] , and therefore ( repeating the argument for all values of @xmath656 ) for all @xmath686 .
the set on the right hand side contains @xmath687 , provided that @xmath688 ( we used that @xmath689 ) . in turn , @xmath690 for all @xmath686 if @xmath111 is sufficiently small .
_ claim ( 2)_. this is proven analogously , taking @xmath691 .
we are going to use a simple consequence of ( * ? ? ?
* theorem 2.2 ) , that gives the convergence of the stochastic evolution @xmath692 to the curve - shortening flow in the case of a smooth convex initial droplet .
[ ptichips ] set @xmath693 , @xmath694 and @xmath76 .
consider the zero - temperature stochastic ising model starting from an initial condition which satisfies @xmath695 then w.h.p .
for all @xmath696 $ ] @xmath697 where @xmath698 denotes the open ball of center @xmath49 and radius @xmath699 if @xmath693 and the empty set otherwise .
according to ( * ? ? ?
* theorem 2.2 ) , for any positive @xmath88 , w.h.p .
for all @xmath700 @xmath701 where @xmath702 is the solution of with initial condition @xmath703 .
the boundary of the ball @xmath704 is instead solution of the _ isotropic _ curve - shortening flow @xmath705 the domain @xmath706 is convex at all times @xcite , so the velocity always points inward .
since @xmath23 is strictly smaller than @xmath707 , equation shrinks convex domains strictly faster than the original equation .
therefore , one can find @xmath88 ( depending on @xmath388 ) such that for all @xmath708 , @xmath709 we first prove the lower bound of .
define @xmath710 . recalling remark [ rem : bij ] ,
note that for every @xmath711 one has @xmath712
choose @xmath713 sufficiently small so that any @xmath714 satisfies @xmath715 in practice , think of @xmath716 .
set @xmath717 .
we have @xmath718 from , that is applicable since @xmath719 as @xmath720 is a compact set , one can extract a finite subset @xmath721 which satisfies @xmath722 for each @xmath723 we have @xmath724 , thus we can apply lemma [ ptichips ] and a union bound to obtain that w.h.p.for all @xmath725 @xmath726 the proof of the other inclusion of , @xmath727 , is similar since the roles of `` @xmath13 '' and `` @xmath10 '' spins are symmetric ; we just have to take care to work with compact sets while the set of `` @xmath13 '' spins is not compact . for @xmath714 ,
let @xmath728 be a rectangle , with sides parallel to the coordinate axes , that contains @xmath729
. from the definition of the dynamics , the spins outside @xmath730 remain `` @xmath13 '' at all times , since they always have four `` @xmath13 '' neighbors .
let @xmath731 be the closure of @xmath732 ( @xmath731 is compact ) .
we have to prove that all spins in @xmath731 stay
@xmath13 up to time @xmath111 .
similarly to one can find a finite subset @xmath733 which satisfies @xmath734 and then the conclusion follows from lemma [ ptichips ] applied to the dynamics where the roles of `` @xmath13 '' and `` @xmath10 '' are reversed .
we give full details only for the proof of the first claim , the proof of second one being very similar .
for notational convenience we prove the result for @xmath735 , and explain briefly in remark [ rem : genj ] why the proof remains valid for all @xmath656 such that @xmath670 .
given @xmath736 we set @xmath737 .
let @xmath738 denote the boundary of @xmath739 and @xmath740 the curvature function associated to it .
by monotonicity , it is sufficient to prove that if one starts with the initial condition `` @xmath10 '' in @xmath741 and `` @xmath13 '' outside , then for @xmath111 smaller than @xmath742 , w.h.p . @xmath743
the parameter @xmath676 will be chosen small in lemma [ voldechi ] .
we can use proposition [ chips ] which says that w.h.p .
@xmath744 is filled with `` @xmath10 '' spins at time @xmath111 .
hence we just have to prove that w.h.p . @xmath745
we denote by @xmath746 the compact closure of @xmath747 . roughly speaking
, @xmath746 is the collection of points in @xmath748 that are at distance @xmath749 from the boundary of @xmath750 . _
* sketch of the strategy . *
the main idea to prove is to control the motion of the boundary of @xmath751 around a given point @xmath752 via a comparison with @xmath753 : this is the stochastic droplet starting from a circular shape @xmath754 which is tangent to @xmath755 at @xmath49 and whose curvature is close to that of @xmath755 at @xmath49 . roughly speaking
, @xmath754 will sit in the interior of @xmath755 when the curvature at @xmath49 is positive and outside it when the curvature is negative ( some care will be needed when the curvature is close to zero ) .
given that the two initial curves @xmath755 , @xmath756 have the same slope and almost the same curvature at @xmath49 , at time zero and close to @xmath49 the boundaries of @xmath52 and @xmath753 feel almost the same drift .
our work will then consist in proving that , locally around @xmath49 and for small positive times , they do remain close . on the other hand , @xmath754 being convex , the evolution of @xmath757 in the scaling limit for every @xmath758 is precisely controlled by theorem 2.2 of @xcite . _ as we will see , in practice we need @xmath759 to be slightly more curved ( in the @xmath760 direction ) than @xmath755 : we define therefore @xmath759 to be the disk tangent to @xmath755 and whose curvature vector at @xmath49 is equal to @xmath761 , with @xmath762 defined by @xmath763 the second condition is there to prevent the curvature radius of @xmath759 from diverging . note that with this definition @xmath740 and @xmath762 do not necessarily have the same sign but we have in any case @xmath764 we set @xmath765 and we note that @xmath766 is unbounded if @xmath767 .
define then @xmath768 to be the dynamics starting from `` @xmath10 '' in @xmath766 and `` @xmath13 '' in @xmath769 at time zero , and @xmath770 as in with @xmath771 replaced by @xmath768 . as it is not true in general that @xmath772 , there is no immediate comparison between @xmath770 and @xmath102 .
however , the comparison works if we restrict ourselves to a sufficiently small box around @xmath49 .
this is what we do next .
consider @xmath773 to be a square of side - length @xmath774 , centered at @xmath775 , whose sides are parallel to the normal and tangent vectors to @xmath755 at @xmath49 .
if @xmath776 is chosen small enough ( depending on @xmath777 , @xmath676 , the sup norm and the modulus of continuity of @xmath740 , but neither on @xmath111 nor @xmath49 ) , the intersection of @xmath755 with @xmath773 has only one connected component and @xmath778 furthermore , if @xmath111 is small enough ( depending on @xmath779 and on the constant @xmath649 of proposition [ chips ] ) then @xmath780 = 8 cm these properties , together with monotonicity of the dynamics ( cf .
section [ monoton ] ) , allow us to establish a _
local _ comparison , that holds w.h.p , between the two stochastic droplets @xmath692 and
@xmath781 : [ lem : domsto ] if @xmath776 and @xmath111 are such that and are satisfied for the constant @xmath649 corresponding to proposition [ chips ] then for all @xmath49 in @xmath755 , w.h.p .
@xmath782 let @xmath783 denote the evolution of @xmath759 by the anisotropic curve shortening flow and set @xmath784 the second important ingredient of the proof is to show : [ inclus ] for @xmath111 small enough , there exists @xmath88 such that @xmath785 where @xmath746 was defined just after .
furthermore the inclusion remains valid if the closed sets on the right - hand side are replaced by their interiors . as @xmath746 is a compact set , in the inclusion we can extract a finite set @xmath786 which satisfies @xmath787 then combining lemma [ lem : domsto ] and ( * ? ? ?
* theorem 2.2 ) , that guarantees that @xmath788 , we notice that w.h.p . for every @xmath723 , @xmath789 as a side remark , note that when @xmath767 we need to reverse the role of `` @xmath10 '' and `` @xmath13 '' spins when applying ( * ?
* theorem 2.2 ) : this is because the bounded initial droplet @xmath759 is filled with `` @xmath13 '' and not `` @xmath10 '' spins .
hence , as @xmath790 is finite , we can take the union over @xmath723 to get that w.h.p .
@xmath791 we observe the dynamics @xmath768 and @xmath771 restricted to the square window @xmath773 .
we let @xmath792 and @xmath793 denote the restriction of @xmath768 and @xmath771 to the boundary @xmath794 of the box ( i.e. the set of lattice sites outside @xmath773 that are at distance @xmath4 from some lattice site in @xmath773 ) . from monotonicity of the dynamics , to show it is sufficient to show that w.h.p.@xmath795 , y\in \partial { { \ensuremath{\mathbf b } } } _ x.\end{aligned}\ ] ] the first point ( domination between the initial conditions ) is just an immediate consequence of .
the second one ( domination between time - dependent boundary conditions ) is a consequence of proposition [ chips ] combined with .
_ proof of lemma [ inclus ] .
_ for all @xmath49 in @xmath755 , we define a local coordinate system @xmath796 .
given @xmath449 ( which will depend on @xmath111 ) we define @xmath797 to be the rectangular box @xmath798\times[-d , d]$ ] in the frame @xmath799 ( it is much narrower than the @xmath800 square @xmath801 , i.e. we have to think @xmath802 ) .
note that given @xmath776 , if @xmath111 is small enough , for all @xmath235 we have @xmath803 hence to prove it is sufficient to prove that for all @xmath775 , @xmath804 we first rewrite this inclusion as an inequality between functions .
if @xmath776 and @xmath111 are sufficiently small ( with @xmath111 that depends on @xmath776 ) , the restriction of @xmath805 and of @xmath806 to @xmath807 can be considered as the graphs of functions in @xmath799 , for all @xmath808 $ ] .
we denote these functions by @xmath809 and @xmath810 respectively , with @xmath811 $ ] .
recall that the reference frame @xmath812 is such that the @xmath174-axis is directed along the _ inward pointing _ normal vector @xmath63 .
the proof of will thus be complete if we can prove the following : [ lem : ratrape ] there exists @xmath813 so that for all @xmath814 there exists @xmath815 and @xmath816 such that for all @xmath775 @xmath817.\ ] ] in order to simplify notations we drop the exponent @xmath49 in the following .
the core of the proof is to show that the drift of @xmath818 in the @xmath63 direction is stronger than the one of @xmath819 .
thus , even though @xmath188 starts above @xmath353 initially ( cf . ) , it has time to catch up .
note that , since @xmath820 is more curved than @xmath821 at time zero ( cf . )
, it would look like the drift of @xmath820 should be larger : however , we will see that the boundary of @xmath822 solves the curve shortening flow with an extra term , proportional to @xmath823 , in the normal velocity .
this extra term guarantees the desired inequality between drifts .
observe first of all that , provided that @xmath824 is small enough , for all @xmath825 $ ] @xmath826 the @xmath827 term is uniform in @xmath49 and just depends on the maximal curvature .
we have simply approximated locally the curves with suitable parabolas , and the second inequality is valid provided that @xmath828 on @xmath829 ( this is true if @xmath824 is small ) .
hence @xmath830 where the second inequality holds if @xmath824 is chosen small enough .
what we want to show is that for all @xmath831 $ ] , @xmath808 $ ] @xmath832 then the equation will easily be derived by integrating starting from , provided that @xmath833 ( for instance on can take @xmath834 and @xmath835 ) . to have an estimate on the time derivatives we need to use the equation that are satisfied by @xmath836 and @xmath837 respectively .
[ voldechi ] the curve @xmath738 is solution of the modified curve - shortening equation @xmath838\bf n,\\\ ] ] where @xmath223 is the normal vector oriented inside the curve and @xmath839 are the slope and curvature at the point in question . on the other hand , @xmath840 solves the usual curve shortening flow , that does not depend on @xmath823 . using simple trigonometry
we then obtain that @xmath841 where @xmath740 ( resp .
@xmath842 ) denotes the curvature on @xmath843 ( resp .
@xmath844 ) , @xmath24 ( resp .
@xmath845 ) the tangent angle on @xmath843 ( resp .
@xmath844 ) at the point of coordinate @xmath846 , and @xmath847 the tangent angle to @xmath755 at @xmath49 . by continuity of curvature in space and time ( cf . proposition [ th : continua ] ) , if one takes @xmath111 and @xmath449 small enough , @xmath24 can be assumed to be arbitrarily close to @xmath847 , @xmath740 arbitrarily close to @xmath848 and @xmath849 arbitrarily close to @xmath850 .
hence one has for all @xmath811 $ ] , @xmath808 $ ] @xmath851 then we conclude the proof of using and the fact that @xmath852 .
the proof of lemma [ inclus ] is also concluded .
as mentioned in remark [ rem : bij ] , there is a natural bijection between @xmath853 and @xmath854 , given by @xmath855 .
it is here that one needs to have @xmath676 smaller than some @xmath856 , to guarantee that @xmath857 . as observed in , for points that are in correspondence one has @xmath858 where we remind that @xmath58 is the curvature of @xmath22 and @xmath740 the curvature of @xmath738 .
moreover equation and the definition of @xmath738 give @xmath859 the desired equation immediately follows from and .
[ rem : genj ] the only properties of @xmath93 we used in the proof of the @xmath860 step of proposition [ infinit ] are @xmath861 and the fact that the curvature function is bounded and uniformly continuous .
since this continues to hold up to time @xmath677 , the proof for @xmath862 such that @xmath863 works exactly the same , and the small parameters @xmath111 and @xmath676 can be chosen to be independent of @xmath656 .
the authors would like to thank jimmy lamboley for enlightening discussions and valuable comments on various analytical aspects of the present work .
f. t. was partially supported by anr project shepi : `` systmes hors quilibre de particules en interaction '' .
p. caputo , f. martinelli , f. simenhaus and f. l. toninelli , _ `` zero '' temperature stochastic 3d ising model and dimer covering fluctuation : a first step towards mean curvature motion _
pure appl . math .
* 64 * ( 2011 ) , 778831 .
m. a. katsoulakis , p. e. souganidis , _ generalized motion by mean curvature as a macroscopic limit of stochastic ising models with long range interactions and glauber dynamics _ , commun .
* 169 * ( 1995 ) , 6197 .
e. lubetzky , f. martinelli , a. sly and f. l. toninelli , _ quasi - polynomial mixing of the 2d stochastic ising model with `` plus '' boundary up to criticality _ , j. eur .
soc . * 15 * no . 2 ( 2013 ) , 339386 . | let @xmath0 be a bounded , smooth enough domain of @xmath1 . for @xmath2
consider the continuous time , zero - temperature heat bath dynamics for the nearest - neighbor ising model on @xmath3 ( the square lattice with lattice spacing @xmath4 ) with initial condition such that @xmath5 if @xmath6 and @xmath7 otherwise .
we prove the following classical conjecture @xcite due to h. spohn : in the diffusive limit where time is rescaled by @xmath8 and @xmath9 , the boundary of the droplet of `` @xmath10 '' spins follows a _
deterministic _
anisotropic curve - shortening flow , such that the normal velocity is given by the local curvature times an explicit function of the local slope .
locally , in a suitable reference frame , the evolution of the droplet boundary follows the one - dimensional heat equation . to our knowledge , this is the first proof of mean curvature - type droplet shrinking for a lattice model with genuine microscopic dynamics .
an important ingredient is our recent work @xcite , where the case of convex @xmath0 was solved .
the other crucial point in the proof is obtaining precise regularity estimates on the deterministic curve shortening flow .
this builds on geometric and analytic ideas of grayson @xcite , gage - hamilton @xcite , gage - li @xcite , chou - zhu @xcite and others . |
the two - dimensional spins glasses , especially the @xmath0 ising model and the random @xmath1-state potts model , have been studied both by theoretical and numerical methods .
some exact results have been derived @xcite , the phase diagram has been drawn @xcite and recently the exact location of the multicritical point has been predicted @xcite .
however the characteristics of the multicritical point are not well understood except that the critical exponents have different values than those along the ferro - para critical curve @xcite .
the properties of the multicritical point are of great interest because this point represents one of the simplest non - trivial renormalization group fixed points in random spin systems @xcite and also because we may be able to understand the significance of the nishimori line from the study of the multicritical point . in the present paper
we investigate the sample - to - sample fluctuations of the internal energy near the multicritical point .
the thermal average of the hamiltonian , namely the internal energy , has different values from one bond configuration to another , given a fixed probability parameter ( @xmath2 for the ferromagnetic bonds in the @xmath0 model , for example ) .
although the internal energy is self - averaging , the variance of its distribution is of order @xmath3 , the system size .
if the variance is anomalously large and singular at some point , it would mean that a non - trivial ( and possibly drastic ) change happens in the bond configuration at that point because the value of the internal energy fluctuates strongly from sample to sample .
this is exactly what we shall show in the present paper to exist at the multicritical point by observing the system along the nishimori line .
the nishimori line plays an important role in the determination of the phase diagram .
in particular it is believed that the multicritical point lies on the line @xcite .
we have another advantage to study the sample - to - sample fluctuations along the nishimori line : these fluctuations are directly related to the average specific heat on the line .
this fact reduces numerical efforts considerably because it is not necessary to generate very many samples to calculate the variance .
only the average of the specific heat needs to be computed .
we shall make full use of this relation later . in the next section ,
we show that the sample - to - sample fluctuations of the internal energy are very simply related to the specific heat on the nishimori line .
this facilitates numerical calculations enormously as mentioned above . in section 3
we investigate numerically the behaviour of these fluctuations near the multicritical point on the nishimori line for three models : the @xmath0 ising model , three - state potts model and four - state potts model .
we will see that these fluctuations show a maximum at the multicritical point .
the physical significance of the results is discussed in the final section .
let us consider the random @xmath4 model with gauge symmetry which includes the @xmath0 ising model and the random @xmath1-state potts model .
the general hamiltonian is written as follows : @xmath5 where @xmath6 is the @xmath1-state spin variable , @xmath7 is the quenched random bond variable and @xmath8 is an interaction function of period @xmath1 . the summation is taken over the first - neighbouring sites of the square lattice .
the @xmath9 variable follows the probability distribution @xmath10 the hamiltonian ( 1 ) is invariant under the gauge transformation @xmath11 , where @xmath12 , but @xmath13 is not . using the gauge theory @xcite some exact results can be derived such as the average internal energy on the so - called nishimori line defined by @xmath14 , @xmath15 \equiv [ \langle h \rangle ] = -n_b \frac{\sum_{\eta}v(\eta ) \rme^{\beta v(\eta)}}{\sum_{\eta}\rme^{\beta v(\eta)}}.\ ] ] here the square brackets denote the configurational average , the angular brackets are for the thermal average and @xmath16 is the number of bonds . if we denote @xmath17 equation ( 3 ) is expressed as @xmath15 = -n_b\{v\}.\ ] ] in the case of the @xmath0 ising model , @xmath18 , @xmath19 and @xmath20 , the average internal energy ( 3 ) reduces to @xmath15 = -n_bj\tanh(\beta j).\ ] ] in the case of the @xmath1-state potts model with @xmath21 and @xmath22 , @xmath15 = -n_bj\frac{1}{1+(q-1)\rme^{-\beta j}}.\ ] ] we also find from the gauge theory that @xmath23=n_b\{v^2\ } + n_b(n_b-1)\{v\}^2,\ ] ] when @xmath24 , thus , with equation ( 4 ) , @xmath25 - [ \langle h \rangle ] ^2.\ ] ] since the specific heat is defined by @xmath26 = [ \langle h^2 \rangle - { \langle h \rangle}^2],\ ] ] we obtain , combining equations ( 6 ) and ( 7 ) , @xmath27 = [ e^2 ] - [ e]^2 ( \ge 0 ) .\ ] ] the right - hand side of equation ( 8) is the sample - to - sample fluctuations of the internal energy , @xmath28 $ ] .
note that equation ( 8) holds only on the nishimori line @xmath24 . in the case of the @xmath0 ising model equation (
8) reduces to the well - established inequality @xmath26
\le n_bj^2 { \rm sech}^2 ( \beta j).\ ] ] equation ( 8) indicates that the sample - to - sample fluctuations of the internal energy can be evaluated from the average of the specific heat because the quantity @xmath29 is trivially calculated .
this fact is a great advantage since it eliminates the necessity to evaluate the variance directly by using a large number of samples ; the specific heat is a self - averaging quantity and therefore can be calculated with high precision from a small number of samples with sufficiently large system size . away from the nishimori line , we do not have such a simple relation because equations ( 4 ) and ( 5 ) do not hold .
since the actual distribution of bond values ( and therefore of frustration ) changes from sample to sample in finite - size systems with given @xmath30 , the internal energy takes various values from one set of bonds to another around the average value . the relation ( 8) helps us to estimate the variance of the internal energy @xmath28 $ ] , and
directly depends on the distribution of frustration , which has a geometrical nature : the sample - to - sample fluctuations of the internal energy should reflect the geometrical structure of the distribution of frustration , which is a quantity independent of thermal variables but depends only on the bond configuration .
this argument suggests that we may be able to investigate the properties of the distribution of frustration by looking at the sample - to - sample fluctuations of the internal energy , or equivalently , the average specific heat on the nishimori line .
it is believed that the nishimori line intersects the phase boundary between paramagnetic and ferromagnetic phases @xcite ( see figure [ diagram ] ) .
if the distribution of frustration goes under a sharp change of its behaviour , it would be reflected in anomalously large sample - to - sample fluctuations of physical quantities including the internal energy . such an anomaly , if it exists on the right - hand side of equation ( 8) , is related to the singularity of the specific heat at the multicritical point on the nishimori line due to equation ( 8) .
since the specific heat is singular at the multicritical point , we may conclude that the sample - to - sample fluctuations are also singular there .
hence , by calculating the average specific heat near the multicritical point , we are able to investigate the structure of the distribution of frustration , which is of purely geometrical nature .
a similar argument has been used to relate the free energy and the entropy of frustration @xcite .
the next section gives us some numerical supports to confirm this idea .
phase diagram of the @xmath0 ising model in two dimensions
. there exist ferromagnetic ( f ) and paramagnetic ( p ) phases , and no spin glass phase exists .
the nishimori line is shown dashed . ]
we have tested the ideas of the previous section by numerically calculating the average specific heat on long strips by transfer matrix method . we have studied three models : the @xmath0 ising model , three - state potts model and four - state potts model .
we have used long strips of length 5000 and of various widths with free boundary conditions and without external field . by the transfer matrix method we calculated the free energy along the nishimori line .
the specific heat was obtained from numerical second derivative of the free energy by the inverse of the temperature .
we averaged the specific heat over ten samples .
for the @xmath0 ising model the widths of the lattice are @xmath31 . the lattice size ( up to @xmath32 ) and number of samples ( ten )
are much smaller than the existing numerical transfer matrix studies @xcite . however , since our purpose is much more modest than these investigations ( not trying to evaluate the critical exponents , for instance ) , our system size turns out to be sufficient as we shall show .
another reason for the small scale of our computation is that the use of the formula ( 8) greatly reduces the necessary number of samples as discussed in the previous section .
we define the parameter @xmath2 for the @xmath0 ising model by @xmath33 thus @xmath2 is the probability that @xmath34 @xmath35 @xmath36 .
takes the values 0 and 1 .
this notation is completely equivalent to the usual hamiltonian @xmath37 where @xmath38,@xmath39 and @xmath40 take the values @xmath41 . ]
each bond follows the probability distribution @xmath42 and we restrict our studies to @xmath43 . for the three - state potts model , @xmath44 and for the four - state potts model , @xmath45 . for the @xmath1-state potts model
, we define @xmath46 and the probability distribution ( 2 ) is then @xmath47 with @xmath48 . for each point on the nishimori line @xmath24 we extrapolate the specific heat to the limit @xmath49 .
we have found that the extrapolation by the @xmath50-law works quite well in almost all cases ( see figures [ pmj1 ] , [ potts31 ] and [ potts41 ] ) .
this is natural because we are mostly treating values of @xmath2 away from the multicritical point .
exactly at the multicritical point , the extrapolation needs some care , in particular if one wishes to estimate the critical exponents @xcite .
however , for our purpose to confirm the existence of a peak in the sample - to - sample fluctuations of the internal energy at the multicritical point , the method of extrapolation is rather irrelevant , which we confirmed by using other powers such as @xmath51 in some cases .
the error bars in the extrapolated values have been estimated by changing the range of extrapolation fitting , for example , from comparison of extrapolation using @xmath31 with the one using @xmath52 .
the resulting values of the energy fluctuations are plotted in figures [ pmj2 ] , [ potts32 ] and [ potts42 ] for the @xmath0 ising model , three - state potts model and four - state potts model , respectively . the position of the multicritical points are predicted in reference @xcite : for the @xmath0 ising model @xmath53 , for the three - state potts model @xmath54 and for the four - state potts model @xmath55 . by comparison of those values with figures [ pmj2 ] , [ potts32 ] and [ potts42 ]
we see clearly that the sample - to - sample fluctuations of the internal energy have a maximum at ( or at least very near ) the multicritical point for each model .
it seems reasonable to expect similar behaviour in other random models with @xmath4 symmetry .
we have studied the @xmath0 ising model , three - state potts model and four - state potts model by the numerical transfer matrix method .
we have first shown that the sample - to - sample fluctuations of the internal energy are , on the nishimori line , related to the specific heat and are expected to have a peculiar behaviour at the multicritical point .
it was found indeed from numerical calculations that the sample - to - sample fluctuations become anomalously large at or very near the multicritical point in all the models we investigated .
since these fluctuations are related to the specific heat by equation ( 8) , they should have a singularity at the multicritical point ( as the specific heat is singular there ) although it is difficult to see this effect within our numerical precision .
the existence of a peak at the multicritical point in conjunction with the existence of a singularity , anyway , suggests that the bond distribution and the distribution of frustration go under an anomalous change of the behaviour at @xmath56 .
this point of view is in agreement with the argument in reference @xcite that the phase transition at the multicritical point along the nishimori line is induced by geometrical anomaly . since the rate of the thermal excitations ( @xmath57 ) coincides with the ratio of the bond configurations ( @xmath58 ) on the nishimori line , the physical situations are very special there , which may be one of the reasons for the existence of a maximum in the configurational variance of the thermal quantity at the multicritical point along the nishimori line .
it is an interesting future problem to investigate what happens on the sample - to - sample fluctuations of physical quantities away from the nishimori line .
_ note added in proof .
_ we have found after submission of the paper that the sample - to - sample fluctuations of the energy , equation ( 8) , is proportional to the derivative of the energy by @xmath59 with @xmath60 fixed .
thus a maximum in these fluctuations means that the energy changes most dramatically at the multicritical point as @xmath59 changes .
the exact upper bound on the specific heat is the sum of two derivatives by @xmath60 and @xmath59 and corresponds to the derivative along the nishimori line .
5 nishimori h 2001 _ statistical physics of spin glasses and information processing : an introduction _ ( oxford : oxford univ .
press ) jacobsen j l and picco m 2002 e * 65 * 026113 nishimori h and nemoto k 2002 _ _ * 71 * 1198 aaro reis f d a , de queiroz s l a and dos santos r r 1999 b * 60 * 6740 le doussal p and harris a b 1988 625 nishimori h 1986 _ _ * 55 * 3305 merz f and chalker j t 2002 b * 65 * 054425 | we study the two - dimensional @xmath0 ising model , three - state potts model and four - state potts model , by the numerical transfer matrix method to investigate the behaviour of the sample - to - sample fluctuations of the internal energy on the nishimori line .
the result shows a maximum at the multicritical point in all the models we investigated .
the large sample - to - sample fluctuations of the internal energy as well as the existence of a singularity in these fluctuations imply that the bond configuration ( or , equivalently , the distribution of frustrated plaquettes ) may be experiencing a non - trivial change of its behaviour at the multicritical point .
this observation is consistent with the picture that the phase transition at the multicritical point is of geometry - induced nature . |
the notion of quantum entanglement plays a key role in the current study of quantum information and quantum computation theory .
there are two main criteria to distinguish entanglement from separable states : the ppt criterion @xcite tells us that the partial transpose of a separable state is positive , that is , positive semi - definite .
the converse is not true in general by a work of woronowicz @xcite who gave an example of a @xmath2 ppt entangled state .
such examples were also given in @xcite for the @xmath0 cases , in the early eighties .
another complete criterion was given by horodecki s @xcite using positive linear maps between matrix algebras , and this was formulated as the notion of entanglement witnesses @xcite .
this is equivalent to the duality theory @xcite between positivity of linear maps and separability of block matrices , through the jamiokowski - choi isomorphism @xcite . through this isomorphism ,
an entanglement witness is just a positive linear map which is not completely positive .
we refer to @xcite for systematic approaches to the duality using the jc isomorphism . for a linear map @xmath3 from the @xmath4-algebra @xmath5 of all @xmath6 matrices into @xmath7 , the choi matrix @xmath8 of @xmath3
is given by @xmath9 where @xmath10 is the usual matrix units in @xmath5 .
the correspondence @xmath11 is called the jc isomorphism .
it is known that @xmath3 is positive if and only if @xmath8 is block - positive , and @xmath3 is completely positive if and only if @xmath8 is positive . for a linear map @xmath12 and a block matrix @xmath13 , we define the bilinear pairing by @xmath14 it turns out that @xmath15 is separable if and only if @xmath16 for every positive map @xmath3 , and @xmath15 is of ppt if and only if @xmath16 for every decomposable positive map @xmath3 .
therefore , every entangled state @xmath15 is detected by a positive linear map @xmath3 in the sense that @xmath17 , and every ppt entangled state is detected by an indecomposable positive linear map .
a positive linear map is said to be an optimal entanglement witness if it detects a maximal set of entanglement , and an optimal pptes witness if it detects a maximal set of ppt entanglement , as was introduced in @xcite .
see also @xcite for the terminology . for a given entangled state ,
it is easy to find an entanglement witness to detect it , as it was suggested in @xcite .
but , it is not clear at all how to construct an optimal entanglement witness to detect a given entangled state .
the primary purpose of this note is to construct optimal pptes witnesses which detect ppt entangled edge states constructed in @xcite .
these states are the first examples of two qutrit ppt entangled edge states of type @xmath1 , whose existence had been a long standing question @xcite .
we also suggest a method to check the optimality of entanglement witness without the spanning property . for nonnegative real numbers @xmath18 and @xmath19 , we consider the map @xmath20 $ ] between @xmath21 defined by @xmath22(x)=\\ \begin{pmatrix } ax_{11}+bx_{22}+cx_{33 } & -e^{i\theta}x_{12 } & -e^{-i\theta}x_{13 } \\
-e^{-i\theta}x_{21 } & cx_{11}+ax_{22}+bx_{33 } & -e^{i\theta}x_{23 } \\
-e^{i\theta}x_{31 } & -e^{-i\theta}x_{32 } & bx_{11}+cx_{22}+ax_{33 } \end{pmatrix},\ ] ] for @xmath23 .
note that the choi matrix @xmath24 of the map @xmath20 $ ] is given by @xmath25=\left ( \begin{array}{ccccccccccc } a & \cdot & \cdot & \cdot & -e^{i\theta } & \cdot & \cdot & \cdot & -e^{-i\theta } \\ \cdot & c & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\ \cdot & \cdot & b & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\cdot & \cdot & \cdot & b & \cdot & \cdot & \cdot & \cdot & \cdot \\ -e^{-i\theta } & \cdot & \cdot & \cdot & a & \cdot & \cdot & \cdot & -e^{i\theta } \\ \cdot & \cdot & \cdot & \cdot & \cdot & c & \cdot & \cdot & \cdot \\ \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & c & \cdot & \cdot \\ \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & b & \cdot \\ -e^{i\theta } & \cdot & \cdot & \cdot & -e^{-i\theta } & \cdot & \cdot & \cdot & a \end{array } \right).\ ] ] the map of the form @xmath26 $ ] and its variants have been investigated by many authors in various contexts , as it was summarized in @xcite .
we just note here that @xmath27 $ ] is separable if and only if it is of ppt @xcite . on the other hand
, many interesting examples of ppt states are of the form @xmath28 $ ] .
for example , the ppt entangled states @xcite considered in the early eighties are just @xmath29 $ ] which turn out to be ppt entangled edge states of type @xmath30 .
these states were reconstructed systematically from indecomposable positive linear maps together with other types of ppt entangled edge states @xcite .
the ppt entangled edge states of type @xmath1 constructed in @xcite are given by @xmath31 $ ] for @xmath32 and @xmath33 .
recently , the authors @xcite analyzed @xmath34 $ ] to understand the boundary structures between separability and inseparability among ppt states . for indecomposable positive linear maps
, we have to be very careful to use the term optimalas was noticed in @xcite .
we denote by @xmath35 the convex cone of all positive linear maps .
recall that a positive map @xmath3 is an optimal ( respectively co - optimal ) entanglement witness if and only if the smallest face of @xmath35 determined by @xmath3 has no completely positive ( respectively completely copositive ) map .
we also note that @xmath3 has the spanning property ( respectively co - spanning property ) if and only if the smallest exposed face of @xmath35 determined by @xmath3 has no completely positive ( respectively completely copositive ) map .
we say that @xmath3 is bi - optimal if it is both optimal and co - optimal .
the term bi - spanning is defined similarly .
after we give conditions on parameters @xmath18 and @xmath36 for which the map @xmath20 $ ] is a positive linear map in the next section , we characterize for each fixed @xmath36 the facial structures of the @xmath37-dimensional convex body representing the positivity in section 3 . from this facial structures ,
it is clear that some of positive maps are not optimal , and/or not co - optimal .
we examine the spanning and co - spanning properties for the map @xmath20 $ ] in section 4 , and check various notions of optimality for all cases in section 5 . to do this , we suggest a more efficient method to check the optimality of an entanglement witness when it has not the spanning property . in the section 6 , we find optimal entanglement witnesses to detects ppt entangled edge states @xcite of type @xmath1 .
we conclude this note to report that our constructions give counter - examples to the spa conjecture @xcite .
to begin with , we first find the conditions for complete positivity and complete copositivity .
we note that the map @xmath20 $ ] is completely positive if and only if @xmath28 $ ] is positive if and only if the following @xmath38 matrix @xmath39= \left ( \begin{matrix } a & -e^{i\theta } & -e^{-i\theta}\\ -e^{-i\theta } & a & -e^{i\theta}\\ -e^{i\theta } & -e^{-i\theta } & a \end{matrix } \right)\ ] ] is positive .
we mention again that positivity of matrices means the positive semi - definiteness , throughout this note .
we see that the polynomial @xmath40 & = a^3 - 3a-(e^{3i\theta}+e^{-3i\theta})\\ & = [ a-(e^{i\theta}+e^{-i\theta})]\,[(a^2+a(e^{i\theta}+e^{-i\theta})+(e^{2i\theta}+e^{-2i\theta}-1 ) ] \end{aligned}\ ] ] has the following three real zeroes : @xmath41 we denote by @xmath42 . ]
then we see that @xmath43 for each @xmath36 , and @xmath44 $ ] is positive if and only if @xmath45 if and only if @xmath20 $ ] is completely positive .
we note that @xmath46 if and only if @xmath47 , and @xmath48 if and only if @xmath49 .
it is easy to see that the map @xmath20 $ ] is completely copositive if and only if @xmath50 the map @xmath20 $ ] is completely positive if and only if the condition ( [ cp ] ) holds , and completely copositive if and only if the condition ( [ ccp ] ) holds . in order to get a necessary condition for the positivity of @xmath20 $ ] , we note that a linear map @xmath3 is positive if and only if @xmath51 for every product vector @xmath52 . here , @xmath53 is considered as a column vector , and so @xmath54 belongs to @xmath55 .
we write @xmath56 by a direct calculation , we see that the pairing @xmath57{\rangle}$ ] is equal to @xmath58 from now on , we suppose that @xmath20 $ ] is positive , and put the product vectors @xmath59 in the above quantity , to get the following necessary condition @xmath60 we also take product vectors @xmath61 for @xmath62 , to get the condition @xmath63 for each @xmath62 if and only if @xmath64 therefore , we get necessary conditions ( [ p1 ] ) and ( [ p2 ] ) for the positivity of @xmath20 $ ] .
now , we proceed to show that two conditions ( [ p1 ] ) and ( [ p2 ] ) are sufficient for positivity .
note that @xmath20 $ ] is positive if and only if the matrix @xmath65 is positive semi - definite for any @xmath66 .
we first consider the determinant @xmath67 we may replace @xmath68 and @xmath69 by nonnegative @xmath70 and @xmath53 to get @xmath71 first of all , we check that all the @xmath72 principal minors are nonnegative . for example , the third @xmath73 minor is @xmath74 and we have @xmath75 this is nonnegative when @xmath76 .
if @xmath77 then we use the condition ( [ p2 ] ) to see easily that this quadratic form is nonnegative for each @xmath78 . in the same way
, we see that all @xmath73 minors are nonnegative , and @xmath79 whenever one of @xmath70 or @xmath53 is zero .
now , we show that @xmath79 on the region @xmath80 .
first , we note that all of @xmath81 , @xmath82 and @xmath83 are quadratic forms associated with the following symmetric matrices ; @xmath84 where @xmath85 that is , @xmath86 is expressed by @xmath87 for example . in the case of @xmath76 , we have @xmath88 and the equality holds if and only @xmath89 . in the case of @xmath90
, we have @xmath91 by ( [ p2 ] ) .
hence , it follows that @xmath92 and the equality holds if and only if @xmath93 .
consequently , we have @xmath94 for all cases .
first , we consider the case of @xmath95 , from which we have the following two cases : * @xmath96 , * @xmath97 . in the first case
, we already know that the map is completely copositive . in the second case
, we have @xmath98 by ( [ p1 ] ) , and so the map is completely positive . therefore ,
if @xmath95 then we see that @xmath20 $ ] is positive .
now , we assume that @xmath99 . in this case
, we have @xmath100 except for the following two cases : * @xmath101 and @xmath102 , * @xmath103 and @xmath104 . for the case of @xmath101 and @xmath102 , we have @xmath76 from the condition , and so @xmath105 which is nonnegative by the condition .
similarly , one can show that @xmath106 is nonnegative for the case of @xmath103 and @xmath104 .
now , we consider the case of @xmath100 . in this case , the coefficients of @xmath107 and @xmath108 in the polynomial @xmath109 are positive , and so there exists a sufficiently large cube @xmath110 so that @xmath79 outside of @xmath111 .
furthermore , we already know that @xmath79 if @xmath112 .
therefore , it suffices to show that the local minimums of @xmath109 in the region @xmath80 are nonnegative .
if @xmath113 is a nontrivial common solution of @xmath81 , @xmath82 and @xmath83 then it is also a nontrivial solution of homogeneous quadratic equation given by @xmath114 this means that a nontrivial common solution @xmath113 of @xmath81 , @xmath82 and @xmath83 satisfies @xmath115 if @xmath116 then common solutions are on the plane @xmath117 , and so there is no nonzero common solution in the region @xmath118
. therefore , we may assume @xmath119 .
in this case , the @xmath72 minors is not zero , and so the rank of the matrix is more than @xmath120 .
we also note that the determinant of this matrix is @xmath121\ge 0\ ] ] by the condition ( [ p1 ] ) .
if @xmath122 then we see that the above matrix is positive semi - definite , and so it must be singular with rank two .
therefore , a common solution must belong to the @xmath123-dimensional kernel space of the matrix .
consequently , all common solutions are of the form @xmath124 .
we consider the case @xmath125 . in this case , we see that common solution satisfies @xmath126 which is impossible in the region @xmath118 . summing up
, we see that if @xmath109 takes a local minimum at @xmath113 with @xmath127 then @xmath128 .
we note that @xmath129 which is nonnegative by ( [ p1 ] ) when @xmath130 .
this completes the proof for the following : the map @xmath20 $ ] is positive if and only if both conditions ( [ p1 ] ) and ( [ p2 ] ) hold .
for each @xmath36 , we denote by @xmath131 the @xmath37-dimensional convex body determined by ( [ p1 ] ) and ( [ p2 ] ) . the facial structures of the convex body @xmath132 has been analyzed in @xcite for the case of @xmath133 . facial structures of @xmath131 is similar as those of @xmath132 except several differences .
we first consider the case @xmath134 . in this case , the convex body @xmath131 has the following four @xmath120-dimensional faces : * @xmath135 , * @xmath136 , * @xmath137 , * @xmath138 . in the plane @xmath139 . ] in the case of @xmath48 , the face @xmath140 shrinks to a single point @xmath141 . in order to figure out the shape of the face @xmath140 , we modify the parametrization in @xcite to put @xmath142 for @xmath143
. then we have @xmath144 if @xmath46 then this face touches the face @xmath145 at the point @xmath146 which gives rise to a completely copositive map . on the other hand ,
if @xmath147 then this face does not touch the face @xmath145 .
the convex body @xmath131 also has the following @xmath123-dimensional faces : * @xmath148 , * @xmath149 , * @xmath150 , * @xmath151 , * @xmath152 , * @xmath153 for @xmath154 .
we note that @xmath155 is the line segment from the point @xmath156 to the point @xmath157 , and lies on the surface @xmath158 for @xmath159 .
we also note that @xmath160 shrink to a single point @xmath146 if @xmath46 .
it remains to list up @xmath161-dimensional faces as follows : * @xmath162 , * @xmath163 , * @xmath164 for @xmath154 , * @xmath165 for @xmath154 .
it @xmath48 then all of the following faces @xmath166 shrink to the single point @xmath141 .
furthermore , the face @xmath155 connects the two points @xmath141 and @xmath157 which represent completely positive and completely copositive maps , respectively .
therefore , every positive linear map @xmath20 $ ] is decomposable .
since we are interested in indecomposable cases , we assume @xmath167 throughout this note . by the exactly same argument as in @xcite , we have the following : * interior points of @xmath168 , @xmath169 , @xmath170 , @xmath171 , @xmath172 and @xmath173 are neither optimal nor co - optimal . * interior points of @xmath140 , @xmath174 , @xmath175 , @xmath162 are not optimal . *
interior points of @xmath155 and @xmath165 are not co - optimal . *
@xmath176 and @xmath177 are not spanning .
we recall that if two positive map @xmath178 and @xmath179 determine a common smallest face containing them , then they are interior points of the common face , and share the above properties related with the optimality . and @xmath180.,title="fig : " ] and @xmath180.,title="fig : " ] and @xmath180.,title="fig : " ]
in this section , we determine which positive linear maps have the spanning property and/or the co - spanning property .
we remind the readers that we are assuming that @xmath167 .
we also assume that @xmath181 , since the case of @xmath133 is already considered in @xcite .
we note that the spanning property ( respective co - spanning property ) implies the optimality ( respectively co - optimality ) . by the discussion of the previous section
, it remains to consider the following cases : 1 .
@xmath182 , 2 .
@xmath183 , 3 .
@xmath184 , 4 .
@xmath185 .
we recall @xcite that @xmath186 has the spanning property if and only if the set @xmath187:=\{z=\xi{\otimes}\eta:{\langle}zz^*,\phi{\rangle}=0\}\ ] ] spans the whole space @xmath188 , and @xmath189 therefore , we see that @xmath190 $ ] if and only if @xmath191 . in order to determine the set @xmath192 $ ] , we first find vectors @xmath66 such that the matrix ( [ mat ] ) is singular . in other words ,
we look for @xmath113 for which @xmath193 .
the only possibility of @xmath193 with nonzero @xmath194 is @xmath195 , and this happens only when the equality holds in ( [ p1 ] ) .
now , we consider the case ( i ) . in this case
, we see that @xmath193 holds only if @xmath112 .
we first consider the case @xmath196 , for which we have @xmath197 therefore , the matrix ( [ mat ] ) is singular with @xmath196 if and only if @xmath198 with complex numbers @xmath199 with modulus one . in this case ,
( [ mat ] ) is given by @xmath200 and the kernel is @xmath201 . in the same way
, we see that @xmath53 belongs to @xmath192 $ ] if and only if @xmath53 is one of the following : @xmath202=&(b^{1/4}\alpha , c^{1/4}\beta,0)^{\rm t}\otimes ( \bar{\alpha } e^{-i\theta}\sqrt{1-a},\bar{\beta}\sqrt b,0)^{\rm t},\\ z_2[\alpha,\beta]=&(c^{1/4}\beta,0,b^{1/4}\alpha)^{\rm t}\otimes ( \bar{\beta}\sqrt b,0,\bar{\alpha } e^{-i\theta}\sqrt{1-a})^{\rm t},\\ z_3[\alpha,\beta]=&(0,b^{1/4 } \alpha , c^{1/4}\beta)^{\rm t}\otimes ( 0,\bar{\alpha } e^{-i\theta}\sqrt{1-a},\bar{\beta}\sqrt b)^{\rm t } , \end{aligned}\ ] ] with complex numbers @xmath203 with modulus one .
it is clear that these vectors do not span the whole space if @xmath204 which implies @xmath205 in this case .
we consider the case @xmath206 .
we take @xmath207 and @xmath208 , and consider the @xmath209 matrix whose columns are nine vectors @xmath210 $ ] for @xmath211 .
then the determinant of @xmath212 is given by @xmath213 which is nonzero , since @xmath214 , and @xmath215 implies that @xmath216 .
therefore , we conclude that @xmath20 $ ] has the spanning property if and only if @xmath215 for the case ( i ) .
it is clear that it has not the co - spanning property from the facial structures in the previous section .
now , we consider the case ( ii ) .
first of all , we note that product vectors in ( [ prod ] ) already belong to @xmath192 $ ] , and so we see that @xmath20 $ ] has the spanning property if @xmath215 .
we will see in the next section that @xmath217 $ ] and @xmath218 $ ] are optimal , but do not have the spanning property .
we look for another product vectors in @xmath192 $ ] to determine if they have the co - spanning property .
if @xmath219 with @xmath220 , then @xmath20(xx^*)$ ] is given by @xmath221 for which * @xmath222 is a kernel vector if @xmath223 .
* @xmath224 is a kernel vector if @xmath225 .
* @xmath226 is a kernel vector if @xmath227 .
therefore , we see that @xmath228 $ ] if and only if @xmath53 is either one of the vectors in or one of the following form : @xmath229= & ( \alpha,\beta,\gamma)^{\rm t}\otimes ( \bar{\alpha } , \bar{\beta}e^{-i \frac{2}3\pi},\bar{\gamma}e^{i \frac{2}3\pi})^{\rm t } \quad \text { if } { \textstyle -\pi < \theta < -\frac \pi 3 } , \\
w[\alpha,\beta,\gamma]= & ( \alpha,\beta,\gamma)^{\rm t}\otimes ( \bar{\alpha},\bar{\beta},\bar{\gamma})^{\rm t } \quad \text { if } { \textstyle -\frac \pi 3<\theta<\frac \pi 3},\\ w[\alpha,\beta,\gamma]= & ( \alpha,\beta,\gamma)^{\rm t}\otimes ( \bar{\alpha } , \bar{\beta}e^{i \frac{2}3\pi},\bar{\gamma}e^{-i \frac{2}3\pi})^{\rm t } \quad \text { if } { \textstyle \frac \pi 3<\theta<\pi } , \\ \end{aligned}\ ] ] with @xmath230
. now , we take product vectors in @xmath192 $ ] as follows : @xmath231,\,z_1[1,-1],\,z_2[1,1],\,z_2[1,-1],\,z_3[1,1],\,z_3[1,-1],\,w[1,1,1],\,w[1,-1,1],\,w[1,i ,- i].\ ] ] we consider the @xmath209 matrix whose columns the partial conjugates of the above nine vectors , then the determinant is given as follows : @xmath232 we note that @xmath233 implies @xmath234 and @xmath235 , which is not possible by the assumption @xmath236 .
therefore , partial conjugates of the product vectors in span the whole space @xmath237 , and @xmath28 $ ] has the co - spanning property for the case ( ii ) .
now , we consider the case ( iii ) . in this case
, we note that @xmath238 $ ] is completely copositive , and so they never satisfies the co - spanning property .
note that the matrix is given by @xmath239 and the kernel is @xmath240 .
therefore , the following vectors @xmath241=&(\alpha,\beta,0)^{\rm t}\otimes ( \bar{\alpha},e^{i\theta}\bar{\beta } b,0)^{\rm t},\\ w_2[\alpha,\beta]=&(\beta,0,\alpha)^{\rm t}\otimes ( e^{i\theta}\bar{\beta}b,0,\bar{\alpha})^{\rm t},\\ w_3[\alpha,\beta]=&(0,\alpha,\beta)^{\rm t}\otimes ( 0,\bar{\alpha},e^{i\theta}\bar{\beta}b)^{\rm t } \end{aligned}\ ] ] belong to @xmath242 $ ] .
we see that the set of the following vectors @xmath243,\,w_1[1,-1],\,w_1[1,i],\,w_2[1,1],\,w_2[1,-1],\,w_2[1,i],\,w_3[1,1],\,w_3[1,-1],\,w_3[1,i]\ ] ] span the whole space @xmath237 , because the determinant of @xmath209 matrix @xmath212 whose columns are the above vectors is given by @xmath244 which is nonzero by the assumption @xmath167 . therefore , we see that the map @xmath20 $ ] has the spanning property for the case ( iii ) .
it remains to consider the case ( iv ) . in this case , they never have the spanning property , since @xmath245 $ ] is completely positive .
we consider interior points of the @xmath120-dimensional face @xmath140 on the plane @xmath246 . in this case , the only possible product vectors in @xmath192 $ ] are of the form ( [ prod2 ] ) .
it is clear that the partial conjugate of them do not span the whole space , and so they do not have the co - spanning property . in the next section
, we will see that they are not co - optimal . finally , we consider the line segment @xmath175 between two points @xmath247 and @xmath248 .
we note that the smallest exposed face @xmath109 containing @xmath217 $ ] is bigger than @xmath175 .
we already shown that @xmath217 $ ] has the co - spanning property , and so @xmath109 has no completely copositive map .
see @xcite for the case of @xmath133 .
this shows that the line segment @xmath175 has the co - spanning property .
[ thm : spanning ] suppose that the map @xmath20 $ ] is positive , and @xmath147 .
then we have the following : 1 . @xmath20 $ ] has the spanning property if and only if @xmath249 2 . @xmath20 $ ] has the co - spanning property if and only if @xmath250 holds or @xmath251 we summarize the results in terms of faces for @xmath147 as follows : * @xmath155 , @xmath164 and @xmath165 have the spanning property .
* @xmath176 and @xmath177 have not the spanning property
. it should be checked if they are optimal or not . *
@xmath174 , @xmath175 , @xmath162 , @xmath176 , @xmath177 and @xmath164 have the co - spanning property .
* @xmath140 have not the co - spanning property
. it should be checked if it is co - optimal or not .
it remains to check the optimality of @xmath217 $ ] and @xmath218 $ ] , and co - optimality of interior points of the face @xmath140 .
we recall that @xmath252 $ ] and @xmath253 $ ] , which are usually called the choi map , are extremal when @xmath133 by @xcite , and so they are optimal . in order to check the optimality of a positive map @xmath3 , we first find all extremal completely positive maps @xmath254 in the smallest exposed face determined by @xmath3 , and check if they belong to the smallest face determined by @xmath3 . recall that @xmath254 is the completely positive map given by @xmath255 where @xmath256 is an @xmath257 matrix .
we also recall @xcite that @xmath254 belongs to the smallest exposed face determined by @xmath3 if and only if @xmath256 is orthogonal to @xmath258 $ ] when we identify an @xmath257 matrix and a vector in @xmath259 .
we proceed to show that @xmath218 $ ] is optimal .
we first consider the case @xmath32 . in this case ,
@xmath260 $ ] consists of @xmath261 with complex numbers @xmath199 and @xmath262 with modulus one , from ( [ prod ] ) and ( [ prod2 ] ) .
every vector orthogonal to all of these vectors is of the form @xmath263 and the choi matrix of the completely positive map associated with this vector is given by @xmath264 = \left ( \begin{array}{ccccccccccc } \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\eta\bar\xi & \cdot & \cdot & \cdot & |\eta|^2 & \cdot & \cdot & \cdot & \eta\bar\zeta \\
\cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\
\zeta\bar\xi & \cdot & \cdot & \cdot & \zeta\bar\eta & \cdot & \cdot & \cdot & |\zeta|^2 \end{array } \right).\ ] ] in order to show the optimality of @xmath218 $ ] , we show that if @xmath265-pv[\xi,\eta,\zeta]\ ] ] is block - positive for a nonnegative @xmath266 with @xmath267 then @xmath268 .
first , we take product vectors @xmath269 for @xmath154 , then we have @xmath270- pv[\xi,\eta,\zeta]{\rangle}=t^2\left ( t(p_{\theta}-1)-p|\xi e^{-i\theta}+\eta|^2\right)\ge 0\ ] ] for all @xmath154 if and only if @xmath271 .
this implies @xmath272 from the relation @xmath267 .
now , we take product vectors @xmath273 for @xmath154 .
then we have @xmath274- pv[\xi ,- e^{-i\theta}\xi,(e^{-i\theta}-1)\xi]{\rangle}=t^2\left ( t(p_{\theta}-1)-p|\xi|^2(1 - 2\cos \theta)^2\right)\ge 0\ ] ] for all @xmath154 if and only if @xmath275 .
since @xmath276 for @xmath277 , we conclude that @xmath278 .
consequently , we have @xmath268 , and this completes the proof of the optimality of @xmath218 $ ] . for another ranges of @xmath36 , we have the similar argument .
now , we show that an interior point @xmath279 $ ] , with @xmath280 , of the face @xmath140 is not co - optimal . in the case of @xmath133
, it is clear .
we first consider the case @xmath281 . to see this
, we consider the completely copositive linear map whose choi matrix is given by @xmath282 = \left ( \begin{array}{ccccccccccc } \cdot & \cdot & \cdot & \cdot & -1 & \cdot & \cdot & \cdot & -1 \\ \cdot & 1 & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\ \cdot & \cdot & 1 & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot \\ \cdot & \cdot & \cdot & 1 & \cdot & \cdot & \cdot & \cdot & \cdot \\ -1 & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & -1 \\ \cdot & \cdot & \cdot & \cdot & \cdot & 1 & \cdot & \cdot & \cdot \\ \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & 1 & \cdot & \cdot \\ \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & \cdot & 1 & \cdot \\
-1 & \cdot & \cdot & \cdot & -1 & \cdot & \cdot & \cdot & \cdot \end{array } \right),\ ] ] and look for a small positive number @xmath266 so that @xmath283-pw[0,1,1;0]$ ] is block - positive .
we note that @xmath284-pw[0,1,1;0 ] = |e^{i\theta}-p|w\left[\frac 1{|e^{i\theta}-p|},\frac{b - p}{|e^{i\theta}-p|},\frac{c - p}{|e^{i\theta}-p|};\theta^\prime\right],\ ] ] where @xmath285 is the argument of @xmath286 .
first , we take the positive number @xmath287 so that @xmath288 , and then pick a positive number @xmath289 . [ fig : find_p ] is less than that of @xmath290 for @xmath291,title="fig : " ] now , it is clear that @xmath292 , and thus the condition is automatically satisfied .
we alse see that the argument @xmath285 of @xmath286 satisfies the condition @xmath293 .
then we have that @xmath294 therefore , the condition is also satisfied , and interior points of @xmath20 $ ] are not co - optimal .
the other cases for @xmath36 are similar .
we summarize the results in table 1 .
.summary of ( co-)optimality and ( co-)spanning property for faces of the convex body @xmath295 [ cols="^,^,^,^,^,^,^,^,^,^,^ " , ] we note that the the bi - optimality automatically implies indecomposability , and so we see that @xmath176 , @xmath177 and @xmath164 give rise to an indecomposable maps .
this can be seen directly .
we note that @xmath296 , \phi[a , b , c;\theta]{\rangle}=3 ( ap_{\pi-\theta}+bt+\frac ct -2).\ ] ] assume @xmath297 and take @xmath298 , to get @xmath299 , \phi[a , b , c;\theta]{\rangle}=3(ap_{\pi-\theta}+2\sqrt { bc}-2)=3a(p_{\pi-\theta}-2).\ ] ] we note that @xmath300 if and only if @xmath301 . therefore , we see that @xmath20 $ ] is an indecomposable positive map whenever the condition @xmath302 holds .
recall that we have already seen that positivity of @xmath20 $ ] implies decomposability in the case of @xmath303 for which @xmath48 .
we proceed to find an optimal entanglement witness which detects the ppt entangled edge state @xcite @xmath304 $ ] for @xmath305 with @xmath33 and @xmath306 .
we note that if we put @xmath307 and @xmath308 then we see that @xmath309{\rangle}= { \langle}w_iw_i^*,w(p_\theta , b,\textstyle\frac 1b;\theta]^\gamma{\rangle}=0.\ ] ] therefore , the most natural candidate is @xmath310 which equal to @xmath311\ ] ] since @xmath312
. to begin with , we note that @xmath313 , w\rangle & = p_\theta(1-\alpha)+\frac 1b ( b-\beta)+b ( \frac 1b-\gamma)-p_\theta-2\\ & = -p_\theta\alpha-\frac\beta b - b\gamma<0 . \end{aligned}\ ] ] we search for @xmath314 so that @xmath315 is bi - optimal . to do this
, we look for @xmath199 and @xmath262 satisfying the conditions @xmath316 from theorem [ thm : spanning ] . for the simplicity
, we put @xmath317 then we have @xmath318 and @xmath319 .
now , the problem is reduced to look for @xmath320 and @xmath321 satisfying the conditions @xmath322 it is easy to see that @xmath323 for @xmath324 . therefore ,
if we choose @xmath325 satisfying the condition for each @xmath318 , then we see that @xmath326 in and @xmath327[3\tilde \alpha-2t(2+t+\sqrt{3(1-t^2)})]\ge 0.\ ] ] consequently , we can find @xmath328 and @xmath321 as positve roots of the quadratic equation @xmath329 x+(2t-\tilde \alpha)^2=0.\ ] ] this completes the proof of the following : for each @xmath330 and @xmath306 , let @xmath325 be a positve number satisfying the condition with @xmath331 , and @xmath328 and @xmath321 be roots of the quadratic equation
. then @xmath332\ ] ] is an optimal pptes witness which detects @xmath333 $ ] .
we note that a general method had been suggested in @xcite to construct entanglement witnesses detecting a given entangled state .
if we follow this method for @xmath334 $ ] , then we get the above @xmath315 with @xmath335 .
but , it turns out that this method does not give us an _ optimal _ pptes witness in general .
in fact , one can show that if @xmath315 is an optimal pptes witness with positive @xmath335 then we have the restriction @xmath336 that is why we consider the above @xmath315 with different @xmath199 and @xmath262 for the full generality .
we determined positivity of linear maps with four parameters which are of choi map types involving complex entries .
we also determined their optimality , co - optimality , spanning property and co - spanning property . in this way , we found parameterized examples of indecomposable positive linear maps with the bi - spanning properties .
they are optimal pptes witnesses , which are nd - owes in the sense of @xcite .
optimality is not so easy to determine for a given positive linear map , because we do not know the whole facial structures of the convex cone @xmath35 consisting of all positive maps .
the spanning property is stronger than optimality and relatively easy to check .
we suggest a general method to check optimality of a positive map @xmath3 : we first find all extremal completely positive maps in the smallest exposed face of @xmath35 containing @xmath3 , and check if they belong to the smallest face containing @xmath3 .
the optimal pptes witnesses we constructed detect two qutrit ppt entangled edge states of type @xmath1 in @xcite , whose existence had been a long standing question @xcite .
we report here one interesting byproduct of our construction .
our constructions give counter - examples to the conjecture @xcite regarding the structural physical approximations , which claims that the spa of an optimal entanglement witness is separable .
several authors @xcite checked recently various kinds of entanglement witnesses to support the conjecture . in the forth - coming paper @xcite
, the authors will consider the spa conjecture in a systematic way .
we introduce the notions of positive type and copositive type for entanglement witnesses depending on the distances to the positive part and the copositive part .
we will show that if the spa of an entanglement witness is separable then it must be of copositive type , and so the spa is meagingful only for those of copositive types .
our construction in this paper shows that the spa conjecture does not hold even in cases of copositive types .
r. augusiak , j. bae , .
czekaj and m. lewenstein , _ on structural physical approximations and entanglement breaking maps , j. phys .
a * 44 ( 2011 ) , 185308 .
choi and s .- h .
kye , _ facial structures for separable states , j. korean math .
* 49 * ( 2012 ) , 623639 .
_ * _ d. chruciski and j. pytel , _ optimal entanglement witnesses from generalized reduction and robertson maps , j. phys .
a * 44 ( 2011 ) , 165304 .
m .- h . eom and s .- h .
kye , _ duality for positive linear maps in matrix algebras , math .
* 86 ( 2000 ) , 130142 . *
* _ k .- c .
ha and s .- h .
kye , _ construction of @xmath337 entangled edge states with positive partial transposes , j. phys . a * 38 ( 2005 ) , 90399050 .
ha and s .- h .
kye , _ one parameter family of indecomposable optimal entanglement witnesses arising from generalized choi maps , phys .
a , * 84 ( 2011 ) , 024302 .
ha and s .- h .
kye , _ geometry of the faces for separable states arising from generalized choi maps , preprint , arxiv:1203.6262 .
_ * _ * _ m. horodecki , p. horodecki and r. horodecki , _ separability of mixed states : necessary and sufficient conditions , phys .
lett . a * 223 ( 1996 ) , 18 .
a. jamiokowski , _ an effective method of investigation of positive maps on the set of positive definite operators , rep .
phys . * 5 ( 1974 ) , 415424 . * _ * _ j. k. korbicz , m. l. almeida , j. bae , m. lewenstein and a. acin , _ structural approximations to positive maps and entanglement breaking channels , phys .
rev . a * 78 ( 2008 ) , 062105 . s .- h . kye , _ facial structures for various notions of positivity and applications to the theory of entanglement , preprint , arxiv:1202.4255 .
* _ m. lewenstein , b. kraus , j. cirac and p. horodecki , _ optimization of entanglement witness , phys .
rev . a * 62 ( 2000 ) , 052310 .
a. peres , _ separability criterion for density matrices , phys .
* 77 ( 1996 ) , 14131415 .
x. qi and j. hou , _ a characterization of optimal entanglement witnesses , phys .
rev . a * 85 ( 2012 ) , 022334 * _ * _ * _ a. sanpera , d. bru , and m. lewenstein , _ schmidt - number witnesses and bound entanglement , phys .
a * 63 ( 2001 ) , 050301 .
skowronek , e. strmer , and k. zyczkowski , _ cones of positive maps and their duality relations , j. math . phys . * 50 * , ( 2009 ) , 062106 .
e. strmer , _ decomposable positive maps on @xmath4-algebras , proc . amer .
* 86 ( 1982 ) , 402404 . *
* _ b. m. terhal , _ bell inequalities and the separability criterion , phys .
lett . a * 271 ( 2000 ) , 319326 .
s. l. woronowicz , _ positive maps of low dimensional matrix algebras , rep . math
. phys . * 10 ( 1976 ) , 165183 . * _ * _ | we construct optimal pptes witnesses to detect @xmath0 ppt entangled edge states of type @xmath1 constructed recently @xcite .
to do this , we consider positive linear maps which are variants of the choi type map involving complex numbers , and examine several notions related to optimality for those entanglement witnesses . through the discussion ,
we suggest a method to check the optimality of entanglement witnesses without the spanning property .
6.0 truemm 1.5 true pc |
heavy quark effective theory ( hqet ) or , more generally , the expansion in inverse powers of the heavy quark mass @xmath0 , has become a generally accepted and widely used tool in heavy quark physics@xcite . based on the infinite mass limit @xmath1 of qcd it provides a model independent starting point for the description of weak transitions involving heavy quarks . the idea to exploit the fact that the mass of a heavy quark is large compared to the typical scale @xmath2 of the light qcd degrees of freedom ( e.g. the constituent mass of a light quark or the scale of the qcd coupling constant @xmath3 ) is in fact quite old@xcite .
however , in the late eighties a breakthrough was achieved by mainly two observations .
first , in the infinite mass limit qcd exhibits an additional flavour symmetry and a spin symmetry@xcite , the group theory of which allow model independent statements concerning weak decays of heavy hadrons .
second , it was noted that the @xmath4 expansion of qcd can be formulated as an effective field theory @xcite , which allows to access the corrections to the infinite mass limit in a systematic way .
since then the field of heavy quark physics has attracted a lot of attention , documented by an enormous number of papers that have been published using these methods .
in addition , from the experimental side a large effort is made to investigate the decays of bottom hadrons in order to pin down the origin of quark mixing and cp violation . in the standard model ( sm )
all of this is encoded in the ckm matrix , and the measurement of those elements of this matrix , which are only poorly known , involves mainly weak processes of @xmath5 quarks .
hqet , or more generally the systematic application of the @xmath4 expansion in qcd , has brought some progress in the determination of these ckm matrix elements , since heavy quark symmetries allow a drastic reduction of the hadronic uncertainties which enter the game through our ignorance to deal with the qcd bound state problem from first principles . in particular , the @xmath6 semileptonic decay may considered as a heavy @xmath7 heavy transition , where heavy quark symmetries work very efficiently ; consequently , the @xmath4 expansion allows for an almost model independent determination of the ckm matrix element @xmath8 . in the next section a brief summary on hqet and heavy quark symmetries
section 3 deals with the `` picture book application '' , namely the exclusive @xmath9 semileptonic decays .
heavy quark symmetries are also of some use in the case of a heavy hadron decaying into something light , and in section 4 this is considered for the case of heavy to light transitions .
finally a summary and a few conclusions are given .
hqet is an effective field theory which may be obtained from qcd by performing a @xmath4 expansion .
the leading term corresponds to the infinite mass limit in which the heavy quark acts as a static color source .
the momentum @xmath10 of the heavy quark scales with its mass and in order to perform the infinite mass limit it is convenient to use the velocity @xmath11 of the heavy quark as the basic kinematic quantity . to this
end the heavy quark momentum is split into a large part @xmath12 and a residual part @xmath13 , which is assumed not to scale with the heavy mass . thus @xmath14 we shall consider exclusively hadrons containing only a single heavy quark such that in the infinite mass limit the velocity @xmath11 of the heavy quark becomes simply the velocity of the heavy hadron . in order to write down a field theory which describes the static heavy quark
one may go through the usual steps of the construction of an effective field theory , namely to integrate out the heavy degrees of freedom @xcite .
an alternative method @xcite is to perform a foldy wouthuysen transformation as it is used in the standard non - relativistic reduction of the dirac equation .
although the @xmath4 expansion of the lagrangian and the corresponding expansion of the fields look completely different , the results for physical matrix elements will be the same since the @xmath4 expansion of the qcd greens functions constructed from this effective theory has to be unique . in the notation of @xcite one obtains @xmath15 h_v \\ & = & e^{-im_qvx } \left [ 1 + \frac{1}{2m_q }
\fmslash{d}_\perp + \nonumber \left(\frac{1}{2m_q}\right)^2 ( -ivd ) \fmslash{d}_\perp + \cdots \right ] h_v \\ \label{harvlag } { \cal l } & = & \bar{h}_v ( iv d ) h_v + \bar{h}_v i \fmslash{d}_\perp \left ( \frac{1}{2 m + ivd } \right ) i \fmslash{d}_\perp h_v \\ & = & \bar{h}_v ( iv d ) h_v + \frac{1}{2 m } \bar{h}_v ( i\fmslash{d}_\perp)^2 i h_v + \left ( \frac{1}{2m}\right)^2 \bar{h}_v ( i\fmslash{d}_\perp ) ( -ivd ) ( i\fmslash{d}_\perp ) h_v + \cdots \nonumber\end{aligned}\ ] ] where @xmath16 is the covariant derivative of qcd and @xmath17 is the heavy quark field in full qcd , while @xmath18 is the static heavy quark moving with the velocity @xmath11 . note that @xmath18 corresponds to the upper components of the full field since @xmath19 the leading terms of these expansions define the static limit ; the static lagrangian @xmath20 is a dimension - four operator and defines ( in combination with the usual lagrangian for the light degrees of freedom ) a renormalizable field theory . in the case in which the bottom and the charam quark are assumed to be heavy
one would write a static lagrangian for both quarks @xmath21 where @xmath22 ( @xmath23 ) is the field operator for the @xmath5 ( @xmath24 ) quark moving with velocity @xmath11 ( @xmath25 ) .
in particular , the masses of the heavy quarks do not appear in the lagrangian ( [ lbc ] ) , and as a consequence ( [ lbc ] ) in the case @xmath26 exhibits an @xmath27 heavy flavour symmetry which rotates the @xmath22 field into the @xmath28 field . the static heavy quark field @xmath18 is still a two component object corresponding to the upper component of the full heavy quark field @xmath29 .
however , both spin directions couple in the same way to the gluons ; we may rewrite the leading - order lagrangian as @xmath30 where @xmath31 are the projections of the heavy quark field on a definite spin direction @xmath32 @xmath33 this lagrangian has a symmetry under the rotations of the heavy quark spin and hence all the heavy hadron states moving with the velocity @xmath11 fall into spin - symmetry doublets as @xmath1 . the simplest spin - symmetry doublet in the mesonic case
consists of the pseudoscalar meson @xmath34 and the corresponding vector meson @xmath35 , since a @xmath36-spin rotation @xmath37 around the rotation axis @xmath38 ( @xmath39 ) yields @xmath40 in the heavy - mass limit the spin symmetry partners have to be degenerate and their splitting has to scale as @xmath4 .
in other words , the quantity @xmath41 has to be the same for all spin symmetry doublets of heavy ground state mesons .
this is well supported by data : for both the @xmath42 and the @xmath43 doublets one finds a value of @xmath44 gev@xmath45 .
this shows that the spin - symmetry partners become degenerate in the infinite mass limit and the splitting between them scales as @xmath4 . in the infinite mass limit
the symmetries imply relations between matrix elements involving heavy quarks .
for a transition between heavy ground - state mesons @xmath46 ( either pseudoscalar or vector ) with heavy flavour @xmath47 ( @xmath48 ) moving with velocities @xmath11 ( @xmath25 ) , one obtains in the heavy - quark limit @xmath49 where @xmath50 is some arbitrary dirac matrix and @xmath34 are the representation matrices for the two possibilities of coupling the heavy quark spin to the spin of the light degrees of freedom , which are in a spin-1/2 state for ground state mesons
@xmath51 due to the spin and flavour independence of the heavy mass limit the isgur wise function @xmath52 is the only non - perturbative information needed to describe all heavy to heavy transitions within a spin - flavour symmetry multiplet .
similar statements may be derived for the spin symmetry doublets of excited heavy mesons@xcite and also for baryons@xcite .
corrections to the infinite mass limit may be considered in a systematic way .
they fall into two classes : the recoil or @xmath4 corrections and the qcd radiative corrections . in order to discuss the corrections we shall consider a specific example , namely the matrix element of a current @xmath53 mediating a transition between a heavy meson and some arbitrary state @xmath54 . using the expansions ( [ harvlag ] ) and ( [ harvfield ] ) one obtains up to order @xmath4 @xmath55 where @xmath56 are the first - order corrections to the lagrangian as given in ( [ harvlag ] ) .
furthermore , @xmath57 is the state of the heavy meson in full qcd , including all its mass dependence , while @xmath58 is the corresponding state in the infinite mass limit . expression ( [ exp ] ) displays the generic structure of the higher - order corrections as they appear in any hqet calculation .
there will be local contributions coming from the expansion of the full qcd field ; these may be interpreted as the corrections to the currents .
the non - local contributions , i.e. the time - ordered products , are the corresponding corrections to the states and thus in the r.h.s . of ( [ exp ] ) only the states of the infinite - mass limit appear .
although the @xmath4 corrections need in general additional input beyond hqet , there is one important result on the corrections linear in @xmath4 , which is called luke s theorem@xcite and which is the application of the ademollo gatto theorem@xcite to the case of heavy flavour symmetry . in its general form
the theorem states that in the presence of explicit symmetry breaking the matrix elements of the symmetry generating currents , which are normalized due to the symmetry , do not receive corrections linear in the symmetry breaking .
applied to the case at hand this means that some of the form factors in weak decays , namely the ones proportional to the isgur wise function , receive only corrections quadratic in @xmath4 .
this result has important phenomenological consequences , which we shall discuss below .
all the relations given up to now are tree level relations .
going beyond tree level will induce qcd radiative corrections of order @xmath59 , @xmath60 . as in any field theory these corrections are perturbatively calculable in terms of feynman diagramms .
the effective theory has two additional feynman rules ( the propagator of the heavy quark and the coupling of the heavy quark to the gluons ) which may be read off from the static lagrangian ( [ lstat ] ) . for the sake of clarity we shall stick to our example of a heavy light current considered above . to leading order in the @xmath4 expansion one may evaluate the radiative corrections to such a matrix element using the above feynman rules and finds a divergent result with a divergence related to the short distance behavior . since hqet is an effective theory , the machinery of effective theory guarantees the factorization of long distance effects from the short distance ones , which are related to the large mass @xmath0 . neglecting @xmath4 corrections
, this factorization takes the form @xmath61 from feynman rule calculation one obtains the perturbative expansion of the renormalization constant @xmath62 which generically looks like @xmath63 where @xmath64 .
this factorization theorem corresponds to the statement that the ultraviolet divergencies in the effective theory have to match the logarithmic mass dependences of full qcd .
the factorization scale @xmath65 is an arbitrary parameter , and the physical quantity @xmath66 does not depend on this parameter .
however , calculating the matrix element of this operator in the effective theory and studying its ultraviolet behavior allows us to access the mass dependence of the matrix element @xmath66 .
the ultraviolet behavior of the effective theory is investigated by the renormalization group equation ( rge ) , which is obtained in the usual way from differentiating ( [ expr ] ) with respect to the factorization scale @xmath65 .
the rge for the short distance coefficient @xmath62 becomes @xmath67 where @xmath68 is the anomalous dimension of the current @xmath69 which is related to the ultraviolet behavior of the matrix elements of @xmath69 .
the function @xmath70 defines the running of the coupling constant @xmath71 both functions @xmath68 and @xmath70 are calculable in perturbation theory using a loopwise expansion , where the first term of the @xmath72 function of qcd is well known @xmath73 where @xmath74 is the number of flavors with a mass less than @xmath0 . with this input
the renormalization group equation may be solved to yield @xmath75 where @xmath76 is the first coefficient in the perturbative expansion of the anomalous dimension @xmath77 and @xmath78 is the one loop expression for the running coupling constant of qcd @xmath79 which is obtained from solving ( [ gml ] ) using ( [ beta0 ] ) .
this expression corresponds to a summation of the leading logarithms @xmath80 which is achieved by a one - loop calculation of the renormalization group functions @xmath72 and @xmath81 ; in other words , in this way a resummation of the first column of the expansion ( [ zexp ] ) is achieved . in a similar way one may also resum the second column of ( [ zexp ] ) ,
if the renormalization group functions @xmath72 and @xmath82 are calculated to two loops and the non - logarithmic terms of the one loop expression are included .
eq.([rge ] ) describes the renormalization group scaling in the effective theory .
it allows to shift logarithms of the large mass scale from the matrix element of @xmath69 into the coefficient @xmath62 : if the matrix element is renormalized at the large scale @xmath0 the logarithms of the type @xmath83 will apear in the matrix element of @xmath69 while the coefficient @xmath62 at this scale will simply be @xmath84 the renormalization group equation ( [ rge ] ) allows to lower the renormalization point from @xmath0 to @xmath65 ; the matrix element renormalized at @xmath65 will not contain any logarithms of @xmath0 any more , they will appear in the coefficient @xmath62 in the way shown in ( [ zexp ] ) . in all cases relevant in the present context
the matrix elements will be matrix elements involving hadronic states , which are in most cases impossible to calculate from first principles .
however , eq.([rge ] ) allows to extract the short distance piece , i.e. the logarithms of the large mass @xmath0 and to separate it into the wilson coefficients .
finally , the case we have considered as an example is indeed very simple ; in general all operators of a given dimension may mix under renormalization , i.e. instead of a simple anomalous dimension a matrix of anomalous dimensions may occur and the renormalization group equation [ rge ] becomes a system of differential equations .
in this section we shall discuss the transitions of the heavy to heavy type , i.e. the @xmath6 decays .
the implications of heavy quark symmetry have been given already in the form of the wigner eckart theorem ( [ wet ] ) in the last section , so we shall focus here on the status of the corrections to the infinite mass limit and give the phenomenological applications of the results .
the matrix elements relevant for the processes under consideration are @xmath86 where the point @xmath26 is of particular interest , since the absolute normalization of the matrix elements is known at this point due to heavy quark symmetries . in order to evaluate the qcd radiative corrections to these matrix elements we shall make use of the renormalization group ( rg ) machinery as outlined in the last section . at a large scale @xmath87
both currents are conserved in the limit of vanishing @xmath5 and @xmath24 quark masses and hence their anomalous dimension vanishes .
running down from @xmath88 to @xmath89 , i.e. lowering the renormalization scale of the matrix elements ( [ currents ] ) from @xmath88 to @xmath89 induces no large logarithms of the form @xmath90 , rather the corrections will be small , of the order @xmath91 . similarly , at scales @xmath65 below the charm
quark mass @xmath92 both quarks may be taken to be infinitely heavy , and at the non - recoil point @xmath26 again the two currents are conserved and thus their anomalous dimension vanishes . running below the charm quark mass will thus induce only small corrections of the order @xmath93 .
thus the main corrections originate from scales @xmath65 between @xmath89 and @xmath92 . in the effective theory where the @xmath5 is taken to be infinitely heavy and the @xmath24 is still light the one@xcite and two loop@xcite anomalous dimensions
have been calculated and allow a resummation of terms of order @xmath94 , and @xmath95 respectively .
furthermore , in this theory the subleading terms of order @xmath96 @xcite and @xmath97 @xcite have been considered at the one loop level , and the matching at the scale @xmath92 yields a resummation of terms of order @xmath98 and @xmath99 .
the procedure described here has the disadvantage that numerically it is useful only in the limit @xmath100 such that @xmath101 is a large logarithm and terms of order @xmath102 may be neglected . in real life
we have @xmath103 and one may think of simply performing the one loop calculation including the masses in full qcd and use this to match directly qcd to a theory with two static quarks , where heavy quark symmetries hold . in this way one obtains all terms of order @xmath104 . however , the price to pay is a scale ambiguity in the scale of @xmath105 which has to be taken at some scale @xmath106 between @xmath89 and @xmath92 .
numerically this is not a problem since @xmath105 does not run much between @xmath92 and @xmath89 .
recently the latter procedure has ben applied at the two loop level @xcite such that the terms of order @xmath107 are now known .
the usual way to parametrize the corrections is by introducing corrections factors @xmath108 which are known up to terms of order @xmath109 . taking the numbers of @xcite one
has @xmath110 the uncertainty given here is the uncertainty due to the terms of order @xmath109 and is conservatively estimated by the size of the calculated @xmath111 corrections .
given the fact that the calculated qcd corrections are already below the level of @xmath112 one needs to worry also about qed corrections .
in fact , the qed corrections to processes of this type have been calculated already some time ago in the context of @xmath65 decay@xcite .
similar to the qcd corrections they may be factorized and in principle one could apply also renormalization group methods .
however , the running of the qed coupling is negligible at this level and the qed corrections factor is given by @xmath113 \sim 1.013 \,\
, \eta_{a / v}\ ] ] hence short distance qed corrections enhance the matrix elements by @xmath114 .
in general the recoil corrections are more complicated , since they may not be calculated from the effective theory , rather they can only be parametrized in terms of new matrix elements .
the calculation of these matrix elements in any case needs additional input ( such as models or a lattice calculation ) beyond the framework of hqet . however , in some cases luke s theorem applies . in particular ,
the axial vector current taken at the non - recoil point @xmath26 is protected by this theorem , such that @xmath115 among the first non - trivial corrections the ones of the order @xmath116 will be the most important , and hqet allows only to parametrize them in terms of new matrix elements .
one obtains @xmath117 | d^*(v,\epsilon ) \rangle \right ) \\ & + & \vphantom{\int}{\cal o } ( 1/m_c^3,1/m_b^2 , 1/(m_c m_b ) ) , \nonumber\end{aligned}\ ] ] where @xmath118 is the first order lagangian for the quark @xmath29 as given in ( [ harvlag ] ) .
furthermore , the parameters @xmath119 and @xmath120 are the kinetic energy and the chromomagnetic moment of the heavy quark .
they also appear in the @xmath4 expansion of the heavy meson mass @xmath121 where @xmath122 for the @xmath123 and @xmath124 for the @xmath125 meson , and may be related to the matrix elements @xmath126 the only parameter which is easy to access is @xmath120 , since it is related to the mass splitting between @xmath34 and @xmath127 . from the @xmath128-meson system
we obtain @xmath129 and from the charm system ( including the renormalization group scaling of @xmath119 ) the same value is obtained .
this shows that indeed the spin - symmetry partners are degenerate in the infinite mass limit and the splitting between them scales as @xmath4 .
the other parameters appearing in ( [ massrel ] ) are not simply related to the hadron spectrum .
furthermore , they exhibit renormalon ambiguities , which imply that a proper prescription has to be given how to extract these quantities from data using renormalized perturbation theory .
recently there has been such an attempt @xcite , namley to extract @xmath130 and @xmath119 from the shape of the lepton energy spectrum in inclusive semileptonic @xmath128 decays .
the values obtained from this analysis are @xmath131 gev and @xmath132 gev@xmath45 , where the @xmath133 definition of the mass has been used .
the uncertainties quoted are only the @xmath134 statistical ones ; the systematical uncertainties of this approach are difficult to estimate .
these considerations only fix the local matrix elements in ( [ oom2 ] ) , while the nonlocal terms involving the time - ordered products are much harder to estimate .
the estimates found in the literature have to rely on some model estimates and span a range which implies roughly a @xmath135 theoretical uncertainty .
a commonly accepted number was given in @xcite @xmath136 in the last few years a lot of data has been accumulated which may be used to test heavy quark symmetry and to extract ckm matrix elements with the use of hqet .
the exclusive decays of prime interest are the transitions @xmath137 and @xmath138 .
the relevant matrix elements are @xmath139 \\
\langle d^ * ( v',\epsilon ) | \bar{c } \gamma_\mu b | b(v ) \rangle & = & i \sqrt{m_b m_{d^ * } } \xi_v ( y ) \varepsilon_{\mu
\alpha \beta \rho } \epsilon^{*\alpha } v^{\prime \beta } v^\rho \\
\langle d^ * ( v',\epsilon ) | \bar{c } \gamma_\mu \gamma_5 b | b(v ) \rangle & = & \sqrt{m_b m_{d^ * } } \left [ \xi_{a1 } ( y ) ( vv'+1 ) \epsilon^*_\mu - \xi_{a2 } ( y ) ( \epsilon^ * v ) v_\mu \right
. \nonumber \\ & & \qquad \qquad \left
. - \xi_{a2 } ( y ) ( \epsilon^ * v ) v'_\mu \right ] , \end{aligned}\ ] ] where we have defined @xmath140 .
thus in general there are six form factors , which in the heavymass limit for both the @xmath5 and the @xmath24 quark may be related to the isgur wise function as introduced in ( [ wet ] ) @xmath141 in particular , at the non - recoil point @xmath142 we have due to heavy quark symmetry and lukes theorem @xmath143 note that the form factors for which there is no normalization at @xmath142 there is also no protection against corrections of lienar order in @xmath4 .
the differential rates for the exclusive semiletonic @xmath6 transitions may be expressed in terms of the six form factors as @xmath144 with @xmath145 ^ 2 \\
\left[1-r_2 ( y ) \right ] \right)^2\end{aligned}\ ] ] where we have defined the form factor ratios @xmath146 in the heavy mass limit these differential rates depend only on the isgur - wise function @xmath147 of heavy quark symmetry , since the ratios of the differential rates do not depend on any unknown form factor any more .
in particular the ratios @xmath148 ( @xmath149 ) measures the ratio of the differential transverse ( longitudinal ) rate and the total differential rate . in the heavy mass limit
both @xmath148 and @xmath149 are unity ; this has to be compared to the measurements by cleo @xcite
@xmath150 from the measured lepton invariant mass spectrum one may determine @xmath8 in a model independent way by extrapolating to the kinematical endpoint of maximal momentum transfer to the leptons , corresponding to the point @xmath26 .
at this point heavy quark symmetries determine the absolute normalization of some of the form factors and the corrections to this normalization have been discussed in section 2 .
the mode @xmath138 has the advantage of a higher branching fraction and hence we shall start the discussion with this decay . the relevant formula may be derived from ( [ bds ] ) and reads @xmath151 the form factor @xmath152 is normalized due to heavy quark symmetries and is hence protected against @xmath4 corrections at @xmath26 by lukes theorem .
hence we have @xmath153 including qed corrections and the estimate of the @xmath154 corrections in the way discussed in section 2 one obtains @xmath155 from this value one may extract from the extrapolations shown in figs.1 and 2 a value for @xmath8 @xmath156 where the last error reflects the theoretical uncertainty of the @xmath4 corrections .
recently cleo has also measured the leptonic invariant mass spectrum for the decay @xmath137 . in a similar way one may perform an extrapolation to obtain @xmath8 . here
one gets from ( [ bd ] ) @xmath157 in this case a form factor enters which is not protected by lukes theorem , @xmath158 .
however , this does not spoil the possibility to determine @xmath159 from this mode , since the @xmath4 corrections are kinematically suppressed by the factor @xmath160 .
here we have @xmath161 where @xmath162 are the @xmath4 corrections induced by @xmath163 .
these corrections have been estimated recently @xcite @xmath164 the data from cleocleovcb1 are shown in fig.3 . from the extrapolation shown in fig.[fig2 ] one obtain the value @xmath165 where the last error again reflects the theoretical uncertainty .
both values of @xmath8 are in very good agreement and are also consistent with values obtained form other methods , in particular with the results from inclusive semileptonic decays@xcite .
the data on the leptonic invariant mass spectrum extend over the whole rage of @xmath166 and the extrapolation to @xmath167 has to rely on some ansatz for the isgur wise function .
this may be used in turn to extract a value also for the slope of the isgur wise function .
close to the point @xmath168 one thus parametrizes the isgur wise function as @xmath169 the theoretical predictions@xcite for the slope parameter @xmath170 depend on matrix elements involving excited @xmath16 mesons which are hard to estimate ; thus the theoretical value is quite uncertain and ranges between @xmath171 data on the slope parameter come from the lep experiments as well as from argus and cleo , the results obtained from experiment are @xmath172 which are compatible with the theoretical expectation .
heavy quark symmetries may also be used to restrict the independent form factors appearing in heavy to light decays . for the decays of heavy mesons into light @xmath123 and @xmath125 particles
heavy quark symmetries restrict the number of independent form factors to six , which is just the number needed to parametrize the semileptonic decays of this type .
furthermore , no absolute normalization of form factors may be obtained from heavy quark symmetries in the heavy to light case ; only the relative normalization of @xmath128 meson decays heavy to light transitions may be obtained from the corresponding @xmath16 decays . in general
we shall discuss matrix elements of a heavy to light current which have the following structure @xmath174 where @xmath50 is an arbitrary dirac matrix , @xmath175 is a light quark ( @xmath176 , @xmath177 or @xmath32 ) and @xmath178 is a state involving only light degrees of freedom .
spin symmetry implies that the heavy quark index hooks directly to the heavy quark index of the dirac matrix of the current . thus one may write for the transition matrix element ( [ htlps ] ) @xmath179 where the matrix @xmath34 representing the heavy meson has been given in ( [ mesonrep ] ) .
the matrix @xmath180 describes the light degrees of freedom and is the most general matrix which may be formed from the kinematical variables involved .
furthermore , if the energies of the particles in the state @xmath178 are small , i.e. of the order of @xmath3 , the matrix @xmath180 does not depend on the heavy quark ; in particular it does not depend on the heavy mass @xmath181 . in the following
we shall discuss some examples .
the first example is the heavy meson decay constant , where the state @xmath178 is simply the vacuum state .
the heavy meson decay constant is defined by @xmath182 and since @xmath183 the matrix @xmath184 is simply the unit matrix times a dimensionful constant may be eliminated using @xmath185 ] and one has , using ( [ htltrg ] ) @xmath186 as discussed above the constant @xmath187 does not depend on the heavy mass and thus one infers the well - known scaling law for the heavy meson decay constant from the last two equations @xmath188 including the leading and subleading qcd radiative corrections one obtains a relation between @xmath189 and @xmath190 @xmath191 f_d \sim 0.69 f_d .\ ] ] the second example are transitions of a heavy meson into a light pseudoscalar meson , which we shall denote as @xmath192 .
the matrix element corresponding to ( [ htlg ] ) is @xmath193 where @xmath194 is the momentum of the light quark , the dirac marix @xmath195 for the light degrees of freedom appearing now in ( [ htltrg ] ) depends on @xmath194 and @xmath11 .
it may be expanded in terms of the sixteen independent dirac matrices @xmath196 , @xmath197 , @xmath198 , @xmath199 , and @xmath200 taking into account that it has to behave like a pseudoscalar . the form factors appearing in the decomposition of @xmath195 depend on the variable @xmath201 , the energy of the light meson in the rest frame of the heavy one . in order to compare different heavy to light transition by employing heavy flavor symmetry
this energy must be sufficiently small , since the typical scale for the light degrees of freedom has to be of the order of @xmath3 to apply heavy quark symmetry ranges between @xmath202 and @xmath203 where we have neglected the pion mass .
thus at the upper end of phase space the variable @xmath201 scales with the heavy mass and heavy quark symmetries are not applicable any more . ] . for the case of a light pseudoscalar meson
the most general decomposition of @xmath195 is @xmath204 where we have defined the dimensionless variable @xmath205 the form factors @xmath178 and @xmath128 are universal in the kinematic range of small energy of the light meson , i.e. where the momentum transfer to the light degrees of freedom is of the order @xmath3 ; in this region @xmath206 is of order unity .
this universality of the form factors may be used to relate various kinds of heavy to light transitions , e.g. the semileptonic decays like @xmath207 , @xmath208 or @xmath209 and also the rare decays like @xmath210 or @xmath211 where @xmath212 denotes an electron or a muon . as an example
we give the relations between exclusive semileptonic heavy to light decays .
the relevant hadronic current for this case may be expressed in terms of two form factors @xmath213 where @xmath214 inserting this into ( [ htlps ] )
one may express @xmath215 in terms of the universal form factors @xmath178 and @xmath128 @xmath216 from these relations one may read off the scaling of the form factors with the heavy mass which was already derived in @xcite .
this may be used to normalize the semileptonic @xmath128 decays into light mesons relative to the semileptonic @xmath16 decays .
one obtains @xmath217 note that @xmath218 for the @xmath128 decay is expressed in terms of @xmath218 _ and _
@xmath219 for the @xmath16 decays . in the limit of vanishing fermion masses
only @xmath218 contributes , which means that the @xmath219 contribution to the rate is of the order of @xmath220 .
thus it will be extremely difficult to determine experimentally .
the case of a heavy meson decaying into a light vector meson may be treated similarly . the matrix element for the transition of a heavy meson into a light vector meson ( denoted generically as @xmath221 in the following )
is given again by ( [ htlg ] ) and is in this case @xmath222 using ( [ htltrg ] ) one has @xmath223 where now the dirac matrix @xmath224 has to be a linear function of the polarization of the light vector meson .
the most general decomposition is given in terms of four dimensionless form factors @xmath225 where the variable @xmath206 has been defined in ( [ eta ] ) .
similar to the case of the decays into a light pseudoscalar meson ( [ htlvtr ] ) may be used to relate various exclusive heavy to light processes in the kinematic range where the energy of the outgoing vector meson is small .
for example , the semileptonic decays @xmath226 , @xmath227 and @xmath228 are related among themselves and all of them may be related to the rare heavy to light decays @xmath229 and @xmath230 with @xmath231 .
data on these decays are still very sparse ; there are first measurements of the decays @xmath232 and @xmath233 from cleo @xcite , from which total rates may be obtained . from this
one may extract a value of @xmath234 by employing form factor models , and the value given by cleo is @xmath235 where the errors are purely experimental , while the range in the central value indicates the span obtained from a representative set of models . in order to perform a model independent determination along the lines discussed above a good measurement of the lepton energy spectra in these decays is needed .
the standard model has turned out to be suprisingly successful and has passed many tests , in particular the very precise tests performed at the lep collider .
however , these tests mainly concern the coupling of the gauge bosons to the fermions while the ckm sector of the standard model has not been tested with comparable accuracy . to test this part of the standard model one has to investigate weak processes among quarks which are in general plagued with strong interaction uncertainties . in this respect
the heavy mass limit has brought an enormous success ; hqet opens the unique possibility to determine some of the ckm matrix elements in a model independent way , thereby reducing uncertainties through models considerably .
in particular , it allows to at least give the order of magnitude of the uncertainties involved , since hqet relies on an expansion in @xmath236 and @xmath4 .
since the discovery of the heavy quark symmetries their phenomenological applications as well as the theoretical background have been studied intensively .
the most prominent example is the determination of @xmath8 , which corresponds to a heavy to heavy transition . combining the method
as described in this mini - review fo exclusive decays with the @xmath4 expansion for inclusive decays @xcite one may by now determine @xmath8 up to an uncertainty significantly less than ten percent . in heavy to light decays heavy quark symmetries do not work as efficiently ; in this case only the relative normalization of @xmath128 decays versus the corresponding @xmath16 decays may be obtained . from the experimental side
there are first measurements of @xmath232 and @xmath233 from the cleo collaboration and an extraction of the ckm matrix element @xmath234 from these processes is still to some extent model dependent .
a more model independent extraction of this matrix element has to wait for more data , in particular a measurement of the lepton energy spectrum is needed to exploit the relative normalization from heavy quark symmetry .
hqet does not yet have much to say about exclusive non - leptonic decays ; even for the decays @xmath237 , which involves three heavy quarks , heavy quark symmetries are not sufficient to yield useful relations between the decay rates @xcite .
of course , with additional assumptions such as factorization one can go ahead and relate the non - leptonic decays to the semileptonic ones ; however , this is a very strong assumption and it is not clear in what sense factorization is an approximation . on the other side ,
the data on the non - leptonic @xmath128 decays support factorization , and first attempts to understand this from qcd and hqet have been untertaken @xcite ; however , the problem of the exclusive non - leptonic decays still needs clarification and hopefully the heavy mass expansion will also be useful here .
99 some recent reviews include : n. isgur and m. wise : contribution to _ heavy flavors _ , a. buras and m. lindner ( eds . ) ( world scientific , singapore , 1992 ) ; m. neubert , phys
. rep .
bf 245 ( 1995 ) 259 ; int .
j. mod .
* a11 * ( 1996 ) 4173 ; t. mannel , j. phys .
* g 21 * ( 1995 ) 1007 .
m. voloshin and m. shifman , sov . j. nucl
* 45 * ( 1987 ) 292 and * 47 * ( 1988 ) 511 ; e. eichten and b. hill , phys .
* b234 * ( 1990 ) 511 ; applications to inclusive decays have been discussed in b. guberina , r.d .
peccei and r. rckl , phys .
lett . * 91b * ( 1980 ) 116 and phys .
lett . * 90b * ( 1980 ) 169 . n. isgur and m. wise , phys .
* b232 * ( 1989 ) 113 and phys . lett . *
b237 * ( 1990 ) 527 b. grinstein , nucl . phys .
* b339 * ( 1990 ) 253 ; h. georgi , phys . lett . * b240 * ( 1990 ) 447 ; a. falk , h. georgi , b. grinstein and m. wise , nucl .
* b343 * ( 1990 ) 1 .
t. mannel , w. roberts and z. ryzak , nucl . phys . * b 368 * ( 1992 ) 204 . j. krner and g. thompson , phys .
* b264 * ( 1991 ) 185 .
a. falk , nucl .
* b 378 * ( 1992 ) 79 .
n. isgur and m. wise , nucl .
* b348 * ( 1991 ) 276 .
h. georgi , nucl .
* b348 * ( 1991 ) 293 .
t. mannel , w. roberts and z. ryzak , nucl . phys .
* b355 * ( 1991 ) 38 .
m. luke , phys .
b252 * ( 1990 ) 447 . m. ademollo and r. gatto , phys . rev .
* 13 * ( 1964 ) 264 .
h. politzer and m. wise , phys .
* 208b * ( 1988 ) 504 and phys .
* 206b * ( 1988 ) 681 ; d. broadhurst and a. grozin , phys .
* b267 * ( 1991 ) 105 and phys .
* d52 * ( 1995 ) 4082 .
a. falk and b. grinstein , phys .
* b247 * ( 1990 ) 406 .
c. balzereit and t. ohl , to be published .
a. czarnecki , phys .
* 76 * ( 1996 ) 4124 .
a. sirlin , nucl .
* b100 * ( 1975 ) 291 . m. gremm et al . ,
* 77 * ( 1996 ) 20 .
m. neubert , cern - th-95 - 107 , contribution to 30th rencontres de moriond : electroweak interactions and unified theories , meribel les allues , france , 11 - 18 mar 1995 .
duboscq et al .
( cleo collaboration ) , phys .
rev . lett .
* 76 * ( 1996 ) 3898 .
d. cassel , these proceedings b. barish et al .
, ( cleo collaboration ) , phys . rev . * d 51 * ( 1995 ) 1014 .
d. buskulic et al . , ( aleph collaboration ) , phys .
* b 359 * ( 1995 ) 236 . i. caprini and m. neubert phys .
* b380 * ( 1996 ) 376 .
t. bergfeld et al . , ( cleo collaboration ) , contribution to the 27th international conference on high energy physics ( ichep ) , glasgow , scotland , 20 - 27 jul 1994 .
m. voloshin , phys . rev . *
b 46 * ( 1992 ) 3062 .
i. scott , ( aleph collaboration ) , contribution to the 27th international conference on high energy physics ( ichep ) , glasgow , scotland , 20 - 27 jul 1994 .
t. browder , ( cleo collaboration ) , contribution to the 27th international conference on high energy physics ( ichep ) , glasgow , scotland , 20 - 27 jul 1994 .
h. albrecht et al . , ( argus collaboration ) , z. phys . *
c57 * ( 1993 ) 533 .
p. abreu et al . , ( delphi collaboration ) z. phys . *
c71 * ( 1996 ) 539 . n. isgur and m. wise , phys
. rev .
* d42 * ( 1990 ) 2388 . k. berkelman ( cleo collaboration ) , talk given at the iii .
german russion workshop on heavy quark physics , dubna , russia , 20 - 22 may 1996 .
i. bigi , these proceedings .
t. mannel , w. roberts and z. ryzak , phys . lett .
* b248 * ( 1990 ) 392 .
m. dugan and b. grinstein , phys .
b255 * ( 1991 ) 583 .
* cleo data@xcite and the extrapolation used to obtain @xmath8 .
the upper figure corresponds to a linear extrapolation , the lower one includes also a curvature . *
aleph data@xcite and the extrapolation used to obtain @xmath8 . * data and the extrapolation used to obtain @xmath8 from @xmath137 .
the figure is taken from @xcite | a short review of a few selected topics in heavy quark effective theory is given .
applications to exclusive decays are discussed .
@sh[#1]#2 fmsl@sh fmsl@sh fmsl@sh fmsl@sh#1#2#3@th -3 cm * ttp 96 56 * * review of heavy quark effective theory * thomas mannel + _ institut fr theoretische teilchenphysik , university of karlsruhe + d
76128 karlsruhe , germany . _
+ _ contribution to the workshop on heavy quakrs at fixed target , + rheinfals castle , st .
goar , germany , october 3 - 6 , 1996 _ |
network and security researchers rely on topological maps of the logical internet to address problems ranging from critical infrastructure protection to policy . production active measurement systems that continually gather and curate internet topology , e.g. @xcite , are thus important to many longitudinal analyses and to shedding light on network events of interest .
obtaining ip , router , and provider - level network topologies has been a continual research focus for more than two decades @xcite .
while significant progress has been made , topology mapping at internet scale remains a challenge . both the accuracy of the inferred network topologies @xcite , and the speed at which they can be recovered @xcite , present obstacles to current mapping efforts . in this work ,
we focus on the _ speed _ and _ scale _ of internet - wide active topology mapping . given its scale , experience and popular belief
dictates that obtaining even partial internet topologies via active network probing is a time - intensive process .
for instance , caida s archipelago ( ark ) @xcite system uses dozens of vantage points and at least a day to traceroute to a single address in each routed /24 ipv4 prefix .
a recent topology cycle gathered by ark from april , 2016 @xcite sent approximately 11 m traceroutes from 37 monitors over the course of 31 hours in order to discover @xmath21 m distinct router interfaces , and @xmath22 m links .
we re - examine some of the assumed fundamental limits of active topology mapping to consider whether such probing could be performed in minutes rather than hours .
taking inspiration from recent stateless and randomized high - speed network scanners such as zmap @xcite and masscan @xcite , we create yarrp(yelling at random routers progressively ) . to facilitate high - probing rates , yarrpis stateless , reconstituting all necessary information from replies as they arrive asynchronously . to avoid overloading routers or links , yarrprandomly permutes its input @xmath3 space when probing .
yarrpis thus well - suited for internet - scale studies .
our contributions include : 1 .
development of yarrp , a publicly available tool @xcite that permits rapid active network topology discovery .
we run yarrpat @xmath1 to discover more than 400,000 router interfaces in under 30 minutes . 2 .
a comparison of yarrpand caida s existing production topology collection platform , showing recall and speed differences .
3 . as an application of rapid topology discovery , we conduct successive topology snapshots separated by a small time delta and characterize the distribution and causes of observed path differences .
traditional traceroute @xcite obtains the sequence of router interface ip addresses along the forward path to a destination by sending probe packets with varying time to live ( ttl ) values and examining the icmp responses . by maintaining the transmission timestamp of each probe , traceroute can report the round trip time ( rtt ) from the source to each responsive hop .
modern traceroute implementations send batches of concurrent probes to lower tracing time , e.g. linux defaults to 16 simultaneous probes . in order to match probes to
the icmp ttl exceeded responses they generate , the probe must include unique identifiers that are returned as part of the icmp quotation @xcite . because the quote is only required to copy the first 28 bytes of the packet that induced the expiry message @xcite , traceroute typically relies on the first 8 bytes of the transport - layer header to match responses to probes . while various improvements have been proposed and implemented , the core behavior of traceroute and large scale active topology scanning
remains largely unchanged . to prevent false inferences due to load - balanced paths , augustin et al .
created paris traceroute @xcite . to reduce unnecessary probing , donnet et al .
developed doubletree @xcite , a modified traceroute that begins probing from a likely path midpoint outward until it reaches previously discovered hops .
similarly , @xcite proposed several topology primitives empirically shown to reduce the volume of probing while maintaining or increasing topological discovery .
luckie et al .
developed scamper @xcite , a production - quality packet prober .
scamper implements both doubletree and paris traceroute , has an open api , can maintain a configurable probing rate , and can be controlled remotely .
caida s production ark infrastructure @xcite uses scamper to perform continual internet - wide probing @xcite .
traceroute was originally designed as a tool for network administrators to diagnose a small number of paths , not as a means to gather snapshots of the entire internet topology @xcite .
fundamentally , traceroute and its variants all have two properties that limit their scalability and speed .
they : * maintain state for each outstanding probe sent , including some identifier and origination time .
* are sequential , probing all hops along the path to a destination .
while some tools ( e.g. scamper ) can traceroute to multiple targets , this parallelism is limited to a finite window of destinations .
in contrast , yarrpis designed to be stateless and random probing different portions of many different paths simultaneously .
this allows yarrpto send probes at a high per - packet rate , while spreading the load among many destination networks to avoid concentrating load on particular paths , links or routers , thereby avoiding anomaly alarms or icmp rate limiting .
the high - level idea of yarrpis : i ) randomization of the probing order of the domain of network range(s ) and ttls ; and ii ) stateless operation , whereby all necessary state is encoded into the probes such that it can be recovered from the icmp replies .
yarrpis written in c++ , is portable to a variety of unix - like platforms , and is publicly available @xcite .
existing traceroute techniques probe all hops along a path to a destination in sequence . instead
, we employ a keyed block cipher to provide a bijection over the input domain of target ips and ttls ( @xmath4 ) .
this means that yarrpwill e.g. send a probe to ip address @xmath5 with @xmath6 , then toward @xmath7 with @xmath8 , then @xmath9 with @xmath10 , and so on until the entire space of @xmath11 for each target @xmath12 is covered . the symmetric rc5 block cipher with a 32-bit block size is fast and a natural fit for our application . with key @xmath13 , yarrpencrypts the sequence @xmath14 where bits of each ciphertext @xmath15 determine the target ip address and ttl to probe . in this way , yarrprandomizes the order of probed @xmath16 .
yarrpcan permute arbitrarily large or small ipv4 address and ttl domains , or can permute the order of specific targets read from a file .
depending on the size of the domain , we switch between either a prefix - cipher or cycle - walking cipher , as described in @xcite .
to facilitate comparison with caida s ipv4 topology dataset @xcite , yarrphas a mode that probes a random address in each ipv4 24-bit subnet this mimics the targets selected in a full cycle of caida s probing . here ,
yarrpencrypts each @xmath17 with key @xmath13 . for @xmath18 , yarrpprobes the ipv4 address @xmath19 * 2 ^ 8 + ( c_i[0]+c_i[1]+c_i[2 ] ) \% 256 $ ] with ttl @xmath20 $ ] . in this fashion , we permute through the space of @xmath21 possible /24s , and construct the least - significant octet as a function of the subnet such that the same random address in each /24 is used as the destination for each ttl .
an advantage of yarrp s randomization method is that the probing work can easily be distributed among multiple vantage points with negligible coordination or communication overhead .
we discuss distributed yarrpas a future enhancement in [ sec : conclusions ] .
existing traceroute techniques require state to match icmp replies to probes .
in contrast , yarrpdoes not require state .
we overload various fields in the probe packets with specific values such that we can reconstruct the corresponding probe s destination , transmission time , and originating ttl from within the quote of the icmp ttl exceeded messages .
figure [ fig : fields ] depicts the tcp / ip header fields we utilize .
we encode the ttl with which the packet was sent in the ipid and the elapsed time in the tcp sequence number .
we use elapsed time rather than e.g. unix time in order to maintain millisecond resolution with only a 32-bit field .
yarrpcan also encode microsecond resolution , so long as the expected duration of a probing run is less than @xmath224300 seconds .
the destination tcp port is fixed to port 80 to facilitate firewall traversal , while we populate the source tcp port with the checksum of the target ip address . in this fashion , we can detect instances where the destination ip address is modified enroute , a phenomenon malone and luckie observe in 2% of their results @xcite . in order to properly accommodate load - balanced paths , which are common in the internet ,
we ensure that , for a given destination , certain fields remain fixed for all ttl probe values .
for instance , although the tcp source port changes , it is a function of the destination ip address and therefore will contain the same value for all probes sent toward the destination .
this design allows us to maintain the benefits of paris traceroute @xcite .
when icmp ttl exceeded messages arrive , we examine the included quotation to recover the destination probed , the originating ttl ( hop ) , responding interface at that hop , and compute the rtt by taking the difference of the packet arrival time and the probe origination time as encoded in the quoted tcp sequence number .
these values can be computed from the minimum 28 bytes of required quotation @xcite .
finally , yarrpcan source either tcp syn or ack probes .
while syn probes can permit middlebox traversal , we use the ack - only mode to avoid alarms triggered by large volumes of syn traffic .
the benefits of yarrp s design come with several concomitant challenges , namely : i ) reconstructing the unordered responses into paths , ii ) knowing when to stop probing , and iii ) avoiding unnecessary probing .
in following with yarrp s stateless nature , icmp responses are decoded as they arrive and written sequentially to a structured output file .
each entry in the output file corresponds to an icmp response .
an entry includes the target ip address , originating ttl , responding router interface ip address , rtt , and meta - data such as timestamps , ipid , response ttl , packet sizes , and dscp markings .
because of the inherently random probing , the entries for each hop along a path to a given destination will be unordered and intermixed with other responses in the yarrpoutput file .
we must therefore reconstruct complete paths by parsing the entire output file and maintaining state for each destination . while this is a memory and time - intensive task ,
the key point is that it can be performed _ off - line_. in this fashion , we decouple probing from path reconstruction to permit the probing to be as fast as possible . included in the yarrpdistribution
@xcite is a ` yrp2warts ` python script that performs this off - line conversion into the standard warts @xcite binary trace format .
a practical consequence of yarrp s randomization and lack of state is that its probing behavior does not depend on the received responses .
thus , yarrpcannot stop probing once it reaches the destination or when the path contains a sequence of unresponsive hops ( the so - called `` gap limit '' ) . to better understand the optimal range of ttls to probe ( from the possible space 1 - 255 ) , we examine the results from a complete cycle of ark probing from january , 2016 @xcite .
we seek to determine , across each of the ark vantage points , the number of unique router interfaces discovered at each ttl .
figure [ fig : ttlhisto ] shows the inter - quartile range of the number of distinct interfaces found as a function of ttl for each of the vantage points ; the red line in the boxplot displays the median number of interfaces per ttl among each vantage point . because of the internet s tree - like structure , the first few hops reveal only a small number of interfaces regardless of the destination probed .
the bulk of the interfaces are found between ttls 10 to 16 , with an inflection point around a ttl of 14 .
the amount of discoverable topology beyond a ttl of 32 is negligible ( note the log scale y - axis ) . as a result , yarrpdefaults to probing ttls 1 to 32 to minimize unnecessary probing while exploring the majority of the space . for many destinations yarrpwill perform more probing than traditional traceroute methods .
this is both an advantage and a disadvantage : we show in [ sec : results : gaplimit ] that discoverable topology exists beyond multiple unresponsive hops ( where existing methods terminate early ) . in environments sensitive to probing volume , several optimizations can substantially decrease unnecessary probing at the expense of maintaining some state .
this subsection discusses optimizations to the base yarrpdesign to enable different tradeoffs . first ,
yarrpcan read a bgp routing table of network prefixes and build a longest - match patricia trie @xcite . when iterating through the entire permuted ipv4 space , yarrpcan skip destinations that are not routed .
based on current global bgp routing tables @xcite , this optimization avoids probing approximately 1.5b ip addresses ( 35% of the 32-bit space ) that are unlikely to return useful results .
note that the memory required to maintain the bgp table is constant during a probing run ( amounting to approximately 300 mb during runtime ) . in our experiments
, these lookups in the patricia trie did not prevent yarrpfrom running at over @xmath23 .
second , the tree - like structure of the network implies that the set of interfaces near to the vantage point is small relative to the universe of router interfaces @xcite . in figure
[ fig : ttlhisto ] for instance , all of the traces have a single first hop in common and orders of magnitude fewer interfaces at hops 1 - 3 as compared to hops 13 - 15 . to avoid rediscovering the same nearby router interfaces repeatedly , yarrpcan maintain state over the set of responding local `` neighborhood '' interface ip addresses at hops 1 through a run - time configurable @xmath24 . for each ttl in the neighborhood , yarrpmaintains two timestamps : the last time a probe was sent with that ttl , and the last time a new interface at that depth replied . if no new interfaces have been discovered within the past @xmath25 seconds of probing , yarrpskips future probes at that ttl .
the ` yrp2warts ` script can then stitch together these missing hops .
while the amount of state in neighborhood mode can grown unbounded , in practice it is small for small @xmath24 , while avoiding substantial over - probing .
this section examines results from running yarrpon the internet . we compare the topological recall against an existing production system and
then analyze the discovery yield ( i.e. the amount of new topology discovered over time ) . finally , as an application of yarrp s probing speed , we gather two successive snapshots of internet topology separated in time by a small delta to reveal instances of short - lived network dynamics .
we empirically verify yarrp s correctness by evaluating it against caida s use of scamper in the ark infrastructure @xcite .
caida makes their topology traces publicly available @xcite .
we find all 67,045 destinations probed from the san diego vantage point on may 1 , 2016 . from this same network attachment point , we instruct yarrpto probe the same destinations . in this fashion , because the vantage point networks and set of destinations are the same , we expect the paths to be largely congruent , allowing an unbiased comparison of topological recall between the two probing methods . from the yarrpand ark probing , we construct graphs of interface nodes connected by edges when the interfaces appear in consecutive hops of a path .
we ignore anonymous interfaces @xcite such that the graph may be disconnected .
figure [ fig : wartscmp ] displays the resulting graph degree distributions on a log - log scale .
we observe that the distributions match closely , empirically supporting yarrp s ability to discover the responsive topology .
however , yarrpdiscovers 16% fewer interfaces ( 80,134 versus 66,939 ) and 13% fewer edges ( 96,763 versus 84,113 ) than ark in this experiment .
we attribute much of this difference to yarrp s use of tcp ack probes as compared to caida s use of icmp probes . in a survey by key et al .
, icmp probes elicit almost 18% more responses as compared to tcp probes @xcite .
icmp and udp - based probing are future yarrpenhancements ( see [ sec : conclusions ] ) .
[ sec : results : gaplimit ] yarrp s stateless nature implies that it probes all ttls from 1 to 32 , whereas ark s use of scamper ceases probing after encountering five unresponsive hops in a row . in the same caida
san diego probing run , 39,613 traces stopped due to this gap - limit .
for each of these gap - limit traces , we compute the difference of the highest responding ttl hop from yarrpprobing and the highest responding ttl from the ark probing .
figure [ fig : gapdiff ] shows the cumulative distribution of this difference among the gapped traces ; a positive difference means that yarrpdiscovered topology beyond the point where ark stopped probing . for @xmath2674% of the traces ,
there is no difference . in 8% of the traces
, ark discovers one more hop than yarrp .
however , yarrpdiscovers one additional hop in @xmath27% of the targets , and more than 5 additional hops in 4% of the cases . a goal of yarrpis rapid topological discovery . in this subsection , we look specifically at the ability to discover unique router interface addresses rapidly . on may 10 , 2016
, we run yarrpfrom a northeast united states university vantage point at @xmath28 and instruct it to perform the ark - mode randomized probing of the globally routed ipv4 /24 prefixes .
we limit yarrpto this rate , and limit the duration of our experiment , per prior agreement with the local network administrator .
the physical machine is a multi - core intel l5640 processor running at 2.27ghz , with yarrprunning on an ubuntu virtual machine allocated a single core . at this rate ,
the cpu utilization is @xmath2652% .
we enable the `` neighborhood '' optimization , as described in [ sec : design : optimize ] , as we are interested in finding as many distinct router interfaces as possible given the probing rate .
figure [ fig : discovery ] displays the cumulative number of distinct router interfaces discovered as a function of time . as a basis of comparison , we also plot the number of unique interfaces found over time for a single vantage point ( again , using data from the san diego node of caida s continual /24 probing @xcite on may 1 , 2016 ) .
caida s san diego monitor discovers 12,568 unique interfaces in 1,500 seconds ( @xmath29 per second ) .
by contrast , yarrpdiscovers 421,162 unique ipv4 interfaces in the same period , or approximately 280 distinct router interfaces a second .
the interfaces found by yarrpin less than 30 minutes equates to 42% of _ all _ unique interfaces discovered from all ark monitors over the course of probing for more than a day . as an application of rapid topology discovery
, we collect topology snapshots in rapid succession and analyze their properties and differences in this subsection .
we gather 67,045 target destinations from caida s may 1 , 2016 topology probing from their san diego monitor . again using the east coast university vantage point
, we run yarrpto probe ttls 1 - 32 for these same 67,045 targets .
we run yarrpat @xmath30 and invoke yarrpthree times in succession with a minute pause in - between . in this way
, each snapshot takes approximately 8 minutes to gather , and each is separated by a minute .
we term the snapshots @xmath31 and @xmath32 in chronological order .
the interface - level graph resulting from @xmath33 contains 39,968 interfaces and 46,721 edges , while @xmath34 has 40,038 interfaces and @xmath35 edges .
@xmath32 contains @xmath36 interfaces and @xmath37 edges . to better understand the differences between snapshots , we perform a per - target path comparison
for each target in @xmath38 , we compare the discovered path in @xmath38 to the discovered path to that same target in @xmath39 .
we use the levenshtein edit distance to measure the per - target path differences between snapshots .
the edit distance is the minimum number of edits ( insertions , substitutions , or deletions of router interfaces ) .
note that inter - snapshot differences are not attributable to per - flow load balancing as yarrpkeeps the packet header fields which are used for load balancing constant for the same destination between snapshots ( [ sec : design : stateless ] ) . additionally , to better understand the types of path changes , we count the frequency of each edit operation and missing hop substitutions .
these missing hop operations are instances where the path contains a responsive router for a particular ttl for one snapshot , but no response at that ttl when probing the same destination in a subsequent snapshot .
such missing hops may be attributable to routers performing icmp rate limiting , or may be due to packet loss .
a deeper analysis of the most frequent missing hops between @xmath33 and @xmath34 reveals that the large majority ( 92.2% ) come from the first four hops within the local network of the vantage point . specifically , 73% of the missing hops are due to the router at ttl 3 , 18% are due to the router at ttl 1 , and 1% are due to the router at ttl 4 .
in contrast , the router at ttl 2 always responds , suggest that some of the local routers implement icmp rate limiting while one does not .
figure [ fig : ed ] displays the results of the edit distance comparison between @xmath33 and @xmath34 , ignoring differences attributable to the local network ( ttl @xmath40 , as described above ) . the paths to approximately 91% of the destinations are identical between @xmath33 and @xmath34 , while approximately 6% have a single hop difference . less than 1% of the destinations show a difference of @xmath41 hop edits
. separated by the edit operation , we see that @xmath2 4% of the 1 hop differences are due to missing hops , 1% are hop deletions , and fewer than 1% are substitutions . to understand the potential of rapidly collected topology snapshots , we manually investigate and highlight a path exhibiting a significant change between snapshots .
figure [ fig : samplepath ] shows , for each of the three snapshots , the final four responsive hops toward the destination 188.32.230.138 ( in as 42610 ) . in the intermediate snapshot , @xmath34 ,
we see two hop substitutions where the next hop after as 1273 ( 213.185.219.106 , cable and wireless ) changes to a different sequence of hops within as 12389 ( rostelecom ) . by the time of the third snapshot , @xmath32 , the path changes back to that seen in @xmath33 . while the exact cause of this short - lived routing change is unknown , the key point is that it would not have been discovered by the existing topology mapping systems .
we leave a comprehensive analysis of the extent and duration of these short - lived dynamics exposed by yarrpto future work .
yarrpdemonstrates a new technique for internet - scale active topology probing that permits rapid collection of topology snapshots . as with our initial investigation of short - lived topology dynamics ,
our hope is that yarrpfacilitates analyses not previously possible .
yarrpis stable and the code is publicly available @xcite .
that said , there are several enhancements that would be valuable additions .
first , yarrpcurrently only supports ipv4 probing .
given the vastly larger ipv6 address space , and relative topological sparsity @xcite , adding ipv6 support to yarrpcould enable more complete maps of the ipv6 topology to be gathered .
second , yarrpis currently capable of sending only tcp probes .
it is well - known that using different transport protocols yields different responses , due to widespread security and policy filtering @xcite .
we plan to add icmp and udp probing to yarrp , which requires utilizing different transport header fields to encode probe information .
doing so is non - trivial as we must maintain both the paris - traceroute property of keeping certain fields constant to keep packets on a single load - balanced path , while also retaining yarrp s stateless behavior .
third , yarrp s stateless and asynchronous nature implies that a malicious actor could attempt to send bogus responses , while middleboxes are known to mangle packet headers @xcite . in the future
, we wish to use a keyed cryptographic integrity function over multiple probe values . instead of
a simple checksum on the target ip address , we will populate the source port with the value of this keyed integrity check .
yarrpcan then ensure that it both sent the original probe , and that the probe was not modified in - flight such that the response is not useful . finally , an attractive feature of yarrp s design is the ability to easily randomize and distribute the probing to multiple vantage points with negligible coordination and communication overhead .
similar to the rapid scanning worm envisioned by staniford et al . , the permuted domain can be distributed @xcite . for @xmath42 vantage points , and a given domain @xmath4 , each vantage point encrypts @xmath43 of the range @xmath14 .
thus , the @xmath44th vantage point encrypts the range of values @xmath45 to @xmath46 to obtain its sequence of + @xmath16 values to probe from the overall permutation .
the potential speed improvement is linearly proportional then to the number of vantage points . only the values @xmath47 and @xmath44 need be sent to each vantage point to distribute the permuted space and achieve complete randomized coverage .
given our empirical ( and conservative ) @xmath23 yarrprate in this work , we estimate that it is possible to implement a distributed yarrpamong @xmath2128 vantage points to traceroute to every routed ipv4 address ( @xmath48 targets ) in approximately one hour .
yarrpmay thus facilitate rapid collection of _ complete _ internet snapshots in the future . | obtaining a `` snapshot '' of the internet topology remains an elusive task .
existing active topology discovery techniques and production systems require significant probing time time during which the underlying network may change or experience short - lived dynamics . in this work ,
we consider how active probing can gather the internet topology in _ minutes _ rather than days . conventional approaches to active topology mapping face
two primary speed and scale impediments : i ) per - trace state maintenance ; and ii ) a low - degree of parallelism . based on this observation
, we develop yarrp(yelling at random routers progressively ) , a new traceroute technique designed for high - rate , internet - scale probing .
yarrpis stateless , reconstituting all necessary information from icmp replies as they arrive asynchronously . to avoid overloading routers or links with probe traffic , yarrprandomly permutes an input @xmath0 space .
we run yarrpat @xmath1 , a rate at which the paths to all /24
s on the ipv4 internet can be mapped in approximately one hour from a single vantage point .
we compare yarrp s topological recall and discovery rate against existing systems , and present some of the first results of topological dynamics exposed via the high sampling rates yarrpenables . |
the dynamics of general relativity is deeply intertwined with space time diffeomorphisms .
this becomes apparent in the canonical formulation where hamiltonian constraints generate the dynamics and at the same time space time diffeomorphisms @xcite . moreover requiring a canonical theory of spatial geometries and a representation of the dirac algbra the canonical incarnation of diffeomorphisms results uniquely @xcite in canonical general relativity .
hence a quantum theory of geometries with a quantum representation of diffeomorpisms should yield a quantum theory of gravity with the correct semi classical limit .
unfortunately such a representation seems to be out of reach yet .
one reason out of many is that many models , such as regge calculus @xcite and spin foam models @xcite , involve an explicit discretization of the space
time manifold as a regularization method .
generically this seems to result in a breaking of the continuum diffeomorphism symmetries .
this can be taken as an advantage as such symmetries lead to divergencies in the path integral and constraints in the canonical formalism and complicate therefore the construction of a quantum theory .
the diffeomorphism symmetries can however not be completely ignored as these have to be recovered in the continuum limit . even in discrete theories there
should exist at least approximate symmetries , which should become exact in the limit .
path integrals for constrained theories such as general relativity can be understood as providing a projector on the space of physical , that is gauge invariant , states of the canonical theory .
these are states that are annihilated by the constraints .
the heuristic idea is that path integrals implement averagings of the wave functions over the gauge orbits .
this can only be made consistent if the symmetries of the canonical and the covariant theory can be brought to match .
in particular approximate symmetries in the action should correspond to some notion of approximate constraints .
we therefore start in section [ action ] with a review and discussion of symmetries for discretized actions , in particular the regge action
. we will clarify that it is not sufficient to identify infinitesimal variations leaving the action invariant .
a further requirement is that the associated transformations have to act non trivially on the extrema of the action .
we will find that in general the symmetries defined in this way are a mixture of exact and approximate ones . in section [ triang ]
we discuss additional symmetries and problems that can arise if one includes a sum over triangulations into the path integral .
in section [ canon ] we will line out how to obtain canonical formulations that match the symmetries of the discretized covariant theory . to this end
we use the method of ` consistent discretizations ' @xcite together with a local evolution scheme for the dynamics of triangulated manifolds @xcite . in this way we can make the idea of approximate constraints precise . as a simplified example we will consider 3d regge calculus with and without cosmological constant in section [ mini ] . section [ dirac ] discusses notions of dirac s algebra for discrete manifolds . in particular we point out that exact versions exist for 3d gravity and for 4d theories describing flat space . since ` chunks of flat space ' can also appear in the full 4d case and moreover in the continuum limit curvature
should become small compared to the discretization scale such a finding could ensure that an exact 4d dirac algebra can be recovered in the continuum limit .
in section [ lqg ] we review the role of diffeomorphsim symmetries in loop quantum gravity and spin foam models and sketch consequences of our earlier discussions for these theories .
we conclude with a discussion and outlook in section [ laber ] .
the examples we will consider are taken from regge calculus
. we will therefore give a very short introduction and fix notations .
more details can be found in @xcite and references therein .
regge calculus is usually considered on a fixed triangulation @xmath0 of a space time manifold .
the variables appearing in the regge action are given by the edge lengths @xmath1 .
this completely specifies the geometry of the triangulation . from the edge length
one can compute the 4d dihedral angles @xmath2 and 3d dihedral angles @xmath3 . for a 4simplex @xmath4
the dihedral angle @xmath2 gives the ( inner ) angle between the two tetrahedra sharing the triangle @xmath5 .
a similar definition applies to the 3d angles in a tetrahedron @xmath6 at an edge @xmath7 . consider a triangle @xmath5 in a 4d triangulation .
this triangle is shared by several 4simplices .
a ( levi civita ) parallel transport of a vector from one 4simplex to the next around the triangle results in a rotation of this vector by the so called deficit angle @xmath8 in the plane perpendicular to this triangle .
the deficit angle measures the curvature concentrated at the triangle .
accordingly the ( euclidean ) regge action with cosmological constant @xmath9 is given by s=-_t a_t _ t + _
v_where @xmath10 denotes the area of a triangle @xmath5 and @xmath11 the volume of a 4simplex @xmath4 .
we work in units with @xmath12 . if there is a non vanishing boundary the boundary term s_=-_t a_t ( -_t^_t ) has to be added to the action in order to make the boundary value problem ( with prescribed edge lengths ) well defined .
the equations of motion can be obtained by varying the regge action with respect to the length variables and are given by -_t e _
t l_e ^t_e + _ e = 0 where @xmath13 is the angle in the triangle @xmath5 opposite the edge @xmath7 .
derivatives of the deficit angles vanish due to the schlfli identity ensuring that @xmath14 for variations of the dihedral angles in a 4simplex .
this is analogous to the continuum where for the variation of the einstein
hilbert action @xmath15 the term with the variation of the ricci tensor leads to a total divergence . for 3d regge calculus
we have the action [ 3dl ] s=-_e l_e _ e + _ v_+_e l_e(-_e _ e^ ) yielding the equations of motion -_e+_e = 0 . here
@xmath16 is the 3d deficit angle .
if the cosmological constant @xmath17 the equations of motion require the deficit angles to vanish , hence in this case solutions are given by flat space triangulations .
regge calculus seems to circumvent the problem of choosing coordinates and to deal directly with discretized geometries .
hence the title of regge s paper @xcite ` general relativity without coordinates ' .
this however does not guarantee that gauge degrees of freedom will not appear .
also in the continuum the problem is not so much in choosing coordinates , but that different metric fields which seem naively to be expressed in the same coordinates describe the same geometry .
that gauge degrees of freedom appear in regge calculus was first shown by roek and williams @xcite . in a perturbative expansion of the 4d path integral around flat space non propagating modes have been identified and it was realized that these correspond exactly to a discretization of the continuous gauge degrees of freedom .
these modes have a simple explanation : any triangulation of flat space is a solution of regge s equation .
an arbitrary vertex of such a triangulation can be translated in four directions , keeping the triangulation flat but inducing the length of the adjacent edges to change .
these are exactly the gauge modes . for a general non
flat triangulation the issue is naturally more involved .
many authors @xcite pointed out that there should be at least an approximate notion of symmetry if the deficit angles ( determined by the edge length compared to the curvature radius ) are small .
the arguments rest on the fact that the ( linearized ) bianchi identity holds approximately , which can be used to show that the regge action should be approximately invariant under infinitesimal vertex displacements .
some authors argue even for the existence of exact symmetries of the regge action , notably hamber and williams@xcite . instead of trying to define what the equivalent of a continuous diffeomorphism might be for a discrete manifold , we will take the regge action and hence the discrete system it describes as fundamental and just ask for symmetries of this action , as is advocated for instance by williams @xcite .
that is we look for infinitesimal transformation of the length variables that leave the regge action invariant .
indeed one can convince oneself by a counting argument that for @xmath18dimensional triangulations at least @xmath18 infinitesimal such transformations per vertex should exist.(a similar argument applied to 3dimensional triangulations appeared in @xcite . ) the piecewise linear manifold conditions for an @xmath18dimensional triangulation require that every point has a neighborhood homeomorphic to the open ball in @xmath19 . from this
it follows that a vertex in such a triangulation is shared by at least @xmath20 edges .
hence there should be at least @xmath18 variations of the associated length variables that leave the regge action invariant . for the 3dimensional regge action without cosmological constant such variations were found by freidel and louapre @xcite with the help of the _ exact _ bianchi identities .
these transformations were actually defined in a first order formulation , but can be adopted to 3d regge calculus .
we will repeat the argument here , since it will shed some light on the 4d case . to this end consider the simplest case , that is a vertex @xmath21 that is shared by @xmath22 edges @xmath23 .
we will choose some cartesian coordinate system , which we can associate for instance to the tetrahedron @xmath24 , that has @xmath25 as its edges .
the deficit angles are encoded in parallel transport matrices around the edges @xmath26 . here
@xmath27 start all in the tetrahedron @xmath24 and are expressed in the cartesian coordinates associated to this tetrahedron .
with @xmath28 we denote the parallel transport matrix around @xmath29 starting and ending in some other tetrahedron but conjugated with the parallel transport to @xmath24 . in this way all parallel transports can be made to start and end at the same point and are expressed in the same coordinate system .
we will denote the resulting matrices by @xmath30 .
as one can check easily the composition of the loops around the edges @xmath31 to @xmath29 can be contracted to a point without crossing any edges .
hence the associated parallel transport matrix must be trivial , that is we have [ bianchi1 ] _
i u_i : = _ i ( a_i ) = where we defined the lie algebra elements @xmath32 ( assuming that the deficit angles are so small that the @xmath33 are in a neighborhood of unity where the exponential function is invertible ) .
we will need the bianchi identity ( [ bianchi1 ] ) in its logarithmic form , however we have to face the problem that 3d rotations do not commute .
this is the point where for instance miller et al.@xcite apply the ` abelian approximation ' for small deficit angles [ bianchi2 ] _
i a_i 0 , and hence derive an approximate invariance of the action .
freidel and louapre note that one can use the baker campell
hausdorff formula in order to find the exact logarithmic expression for ( [ bianchi1 ] ) in the form [ bianchi3 ] _ i ( a_i + [ _ i , a_i])=0 where @xmath34 are lie algebra elements depending in a complicated way on the elements @xmath35 and therefore the deficit angles @xmath36 .
representing @xmath37 lie algebra elements as @xmath38 with @xmath39 the totally antisymmetric levi - civita tensor , ( [ bianchi3 ] ) turns into [ bianchi32 ] _ i ( a_i^c + a_i^d _
i^e ^dec ) . to find the associated invariance of the action we express the relevant part of the regge action with the help of the cartesian coordinates introduced before as [ bianchi33 ] s_i _ abc a_i^abe_i^c where @xmath40 are the edges expressed in the cartesian coordinates and @xmath41 .
( for the edge @xmath29 we use the coordinates in @xmath24 . )
variations @xmath42 of the coordinates @xmath40 will induce variations in the associated lengths and we obtain [ bianchi34 ] s _ i ^abca_i^abe_i^c .
the terms with @xmath43 vanish .
the first summand is orthogonal to @xmath40 and for the second summand one can use the schlfli identity . with the approximate form of the bianchi identity
we obtain that under the variations [ approx ] _
approx e^c_i : = x^c describing the displacement of the vertex @xmath21 , the action is approximately invariant _
^abcx^c _ i a_i^ab 0 .
but under the variations [ exact ] _
exact e^c_i : = x^c - _ i^cdx^d the action is exactly invariant : _
exacts&&_i ^abca^ab_i ( x^c-_i^cdx^d ) + & & _ i 2x^d(a_i^d + a^c_i_i^e ^ced ) + & = & 0 due to the exact bianchi identity ( [ bianchi32 ] ) .
note that the ` exact ' variations depend via the lie algebra elements @xmath34 on the deficit angles and therefore on the length variables whereas the ` approximate ' variations do not . in the case of a flat vertex , where the @xmath44 the two transformations agree .
as the counting argument presented above suggest it should also be possible to find variations that leave the 3d action with cosmological constant term exactly invariant ( or any other action for instance with higher order terms ) . for the 4d regge action the exact ( contracted ) bianchi identities could be used in a similar way as in 3d to derive invariant variations corresponding to vertex displacements , for alternative actions again the counting argument applies . indeed invariant variations can be always found .
the problem is however that the transformations defined by these invariant variations might turn out to act trivially on _ solutions _ of the regge equations , where by definition all variations vanish .
consider the example of an action depending on two variables and invariant under 2d rotations in these variables .
due to rotational invariance there will be an extremum at the center of rotation , hence this solution is a fixed point of the symmetry of the action .
there could be of course other additional sets of solutions with a non trivial action of the symmetry . to test whether candidate symmetries act non trivially on solutions one can consider the hesse matrix of second derivatives of the regge action evaluated on these solutions possesses vanishing eigenvalues or not .
the hesse matrix also appears in the perturbative treatment of quantum regge calculus . for the expansion around flat space solutions one can indeed find zero eigenmodes corresponding to vertex displacements @xcite . to study an ( accessible ) example with
non flat solutions , we consider the 3d regge action with non vanishing cosmological constant . with the additional volume term the exact invariance ( [ exact ] ) does not need to hold any more , but as discussed before there should exist three independent variations per vertex that leave the action ( [ 3dl ] ) exactly invariant .
in the case that these variations lead to non trivial transformations also for solutions , we should find zero eigenvalues for the hesse matrix evaluated on these solutions . to simplify the problem as much as possible we consider the boundary of an equilateral tetrahedron with edge length @xmath45 and subdivide this tetrahedron by connecting one inner vertex @xmath21 to the four vertices of the tetrahedron by edges with length @xmath46 . we have to find the extrema of the action as a function of the @xmath47 .
there exist one extremum where all the inner edge lengths are equal @xmath48 .
the length @xmath49 can be found by solving [ eom ] -(2 - 3 ) + = 0 for @xmath17 we obtain @xmath50 .
the equation can not be solved exactly for non
vanishing @xmath9 but one can obtain a perturbative solution in powers of @xmath9 .
the @xmath51 hesse matrix around a configuration with @xmath48 can be calculated @xcite and has a very simple form due to the symmetry of our problem .
all the diagonal entries are equal to a rational function a= ( 2- ) .
also all the non diagonal elements coincide and are given by b= ( - + ) .
the hessian matrix will have three null eigenvalues if and only if @xmath52 . with the equation of motion ( [ eom ] ) and the requirement that @xmath52 we have two conditions for the variable @xmath49 , which can only be satisfied simultaneously for vanishing cosmological constant @xmath17 . for non
vanishing @xmath9 the hesse matrix evaluated at a solution of the equation of motion will not have zero eigenvalues , from which we conclude that the candidate symmetries of the action act trivially on ( at least these ) solutions .
in 4d ( without cosmological constant ) there are zero modes of the hessian at least for every flat vertex . to the best knowledge of the author
there is no explicit test for vertices with curvature in the @xmath17 case , but numerical results @xcite ( involving convergence properties of constraints for flat and non flat solutions ) indicate that generically there are no zero modes
. then different solutions might have symmetry orbits of different size , according to whether they have flat vertices or not . as will be explained in the next section
, there are many curved solutions with such flat vertices , moreover the number of such flat vertices might become large for very fine grained triangulations . additionally one would expect that with fine graining and hence decreasing deficit angles gauge symmetry will be restored also for the non flat vertices . in our example
the only scale available is the cosmological constant @xmath9 ( with @xmath17 the regge equations are invariant under global rescalings ) hence this corresponds to the @xmath53 limit .
the hessian matrix will exhibit a set of smaller and smaller eigenvalues .
the corresponding eigenvectors generate an approximate gauge orbit , along which the action is almost constant .
note that the inverse of the hessian appears in saddle point approximations to the path integral .
indeed in the same way diffeomorphism symmetry is gradually restored the path integral should diverge due to the integration over the approximate gauge orbits .
along such an orbit the true extremum ( or even several extrema if the solution is not unique ) is difficult to locate . even for classical numerical treatments solving for the exact extrema of the action becomes so complicated , that instead one ( gauge ) fixes four length variables per vertex and does not take the corresponding number of equations into account @xcite .
one way to quantize regge calculus is to fix a triangulation and to perform the path integral over all allowed length assignments .
one criticism towards this procedure is that the final results may depend on the choice of triangulation , which is rather an auxiliary structure .
if the triangulation serves as regulator for the continuum path integral the induced continuum path integral measure might be concentrated around geometries which arise by assigning roughly the same length to all the edges @xcite .
the dependence on one choice of triangulation can be avoided by summing over all triangulations .
that is one considers a configuration space @xmath54 labeled by triangulations and length assignments to the edges of the triangulations ( or more generally some set of variables on the different k
simplices ) .
also the space of triangulations needs to be specified in more detail , for instance whether one considers fixed topology or not , or whether the triangulations have to specify the piecewise linear manifold conditions or not .
all these discussions can be summarized in a choice of measure on the configuration space @xmath54 .
regge calculus on a fixed triangulation has then only non vanishing measure on one triangulation .
in contrast the dynamical triangulation measure @xcite is concentrated on triangulations with fixed edge length , that is all the geometric degrees of freedom are put into the choice of triangulation .
group field theories @xcite generate sums over topologies , ` generalized ' triangulations and variables attached to the simplices of these triangulations . with the enlargement of the configuration space
the question arises whether some form of diffeomorphism or more general gauge symmetries act on this space involving configurations based on different triangulations .
for instance the authors @xcite claim that this is actually the case with romer et al.@xcite arguing for a gauge fixing procedure on this enlarged configuration space .
indeed for every solution of the regge equations ( with vanishing cosmological constant ) one can introduce trivial refinements of the underlying triangulation and in this way obtain again solutions of the regge equations .
these refinements are trivial since the new edges or triangles do not carry additional curvature , hence the actual geometry is invariant under these refinements . to be more precise , given a solution of the ( four dimensional ) regge equations choose a 4simplex @xmath4 and subdivide this simplex by placing a vertex @xmath21 into this simplex .
connect the vertex @xmath21 to the vertices of @xmath4 .
this introduces five new edges , ten new triangles and five new 4simplices . to
the new edges there exist a four parameter space of length assignments such that the deficit angles at the new triangles vanish , corresponding to the possible positions of the new vertex @xmath21 in the simplex @xmath4 .
also it is easy to see that the regge equations of motion [ reggeb ] _ t e _
t ^t_e=0are still satisfied : for the new edges the sum in ( [ reggeb ] ) involves only deficit angles of new triangles .
as these are vanishing the equations of motion are satisfied for the new edges .
the equation of motions for the other edges do not change either , again because the deficit angles at the new triangles are vanishing .
hence if one attempts to include all possible triangulations into the path integral one has to face these redundant configurations .
one could try to exclude triangulations with trivial subdivisions from the path integral , for instance by summing only over some class of ` simplest ' bulk triangulations that interpolate a given boundary triangulation @xcite .
indeed this works in the ( topological ) 3d case @xcite . to this end , however , it would be valuable to know more about the uniqueness properties of the regge solutions : for instance , given a solution with a subdivided 4simplex , is this 4simplex necessarily flat ?
if this is not the case , restricting to the ` simplest triangulations ' could miss out on physical degrees of freedom .
so far some available results @xcite indicate that solutions to the regge equations are not necessarily unique and can vanish or appear under subdivisions or coarse graining .
a more systematic study of the behavior of regge solutions under subdivisions would be valuable , as this could be helpful for obtaining the continuum limit .
if the conjecture holds , that subdivisions beyond some scale ( determined by either the curvature scale of the boundary triangulation or the cosmological constant ) are trivial , that is only add gauge equivalent solutions , it might not be necessary to include all triangulations into the path integral to capture essential physical properties
. it might be even sufficient to restrict to a fixed triangulation , which is fine enough , i.e. solutions should have mostly vanishing deficit angles . summing over all kinds of length assignments on
this triangulation would include configurations that are equivalent to geometries defined on coarse grained triangulations .
these terms should also arise if one chooses to start with another equally fine triangulation .
hence one would expect physical observables only including scales above the discretization scale to be independent from the concrete choice of this triangulation .
there are several approaches to obtain a canonical or ( 3 + 1 ) version of regge calculus , on the one hand in order to develop a numerical scheme for solving the initial value problem and on the other for canonical quantization . the main difficulty in developing a canonical scheme is firstly the special status of the time variable and furthermore the question of the existence or non existence of gauge symmetries .
both points reappear in some form in loop quantum gravity and spin foam models , so it might be worthwhile to discuss these .
there are several options to define a canonical version of regge calculus .
one is to start from the discrete 4dimensional regge action and to come up with a canonical , that is initial value formulation that reproduces the 4dimensional regge solutions .
this can be summarized as first discretizing and then performing a @xmath55splitting .
another option is to perform first the @xmath55 splitting and then to discretize . in the latter case
the additional question arises whether to aim for a continuous time formulation or an evolution in discrete time steps . in @xcite the second viewpoint
is adopted , that is to consider the time evolution of a simplicial 3manifold in continuous time .
basically the @xmath55 form of the action is evaluated on a three dimensional triangulation .
the resulting canonical variables are the length of the spatial edges , conjugated momenta ( which in the version @xcite are in a non local relation with respect to the extrinsic curvature leading to a non local expression for the hamiltonian ) and lapse and shift functions associated to the tetrahedra .
accordingly for each tetrahedron there is one hamiltonian and three momentum constraints . whereas in the continuum
these constraints are first class , indicating the gauge freedom to choose lapse and shift freely , this is not the case for the discretized constraints , which are second class . to obtain a consistent evolution in which the constraints are preserved during time evolution one has to fix@xcite the values of lapse and shift , hence there is no gauge freedom left .
that first class constraints turn under discretization into second class ones is a general phenomenon and appears not only in regge calculus .
for instance this happens also for the spatial diffeomorphism constraints@xcite in a lattice version of loop quantum gravity .
as mentioned in the introduction it would be valuable to have a canonical formalism that matches the amount of symmetries inherent in the action . a strategy to ensure this is to construct a canonical formalism that reproduces exactly the solutions of regge calculus .
to this end we consider 4d triangulations and as in a continuum have to choose a slicing of these triangulations into constant time surfaces .
such slices can be defined so that constant time surfaces are built from the 3d boundaries of the 4simplices , and such that every vertex of the 4d triangulation is part of at least one constant time surface .
note that there might be many choices of possible slicings for a given triangulation .
if the canonical formalism could reproduce all such slicings it would ensure a discrete version of the multi fingered time picture @xcite in continuum general relativity .
but generically one will encounter the problem that constant time surfaces in the same triangulation might have a different number of edges and hence would lead to phase spaces of different size .
the time evolution map would then have to act between different ( sized ) phase spaces .
it would be of interest to have a formalism that can accommodate such a feature in a canonical formalism , as it might be relevant in particular for cosmological application .
for the time being we restrict ourselves to cases where the 3d triangulation is the same for every time step .
this restriction could lead to very rigid time evolution .
fortunately this is not the case , as in the so called puff pastry evolution scheme @xcite an arbitrary 3d triangulation can be evolved in many different ways , so that consecutive 3d triangulations do not change .
that is assume that we have a 4d solution of the regge equations with a 3d boundary representing the equal time surface .
this solution can be evolved further by choosing a vertex @xmath21 in the boundary to which an edge ( the ` tent pole ' ) is glued .
the other vertex of this edge has then to be connected to the vertices in the 3d boundary adjacent to @xmath21 . in this way
a tent is erected and we pushed the hypersurface locally forward in time .
since we can choose ( apart from restrictions arising from causality in lorentzian space - times ) at which vertices to evolve the hypersurface this evolution scheme is very reminiscent of a discrete version of multi fingered time .
the question is whether we can find also a continuous or approximate version , that is the freedom to choose lapse and shift at each vertex .
indeed there are numerical indications that this is the case @xcite . evolving a vertex amounts to solving the regge equations associated to the new inner edges , that is the pole edge and the edges that are adjacent to the vertex @xmath21 in the initial hypersurface .
these equations have to be solved for the length of the pole edge and the length of the new boundary edges , hence we have as many equations as variables to solve for .
the approximate freedom in choosing lapse and shift means that these equations are ` approximately ' dependent , i.e. values that satisfy all of these but four , will satisfy approximately the remaining four .
note that these equations for the new inner edges depend on the initial data , that is the edge length in the initial hypersurface and to the past of it . as in the continuum
one would expect that these data satisfy constraints ( at least in some approximate sense ) , and only if the initial constraints satisfy these constraints the equations for the new edge length will be approximately dependent .
these constraints describe ` initial data ' on 3d hypersurfaces that can arise as 4d boundaries of regge solutions . in the continuum the hamiltonian and diffeomorphism constraints
are given as the contraction of the einstein tensor with the normal to the hypersurface as @xmath56 and @xmath57 respectively .
the regge equations associated to an edge @xmath58 can be understood as the double projection of the einstein tensor @xmath59 to this edge .
therefore we will expect that one of the four constraints is given by the regge equation associated to the tent pole . to make this more explicit we have to first find a suitable set of initial data , that is phase space variables that can be associated to a 3d triangulation and then have to be evolved in discrete time steps to reproduce a regge solution . to this end we can adopt the formalism of consistent discretization @xcite as we will see it amounts essentially to a rewriting of the regge equations of motion as canonical transformations on a phase space with length and conjugated momenta . the regge action ( with boundary terms ) for the 4d tent region
serves as a generating function for the canonical transformation implementing the discrete time evolution .
this will ensure that the symmetries of the regge action are reflected in the canonical formalism .
the consistent discretization scheme has been applied to 3d regge calculus ( with non vanishing cosmological constant ) by gambini and pullin @xcite .
but there a hyper - cubical lattice was chosen making the analysis very complicated .
the advantage of the puff pastry evolution scheme is that we can reduce our considerations to a very small region , in this way we have to deal only with a very small number of equations .
a detailed treatment of the 4d dimensional case will appear elsewhere @xcite , here we will consider a simplified 3d example corresponding to our earlier discussion in section [ action ] . as we will see the exact and approximate symmetries discussed there are reflected in the canonical formalism .
the local evolution scheme describes above allows us to resort to a very simple example in order to illustrate the mechanism and the main points .
consider 3d regge calculus with and without cosmological constant .
( the results do not change for the analogous example in 4d ) .
the equal time surfaces in our example will be given by boundaries of tetrahedra , moreover to simplify calculations we require that the three edges of the ` base ' of the tetrahedron have all length @xmath45 and the three edges meeting at the ` tip ' of the tetrahedron have all length @xmath49 .
we evolve these surfaces by the evolution described above , that is we join an edge ( the pole ) with length @xmath5 to the tip of the tetrahedron and connect the other end of this edge to the vertices of the base , see picture [ fig1 ] .
a three valent vertex is evolved in discrete steps by erecting a pole and connecting the other vertices to the end of the pole .
the different types of line indicate the different time steps.,width=151 ] hence we have two configuration variables @xmath49 and @xmath5 , the edge length @xmath45 does not change in time and therefore is merely a parameter of the system . performing a step from time @xmath18 to time @xmath60 three simplices are added that hinge together at the pole .
the action ( including boundary terms ) for this simplicial complex is given by s_n(b_n , b_n+1,t_n)&= & -b_n ( -2 ( b_n , b_n+1,t_n ) ) + & & -b_n+1 ( -2(b_n+1,b_n , t_n ) ) + & & -t_n ( b_n+1,b_n , t_n ) + & & + v(b_n , b_n+1,t_n ) where @xmath61 are the dihedral angles at the spatial edges ( note that due to the symmetry of our problem we can use the same functions at two consecutive time steps , but we have to exchange the arguments ) and @xmath62 is the deficit angle at the pole .
the volume of the three tetrahedra is denoted by @xmath63 .
these geometric quantities are given by ( b , b,t)&= & ( ) + ( b , b,t ) & = & 2 - 3 ( ) + v(b , b,t ) & = & .
we expect the length @xmath5 of the time like edge to act as the lapse , the shift is vanishing due to our symmetry requirements . following the consistent discretization scheme @xcite we rewrite the equations of motion for the spatial edge length @xmath64 and the tent pole @xmath65 0 & = & ( s_n-1(b_n-1,b_n , t_n-1)+s_n(b_n , b_n+1,t_n ) ) + 0 & = & s_n(b_n , b_n+1,t_n ) as a canonical transformation given implicitly by p_n^b & & - [ e1 ] + & = & 3(-2 ( b_n , b_n+1,t_n ) ) - + p_n^t & & - [ e2 ] + & = & ( b_n , b_n+1,t_n ) - + p_n+1^b & & [ e3 ] + & = & -3(- 2 ( b_n+1,b_n , t_n ) ) + + p_n+1^t & & [ e4 ] + & = & 0 . derivatives of the angles vanish due to the schlfli identity .
equation ( [ e4 ] ) demanding vanishing of the @xmath5momentum , is a constraint .
it corresponds to the continuum result that the momentum conjugated to the lapse is vanishing . as in the continuum requiring that this constraint is preserved in time results in a secondary constraint , namely that the right hand side of ( [ e2 ] ) should be equal to zero . in the continuum
this secondary constraint is the hamiltonian constraint . to obtain a constraint on the canonical data with the same evolution label @xmath18
, we have to solve equation ( [ e1 ] ) for @xmath66 as a function of @xmath67 and @xmath65 and to use this expression in ( [ e2 ] ) . [ con1 ] c_(b , p^b , t)=c_0(b , p^b)+f_(b , p^b , t ) = 0 . in the case
@xmath17 this constraint does not depend on @xmath5 [ con2 ] c_0(b , p^b)=2 - 3 ( ) . that the constraint depends only on @xmath49 and @xmath68 expresses the dynamical fact that the equal time surfaces always bound a flat tetrahedron .
this tetrahedron might be subdivided into smaller ones , but since the deficit angles have to vanish the details of the subdivision do not matter .
the momenta conjugated to the variable @xmath49 are given by the dihedral angles of this tetrahedron , which can be computed knowing only the edge length @xmath49 ( and the edge length @xmath45 ) .
the constraint ( [ con1 ] ) is preserved under time evolution , that is if @xmath69 vanishes this holds also for @xmath70 with @xmath65 arbitrary .
we obtain a consistent evolution , where @xmath71 is constrained as a function of @xmath64 but the lapse @xmath65 can be chosen arbitrarily at every time step . as in the continuum where the lapse function can be dropped from phase space
we can omit the variable @xmath5 and are left with the edge length @xmath49 and conjugated momentum @xmath68 .
there is one constraint between the variables generating a one dimensional gauge orbit .
hence the number of physical degrees of freedom is zero .
the same result holds for arbitrary complicated triangulations ( of spheres ) in 3d gravity without cosmological constant @xcite and can also be extended to four dimensions for a special class of hypersurface triangulations leading to a dynamics of flat 4dimensional space @xcite , see the discussion in the next section . the time steps @xmath65 can be chosen arbitrarily , that is also arbitrarily small .
in this way we can recover a continuous time formalism in which the constraint ( [ con2 ] ) does generate time evolution with arbitrary choice of lapse .
this reflects the continuous group of symmetries acting on the space of solutions of the covariant theory . for non
vanishing cosmological constant the constraint ( [ con1 ] ) does depend on the variable @xmath65 .
an analytical solution is not possible for @xmath72 , but one can make a perturbative ansatz @xmath73 for the solution of equation ( [ e1 ] ) and use it to obtain the constraint ( [ con1 ] ) .
the constraint depends on @xmath65 starting with the quadratic order in @xmath9 .
given some initial data @xmath74 and @xmath75 one can solve the constraint for @xmath76 and evolve to the next time step where again one has to solve the constraint for @xmath77 and so on .
the time intervals @xmath65 are therefore determined by the chosen initial data @xmath74 and @xmath75 .
this allows to define a reduced system by solving the constraints at all time steps for the time intervals @xmath65 as functions of @xmath64 and @xmath78 and to insert these solutions in the remaining time evolution equations .
this has the advantage of getting rid of the constraints as well as some of the technical and conceptual difficulties connected to these .
this ` consistent discretization ' approach has been proposed and developed by gambini , pullin and collaborators @xcite .
in particular the application of this scheme to 3d regge calculus ( with cosmological constant ) has been outlined in @xcite , choosing a rigid slicing based on a hyper - cubical lattice .
note however that this strategy does not work in instances where the discrete action exhibits symmetries , as in the case @xmath17 ( even partly in four dimensions ) . for 3d gravity
the question arises how to obtain the continuum limit , which has a finite number or even zero physical degrees of freedom , starting with a discrete theory with many physical degrees of freedom .
moreover this limit would involve an increasing number of edges and in the scenario without constraints therefore degrees of freedom .
on the other hand from the discussion of approximate symmetries one would expect that these symmetries should become exact in the continuum limit , with the constraints reappearing in some way .
this indicates that choosing the ( spatial ) edge lengths and conjugated momenta as free variables by solving for the time like edge lengths may lead to a bad parametrization of the dynamically allowed configurations .
the constraints ( [ con1 ] ) could either have no solutions @xmath65 at all or lead to rather large edge lengths @xmath65 , spoiling the continuum limit .
there is only a small range of variables @xmath79 that will cause small time intervals .
for this reason we advocate to keep the constraints explicitly and use them as a controlling device for the continuum limit .
that is at the initial time step we keep @xmath80 as a free parameter and find initial values @xmath74 and @xmath81 satisfying the constraint .
these data have to be evolved using the equations ( [ e1]-[e4 ] ) .
the difference to an exact reparametrization invariant theory ( as the @xmath17 case ) is that now @xmath82 _ and _ @xmath65 are fully determined by the evolution equations for @xmath83 . in the case that @xmath84 , that is if the time intervals remain small , a ( time ) continuum limit @xmath85 can be obtained .
numerical evolution schemes @xcite for regge equations follow an opposite strategy in choosing the equivalents of lapse and shift freely and ignoring in every time step four equations of motion which are therefore treated as constraints .
one hopes that the violation of these constraints does not grow during evolution and tends to zero in the continuum limit .
indeed numerical evidence for this scenario has been found . in this example
the constraints are implemented exactly and the question is how the time steps grow .
the same numerical evidence can be taken as an indication that if the initial data are chosen carefully , these remain bounded through evolution , although this deserves further study .
to keep the constraints explicitly is closer to the continuum time canonical formalisms with dynamically determined lapse and shift mentioned before and to the viewpoint developed in the work @xcite which argues that discretizations will in general induce complicated gauge fixings .
but the examples considered in this paper differ from the approach followed here . the work @xcite starts directly with a hamiltonian formalism in which there are explicit constraints apriori not involving the time intervals @xmath65 .
evolution of these constraints under a finite canonical transformation leads however to the appearance of @xmath65 in the time evolved constraints .
conservation of the constraints in time require the @xmath65 to be ( gauge ) fixed .
here we start from a discrete action and end up with constraints @xmath86 depending on the time intervals directly .
we want to stress that this is the main difference to a continuous time canonical formalism , in which one assumes exact constraints , independent of lapse and shift .
the discrete time scenario makes the notion of ` approximate constraints ' precise : we can see these as a family of exact constraints labeled by lagrange multipliers . choosing these lagrange multipliers in a small interval ( around zero ) we obtain as allowed configurations for the remaining phase space variables a ` thickened constraint hypersurface ' .
if in the continuum limit the approximate symmetries are restored also the constraints should become exact : one can keep either the range for the lagrange multipliers fixed , then the ` thickness ' of the constraint hypersurface should decrease . or we keep the ` thickness ' constant , then the range of lagrange multipliers should increase until the full freedom of their choice is restored . in the quantum theory
it is therefore consistent to impose the constraints only in an approximate sense .
for instance in a path integral the integration over the lagrange multipliers should result in an exact projector on the constraints only if there is a solution ( saddle point in a semi - classical approximation ) to every choice of lagrange multiplier .
a path integral with approximate symmetries can only impose constraints approximately on the canonical states .
dirac s hypersurface deformation algebra [ alg ] \{h[m],h[n]}&=&d[g^ab(m_b n - n_b m ) ] + \{d[m^a],h[n]}&=&h[m^b_b n ] + \{d[m^a],d[n^b]}&= & d[m^b_b n^a - n^b_b m^a ] can be understood as the canonical incarnation @xcite of the space
time diffeomorphism group . here
@xmath87=\int_\sigma n(\sigma ) h(\sigma){\mathbf d}\sigma$ ] is the hamiltonian constraint integrated with a lapse function @xmath88 over a spatial hypersurface @xmath89 whereas @xmath90=\int_\sigma
n^a h_a(\sigma){\mathbf d}\sigma$ ] are the spatial diffeomorphism or vector constraints integrated with the shift vector field @xmath91 .
the algebra is very complicated as we have in the first relation of ( [ alg ] ) a structure function appearing that involves the inverse @xmath92 of the spatial metric .
the hypersurface deformation algebra is describing a universal @xcite relationship between infinitesimal deformations of a hypersurface in the embedding space .
the hamiltonian constraints generate deformations of the hypersurface normal to itself whereas the vector constraints generate tangential deformations , that is spatial diffeomorphisms .
the universality , i.e. independence from the field content of the theory , follows from the fact that the algebra can be derived by purely geometric considerations @xcite .
we want to stress that therefore such a notion of hypersurface deformation algebra should also exist for discretized geometries . also the structure constants and structure functions should be derivable by geometrical considerations .
another question is whether it has universal properties and whether there are classical and quantum representations of this algebra realized by phase space generators and operators respectively .
it has been shown @xcite , that under certain assumptions including the nature and number of phase space variables , canonical general relativity provides the unique phase space representation of the algebra ( [ alg ] ) .
a quantum representation of this algebra could therefore ensure the correct classical limit of the theory . because of the intricacies and unusual form of the algebra this however has not been achieved so far for any theories , even free parametrized field theories@xcite in space time dimension larger than two .
given the significance of dirac s algebra for the continuous theory it would be valuable to have a notion of a hypersurface deformation algebra for discrete geometries @xcite . also here
two different notions are possible , one with continuous and one with discrete time or deformation parameters .
the next question is to construct ( at least classical ) representations , that is canonical theories exhibiting these symmetry algebras .
of course it would be interesting to know in which sense the universality features and the uniqueness proof can be carried over to the discrete cases . for continuous deformation parameters the easiest case to consider are hypersurface deformation algebras describing embeddings of piecewise linear hypersurfaces into flat space .
as in the continuum the algebras itself can be defined in any dimension by purely geometric considerations , that is by deriving the commutation relations of infinitesimal deformations of such hypersurfaces .
one question would be how to define such deformations .
the most elementary notion ( and the one matching the symmetries encountered so far in covariant regge calculus ) is to use vertex displacements , i.e. to change the position of one vertex in the embedding space and to accordingly adjust the length of the adjacent edges .
displacements of higher dimensional simplices can be combined out of vertex displacements . in the continuum
dirac s algebra is defined by using two types of geometric deformations , the ( normalized ) deformations normal to the hypersurface and tangential ones .
such a splitting is ambiguous at a vertex in a piecewise linear geometry , so some choice for a normal to the vertex has to be done .
the final hypersurface deformation algebra will depend on this choice , details will appear in @xcite . note that even a tangential vertex deformation will change the geometry of the spatial hypersurface ( with the exception of intrinsically flat vertices ) , whereas in the continuum the spatial diffeomorphism constraints do not change the intrinsic geometry . nevertheless for instance for three dimensional gravity all three vertex deformations are realized as constraints and are needed to obtain the correct counting of physical degrees of freedom for 3d gravity .
the naive viewpoint that the spatial triangulation represents the geometry of the hypersurface , i.e. that the spatial diffeomorphism have been already factored out completely , leads to an incorrect result , see also the discussion in @xcite .
this leads us to the next question , namely if there exist any representations of these algebras . in three dimension
such representations are provided by several canonical formulations of discretized three dimensional gravity @xcite , which can be understood as the dynamics of hypersurfaces embedded into flat 3d space .
one formulation for ( euclidean ) 3d gravity is in terms of @xmath93 connection variables , see for instance @xcite .
the phase space variables are given by @xmath93 holonomies and triad variables associated to the edges of the triangulations .
the dynamics is given by two sets of first class constraints , the gau constraints and the so called flatness constraints .
the gau constraints can be used to reduce the phase space to @xmath93 invariant variables giving the length of the edges and conjugated momenta providing a canonical formulation of 3d regge calculus@xcite .
( the example in section [ mini ] with vanishing cosmological constant is a symmetry reduction of these theories . )
the flatness constraints generate vertex displacements .
a priori the flatness constraints are abelian and do not have a geometric invariant meaning as normal or tangential generators .
such a split can however be introduced by using appropriate linear combinations ( with phase space dependent coefficients ) of the flatness constraints .
the new algebra is still first class and reflects the geometric commutation relations of the infinitesimal deformations . in four dimension
a representation for such an algebra can be defined for a restricted class of spatial three dimensional triangulations @xcite .
this subclass can be understood as a sub - sector of regge calculus , whose triangula tions only allow flat solutions ( see also the papers @xcite for defining a flat geometric sub - sector in a canonical formulation of bf theory ) .
the reason for the restricted class of triangulations is the following : a 2d triangulation with prescribed edge lengths can be generically ( ignoring triangle inequalities , self intersections and similar issues ) embedded into 3d flat space .
this reflects the dynamical content of 3d gravity : given a 2d metric the constraints can be solved ( locally in phase space ) for the momenta which define an embedding of the corresponding surface in flat 3d space .
a generic 3d triangulation with prescribed length can not be embedded into flat 4d space , as this would require constraints on the length variables @xcite .
however there is a special class of triangulations , such as the boundary of a 4simplex , which can be embedded into 4d flat space .
more complicated triangulations with this property can be generated by applying @xmath94 pachner moves to this triangulation , i.e. by placing a vertex into a tetrahedron , subdividing it thus into four new tetrahedra .
a more complete discussion can be found in the paper @xcite here we will just give the canonical theory describing the ` dynamics ' of a 4simplex boundary embedded in flat space .
the canonical variables are given by the ten edge length @xmath95 denoting vertices and conjugated momenta @xmath96 .
a canonical transformation can be performed ( in local patches of phase space ) to area variables @xmath97 and conjugated momenta @xmath98 .
the dynamics is then defined by the ten constraints [ triangle ] c_t:=p_t - _ t(l ) where @xmath99 is defined to be the dihedral angle at the triangle @xmath5 in a flat 4simplex . due to the schlaefli identity
the constraints are first class and hence generate gauge transformations .
these gauge transformations correspond to the vertex deformations of the 4simplex .
since there are 5 vertices in a 4simplex we have 20 possible deformations , however there are 10 combinations corresponding to rigid translations and rotations of the 4simplex which do not change the geometry of the 4simplex .
taking the following linear combinations of constraints
c_e:=_t c_t = p_e-_t _ t ( l ) we obtain generators that change only one length of one edge at a time ( and therefore commute ) .
the constraint @xmath100 induce therefore a displacement of either of the vertices in the direction normal to the other edges adjacent to that vertex .
( the ambiguity in associating a vertex stems from the fact that a transformation at both vertices amounts to a rigid transformation of the 4simplex ) .
taking linear combinations of these constraints we can define generators leading to arbitrary vertex displacements and compute a deformation algebra .
note that in equation ( [ triangle ] ) the triangle areas seem to appear as an easier set of variables to deal with .
the conjugated momenta are directly related to the dihedral angles , encoding the extrinsic curvature of the triangulated 3d surface .
indeed also in loop quantum gravity areas and 3d dihedral angles between triangles arise as the simplest kind of geometrical variables .
these variables provide however an over - complete set and need therefore to be constrained in order to yield consistent length assignments to the edges of the triangulations .
fortunately such constraints are available for 4d triangulations @xcite and can be used to define an area
angle regge action .
the same constraints arise for 3d triangulations @xcite and can therefore be used in a canonical analysis of the area
angle regge action @xcite .
the advantage of such a formalism is its closeness to spin foams and loop quantum gravity .
in fact the variables used there are related by a gau constraint reduction @xcite to the area and angle variables .
the incidences where one can define representations of an ( infinitesimal ) hypersurface deformation algebra correspond exactly to the known appearances of non trivial symmetries in the corresponding covariant versions , that is regge calculus .
if there are no corresponding symmetries in the action , it is questionable that a first class constraint algebra reflecting infinitesimal hypersurface deformations of discretized hypersurfaces can be defined exactly , as this would generate solutions not present in the covariant theory .
it might exist in an approximate sense , that is reflecting the algebra ( [ alg ] ) modulo anomalous terms vanishing in the continuum limit .
this has been shown for the spatial diffeomorphism constraints in a lattice approach to loop quantum gravity @xcite . in a quantum theory
a deformation algebra might rather involve finite transformations .
since the spectra of spatial geometric quantities in loop quantum gravity are discrete @xcite there might appear a similar discreteness in time direction .
in the examples discussed so far the choice at which vertices to apply a tent move in the puff pastry evolution scheme @xcite can be understood as choosing lapse to be either equal to zero or equal to some dynamically determined value . in a quantization
such a move could give rise to a finite ( unitary ) evolution instead of a hamiltonian viewed as infinitesimal generator .
similarly the spatial diffeomorphism constraints in loop quantum gravity do not exist as infinitesimal generators but only as finite unitary operators .
this understanding of the hypersurface deformation algebra corresponds to different choices of triangulations in the covariant theory .
the puff pastry evolution scheme will generate a certain subclass of triangulations .
a complete equivalence can be obtained if one considers all the allowed pachner moves in the @xmath101dimensional hypersurface .
these can be related to gluing or removing @xmath18simplices from the hypersurface thus building up an @xmath18dimensional triangulation @xcite .
the discrete time notion of hypersurface deformation algebras therefore depends on a better understanding of any symmetries associated with the choice of triangulations .
in this section we will consider implications of the results collected in the previous section for quantum gravity models , in particular spin foams and canonical loop quantum gravity @xcite .
similar to quantum regge calculus spin foam models are based on a path integral over discretized geometries .
what differs from ( traditional ) quantum regge calculus is the choice of variables and the discrete integration measure over these variables .
this leads to a sum over spin labels ( @xmath93group representations ) and group intertwiners , associated to subsimplices of the triangulation .
the discrete integration measure is motivated among other things by canonical loop quantum gravity , where the spectrum of kinematical geometric operators is discrete and encoded in such spin labels and intertwiners . to obtain the amplitude associated to a given set of labels one considers the ( discretized ) plebanski action @xcite that formulates general relativity as a second class constrained @xmath102 theory .
the constraints are imposed by different methods @xcite onto the quantum amplitudes .
furthermore the amplitudes depend on a choice of discretized measure which is often left open to some extend in the definition of spin foam models .
so far the strategy to obtain a semi classical description is to consider a fixed triangulation and to consider the large spin limit .
this is a limit where the discretization scale given by the spectra of geometric operators , i.e. the planck length becomes small compared to edge length .
( there is a second discretization scale defined by the underlying piecewise linear triangulation , in other words the typical edge length . ) in this limit one hopes to find an effective description given by the regge action and indeed there are indications that certain models satisfy this criterion @xcite . in so far one can understand spin foam models as a quantization of regge calculus . for other models constructed in a more direct way from first order versions of regge s action see the work @xcite .
the question arises whether the exact or approximate symmetries of the discretized action are reflected in spin foam models , that is in the amplitudes defined by the action and the choice of measure .
there is another viewpoint @xcite of the path integral which makes it absolutely necessary to study the symmetries of the path integral : a path integral for constrained theories such as gravity can be understood as a projector onto physical states , i.e. wave functions that satisfy the hamiltonian and diffeomorphism constraints .
( this is not a proper projector as it will map normalizable wave functions to unnormalizable ones . )
the heuristic idea behind is that the path integral implements an averaging over the gauge orbits generated by the constraints , resulting in wave functions that are constant along these orbits .
this can only work , if the path integral respects the symmetries , in particular if the measure is gauge invariant .
so far this idea is rather heuristic for 4d gravity and has been made only rigorous for 3d gravity @xcite , where the path integral has all the necessary symmetries @xcite .
as was mentioned before a discretized path integral might have a complicated symmetry structure : diffeomorphism or translation symmetries associated to trivial vertices can be implemented exactly , for non trivial vertices there might be only approximate symmetries . here
it would be valuable to know whether this can lead at least to an approximate ` projector ' , that is a gaussian compared to a delta function .
this would be consistent to the discussion in section [ mini ] , where we pointed out that in a discrete theory the allowed initial data are described by a ` thickened ' constraint hypersurface . in the continuum limit , in which the symmetries become exact
, the projector should also become exact .
approximate symmetries are reflected by very small eigenvalues in the hessian of the action , which should converge to zero in the continuum limit .
the integral over the directions corresponding to these eigenvalues is an integral over the lagrange multipliers parameterizing the ( approximate ) gauge orbit .
if the action and the path integral measure is approximately constant along these gauge orbits it should result in a projector onto approximately gauge invariant states .
this behavior should be studied in toy models .
other ideas how to construct a projector on physical states will be mentioned below .
so far we discussed the symmetries of the action , but to obtain a ` projector onto physical states ' also the path integral measure has to respect the symmetries . in the context of spin foam models
this has been discussed extensively by bojowald and perez @xcite .
they derive two kinds of requirements .
the first requirement can be understood to arise from the continuum diffeomorphism symmetry leading to vertex translations in the discretized action .
together with the second class constraints appearing in the plebanski action the constraints associated to the vertex translations ( either second or first class ) form a complicated constraint algebra , which has been so far only computed in the continuum case @xcite .
this constraint algebra is however needed to compute the fadeev
popov determinant to obtain the correct path integral measure .
the second requirement is related to symmetries which involve a change of triangulation and labels ( so called moves ) . in spin foam models labels
are allowed to have zero spin leading to vanishing edge length therefore there are more moves that leave the encoded state unchanged .
the amplitudes associated to two configurations related by such moves should be the same .
this leads to a number of consistency conditions for the amplitudes , in particular restricting the choice of face and edge amplitudes , that are often left open to choice .
furthermore bojowald and perez argue that a consistent choice of amplitudes has to lead to divergencies .
this has to be expected as the path integral includes a sum over non compact gauge orbit . for the 3d spin foam model this has been discussed by freidel and louapre @xcite and the symmetries gauge fixed . for the 4d case ,
if the spin foam model reproduces the symmetries of the regge action , one would expect at least divergencies associated to the exact translation symmetry of trivial vertices , whereas approximate symmetries associated to non trivial vertices could still lead to finite results ( however diverging in the continuum limit ) .
here trivial vertices are those for which saddle points in the regge action exist such that the deficit angles at adjacent triangles vanish .
this depends on the boundary conditions for generic vertices but will always be the case for five valent vertices .
spin foam models are hoped for to provide a projector on diffeomorphism invariant states in canonical loop quantum gravity . on the other hand
such states are defined to be annihilated by the diffeomorphism and hamiltonian constraints , and we will discuss certain aspects of these constraints in the remainder of this section .
loop quantum gravity provides a rigorous construction of a kinematical hilbert space which carries a representation of ( kinematical @xcite ) geometric operators .
as was already mentioned these turn out to have a discrete @xcite spectrum , which is one reason for the choice of a discrete path integral measure for the spin foam models .
this makes it very tempting to associate a discretized piecewise linear geometry to the eigenstates , the so called spin networks , of these operators , a picture close to the construction of spin foam models as discretized path integrals . but loop quantum gravity is a quantization of the continuum theory and does _ not _ proceed by considering discretized classical geometries and then quantizing .
hence the hilbert space spanned by spin networks is the state space associated to continuum geometries . as such one would expect the implementation of the continuum constraints and continuum dirac algebra . indeed
the celebrated flost uniqueness theorem @xcite , showing that loop quantum gravity provides a unique kinematical set up satisfying certain assumptions , requires a representation of the continuous ( more specifically semi
analytic ) spatial diffeomorphisms .
another route is to first discretize the geometry , obtain a canonical description and then quantize using loop quantum gravity techniques , see for instance @xcite where this approach is followed or advocated for . here
one would expect the implementation of the ( possibly second class ) constraints for a discretized geometry .
the relations on the kinematical and dynamical level between these two approaches are not quite clear @xcite and should be explored in more detail .
as spin foam models are derived from a discretized path integral one would expect that boundary states of this path integral are states obtained from the second route of canonical quantization .
also at least before the continuum limit spin foams should project onto the physical states defined by the constraints in the explicitly discretized theory .
group field theories generate sums over all possible triangulations ( or more general complexes ) , in a certain sense these theories should include information about continuum geometries and dynamics .
hence a canonical formulation of these could be connected to a continuum canonical theory related to loop quantum gravity .
the immediate difficulties in obtaining such canonical formulations as well as its potential significance are sketched in the report @xcite .
the full set of symmetries in the covariant version has still to be identified . for the continuum theory of loop quantum gravity a uv
finite quantization of the hamiltonian constraints has been obtained by thiemann @xcite .
the status of the quantum dirac algebra in the ( continuum ) loop quantum gravity is however shrouded by several severe technical difficulties @xcite .
that is the algebra of the hamiltonian constraints is anomaly free , even abelian .
abelianess does not reproduce the constraint algebra ( [ alg ] ) but is consistent as the hamiltonian constraints require for their continuum limit the topology provided by the hilbert space of spatially diffeomorphism invariant states . on this space
the diffeomorphism constraints ( on the right hand side of the classical poisson bracket between two hamiltonians ) vanish by definition .
a further difficulty in obtaining an infinitesimal algebra is that the spatial diffeomorphism constraints are not quantized as infinitesimal generators but are only implemented via a family of non weakly continuous unitary operators . for extended discussions of these aspects we refer to @xcite . for the second route of first discretizing and
then quantizing there is not a full classical definition of the constraint algebra ( apart from the sub - sector yielding flat dynamics ) yet , but this might be obtained with the techniques described in section [ mini ] . as discussed there the results might be a complicated mix of second and first class constraints . additionally one would expect that with decreasing discretization scale the approximate symmetries become exact and hence the second class constraints converted into first class . before quantization one should test whether this is indeed the case , as otherwise ` classical anomalies ' are bound to appear .
if one follows the approach to quantize the discrete theory and then to perform the continuum limit the question arises of how to deal with the constraints .
usually second class constraints have to be taken care off at the classical level as they can not be implemented exactly on quantum states : the uncertainty relations require the fluctuations of the second class constraints to be of the order of the expectation value of the commutator .
again one might hope that heuristically the expectation values of this commutator should become small with the discretization scale ( encoded in the state used to obtain the expectation value ) . to solve the second class constraints at the classical level and to introduce dirac brackets
would be overly complicated and might not allow for a continuum limit .
therefore methods are asked for that can deal with mixed constraint systems on the quantum level .
there are several such proposals @xcite based on different motivations .
but they have in common to consider just one ( ` master ' ) constraint , typically a sum or integral of the squares of the original constraints .
this constraint is used to construct a projector on the physical hilbert space , either via path integral methods @xcite or via spectral analysis . in the latter case ,
since the master constraint operator has to be positive definite , a physical hilbert space can be defined as the ( generalized ) eigenspace corresponding to the minimum of the spectrum .
( alternatively one can consider the generalized eigenspaces from a small interval around zero and in this way obtain a proper projector .
this would correspond to the ` thickened constraint hypersurface ' picture from section [ mini ] . ) as there are second class constraints this minimum will not be zero and may or may not @xcite converge to zero in the continuum limit .
the ` uniform discretization ' method @xcite has been applied to a spherically reduced model with spatial diffeomorphism constraints and it is found that the minimum of the spectrum converges to zero @xcite . following these approaches a physical hilbert space can be constructed even with second class constraints . the question arises how to detect quantum anomalies , i.e. to test whether the ( hypothetical ) quantized second class constraints become first class in the continuum limit .
the minimum of the spectrum going to zero in the continuum limit might be one indication but there are examples @xcite based on non compact lie groups in which the minimum is non
zero despite a non anomalous quantum constraint algebra , therefore one might have to allow for non zero values . furthermore in a classical reduction second class constraints lead to the elimination of only half the number of degrees of freedom as for first class constraints .
one has to make sure that unphysical gauge degrees do not survive the continuum limit .
quantum anomalies might depend on operator orderings and the precise definition of the master constraints but could also arise do to a physical reason , one possibility being a fundamental discretization scale for time like intervals . to differentiate between these issues
it might be necessary to quantize the original constraints and to study the commutator algebra on the family of physical hilbert spaces which arise in taking the continuum limit .
incorporating diffeomorphism symmetry into quantum gravity models is a very difficult task .
it can however ensure the correct semi classical and continuum limit of quantum gravity .
we expect that models involving a discretization of the action will break at least a part of the symmetries to approximate ones , although existent discretizations , as for instance of the plebanski action , have yet to be checked .
one method is to consider the eigenvalues of the hesse matrix evaluated at a solution .
a minimal requirement for discretized actions is that symmetries corresponding to translations of flat vertices are preserved exactly .
this opens the question whether one can construct better actions or discretizations in which all the symmetries are exact .
indeed this is possible for 3d gravity with cosmological constant @xcite . for 4d gravity renormalization flow methods could yield such ( possibly non local ) actions .
lehto et.al.@xcite consider an action with quadratic curvature terms in order to force the dynamics to give nearly flat geometries , as for flat geometries the symmetries are exact .
in addition symmetries corresponding to a free choice of triangulation could appear in covariant and in canonical formalisms .
it would be of interest whether symmetries involving only a change of labels and symmetries involving also a change of triangulation or more general simplicial complex can be related . on the other hand one can proceed with actions where the symmetries are only approximate .
the methods of consistent discretization can be used to obtain canonical theories that reflect the exact and approximate symmetries of such actions .
approximate symmetries lead to approximate constraints these are families of constraints labeled by lagrange multipliers .
the algebra of such constraints should be studied and it remains to be seen whether the continuum constraint algebra can be recovered in the limit .
at least for discretized hypersurfaces embedded in flat space a first class algebra of constraints can be obtained .
it is necessary to have the same symmetries in the covariant action and the canonical formalism in order to make the idea of a path integral acting as a projector on physical states consistent . to this end
also the path integral measure should be taken into account . here
conditions for an approximate invariance have to be formulated .
alternatively a projector on physical states can be constructed using canonical quantization and in most cases some kind of master constraint .
results have to checked carefully for anomalies .
boundary states @xcite arising from path integrals over a region with boundary should reflect the symmetries of the path integral .
the same holds for observables acting on such states .
as most of the symmetries are broken this might not constrain states and observables as much as requiring exact gauge invariance .
note however that non gauge invariant variables might be very sensitive to discretization ambiguities and furthermore reflect unphysical degrees of freedom , as is emphasized in the consistent discretization program @xcite .
indeed in the limit in which the constraints are imposed exactly on the boundary states , operators that do not commute with the constraints will have unbounded fluctuations due to the uncertainty relations .
hence observables commuting either approximately or exactly with the constraints are required . as an exact commutation
might restrict the space of observables to much for anomalous constraint algebras , proper conditions have to be formulated .
techniques to compute observables are available for continuum theories @xcite and should be adapted to discrete ones .
it is a pleasure to thank benjamin bahr , laurent freidel , renate loll , rafael sorkin , simone speziale , thomas thiemann , ruth williams , the organizer daniele oriti and participants of the workshop ` microscopic models of quantum spacetime ' for discussions .
this work was funded by a marie curie fellowship of the european union .
s. a. hojman , k. kuchar and c. teitelboim , `` geometrodynamics regained , '' ann .
* 96 * ( 1976 ) 88 . t. regge , `` general relativity without coordinates , '' nuovo cim .
* 19 * ( 1961 ) 558 .
t. regge and r. m. williams , `` discrete structures in gravity , '' j. math .
* 41 * ( 2000 ) 3964 [ arxiv : gr - qc/0012035 ] .
a. perez , `` spin foam models for quantum gravity , '' class .
grav . * 20 * ( 2003 ) r43 [ arxiv : gr - qc/0301113 ] .
`` introduction to loop quantum gravity and spin foams , '' arxiv : gr - qc/0409061 .
r. gambini and j. pullin , `` canonical quantization of general relativity in discrete space - times , '' phys .
* 90 * ( 2003 ) 021301 [ arxiv : gr - qc/0206055 ] . c. di bartolo , r. gambini , r. porto and j. pullin , `` dirac - like approach for consistent discretizations of classical constrained theories , '' j. math .
* 46 * ( 2005 ) 012901 [ arxiv : gr - qc/0405131 ] .
r. gambini and j. pullin , `` classical and quantum general relativity : a new paradigm , '' gen .
* 37 * ( 2005 )
1689 [ int . j. mod .
d * 14 * ( 2005 ) 2355 ] [ arxiv : gr - qc/0505052 ] .
r. sorkin , `` time evolution problem in regge calculus , '' phys .
d * 12 * ( 1975 ) 385 [ erratum - ibid .
d * 23 * ( 1981 ) 565 ] .
p. a. tuckey , `` the construction of sorkin triangulations , '' class .
* 10 * ( 1993 ) l109 [ arxiv : gr - qc/9303011 ] .
j. w. barrett , m. galassi , w. a. miller , r. d. sorkin , p. a. tuckey and r. m. williams , `` a paralellizable implicit evolution scheme for regge calculus , '' int .
j. theor .
* 36 * ( 1997 ) 815 [ arxiv : gr - qc/9411008 ] .
w. a. miller,``the geometrodynamic content of the regge equations as illuminated by the boundary of a boundary principle , '' found .
* 16 * ( 1986 ) 143 .
a. p. gentle , a. kheyfets , j. r. mcdonald and w. a. miller , `` a kirchoff - like conservation law in regge calculus , '' arxiv:0807.3041 [ gr - qc ] .
a. morse , `` approximate diffeomorphism invariance in near flat simplicial geometries , '' class .
quant . grav .
* 9 * ( 1992 ) 2489 .
h. w. hamber and r. m. williams , `` gauge invariance in simplicial gravity , '' nucl . phys .
b * 487 * ( 1997 ) 345 [ arxiv : hep - th/9607153 ] .
l. freidel and d. louapre , `` diffeomorphisms and spin foam models , '' nucl .
b * 662 * ( 2003 ) 279 [ arxiv : gr - qc/0212001 ] .
h. w. hamber and r. m. williams , `` simplicial quantum gravity in three - dimensions : analytical and numerical results , '' phys . rev .
d * 47 * ( 1993 ) 510 .
b. dittrich , l. freidel and s. speziale , `` linearized dynamics from the 4-simplex regge action , '' phys .
d * 76 * ( 2007 ) 104020 [ arxiv:0707.4513 [ gr - qc ] ] .
m. galassi , `` lapse and shift in regge calculus , '' phys .
d * 47 * ( 1993 ) 3254 .
a. p. gentle and w. a. miller , `` a fully ( 3 + 1)-d regge calculus model of the kasner cosmology , '' class .
grav . * 15 * ( 1998 ) 389 [ arxiv : gr - qc/9706034 ] .
f. david , `` simplicial quantum gravity and random lattices , '' arxiv : hep - th/9303127 .
r. loll , `` the emergence of spacetime , or , quantum gravity on your desktop , '' class .
quant . grav .
* 25 * ( 2008 ) 114006 [ arxiv:0711.0273 [ gr - qc ] ] .
d. oriti , `` the group field theory approach to quantum gravity , '' arxiv : gr - qc/0607032 .
h. romer and m. zahringer , `` functional integration and the diffeomorphism group in euclidean lattice quantum gravity , '' class .
* 3 * ( 1986 ) 897 .
h. hagura , `` dynamical regge calculus as lattice quantum gravity , '' arxiv : hep - lat/0304011 .
l. freidel , `` group field theory : an overview , '' int .
j. theor .
* 44 * ( 2005 ) 1769 [ arxiv : hep - th/0505016 ] .
t. piran and a. strominger , `` solutions to the regge equations , '' class . quant .
grav . * 3 * ( 1986 ) 97 .
j. b. hartle , simplicial minisuperspace .
ii . some classical solutions on simple j. math . phys .
* 27 * ( 1986 ) 287 .
j. b. hartle and z. perjes , `` solutions of the regge equations on some triangulations of cp(2 ) , '' j. math .
* 38 * ( 1997 ) 2577 [ arxiv : gr - qc/9606005 ] .
r. l. gibbs , t. r. larry and s. mokhtari , `` subdivision analysis of solutions to the regge equations , '' phys .
d * 59 * ( 1999 ) 084009 .
t. piran and r. m. williams , `` a ( 3 + 1 ) formulation of regge calculus , '' phys . rev .
d * 33 * ( 1986 ) 1622 .
j. l. friedman and i. jack , `` ( 3 + 1 ) regge calculus with conserved momentum and hamiltonian constraints , '' j. math .
* 27 * ( 1986 ) 2973 .
r. loll , `` on the diffeomorphism - commutators of lattice quantum gravity , '' class .
grav . *
15 * ( 1998 ) 799 [ arxiv : gr - qc/9708025 ] .
k. v. kuchar , `` time and interpretations of quantum gravity , '' _ in * winnipeg 1991 , proceedings , general relativity and relativistic astrophysics * 211 - 314 . _
b. dittrich and j. p. ryan , `` phase space descriptions for simplicial 4d geometries , '' arxiv:0807.2806 [ gr - qc ]
. v. husain and o. winkler , `` discrete hamiltonian evolution and quantum gravity , '' class .
* 21 * ( 2004 ) 941 [ arxiv : gr - qc/0308011 ] .
k. kuchar , `` dirac constraint quantization of a parametrized field theory by anomaly - free operator represnetations of space - time diffeomorphism , '' phys .
d * 39 * ( 1989 ) 2263 . c. g. torre and m. varadarajan , `` quantum fields at any time , '' phys .
d * 58 * ( 1998 ) 064007 [ arxiv : hep - th/9707221 ] .
`` functional evolution of free quantum fields , '' class .
* 16 * ( 1999 ) 2651 [ arxiv : hep - th/9811222 ] . m. varadarajan , `` dirac quantization of parametrized field theory , '' phys .
d * 75 * ( 2007 ) 044018 [ arxiv : gr - qc/0607068 ] .
a. laddha and m. varadarajan , `` polymer parametrised field theory , '' phys .
d * 78 * ( 2008 ) 044008 [ arxiv:0805.0208 [ gr - qc ] ] .
m. bander , `` hamiltonian lattice gravity .
deformations of discrete manifolds , '' phys .
d * 36 * ( 1987 ) 2297 .
a. corichi and j. a. zapata , `` on diffeomorphism invariance for lattice theories , '' nucl .
b * 493 * ( 1997 ) 475 [ arxiv : gr - qc/9611034 ] . h. waelbroeck and j. a. zapata , `` translation symmetry in 2 + 1 regge calculus , '' class .
* 10 * ( 1993 ) 1923 .
`` @xmath103 covariant lattice theory and thooft s formulation , '' class .
* 13 * ( 1996 ) 1761 [ arxiv : gr - qc/9601011 ] . h. waelbroeck and j. a. zapata , `` a hamiltonian formulation of topological gravity , '' class .
* 11 * ( 1994 ) 989 [ arxiv : gr - qc/9311035 ] .
j. a. zapata , `` topological lattice gravity using self - dual variables , '' class .
* 13 * ( 1996 ) 2617 [ arxiv : gr - qc/9603030 ] .
b. dittrich and s. speziale , `` area - angle variables for general relativity , '' new j. phys .
* 10 * ( 2008 ) 083006 [ arxiv:0802.0864 [ gr - qc ] ] .
a. ashtekar , c. rovelli and l. smolin , `` weaving a classical geometry with quantum threads , '' phys .
* 69 * ( 1992 ) 237 [ arxiv : hep - th/9203079 ] . c. rovelli and l. smolin ,
`` discreteness of area and volume in quantum gravity , '' nucl . phys .
b * 442 * ( 1995 ) 593 [ erratum - ibid .
b * 456 * ( 1995 ) 753 ] [ arxiv : gr - qc/9411005 ] .
a. ashtekar and j. lewandowski , `` quantum theory of geometry .
i : area operators , '' class . quant .
* 14 * ( 1997 ) a55 [ arxiv : gr - qc/9602046 ] .
`` quantum theory of geometry .
ii : volume operators , '' adv .
* 1 * ( 1998 ) 388 [ arxiv : gr - qc/9711031 ]
. g. carbone , m. carfora and a. marzuoli , `` hierarchies of invariant spin models , '' nucl .
b * 619 * ( 2001 ) 654 [ arxiv : gr - qc/0008011 ] .
a. ashtekar , j. lewandowski , `` background independent quantum gravity : a status report '' , class .
* 21 * ( 2004 ) r53 ; [ gr - qc/0404018 ] .
l. smolin , `` an invitation to loop quantum gravity '' , [ hep - th/0408048 ] . c. rovelli , `` quantum gravity '' ( cambridge university press , cambridge 2004 ) .
t. thiemann , `` modern canonical quantum general relativity '' , ( cambridge university press , cambridge 2005 ) , [ gr - qc/0110034 ] .
j. engle , r. pereira and c. rovelli , `` the loop - quantum - gravity vertex - amplitude , '' phys .
* 99 * ( 2007 ) 161301 [ arxiv:0705.2388 [ gr - qc ] ] .
e. r. livine and s. speziale , `` consistently solving the simplicity constraints for spinfoam quantum gravity , '' europhys .
* 81 * ( 2008 ) 50004 [ arxiv:0708.1915 [ gr - qc ] ] . l. freidel and k. krasnov , class .
* 25 * ( 2008 ) 125018 [ arxiv:0708.1595 [ gr - qc ] ] . j. j. halliwell and j. b. hartle , `` wave functions constructed from an invariant sum over histories satisfy constraints , '' phys .
d * 43 * ( 1991 ) 1170 . c. rovelli , `` the projector on physical states in loop quantum gravity , '' phys .
d * 59 * ( 1999 ) 104015 [ arxiv : gr - qc/9806121 ] .
e. alesci , k. noui and f. sardelli , `` spin - foam models and the physical scalar product , '' arxiv:0807.3561 [ gr - qc ] .
k. noui and a. perez , `` three dimensional loop quantum gravity : physical scalar product and spin foam models , '' class .
* 22 * ( 2005 ) 1739 [ arxiv : gr - qc/0402110 ] .
m. bojowald and a. perez , `` spin foam quantization and anomalies , '' arxiv : gr - qc/0303026 .
e. buffenoir , m. henneaux , k. noui and ph .
roche , `` hamiltonian analysis of plebanski theory , '' class .
* 21 * ( 2004 ) 5203 [ arxiv : gr - qc/0404041 ] .
s. alexandrov , e. buffenoir and p. roche , `` plebanski theory and covariant canonical formulation , '' class .
* 24 * ( 2007 ) 2809 [ arxiv : gr - qc/0612071 ] .
b. dittrich and t. thiemann , `` are the spectra of geometrical operators in loop quantum gravity really discrete ? , '' arxiv:0708.1721 [ gr - qc ] .
c. rovelli , `` comment on are the spectra of geometrical operators in loop quantum gravity really discrete? by b. dittrich and t. thiemann , '' arxiv:0708.2481 [ gr - qc ] . c. fleischhack , `` representations of the weyl algebra in quantum geometry , '' to appear in commun
. math .
arxiv : math - ph/0407006 .
j. lewandowski , a. okolow , h. sahlmann and t. thiemann , `` uniqueness of diffeomorphism invariant states on holonomy - flux algebras , '' commun .
* 267 * ( 2006 ) 703 [ arxiv : gr - qc/0504147 ] .
j. makela , `` on the phase space coordinates and the hamiltonian constraint of regge calculus , '' phys .
d * 49 * ( 1994 ) 2882 .
g. immirzi , `` quantum gravity and regge calculus , '' nucl .
* 57 * ( 1997 ) 65 [ arxiv : gr - qc/9701052 ] .
m. campiglia , c. di bartolo , r. gambini and j. pullin , `` uniform discretizations : a new approach for the quantization of totally constrained systems , '' phys .
d * 74 * ( 2006 ) 124012 [ arxiv : gr - qc/0610023 ] . c. rovelli and f. vidotto , `` stepping out of homogeneity in loop quantum cosmology , '' arxiv:0805.4585 [ gr - qc ] .
d. oriti , `` group field theory as the microscopic description of the quantum spacetime fluid : a new perspective on the continuum in quantum gravity , '' arxiv:0710.3276 [ gr - qc ] .
t. thiemann , `` anomaly - free formulation of non - perturbative , four - dimensional lorentzian quantum gravity , '' phys . lett .
b * 380 * ( 1996 ) 257 [ arxiv : gr - qc/9606088 ] .
`` quantum spin dynamics ( qsd ) , '' class . quant .
grav . * 15 * ( 1998 ) 839 [ arxiv : gr - qc/9606089 ] .
t. thiemann , `` qsd iii : quantum constraint algebra and physical scalar product in quantum general relativity , '' class .
* 15 * ( 1998 ) 1207 [ arxiv : gr - qc/9705017 ] . j. lewandowski and d. marolf , `` loop constraints : a habitat and their algebra , '' int .
j. mod .
d * 7 * ( 1998 ) 299 [ arxiv : gr - qc/9710016 ] .
r. gambini , j. lewandowski , d. marolf and j. pullin , `` on the consistency of the constraint algebra in spin network quantum gravity , '' int .
j. mod .
d * 7 * ( 1998 ) 97 [ arxiv : gr - qc/9710018 ] .
t. thiemann , `` the phoenix project : master constraint programme for loop quantum gravity , '' class .
* 23 * ( 2006 ) 2211 [ arxiv : gr - qc/0305080 ] . b. dittrich and t. thiemann , `` testing the master constraint programme for loop quantum gravity .
i : general framework , '' class .
* 23 * ( 2006 ) 1025 [ arxiv : gr - qc/0411138 ] .
ii : finite dimensional systems , class . quant .
* 23 * ( 2006 ) 1067 [ arxiv : gr - qc/0411139 ] .
iii : sl(2,r ) models , class .
* 23 * ( 2006 ) 1089 [ arxiv : gr - qc/0411140 ] .
`` testing the master constraint programme for loop quantum gravity .
iv : free field theories , '' class .
* 23 * ( 2006 ) 1121 [ arxiv : gr - qc/0411141 ] .
v : interacting field theories , class .
* 23 * ( 2006 ) 1143 [ arxiv : gr - qc/0411142 ] . j. r. klauder , `` coherent state quantization of constraint systems , '' annals phys .
* 254 * ( 1997 ) 419 [ arxiv : quant - ph/9604033 ] .
`` universal procedure for enforcing quantum constraints , '' nucl .
b * 547 * ( 1999 ) 397 [ arxiv : hep - th/9901010 ] .
`` quantization of constrained systems , '' lect .
notes phys .
* 572 * ( 2001 ) 143 [ arxiv : hep - th/0003297 ] .
r. gambini and j. pullin , `` emergent diffeomorphism invariance in a discrete loop quantum gravity model , '' arxiv:0807.2808 [ gr - qc ] .
m. lehto , h. b. nielsen and m. ninomiya , `` diffeomorphism symmetry in simplicial quantum gravity , '' nucl .
b * 272 * ( 1986 ) 228 . c. rovelli ,
`` graviton propagator from background - independent quantum gravity , '' phys .
* 97 * ( 2006 ) 151301 [ arxiv : gr - qc/0508124 ] . s. speziale , `` background - free propagation in loop quantum gravity , '' arxiv:0810.1978 [ gr - qc ] .
r. gambini and j. pullin , `` discrete quantum gravity : applications to cosmology , '' class .
grav . * 20 * ( 2003 ) 3341 [ arxiv : gr - qc/0212033 ] . c. rovelli ,
`` partial observables , '' phys .
d * 65 * ( 2002 ) 124013 [ arxiv : gr - qc/0110035 ] . b. dittrich , `` partial and complete observables for hamiltonian constrained systems , '' gen .
* 39 * ( 2007 ) 1891 [ arxiv : gr - qc/0411013 ] .
`` partial and complete observables for canonical general relativity , '' class .
* 23 * ( 2006 ) 6155 [ arxiv : gr - qc/0507106 ] . b. dittrich and j. tambornino , a perturbative approach to dirac observables and their space - time class .
* 24 * ( 2007 ) 757 [ arxiv : gr - qc/0610060 ]
. `` gauge invariant perturbations around symmetry reduced sectors of general relativity : applications to cosmology , '' class .
* 24 * ( 2007 ) 4543 [ arxiv : gr - qc/0702093 ] . | we review and discuss the role of diffeomorphism symmetry in quantum gravity models .
such models often involve a discretization of the space
time manifold as a regularization method .
generically this leads to a breaking of the symmetries to approximate ones , however there are incidences in which the symmetries are exactly preserved .
both kind of symmetries have to be taken into account in covariant and canonical theories in order to ensure the correct continuum limit .
we will sketch how to identify exact and approximate symmetries in the action and how to define a corresponding canonical theory in which such symmetries are reflected as exact and approximate constraints .
_ keywords _ : quantum gravity , regge calculus , loop quantum gravity , spin foams , diffeomorphism symmetry |
during the recent decades , several x - ray surveys were carried out and several catalogues of x - ray sources were produced , including most recent xmm - newton and chandra sources . however , until now rosat remains the only enough deep all - sky survey .
rosat catalogues contain thousands of interesting objects , including agn ( especially many blazars ) , hot subdwarfs ( sb ) , white dwarfs ( wd ) , cataclysmic variables ( cv ) , carbon ( c ) stars , etc .
rosat sources are listed in two main catalogues : * bright source catalogue ( bsc * , voges et al . 1999 ) and * faint source catalogue ( fsc * , voges et al .
rosat bsc contains 18,806 sources with cr @xmath10.05 cts@xmath2s@xmath3 in the 0.1 - 2.4 kev energy range , while fsc contains 105,924 fainter sources ( cr @xmath40.05 cts@xmath2s@xmath3 ) .
thus , in total 124,730 sources are available . among the identification
works , the rosat bright sources ( rbs , schwope et al .
2000 ) is well - known ; 2012 bsc sources with cr @xmath1 0.20 and @xmath5b@xmath5>30@xmath6 have been optically identified .
however , most of the identified sources come from the hamburg quasar survey ( hqs , hagen et al .
its low - dispersion spectra allow a preliminary classification of objects into a number of types , giving possibility to make up subsamples of objects for further studies .
two main projects have been carried out : * hamburg rosat catalogue ( hrc * , zickgraf et al .
2003 ) and * byurakan - hamburg rosat catalogue ( bhrc * , mickaelian et al .
both have been carried out by means of hamburg quasar survey ( hqs ; hagen et al . 1995 ) low - dispersion spectra .
hrc is based on rosat bsc and contains 5341 sources at @xmath5b@xmath5>20@xmath6 and @xmath7>0@xmath6 . + * bhrc * used brighter sources of rosat faint source catalogue ( fsc ) , thus extending the sample to count rates ( cr ) of photons @xmath1 0.04 ct / s in the area with galactic latitudes @xmath5b@xmath5@xmath120@xmath6 and declinations @xmath7@xmath10@xmath6 ( the area of hqs).it contains 2791 sources .
main advantages of bhrc compared to hrc were the use of dss1 @xmath8 dss2 red / blue / ir images ( which allowed revealing faint objects , more details , extension , proper motions , variability , and accurate positions ) , the use of 2d hqs images ( for quicker selection and better classification ) , the refinement of classification ( identifying qsos / agn from sed ) , introducing new class of interacting galaxies ( pairs and groups ) , search in real 3@xmath9 positional errors for each source compared to standard search radius in hrc , cross - correlation with maps and usno - b , associations in ned , simbad , catalogues of agn , wds , and cvs . out of 2791 sources , only 95 have no identification ; 2696 are identified with 3187 objects ( some x - ray sources are identified with double or multiple object ) .
+ thus , out of 124,730 rosat sources , only @xmath1010,000 have optically identifications and rosat catalogues still remain as a rich source for new interesting objects .
to improve the accuracy of measurements and collect more data for hrc sources , we have carried out similar to bhrc studies for hrc , thus establishing homogeneous joint sample of rosat identifications .
a * joint hrc / bhrc catalogue * has been created by merging hrc and bhrc , altogether 8132 sources ( paronyan @xmath8 mickaelian 2015 ) , the largest sample of identified rosat sources . together with combining data from the two catalogues , we have checked and added many new ones , as well as corrected many errors and misidentifications .
based on hrc / bhrc sample , we have established the types of objects that appear to be x - ray sources .
main types of optical counterparts are : bright ( typically spiral ) galaxies , blue galaxies ( agn candidates ) , interacting galaxies ( pairs @xmath8 groups , those with red colours ? ) , blue compact galaxies ( bcg ) that appear to be qsos and seyferts , blue stellar objects ( bsos ) that appear to be qsos or white dwarfs ( wd ) , bright stars ( mostly fg type , including pairs and triples ) , k - m stars ( including binary k - stars ) , wds and subdwarfs , cataclysmic variables ( cvs ) .
* table 1 * gives the distribution of the identified objects by types in hrc , bhrc and joint hrc / bhrc catalogues .
+ .distribution of objects by types in hrc , bhrc and joint hrc / bhrc . [ cols="<,^,^,^,^,^,^",options="header " , ] we give several examples of different objects based on our sdss classification .
many bright galaxies show x - ray ( as well as they radiate at all wavelength ranges ) even without having any sign of activity .
one of our tasks is to establish the limit of rosat detection of normal bright galaxies .
then all fainter objects may turn to be relevant agn candidates even without having their spectra .
qsos are most typical and comprise the largest sample among extragalactic x - ray sources . in * figure 3
* we give two examples of hrc / bhrc qsos revealed as qso candidates based on hqs low - dispersion spectra and later on discovered by means of sdss medium resolution spectra . given that bl lacs typically do not show any ( emission or absorption ) lines , it is rather difficult to distinguish them from stars ( as very often stellar absorption lines are weak and not detectable ) . however , due to synchrotron radiation , their continua are linear compared to the characteristic planck curve for the stars .
hence , only high quality spectra and very accurate reduction may reveal their spectral energy distribution and prove their nature even without measuring the redshift . on the other hand , presence of radio
may also be a relevant criterion to distinguish bll against blue stars .
narrow line seyfert 1 galaxies are among the most interesting objects in our sample .
we have found different types of such objects ; nls1.0 , nls1.2 and nls1.5 , as well as nls1.8 and nls1.9 may exist , however due to small intensity of their broad lines , only high signal - to - noise and high resolution spectra may reveal them . in * figure 4
* , two examples of nls1s are given .
.,scaledwidth=80.0% ] other agn include various types of seyferts ( from sy1.0 to sy2.0 in the order of agn luminosity ) and liners .
we give in * figure 5 * sdss spectra for 2 seyfert galaxies ; ( left ) sy1.5 ( broad and narrow components have similar intensity ) and ( right ) sy1.8 ( broad components are weak and observed only for h@xmath11 and h@xmath12 lines ) . and h@xmath12 lines).,scaledwidth=80.0% ] starburst ( or hii ) galaxies have no agn in the center , however due to undergoing intense star formation process ( sfr is a few m@xmath13 yr@xmath3 , but may reach up to 103 m@xmath13 yr@xmath3 ) and presence of heated gas they show strong balmer and some forbidden emission lines ( oiii , oii , oi , nii , sii and others .
they may have enough strong stellar component , as well as strong continuum , which sometimes is very blue . in * figure 6
* , two examples of hii galaxies are given .
absorption line galaxies do not show any sign of activity when studies by means of optical spectra .
however , the presence of strong x - ray and in many cases also radio suggests that they are hidden agn , so that they are completely opaque in optical light , but they may be exactly the same agn similar to classical seyferts .
we give in * figure 7 * two examples of such galaxies that are typically red ones .
we have discovered red interacting galaxies that are strong x - ray sources .
most powerful extragalactic ir sources ( ulirgs and hlirgs ) are interacting / merging galaxies .
however , the x - ray sample among such objects must be investigated in detail .
many remote clusters have been detected in rosat as single point sources .
the question is to check if there are strong x - ray emitters among the cluster galaxies ( e.g. central giant ones ) or x - ray comes from the intergalactic gas in the cluster .
new x - ray missions , such as xmm - newton and chandra may answer this question .
as x - ray sources contain many interesting stars , some fraction of contamination is always present , especially because our former classification was based on low - dispersion spectra .
typical types of stars that appear to be x - ray sources are : white dwarfs ( wd ) , cataclysmic variables ( cv ) , late - type stars ( k , m , c ) and others .
we give the results of sdss spectral classification of these objects in another paper ( mickaelian et al .
a project on study of x - ray selected galaxies has been conducted in bao .
it includes selection , cataloguing , spectral classification , multiwavelength studies and statistical analysis of such objects . + + follow - up programs and future studies based on x - ray selected agn are : * estimation of rosat bsc / fsc extragalactic content * optical variability of hrc / bhrc extragalactic sources ( based on the analysis of dss1/dss2 photometric data from usno - a2.0 , apm , maps , usno - b1.0 and gsc 2.3.2 , as shown in mickaelian et al .
2011 ) , * radio 1400 mhz variability of hrc / bhrc extragalactic sources ( based on nvss - first cross - matching and study of radio fluxes of associated sources ) , * detailed spectral study ( planned , the results will be published soon ) , * creation of rosat - nvss agn sample ( a preliminary list of 6,099 x - ray / radio sources without an optical identification ; brighter objects are normal bright galaxies , while all faint ones are candidate agn with some contamination of distant clusters ) , * study of x - ray selected blazars ( to support their definition in terms of x - ray properties ) , * search for optical counterparts of all rosat sources ( using available catalogues of bright stars and galaxies , agn , wds , cvs , late - type stars , etc . to distinguish different types of objects then using extension , radio , proper motion , optical colours , sdss/2mass / wise diagrams for better understanding of the nature of these objects ) . 1 . what is the fraction of x - ray agn among all ? 2 .
are there x - ray galaxies ( like radio galaxies ) ? 3 .
is there a limit between x - ray luminosities between normal galaxies ( background radiation ) and agn ? 4 .
is there a correlation between optical and radio variability of agn ?
what is the definition of blazars ?
the question of definition of blazars is still open .
there are many parameters that may be regarded as such criteria : high luminosity , flat radio spectrum , strong x - ray and @xmath14-ray , optical and/or radio variability , polarization , etc .
future studies will work on answering to these questions . | the joint catalogue of active galactic nuclei selected from optical identifications of x - ray sources was created as a combination of two samples : hamburg - rosat catalogue ( hrc ) and byurakan - hamburg - rosat catalogue ( bhrc ) .
both are based on optical identifications of x - ray sources from rosat catalogues using low - dispersion spectra of hamburg quasar survey ( hqs ) .
however , hrc and bhrc contain a number of misidentifications and using the recent optical and multiwavelength ( mw ) catalogues we have revised both samples excluding false agn and adding new genuine ones .
thus a new large homogeneous complete sample of 4253 x - ray selected agn was created .
3352 of them are listed in the catalogue of qsos and active galaxies and 387 also are in roma multifrequency catalogue of blazars .
901 candidate agn are subject for further study .
we classified 173 of these objects using their sdss dr12 spectra .
following activity types were revealed : 61 agn , 21 hii galaxies , 12 emission - line galaxies without definite type , 71 absorption - line galaxies , 2 stars , and 6 were classified as @[email protected] special emphasis is made on narrow - line sy1.0-sy1.5 galaxies and qsos , as many of them have soft x - ray , strong feii lines , and relatively narrow lines coming from blr ( @xmath0narrow broad lines@xmath0 ) . as a result ,
the sample of genuine agn was enlarged to 3413 objects .
we have retrieved mw data from recent catalogues and carried out statistical investigations for the whole agn sample .
an attempt to find connections between fluxes in different bands for different types of sources , and identify their characteristics thus confirming candidate agns have been carried out .
we have analyzed x - ray properties of these sources to find a limit between normal galaxies and x - ray agn .
+ * keywords . * x - ray sources , cross - correlations , x - ray : galaxies , galaxies : active , agn , quasars : general , starburst galaxies , interacting galaxies . |
neutron stars are observed as several classes of self - gravitating systems : as radio and x - ray pulsars , as x - ray bursters , as compact thermal x - ray sources in supernova remnants , as rotating radio transients . in general ,
the structure of neutron stars and the relation between the mass and the radius are determined by equations of state ( eos ) of dense matter .
the maximal mass of neutron star is still an open question .
recent observations allows to estimate this limit at least as @xmath7 : the well - measured limit of the pulsar psr j1614 - 2230 is @xmath8 @xcite , while , for the pulsar j0348 + 0432 , it is @xmath9 @xcite .
other examples of massive neutron stars are vela x-1 ( @xmath10 @xcite ) and 4u 1822 - 371 ( @xmath11 , @xcite ) .
there are some indications in favor of the existence of more massive neutron stars with masses @xmath12 ( the possible masses of b1957 + 20 @xcite and 4u 1700 - 377 @xcite ) or even @xmath13 ( j1748 - 2021b @xcite ) .
it is interesting to note that for various eos including hyperons , the maximal mass limit for non - magnetic neutron stars is considerably below than of the two - solar masses limit .
the hyperonization process softens eos and then the maximal allowable mass results reduced @xcite .
there are several ways to approach the solution of this problem ( the so called `` _ hyperon puzzle _ '' ) .
firstly , the extensions of the simple model of hyperonic matter ( with three exchange meson fields - the so called @xmath14-model ) allow to achieve the increasing of maximal mass .
various approaches are proposed following this track .
for instance , larger hyperon - vector couplings ( in comparison with quark counting rule ) require stiffness of the eos @xcite .
similar effect occurs in model with chiral quark - meson coupling @xcite .
the quartic vector - meson terms in the lagrangian @xcite or the inclusion of an additional vector - meson mediating repulsive interaction amongst hyperons @xcite also lead to the increasing of the maximal mass limit .
authors of ref .
@xcite proposed an eos with maximum mass @xmath15 using the quark - meson coupling model , which naturally incorporates hyperons without additional parameters . a model with in - medium hyperon interactions
is considered in @xcite .
another source for increasing the maximal mass limit is the existence of strong magnetic fields inside the star .
the existence of soft gamma - ray repeaters and anomalous x - ray pulsars can be linked to neutron stars with very strong magnetic fields of the order @xmath16 g on the surface . in these cases ,
the maximum magnetic field in the central regions of neutron star can exceed @xmath17 g , according to the scalar virial theorem .
such magnetic fields affect considerably the eos for dense matter and result in increasing the maximal mass of neutron stars .
various models of dense nuclear matter in presence of strong magnetic fields have been considered in literature . the simplest model with interacting @xmath18 gas
is investigated in @xcite .
models with hyperons and quarks are considered in @xcite-@xcite .
it has been demonstrated that the landau quantization leads to the softening of the eos for matter but account for contributions of magnetic field into pressure and density .
this fact leads , on the other hand , to the stiffening of eos .
therefore neutron stars are very peculiar objects for testing theories of matter at high density regimes and in strong magnetic fields .
it is interesting to note that data about neutron stars ( mainly mass - radius ( @xmath19 ) relation ) can be used for investigating possible deviations from general relativity ( gr ) .
the initial motivation for studying modified gravity came from the discovered accelerated expansion of the universe confirmed by numerous independent observations .
these observations include hubble diagram for ia type supernovae @xcite , cosmic microvawe background radiation ( cmbr ) data @xcite , surveys of gravitational weak lensing @xcite and data on lyman alpha forest absorption lines @xcite .
this acceleration takes place at relatively small distances ( `` hubble flow '' ) and requires ( in gr ) non - standard cosmic fluid ( dark energy ) filling the universe with negative pressure but not clustered in large scale structure .
the nature of dark energy is unclear .
although from an observational viewpoint , the so called @xmath20cdm model ( where dark energy is considered as einstein cosmological constant ) is in agreement with data coming from observations there are various problems and shortcomings at theoretical level .
one of this issues is the `` smallness '' of cosmological constant i.e. the difference of 120 orders of magnitude between its observed value and the one predicted by quantum field theory @xcite .
an alternative approach to dark energy problem consists of extending of gr . in this case
, the accelerated expansion can be obtained without using `` dark energy '' but enlarging the gravitational sector @xcite .
therefore theories of modified gravity can be considered as real alternative to gr .
the study of relativistic stars in modified gravity is interesting from several reasons and could constitute a formidable probe for such theories .
firstly one can reject some models that do not allow the existence of stable star configurations @xcite ( however one has to note that stability can be achieved due to `` chameleon mechanism '' @xcite and may depend on the choice of the eos ) .
secondly there is the possibility for the existence of new stellar structures , in the framework of modified gravity , escaping the standard stellar models .
the observation of such self - gravitating anomalous structures could provide strong evidence for the extended gravity ( see e.g. @xcite ) .
the present paper is devoted to neutron stars with strong magnetic fields in framework of analytic @xmath0 gravity . assuming
a simple model for strong interactions , one can obtain the eos for dense matter in magnetic field .
landau quantization , due to magnetic fields , results to have significant effects .
we consider the cases of slowly and fast varying fields .
the paper is organized as follows . in sec .
ii , we briefly consider the the field equations for @xmath0 gravity and the modified tolman
volkoff ( tov ) equations .
then relativistic mean field theory for dense matter in strong magnetic fields is presented ( sec.iii ) . in sec .
iv , the neutron star models for strong magnetic fields in quadratic ( @xmath21 ) and cubic @xmath22 gravity are presented .
the @xmath19 relation is derived and compared with the one in gr .
conclusions and outlooks are reported in sec.v .
the action of @xmath0-gravity is @xmath23 it can be expressed as @xmath24 .
the field equations are @xmath25 here @xmath26 is the determinant of the metric @xmath27 and @xmath28 is the action of the standard perfect fluid matter .
the einstein tensor is @xmath29 and @xmath30 . for star configurations
, one can assumes a spherically symmetric metric with two independent functions of radial coordinate , that is : @xmath31 for the exterior solution , we assume a schwarzschild metric . therefore it is convenient to define the variable @xcite @xmath32 the value of variable @xmath33 on the star surface is the gravitational mass . for a perfect fluid , the energy - momentum tensor is @xmath34 , where @xmath2 is the matter density and @xmath35 is the pressure .
the field equations of interest are @xmath36 \label{f - tt},\\ 8\pi g p / c^4 & = & -r^{-2 } + e^{-2\lambda}(1 + 2r\phi')r^{-2 } + \alpha h_r(-r^{-2 } + e^{-2\lambda}(1 + 2r\phi')r^{-2 } ) \nonumber \\ & & -\frac12\alpha(h - h_{r}r ) + e^{-2\lambda}\alpha h_r'r^{-1}(2+r\phi ' ) , \label{f - rr}\end{aligned}\ ] ] where @xmath37 .
the second tov equation follows from the conservation law @xmath38 and eq.([f - rr ] ) . as result
, the modified tov equations can be written as @xcite @xmath39,\ ] ] @xmath40 @xmath41,\ ] ] here we have introduced the dimensionless variables @xmath42 , @xmath43 , where @xmath44 km .
the third independent equation for ricci curvature scalar is @xmath45h_{{r}}+\alpha r_{g}^{2 } h_{{r}}{r}-2\alpha r_{g}^{2 } h-{r}=-8\pi({\rho}-3p)\,.\ ] ] eqs .
( [ tov-1 ] ) , ( [ tov-2 ] ) , and ( [ tov-3 ] ) can be solved numerically for given eos . in order to get solution
, one can use perturbative approach ( see for details @xcite ) . in the framework of perturbative approach , terms containing @xmath46
are assumed to be of first order in the small parameter @xmath47 , so all such terms should be evaluated at @xmath48 order .
the ricci curvature scalar at zero order is @xmath49 .
therefore the deviation from gr strongly depends from the assumed form of eos .
let us assume a simple model for describing nuclear matter in magnetic field .
the magnetic field @xmath50 is assumed along @xmath51-axis i.e. the 4-potential is @xmath52 . for nuclear matter consisting of baryon octet ( @xmath53@xmath54 , @xmath55 , @xmath20 , @xmath56 , @xmath57 ) interacting with magnetic field and scalar @xmath1 , isoscalar - vector @xmath58 and isovector - vector @xmath59 meson fields and leptons ( @xmath60@xmath61 , @xmath62 ) ,
it is @xcite @xmath63 @xmath64 here the mesonic and electromagnetic field strength tensors are defined by the usual relations @xmath65 , @xmath66 , @xmath67 . for the sake of simplicity , we consider frozen - field configurations of electromagnetic field . also we neglect the anomalous magnetic moments ( amm ) of baryons and leptons because their effect is very small .
the strong interaction couplings @xmath68 , @xmath69 and @xmath70 depend from density .
we use the parameterization adopted in @xcite : @xmath71 where @xmath72 for the isovector field , it is @xmath73.\ ] ] the values of constants @xmath74 , @xmath75 , @xmath76 , @xmath77 are given in @xcite . using the mean - field approximation , one can obtain the following equations for meson fields : @xmath78 here @xmath1 , @xmath79 , @xmath80 are the expectation values of the meson fields in uniform matter .
the quantities @xmath81 , @xmath82 are the scalar and vector baryon number densities , correspondingly . the simplest scalar field potential is defined as @xmath83 where @xmath84 and @xmath85 are dimensionless constants .
the values of nucleon - meson couplings and parameters @xmath84 , @xmath85 are given in table i. from the dirac equations for charged and neutral baryons and leptons , we have the energy spectra : @xmath86 @xmath87 @xmath88 the number @xmath89 denotes the landau levels of the fermions with electric charge @xmath90 , spin number @xmath91 for spin up and spin down cases correspondingly . the spin degeneracy is @xmath92 for lowest landau level ( @xmath93 ) and 2 for all other levels .
the effective mass for baryons is @xmath94 .
the rearrangement self - energy term is defined by @xmath95 here @xmath96 .
the scalar densities for neutral baryons are @xcite @xmath97 and for charged baryons , it is @xmath98 for the vector densities for neutral baryons , we have @xmath99 and for charged baryons and leptons , it is @xmath100 here @xmath101 is the fermi energy . for charged baryon , @xmath102 is related to the fermi momentum @xmath103 as @xmath104 . for neutral baryon ,
it is @xmath105 .
the summation over @xmath106 terminates at value @xmath107 where the square of fermi momenta is still positive . for large magnetic fields @xmath108
g , only few landau levels are occupied . for hyperon - meson couplings
there are no well - defined rule .
one can use for these constants quark counting rule @xcite : @xmath109 and @xmath110 another choice is assuming that the fractions of nucleon - meson couplings , i.e. @xmath111 , is fixed . here , it is @xmath112 , @xmath113 , @xmath114 ( see @xcite ) .
we use this definition for further calculations .
[ table1 ] .the nucleon - meson couplings and parameters of scalar field potential for some models ( gm1 - 3 - @xcite , tw - @xcite ) . the nuclear saturation density @xmath115 , the dirac effective mass @xmath116 and the binding energy ( @xmath117 ) are also given . [ cols="^,^,^,^,^,^,^,^,^ " , ] and in gr for slowly ( upper panel ) and fast ( lower panel ) varying field .
one can see that for the deviation of m - r relation from gr is smaller for larger values of @xmath118.,title="fig : " ] + and in gr for slowly ( upper panel ) and fast ( lower panel ) varying field .
one can see that for the deviation of m - r relation from gr is smaller for larger values of @xmath118.,title="fig : " ] +
let us firstly take into account models with quadratic gravity corrections , that is @xmath119 neutron stars with strong magnetic field in quadratic gravity was considered in @xcite for relatively stiff eos based on model with five meson fields .
we consider the quadratic gravity case for eos based on the above described model . for our calculations ,
the typel - wolter ( tw ) parametrization is used .
let us note the following feature : for @xmath120 , the mass of neutron star increases with decreasing @xmath47 for the various radii ( see fig .
1 ) . for strong magnetic field ,
the @xmath19 relation for @xmath121 differs from such in gr only for masses close to the maximal one ( see figs . 2 , 3 ) .
another interesting feature appears for fast varying field . at high central densities ,
a second `` branch '' of stability can exists ( fig .
3 , lower panel ) .
it is interesting to note that similar effects take place for non - magnetic neutron stars in the framework of a model like @xmath122 @xcite .
the stabilization of star configurations occurs thanks to the cubic term .
the maximal masses and corresponding radii are given in tables ii , iii for some values of @xmath47 and @xmath118 .
the maximal value of central density ( and therefore magnetic field ) decreases with increasing @xmath47 .
the parameters for compact ( in comparison with gr ) neutron stars on second `` branch of stability '' are given in table iv .
for modified gravity with cubic term , @xmath123 , the maximal value of neutron star mass for given eos increases for @xmath124 ( fig .
some results are given in tables v , vi .
the maximal mass of neutron star can exceed @xmath125 .
one can note that stars with magnetic field and cubic curvature corrections result stable for central energy density close to @xmath126 gev/@xmath127 . in principle
, calculations show that , for eos based on gm2-gm3 parameterizations , we have similar results for models with @xmath21 and @xmath128 .
for more stiff eos the deviation from gr is larger .
we presented neutron star models with strong magnetic fields in the framework of power - law @xmath0 gravity models . for describing dense matter in magnetic field , a model with baryon octet interacting through @xmath1@xmath2@xmath3-fields
is used .
although the softening of nucleon eos , due to hyperonization , leads to the decrease of the upper limit mass of neutron star , the strong magnetic field can increase considerably the maximal mass of star . in particular , we investigated the effect of strong magnetic field in models of quadratic , @xmath21 , and cubic , @xmath129 , gravity . for large fields ,
the @xmath19 relation differs considerably from such in gr only for stars with masses close to maximal .
another interesting feature is the possible existence of more compact stable stars with extremely large fields ( @xmath130 g instead @xmath6 g in gr ) in central regions of star . due to the cubic term ,
the significant increasing of maximal mass ( @xmath131 ) is possible .
the central energy density can exceed @xmath126 gev/@xmath127 .
however , it is worth stressing that the @xmath0 models considered here can be related to the presence of strong gravitational fields where higher order curvature terms can emerge .
their origin is related to the effective actions of quantum field theory formulated in curved spacetime @xcite . in the extreme field of neutron stars
, it is realistic supposing the emergence of curvature corrections that improve the pressure effects and could explain supermassive self - gravitating systems . as a next step
, we will consider models of self - bounded quark stars and hybrid stars with quark cores .
the eos for quark matter ( without magnetic field ) is close to @xmath132 and therefore , in the framework of perturbative approach , the deviations from gr occur only for very large values of @xmath47 in comparison with the above considered in quadratic gravity .
however , for large magnetic fields , considerable effects can be induced on eos and therefore the modified gravity effects can appear .
this work is supported in part by projects 14 - 02 - 31100 ( rfbr , russia ) ( ava ) , by mineco ( spain ) , fis2010 - 15640 and by mes project tspu-139 ( russia ) ( sdo ) .
sc is supported by infn ( _ iniziative specifiche _ teongrav and qgsky ) . | neutron stars with strong magnetic fields are considered in the framework of @xmath0 gravity . in order to describe dense matter in magnetic field , the model with baryon octet interacting through @xmath1@xmath2@xmath3-fields
is used .
the hyperonization process results in softening the equation of state ( eos ) and in decreasing the maximal mass .
we investigate the effect of strong magnetic field in models involving quadratic and cubic corrections in the ricci scalar @xmath4 to the hilbert - einstein action . for large fields , the mass - radius relation differs considerably from that of general relativity only for stars with masses close to the maximal one .
another interesting feature is the possible existence of more compact stable stars with extremely large magnetic fields ( @xmath5 g instead of @xmath6 g as in general relativity ) in the central regions of the stars . due to cubic terms ,
a significant increasing of the maximal mass is possible . |
the author is pleased to acknowledge the support from the new frontiers in astronomy and cosmology program at john templeton foundation . | a novel method allowing to compute density , velocity and other fields in cosmological n body simulations with unprecedentedly high spatial resolution is described .
it is based on the tessellation of the three - dimensional manifold representing cold dark matter in six - dimensional phase space .
the density , velocity and other fields are computed by projecting the tessellation on configuration space .
the application of this technique to cosmological n
body simulations in @xmath0cdm cosmology reveals a far more elaborate cosmic web then dot plots or self adaptive sph .
in addition , this method allows to uniquely define physical voids and identify and study the caustic surfaces directly .
cosmological n - body simulations currently represent the major tool for the theoretical studies of the formation and evolution of the cosmic web .
they provide the most accurate statistical information on the distribution of mass and galaxies to be compared with observations .
most of the simulations deal with the growth of the structure only in collisionless dark matter ( hereafter dm ) since it dominates the mass being able to cluster gravitationally . due to a huge number of dark matter particles
the most accurate physical model of dm is collisionless fluid .
a new method@xcite ( see also @xcite ) considers the growth of cold dark matter web as the evolution of a three dimensional manifold ( the phase
space sheet , hereafter pss ) in six dimensional phase space .
the degeneracy of the pss is a result of extremely low thermal velocity dispersion in cold dm .
the pss is approximated by a tessellation with the vertices represented by the particles of n - body simulation .
the tessellation must be created at the linear stage before shell crossing occurs and must remain intact throughout the rest of the evolution . out of many choices
the tetrahedral tessellation is the most sensible because tetrahedra practically always remain convex .
the exceptions happen only at isolated instances of time when four vertices become coplanar and the tetrahedron degenerates into a flat polygon .
computing the volume of the tetrahedron at later times using the initial order of vertices results in a negative value .
if the tetrahedron passes through the coplanar state one more time its volume become positive again .
these metamorphoses can repeat many times .
the vertices aka particles are treated as the flow tracers while the tetrahedra as the mass tracers .
the density of each tetrahedron can be easily computed assuming that the mass of each tetrahedron is conserved and uniformly distributed within its volume .
the tetrahedra tile the eulerian configuration space without overlapping in the linear regime before shell crossing occurs . at later times
they start to overlap resulting in the origin of multi stream flow regions .
various fields ( density , velocity and others ) can be derived by projecting the pss on configuration space .
this techniques uses the full phase
space information available in the simulations . as a result
the derived fields can be computed with unprecedented spatial resolution .
in addition , new fields such as multi - stream@xcite and parity@xcite fields can also be obtained .
these fields provide new invaluable information about the dynamics of the cosmic web which is supplemental to the density and velocity fields routinely used in cosmology .
in particular , the whole volume with one stream flow can be defined as the physical void i.e. the volume devoid of structures that the n body simulation is capable to resolve .
the signs of the tetrahedra volumes make the parity field in lagrangian space@xcite .
this field has a nontrivial topological property : the three dimensional map of parity can be painted by only two different colors .
the interface between regions with different colors approximates the caustic surfaces .
the number of streams is an odd integer at every generic point of configuration space .
the number of stream can be even only on caustic surfaces having measure zero .
the number of streams can is easy to compute at an arbitrary point by counting the number of tetrahedra enclosing it .
similarly the density and other fields can be estimated by summing up the corresponding quantities over all tetrahedra enclosing the chosen point .
we demonstrate the method using an example from n - body simulation of the @xmath0cdm cosmological model in a 512@xmath1mpc - sided cubic box@xcite .
the number of particles is 512@xmath2 and the grid size in the gravitational poisson solver is 1024@xmath2 .
the parameters of the @xmath0cdm model are as follows : @xmath3 , @xmath4 , @xmath5 , @xmath6 , @xmath7 , the initial redshift @xmath8 .
we present a relatively small box cut from a much larger simulation box in order to reduce the obscuration effects .
figure 1 illustrates the change in the appearance of the structure with the growth of spatial resolution of the sampling grid employed for computing the fields .
it is worth stressing that the number of particles and their coordinates are exactly same in all panels .
the structures in density ( top panels ) and multi stream ( bottom panels ) fields obviously have some common features however they are not identical ( see also@xcite ) .
the left panels show the structure at the mass resolution of the n - body simulation .
it is remarkable that the multi
stream field clearly demonstrates pancakes at the lowest spatial resolution while the density field hardly suggests even a hint of the pancake presence .
the explanation of the failure of detecting pancakes in n
body simulations was suggested long time ago@xcite .
the pancakes contain a low fraction of mass and at the same time occupy a relatively large volume that results in a very low density contrast .
this was quantitatively confirmed in the recent analysis of cosmological n body simulations by applying a new elaborate technique called multiscale morphology filter@xcite .
the mean density contrast of pancakes aka walls the @xmath0cdm cosmological model is about 1.1 while that of filaments and clusters is 4.5 and 73 respectively .
the panels in the middle and on the right demonstrate the structure rendered with four and sixteen times higher resolution respectively .
they clearly show all essential components of the cosmic web : clusters , filaments , pancakes , and voids .
summarizing we stress the major points .
the phase space sheet contains significantly more dynamical information than coordinates and velocities of particles if particles are treated as independent items .
the connectivity of the tessellation allows to recover underlying structure much more accurately and with much greater spatial resolution than dot plots or even self
adaptive sph@xcite . instead of using the particle discretization of the pss
the suggested method is based on a piecewise - linear approximation .
the new technique allows to reveal , visualize and quantify the complexity of the cosmic web at a much deeper and more profound level . |
representing data as graphs is becoming increasingly popular , as technological progress facilitates measuring `` connectedness '' in a variety of domains , including social networks , trade - alliance networks , and brain networks . while the theory of pattern recognition is deep @xcite , previous theoretical efforts regarding pattern recognition
almost invariably assumed data are collections of vectors . here , we assume data are collections of graphs ( where each graph is a set of vertices and a set of edges connecting the vertices ) . for some data sets , the vertices of the graphs are _ labeled _ , that is , one can identify the vertex of one graph with a vertex of the others ( note that this is a special case of assuming vertices are labeled , where each vertex has a unique label ) .
for others , the labels are unobserved and/or assumed to not exist .
we investigate the theoretical and practical implications of the absence of vertex labels .
these implications are especially important in the emerging field of `` connectomics '' , the study of connections of the brain @xcite . in connectomics ,
one represents the brain as a graph ( a brain - graph ) , where vertices correspond to ( groups of ) neurons and edges correspond to connections between them . in the lower tiers of the evolutionary hierarchy ( e.g. , worms and flies )
, many neurons have been assigned labels @xcite . however , for even the simplest vertebrates , vertex labels are mostly unavailable when vertices correspond to neurons .
classification of brain - graphs is therefore poised to become increasingly popular .
although previous work has demonstrated some possible strategies of graph classification in both the labeled @xcite and unlabeled @xcite scenarios , relatively little work has compared the theoretical limitations of the two .
we therefore develop a random graph model amenable to such theoretical investigations .
the theoretical results lead to universally consistent graph classification algorithms , and practical approximations thereof .
we demonstrate that the approximate algorithm has desirable finite sample properties via a real brain - graph classification problem of significant scientific interest : sex classification .
a labeled graph @xmath0 consists of a vertex set @xmath1 , where @xmath2 is the number of vertices , and an edge set @xmath3 , where @xmath4 .
let @xmath5 be a _ labeled _ graph - valued random variable taking values @xmath6 , where @xmath7 is the set of labeled graphs on @xmath8 vertices .
the cardinality of @xmath7 is super - exponential in @xmath8 .
for example , when all labeled graphs are assumed to be simple ( that is , undirected binary edges without loops ) , then @xmath9 .
let @xmath10 be a categorical random variable , @xmath11 , where @xmath12 .
assume the existence of a joint distribution , @xmath13 which can be decomposed into the product of a class - conditional distribution ( likelihood ) @xmath14 and a class prior @xmath15 .
because @xmath8 is finite , the class - conditional distributions @xmath16 can be considered discrete distributions @xmath17 , where @xmath18 is an element of the @xmath19-dimensional unit simplex @xmath20 ( satisfying @xmath21 @xmath22 and @xmath23 ) . in the above , it was implicitly assumed that the vertex labels were observed .
however , in certain situations ( such as the motivating connectomics example presented in section [ sec:1 ] ) , this assumption is unwarranted . to proceed , we define two graphs @xmath24 to be isomorphic if and only if there exists a vertex permutation ( shuffle ) function @xmath25 such that @xmath26 .
let @xmath27 be a permutation - valued random variable , @xmath28 , where @xmath29 is the space of vertex permutation functions on @xmath8 vertices so that @xmath30 .
[ def : shuffled ] let @xmath31 be a _ shuffled _ graph - valued random variable , that is , a labeled graph valued random variable that has been passed through a random shuffle channel @xmath27 . extending the above graph - classification model to include this vertex shuffling distribution yields @xmath32
we assume throughout this work ( with loss of generality ) that the shuffling distribution is both _
class independent _ and _ graph independent _ ; therefore , this joint model can be decomposed as @xmath33 as in the labeled case , the shuffled graph class - conditional distributions @xmath34 can be represented by discrete distributions @xmath35 . because @xmath36 can be any of @xmath37 different graphs
, it must be that @xmath38 . when @xmath39 is uniform on @xmath29 , all shuffled graphs within the same isomorphism set are equally likely ; that is @xmath40 for some @xmath41 .
note that one can think of a labeled graph as a shuffled graph for which @xmath27 is a point mass at @xmath42 , where @xmath43 is the identity matrix .
the above shuffling view is natural whenever the vertices of the collection of graphs share a set of labels , but the labeling function is unknown .
however , when the vertices of the collection of graphs have different labels , perhaps a different view is more natural . an _ unlabeled graph _
@xmath44 is the collection of graphs isomorphic to one another , that is , @xmath45 .
let @xmath44 be an element of the collection of graph isomorphism sets @xmath46 .
the number of unlabeled graphs on @xmath8 vertices is @xmath47 ( see @xcite and references therein ) .
an _ unlabeling function _
@xmath48 is a function that takes as input a graph and outputs the corresponding unlabeled graph .
let @xmath49 be an _
unlabeled _ graph - valued random variable , that is , a labeled graph - valued random variable that has been passed through an unlabeled channel .
in other words , @xmath50 , and takes values @xmath51 . the joint distribution over unlabeled graphs and classes is therefore @xmath52 , which decomposes as @xmath53 .
the class - conditional distributions @xmath54 over isomorphism sets ( unlabeled graphs ) can also be thought of as discrete distributions @xmath55 where @xmath56 are vectors in the @xmath57-dimensional unit simplex .
comparing shuffling and unlabeling for the independent and uniform shuffle distribution @xmath39 , we have @xmath58 for all @xmath59 .
we consider graph classification in the three scenarios described above : labeled , shuffled , and unlabeled . to proceed , in each scenario we define three mathematical objects : ( i ) a graph classifier , ( ii ) risk , ( iii ) , the bayes optimal classifier , and ( iv ) the bayes risk . a _ labeled graph classifier _
@xmath60 is any function that maps from labeled graph space to class space .
the risk of a labeled graph classifier @xmath61 under @xmath62 loss is the expected misclassification rate @xmath63 $ ] , where the expectation is taken against @xmath13 . the _ labeled graph bayes optimal classifier _ is given by @xmath64 where @xmath65 is the set of possible labeled graph classifiers .
the _ labeled graph bayes risk _ is given by @xmath66 , where @xmath67 implicitly depends on @xmath13 .
a _ shuffled graph classifier _ is also any function @xmath60 ( note that the set of shuffled graphs is the same as the set of labeled graphs ) .
however , by virtue of the input being a shuffled graph as opposed to a labeled graph , the shuffled risk under @xmath62 loss is given by @xmath68 $ ] , where the expectation is taken against @xmath69 .
the _ shuffled graph bayes optimal classifier _ is given by @xmath70 where @xmath65 is again the set of possible labeled ( or shuffled ) graph classifiers .
the _ shuffled graph bayes risk _ is given by @xmath71 , where @xmath72 implicitly depends on @xmath69 .
an _ unlabeled _ graph classifier @xmath73 is any function that maps from unlabeled graph space to class space .
the risk under @xmath62 loss is given by @xmath74 $ ] , where the expectation is taken against @xmath75 .
the _ unlabeled graph bayes optimal classifier _ is given by @xmath76 the _ unlabeled graph bayes risk _ is given by @xmath77 , where @xmath78 is the set of possible unlabeled graph classifiers and @xmath79 implicitly depends on @xmath75 .
the three bayes optimal graph classifiers can be written explicitly in terms of their model parameters : @xmath80
the result of either shuffling or unlabeling a graph can only degrade , but not improve bayes risk .
this is a restatement of the data processing lemma for this scenario .
specifically , @xcite shows that the data processing lemma indicates that in the classification domain @xmath81 for any transformation @xmath82 and data @xmath83 . in our setting , this becomes : [ thm:1 ] @xmath84 .
assume for simplicity @xmath85 and @xmath86 .
@xmath87 an immediate consequence of the above proof is that the inequality in the statement of lemma [ thm:1 ] strict whenever the inequality in eq .
is strict : [ thm:2 ] @xmath88 if and only if there exists @xmath44 such that @xmath89 the above result demonstrates that even when the labels _ do _ carry some class - conditional signal , it may be the case that shuffling or unlabeling does not degrade performance .
in other words , the following two statements are equivalent : ( i ) the labels contain information with regard to the classification task , and ( ii ) some graphs within an isomorphism set are class - conditionally more likely than others : @xmath90 where @xmath91 for some @xmath92 , @xmath93 , and @xmath94 .
uniform shuffling has the effect of `` flattening '' likelihoods within isomorphism sets , from @xmath18 to @xmath95 , so that @xmath95 satisfies @xmath96 .
but just because the shuffling changes class - conditional likelihoods does _ not _ mean that bayes risk must also change .
this result follows immediately upon realizing that posteriors can change without classification performance changing .
the above results are easily extended to consider non - equal class priors and @xmath97-class classification problems . to see this , ignoring ties
, simply replace each minimum likelihood with a sum over all non - maximum posteriors : @xmath98 prior to concluding this section , we remark that one can achieve bayes optimal risk using graph invariants .
a graph invariant on @xmath7 is any function @xmath99 such that @xmath100 for all @xmath101 and @xmath93 ( note that an unlabeling function @xmath102 is a special case of @xmath99 ) .
a graph invariant classifier is a composition of a classifier with an invariant function , @xmath103 .
the bayes optimal graph invariant classifier minimizes risk over all invariants : @xmath104,\end{aligned}\ ] ] where @xmath105 is the space of all possible invariants and @xmath106 is the space of classifiers composable with invariant @xmath99 .
the expectation in eq . is taken against @xmath13 or equivalently @xmath69 , since invariants are invariant .
let @xmath107 denote the bayes invariant risk .
[ thm:3 ] @xmath108 .
let @xmath99 indicate in which equivalence set @xmath109 resides ; that is , @xmath110 if and only if @xmath111
. then @xmath112
throughout this paper , we consider two distinct `` flavors '' of graph classifiers that we can estimate from the data : ( i ) bayes plug - in and ( iii ) nearest neighbor .
below , we first introduce bayes plug - in graph classifiers . in the following sections , we will discuss their asymptotic properties , as well as the asymptotic properties of @xmath113 nearest neighbor graph classifiers . implementing the above optimal classifiers
requires knowing the model parameters .
when the parameters are unknown ( effectively always ) , we assume that the data are sampled identically and independently from some unknown joint distribution : @xmath114 . for _ labeled _ graph classification , @xmath39 is assumed to be the identity function , therefore , @xmath115}$ ] , because when graphs are labeled @xmath116 . for _ shuffled _ graph classification @xmath39 is assumed to be uniform over the permutation matrices , so that all label information is both unavailable and irrecoverable .
the training data are therefore @xmath117}$ ] , where @xmath118 . for _ unlabeled _ graph classification
the training data are again @xmath119 .
our task is to utilize training data to induce a classifier that approximates a bayes classifier as closely as possible .
a _ labeled _ graph bayes plugin classifier , @xmath120 , estimates the parameters @xmath121 using the training data @xmath115}$ ] , and then plugs those estimates into the labeled bayes classifier , eq . , resulting in @xmath122 where the dependency on the training data is implicit in the @xmath123 notation .
a _ shuffled _ graph bayes plugin classifier , @xmath124 , estimates the parameters @xmath125 using the training data @xmath117}$ ] , and then plugs those estimates into the shuffled bayes classifier , eq . , resulting in @xmath126 an _ unlabeled _ graph bayes plugin classifier , @xmath127 , first determines in which unlabeled set each shuffled graph resides , using @xmath99 as defined in section [ sec : gi ] .
then , it estimates the parameters @xmath128 and @xmath129 using the training data @xmath119 .
finally , it plugs those estimates into the unlabeled bayes classifier , eq . , resulting in @xmath130 for brevity , we will sometimes refer to the above three induced classifiers as simply `` classifiers '' .
moreover , the sequence of classifiers ( for example , @xmath131 ) we will also refer to as a `` classifier '' .
the three parametric classifiers , eqs .
, admit classifier estimators that exist , are unique , and moreover , are universally consistent , although the relative convergence rates and values that they converge to differ .
let @xmath132 be the risk of the induced _ labeled _ graph bayes plugin classifier using the training data @xmath133 to obtain maximum likelihood estimators for @xmath121 .
note that @xmath134 is a random variable , as it is a function of the random training data @xmath133 .
this yields [ thm:4 ] @xmath135 as @xmath136 .
because @xmath137 and @xmath138 are both finite , the maximum likelihood estimates for the categorical parameters @xmath121 are guaranteed to exist and be unique @xcite .
hence , the labeled graph bayes plugin classifier is universally consistent to @xmath139 ( that is , it converges to @xmath139 regardless of the true joint distribution , @xmath140 ) @xcite .
similarly , let @xmath141 be the risk of the induced _ shuffled _ graph bayes plugin classifier using the training data @xmath119 to obtain maximum likelihood estimators for @xmath125 .
this yields [ cor : sh_plug ] @xmath142 as @xmath136 .
the previous proof rests on the finitude of @xmath7 , which remains finite after shuffling ( uniform or otherwise ) , and therefore , the previous proof holds , replacing @xmath67 with @xmath72 .
thus while one could merely plug the shuffled graphs into @xmath95 , such a procedure is inadvisable .
specifically , the above procedure does not use the fact that all @xmath143 whenever @xmath144 for some @xmath145 .
instead , consider the risk @xmath146 of the induced _ unlabeled _ graph bayes plugin classifier upon using the @xmath99 function to map each shuffled graph to its corresponding unlabeled graph , and then obtaining maximum likelihood estimates of the unlabeled graph parameters , @xmath147 .
[ cor : un_plug ] @xmath148 as @xmath136 .
because @xmath149 ( by a factor of approximately @xmath150 ) , it follows that classifying by first projecting the graphs into a lower dimensional space should yield improved performance .
specifically , we have the following result : [ thm : tdomp ] @xmath151 dominates @xmath152 for _ shuffled _ graph data .
consider the scalar @xmath153 decomposed into the vector @xmath154 , where each @xmath155 .
note that each @xmath156 .
yet , the estimators , @xmath157 and @xmath158 are not equal , because the former can borrow strength from all shuffled graphs within the same unlabeled graph , but the latter does not . assuming without loss of generality that the class priors are equal and known , the above domination claim is equivalent to stating that for each @xmath109 , @xmath159 \leq { \mathbb{p}}[{\operatornamewithlimits{argmax}}_{y \in { \mathcal{y } } } { \hat{\theta}}'_{g|y } \neq { \operatornamewithlimits{argmax}}_{y \in { \mathcal{y } } } \theta'_{g|y } | { \mathcal{t}}_s'].\end{aligned}\ ] ] because @xmath160 , the only difference between the two sides of the above inequality is the estimators .
we know that the estimators have the following distributions : @xmath161 where @xmath162 is the number of observations of any @xmath111 in the training data , and @xmath163 is the number of observations of @xmath109 in the training data . from this
, we see that for each @xmath109 , @xmath164 will have a tighter concentration around the truth due to is borrowing strength , because @xmath165 , so our result holds .
corollary [ cor : un_plug ] demonstrates that one can induce a universally consistent classifier @xmath151 using eq . .
lemma [ thm : tdomp ] further shows that the performance of @xmath151 dominates @xmath152 . yet
, using @xmath151 is practically useless for two reasons .
first , it requires solving @xmath166 graph isomorphism problems .
unfortunately , there are no algorithms for solving graph isomorphism problems with worst - case performance known to be in only polynomial time @xcite .
second , the number of parameters to estimate is super - exponential in @xmath8 ( @xmath167 ) , and acceptable performance will typically require @xmath168 .
we can therefore not even store the parameter estimates for small graphs ( e.g. , @xmath169 ) , much less estimate them .
this motivates consideration of an alternative strategy .
a @xmath113 nearest - neighbor ( @xmath170nn ) classifier using euclidean norm distance is universally consistent to @xmath67 for vector - valued data as long as @xmath171 with @xmath172 as @xmath136 @xcite .
this non - parametric approach circumvents the need to estimate many parameters in high - dimensional settings such as graph - classification .
the universal consistency proof for @xmath113nn was extended to graph - valued data in reference @xcite , which we include here for completeness .
specifically , to compare labeled graphs , reference @xcite considered a frobenius norm distance @xmath173 where @xmath174 is the adjacency matrix representation of the labeled graph , @xmath175 .
let @xmath176 denote the frobenius norm @xmath113nn classifier on _ labeled _ graphs using @xmath177 , and let @xmath178 indicate the misclassification rate for this classifier .
reference @xcite showed : [ thm:5 ] @xmath179 as @xmath136 . because both @xmath180 and @xmath138 have finite cardinality , the law of large numbers ensures that eventually as @xmath136 , the plurality of nearest neighbors to a test graph will be identical to the test graph .
let @xmath181 denote the frobenius norm @xmath113nn classifier on _ shuffled _ graphs using @xmath182 , and let @xmath183 indicate the misclassification rate for this classifier .
from the above lemma and corollary [ cor : sh_plug ] , the below follows immediately : [ cor : knn1 ] @xmath184 as @xmath136 .
given shuffled graph data @xmath119 , however , other distance metrics appear more `` natural '' to us .
for example , consider the `` graph - matched frobenius norm '' distance : @xmath185 where @xmath186 and @xmath187 are shuffled adjacency matrices .
let @xmath188 indicate the misclassification rate of the @xmath113nn classifier using the above graph - matched norm @xmath182 _ shuffled _ graphs , and let @xmath183 indicate the misclassification rate for this classifier . given an exact graph matching function
a function that actually solves eq .
we have the following result : [ cor : sh_knn ] @xmath189 as @xmath136 . thus , given shuffled data @xmath119 , one could consider either @xmath190 or @xmath191 .
interestingly , when the data are labeled graphs , @xmath133 , one can outperform @xmath190 by _ shuffling _ , that is , by apparently destroying the label information .
consider an example in which @xmath192 , such that no information is in the labels .
in such scenarios , shuffling can effectively borrow strength from different labeled graphs that are within the same unlabeled graph set .
let @xmath193 indicate the misclassification rate of the @xmath113nn classifier using @xmath182 _ labeled _ graphs , and let @xmath194 indicate the misclassification rate for this classifier .
we therefore state without proof : [ thm : nodom ] neither @xmath195 nor @xmath190 dominates when data are _ labeled _ graphs .
thus , when the training data consists of shuffled graphs , the best universally consistent classifier ( of those considered herein ) is a @xmath113nn that uses @xmath182 as the distance metric .
other universally consistent classifiers that we considered either require estimating more parameters than there are molecules in the universe , or are inadmissible under @xmath62 loss . when vertex labels are available , no classifier dominates .
the above theoretical results consider bayes plug - in and @xmath113nn classifiers . here
we consider other classifiers .
specifically , let @xmath196 be the misclassification rate for some classifier that operates on @xmath119 , that is , only has access to shuffled graphs .
consider the set of seven graph invariants studied in @xcite : size , max degree , max eigenvalue , scan statistic , number of triangles , and average path length . via monte
carlo , @xcite was unable to find a uniformly most powerful graph invariant ( test statistic @xcite ) for a particular hypothesis testing scenario with unlabeled graphs .
the above results , however , indicate that there exists optimal classifiers ( or test statistics ) for any unlabeled or shuffled graph setting . to proceed , let @xmath197 be the _ chance _ classifier , that is @xmath198 and let @xmath199 be the misclassification rate for this classifier .
moreover , let @xmath200 be the risk of the invariant classifier that is equivalent to the unlabeled bayes plug - in classifier ( see lemma [ thm:3 ] ) . from the above results , it follows that : [ thm : order ] in expectation , + @xmath201 as @xmath136 .
while asymptotic results can be informative and insightful , understanding the computational properties of the different classifiers can be as ( or even more ) informative for real applications .
table [ tab : comp ] compares the space and time complexity of the various classifiers considered above .
only the @xmath170nn classifiers have the property that they do not require more space than there are atoms in the universe ( for any @xmath8 bigger than @xmath202 ) . of those ,
the labeled @xmath113nn classifier does not require time exponential in the number of vertices .
therefore , we only found one type of classifier with performance guarantees that has both polynomial space and time .
unfortunately , the finite sample performance of this classifier is abysmal .
this motivates constructing approximate classifiers .
.order of computational properties for training the various shuffled graph classifiers . [ cols=">,^,^,^",options="header " , ] [ tab : comp ]
we buttress the above theoretical results via numerical experiments .
the asymptotic results combined with the computational complexities of the above described algorithm suggest that none of the proposed algorithms have all the properties we effectively require for real world applications , in particular , polynomial space and time complexity , as well as reasonable convergence rates .
we therefore propose a different algorithm , which lacks universal consistency , but can be run on real data with good hope for reasonable performance . in particular , we modify @xmath191 , the _ unshuffled _ @xmath170nn classifier . instead of requiring this classifier to actually solve the graph matching problem , eq . , we use a recently proposed state - of - the - art approximate cubic time algorithm @xcite . denote this classifier @xmath203 .
a `` connectome '' is a brain - graph in which vertices correspond to ( groups of ) neurons , and edges correspond to connections between them .
diffusion magnetic resonance ( mr ) imaging and related technologies are making the acquisition of mr connectomes routine @xcite .
49 subjects from the baltimore longitudinal study on aging comprise this data , with acquisition and connectome inference details as reported in @xcite .
each connectome yields a @xmath204 vertex simple graph ( binary , symmetric , and hollow adjacency matrix ) . associated with each graph is class label based on the sex of the individual ( 24 males , 25 females ) .
because the vertices are labeled , we can compare the results of having the labels and not having the labels .
consider the following five classifiers : * @xmath177-@xmath205nn : a @xmath205-nearest neighbor ( @xmath205nn ) with frobenius norm distance on the _ labeled _ adjacency matrices .
* @xmath182-@xmath205nn : a @xmath205nn with frobenius norm distance on the _ shuffled _ adjacency matrices .
* @xmath206-@xmath205nn : a @xmath205nn with an _ approximate _ graph - matched frobenius norm distance on the shuffled adjacency matrices , as described above .
because graph - matching is @xmath207-hard @xcite , we instead use an inexact graph matching approach based on the quadratic assignment formulation described in @xcite , which only requires @xmath208 time . *
@xmath99-@xmath205nn : a @xmath205nn with euclidean distance using the seven graph invariants described above .
prior to computing the euclidean distance , for each invariant , we rescale all the values to lie between zero and one . *
@xmath209 : use the chance classifier defined above .
performance is assessed by leave - one - out misclassification rate .
figure [ fig:1 ] reifies the above theoretical results in a particular finite sample regime .
we apply the five algorithms discussed above to sub - samples of the connectome data , which shows approximate convergence rates for this data . fortunately
, this real data example supports the main lemmas of this work .
specifically , the @xmath113nn classifier using @xmath177 on the _ labeled _ graphs ( dashed gray line ) achieves the lowest misclassification rate for all @xmath166 , which one would expect if labels contain appropriate class signal .
moreover , the @xmath113nn classifier using the inexact graph - matching frobenius norm on the shuffled adjacency matrices , @xmath206 , performs best of all classifiers using only shuffled graphs ( compare dashed black line with solid black and gray lines ) . on the other hand , while the @xmath113nn classifier using the frobenius norm on shuffled graphs , @xmath182 , must eventually converge to @xmath210 , its convergence rate is quite slow , so the classifier using standard invariants @xmath99 outperforms the simple @xmath182 based @xmath113nn .
, such that errorbars were neglibly small .
five classifiers were compared , as described in main text . note that when @xmath166 is larger than @xmath211 , as predicted by theory , we have @xmath212 .
moreover , @xmath213 . ]
in this work , we address both the theoretical and practical limitations of classifying shuffled graphs , relative to labeled and unlabeled graphs . specifically , first we construct the notion of shuffled graphs and shuffled graph classifiers in a parallel fashion with labeled and unlabeled graphs / classifiers , as we were unable to find such notions in the literature .
then , we show that shuffling the vertex labels results in an irretrievable situation , with a possible degradation of classification performance ( lemma [ thm:1 ] ) .
even if the vertex labels contained class - conditional signal , bayes performance may remain unchanged ( lemma [ thm:2 ] ) .
moreover , although one can not recover the vertex labels , one can obtain a bayes optimal classifier by solving a large number of graph isomorphism problems ( lemma [ thm:3 ] ) .
this resolves a theoretical conundrum : is there a set of graph invariants that can yield a universally consistent graph classifier ? when the generative distribution is unavailable , one can induce a consistent and efficient `` unshuffling '' classifier by using a graph - matching strategy ( corollary [ cor : un_plug ] ) . while this unshuffling approach dominates the more nave approach ( lemma [ thm : tdomp ] ) , it is intractable in practice due to the difficulty of graph matching and the large number of isomorphism sets . instead
, a frobenius norm @xmath113nn classifier applied to the adjacency matrices may be used , which is also universally consistent ( corollary [ cor : sh_knn ] ) .
surprisingly , none of the considered classifiers dominate the other for labeled data ( lemma [ thm : nodom ] ) , yet asymptotically , we can order shuffled graph classifiers ( lemma [ thm : order ] ) .
because graph - matching is @xmath207-hard , we instead use an approximate graph - matching algorithm in practice ( see @xcite for details ) . applying these @xmath113nn classifiers to a problem of considerable scientific interest classifying human mr connectomes
we find that even with a relatively small sample size ( @xmath214 ) , the approximately graph - matched @xmath113nn algorithm performs nearly as well as the @xmath113nn algorithm _ using _ vertex labels , and slightly better than a @xmath113nn algorithm applied to a set of graph invariants proposed previously @xcite .
this suggests that the asymptotics might apply even for very small sample sizes .
thus , this theoretical insight has led us to improved practical classification performance .
extensions to weighted or ( certain ) attributed graphs are straightforward .
this work was partially supported by the research program in applied neuroscience .
l. devroye , l. gyrfi , g. lugosi , and l. gyorfi , _ a probabilistic theory of pattern recognition_.1em plus 0.5em minus 0.4emnew york : springer , 1996 .
[ online ] .
available : http://www.amazon.ca/exec/obidos/redirect?tag=citeulike09-20&path=asin/0387946187 j. white , e. southgate , j. n. thomson , and s. brenner , `` the structure of the nervous system of the nematode caenorhabditis elegans . ''
_ philosophical transactions of royal society london .
series b , biological sciences _ , vol .
1165 , pp . 1340 , 1986 .
r. p. w. duin , e. pkalska , and e. pkalskab , `` the dissimilarity space : bridging structural and statistical pattern recognition , '' _ pattern recognition letters _ , vol . in press ,
april , may 2011 .
[ online ] .
available : http://linkinghub.elsevier.com/retrieve/pii/s0167865511001322 http://www.sciencedirect.com/science/article/pii/s0167865511001322 http://dx.doi.org/10.1016/j.patrec.2011.04.019[http://linkinghub.elsevier.com/retrieve/pii/s0167865511001322 http://www.sciencedirect.com/science/article/pii/s0167865511001322 http://dx.doi.org/10.1016/j.patrec.2011.04.019 ] j. t. vogelstein , r. j. vogelstein , and c. e. priebe , `` are mental properties supervenient on brain properties ? '' _ nature scientific reports _ , vol . in press , p. 11
[ online ] .
available : http://arxiv.org/abs/0912.1672 h. pao , g. a. coppersmith , c. e. priebe , h. p. ao , g. a. c. oppersmith , and c. e. p. riebe , `` statistical inference on random graphs : comparative power analyses via monte carlo , '' _ journal of computational and graphical statistics _ , pp . 122 , 2010 .
[ online ] .
available : http://pubs.amstat.org/doi/abs/10.1198/jcgs.2010.09004 p. hagmann , l. cammoun , x. gigandet , s. gerhard , p. ellen grant , v. wedeen , r. meuli , j .-
thiran , c. j. honey , and o. sporns , `` mr connectomics : principles and challenges , '' _ j neurosci methods _ , vol .
194 , no . 1 ,
pp . 3445 , 2010 .
[ online ] .
available : http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=retrieve&db=pubmed&dopt=citation&list_uids=20096730 w. r. gray , j. a. bogovic , j. t. vogelstein , b. a. landman , j. l. prince , and r. j. vogelstein , `` magnetic resonance connectome automated pipeline : an overview . ''
_ ieee pulse
_ , vol . 3 , no . 2 ,
pp . 428 , mar
[ online ] .
available : http://ieeexplore.ieee.org/xpl/articledetails.jsp?arnumber=6173097 m. r. garey and d. s. johnson , _ computers and intractability : a guide to the theory of np - completeness ( series of books in the mathematical sciences)_.1em plus 0.5em minus 0.4emw .
h. freeman , 1979 .
[ online ] .
available : http://www.amazon.com/computers-intractability-np-completeness-mathematical-sciences/dp/0716710455 j. t. vogelstein , j. c. m. conroy , l. j. podrazik , s. g. kratzer , d. e. fishkind , r. j. vogelstein , and c. e. priebe , `` ( brain ) graph matching via fast approximate quadratic programming , '' _ arxiv preprint _ | we develop a formalism to address statistical pattern recognition of graph valued data .
of particular interest is the case of all graphs having the same number of uniquely labeled vertices . when the vertex labels are latent , such graphs are called
_ shuffled graphs_. our formalism provides insight to trivially answer a number of open statistical questions including : ( i ) under what conditions does shuffling the vertices degrade classification performance and ( ii ) do universally consistent graph classifiers exist ? the answers to these questions lead to practical heuristic algorithms with state - of - the - art finite sample performance , in agreement with our theoretical asymptotics . |
soon after the first explicit formulations of supersymmetric ( susy ) versions of the standard model ( sm ) @xcite , enforcing r - parity conservation to prevent sources of violation of baryon and lepton number at tree - level , it was realized , as an interesting by - product , that these theories could naturally embed a candidate for cold dark matter @xcite .
in fact , it had been shown long before that weakly - interacting massive particles ( wimps ) have a thermal relic abundance that is generically at a level relevant for cosmology , and the stable lsp falls under this category if it happens to be electric and color neutral ( and as it is indeed the case in large portions of the susy parameter space ) .
two decades later , no susy particle has unfortunately been found at particle accelerators and no clean indication of susy has emerged from tests of rare processes or precision measurements . the initial theoretical motivations for susy , although still very appealing , are being questioned , with the scale of physics beyond the sm that is slowly , but steadily , drifting above the weak scale . on the other hand , in these twenty years , the case for non - baryonic cold dark matter ( cdm ) being the building block of the universe has become stronger and stronger , with the determination of the cdm contribution to the universe energy density which has already reached a 20% level of accuracy @xcite and is going to improve further in the upcoming years .
it is then not surprising that in formulating an extension to the sm , sometimes , the issue of incorporating a dark matter candidate has changed from being a by - product of the proposed scenario into being one of the hinges of the theory itself .
this is the case for a recently proposed susy framework , dubbed `` split supersymmetry '' @xcite .
split supersymmetry indicates a generic realization of the susy extension to the sm where fermionic superpartners feature a low mass spectrum ( say at the tev scale or lower ) , while scalar superpartners are heavy , with a mass scale which can in principle range from hundreds of tev up to the gut or the planck scale @xcite . in split susy one
gives up on the idea that susy is the tool to stabilize the weak scale , since some other mechanism should anyway account for other fine tuning issues , such as that of the cosmological constant . on the other hand ,
a number of phenomenological problems appearing in ordinary susy setups , where the whole spectrum of superpartners is supposed to be light , are cured : the heavy sfermions minimize flavor and cp violating effects , alleviate proton decay and avoid an excessively light higgs boson . at the same time , two major features are maintained : the successful unification of gauge couplings , and the central issue in our analysis
the lsp as a viable particle candidate for cdm . finally , it is certainly a well - motivated scenario , as it has been shown that the occurrence of a split susy spectrum is indeed a generic feature of a wide class of theories @xcite .
it arises , among others , in frameworks where the mechanism of susy breaking preserves a r - symmetry and forbids gaugino and higgsino mass terms , or in case of direct susy breaking through renormalizable interactions with the susy breaking sector ( direct mediation ) @xcite . rather than focusing on some specific realizations ,
we consider here a generic split susy setup : we refer to the minimal susy extension to the standard model ( mssm ) , set the scalar sector at a heavy scale , but allow for the most general fermion sector .
we investigate how cosmological constraints affect the definition of such framework and discuss its testability by examining the issue of dark matter detection within this wide class of models , comparing also with the perspectives to test this scenario at accelerators . as a result ,
we work out here a fully general analysis of neutralino dark matter on the basis of the lightest neutralino mass and composition , largely independent of the split supersymmetry framework ; in this respect , we outline a reference guide to the sensitivity of upcoming detection experiments which applies , especially as fas as indirect detection is concerned , to supersymmetric models with a generic ( moderatly ) heavy scalar sector .
once the lsp composition and mass is known , our results allow to get a fair estimate of the detection prospects .
the paper is organized as follows : in section [ sec : paramspace ] we describe our working framework and discuss the current constraints to its parameter space . in section
[ sec : rates ] we introduce the relevant dark matter detection techniques and show rates in extremal regimes of the parameter space . in section
[ sec : casestudy ] we discuss in detail the phenomenology of the split susy setup along a sample slice in the parameter space , show constraints from current data and address detectability in future searches , cross - checking the outreach of the different techniques . in section [ sec : discussion ] we indicate how to generalize the results of the previous section to the full parameter space . finally , we draw an outlook and conclude in section [ sec : conclusion ] .
we consider a mssm setup , with the relevant parameters defined at at the low energy scale and without implementing any unification scheme .
we restrain to the split susy framework by assuming that the sfermion sector is much heavier than the fermion sector and hence is completely decoupled ; as for the higgs sector , we assume there is only one light sm - like higgs , with a mass labeled by the parameter @xmath5 , while the other two neutral higgs and the charged higgs are taken again to be very heavy , and hence decoupled .
finally , we will leave out of most of our discussion the gluino and the gravitino , supposing they are ( moderately ) heavy , just mentioning what are possible consequences in case this is not true . in this scheme , the lightest neutralino , defined as the lightest mass eigenstate from the superposition of the two neutral gaugino and the two neutral higgsino fields
, @xmath6 is automatically the lsp .
the coefficients @xmath7 , obtained by diagonalizing the neutralino mass matrix , are mainly a function of the bino and the wino mass parameters @xmath8 and @xmath9 , and of the higgs superfield parameter @xmath10 , while depending rather weakly on @xmath11 , the ratio of the vacuum expectation values of the two neutral components of the su(2 ) higgs doublets .
the hierarchy between @xmath9 and @xmath10 determines also whether the lightest chargino is wino like or higgsino like , and again the role @xmath11 is minor .
the phenomenology of the scheme we are considering is hence fully defined by only five parameters : @xmath12 in this construction , constraints from accelerator data are rather weak , essentially just in connection with the fact that charginos have not been observed so far : we implement , as a conservative limit on the lightest chargino mass , the kinematic limit from the last phase of lep2 , _ i.e. _ @xmath13103.5 gev @xcite ( notice however that in large portions of our parameter space the lightest neutralino and lightest chargino are nearly degenerate in mass , and this limit should be slightly relaxed ; however , this does not critically enter in our discussion ) .
+ + much more powerful are constraints on the model from cosmology ; to implement the cosmological bounds , we refer to the determination of the cdm component of the universe by wmap @xcite : @xmath14 .
lsp relic densities are computed with the darksusy package @xcite , which allows for high accuracy solutions of the equations describing thermal decoupling .
we study the parameter space varying freely the mass parameters @xmath8 , @xmath9 and @xmath10 ( no gut relations are assumed between @xmath8 and @xmath9 ; we restrict to positive @xmath10 , since in this contest the case of negative @xmath10 is just specular ) , and taking the sample value @xmath15 and @xmath16 gev ( close to the current lower limit on the mass of a sm - like higgs ) . in fig .
[ m1m2 ] we consider slices , in the parameter space , along the planes @xmath17 ( left - hand - side panel ) and @xmath18 ( right - hand - side panel ) and for large values of the third mass parameter ( which we set to 10 tev ) . in the first case , since @xmath8 and @xmath9 are both diagonal entries in the neutralino mass matrix , the transition of the lsp from being bino like ( above the diagonal in the figure ) to being wino like ( below it ) is very sharp .
the two regimes are very different because winos annihilate very efficiently into gauge bosons , while binos can annihilate just into fermions .
the latter annihilation processes are very sharply suppressed , as compared to ordinary susy setups , since diagrams with @xmath19- and @xmath20-channel sfermions exchanges do not contribute here . in general , the relic density of the lsp scales with the inverse of the annihilation rate : it is very large for binos regardless of the lsp mass , while it tends to be too small for winos , unless one considers @xmath9 in the 2 tev range .
the exception is in a narrow strip above the diagonal where the bino lsp is nearly degenerate in mass with the next - to - lightest wino like neutralino and chargino states , and the thermal equilibrium of the lsp in the early universe is enforced through these slightly heavier particles rather than by just lsp pair annihilations into sm particles : this is a very well - studied effect , usually dubbed `` coannihilation '' @xcite , but that has been very seldom considered in this specific realization , since usually @xmath8 and @xmath9 are not taken as independent parameters .
the region in the plane where the lsp relic abundance is too large is shaded . starting from light lsps
( the mass of the lsp is essentially the minimum between @xmath8 and @xmath9 ) in the bottom - left corner , the border of the excluded region approaches the diagonal , since increasing the mass scale the cross sections decreases , and hence coannihilation effects have to become larger and larger , _ i.e. _ mass splitting with coannihilating particles smaller and smaller .
we reach then the mass value at which this process is saturated , and the lsp turns into wino like ; above this scale , even for winos , the annihilation cross section becomes too small and cdm is overproduced . in the figure , since we are plotting results on logarithmic scales , the iso - level curve at @xmath21 is essentially overlapping with the border of the excluded region , while in the remaining part of the plane , cdm is underproduced . in fig .
[ m1m2 ] and [ 3dlabels ] we shade in red the regions with an excessively light chargino . for the right - hand - side panel of fig .
[ m1m2 ] , the picture is analogous .
the wino mass parameter is assumed to be heavy , hence we are left with the possibility that the lsp is either bino like or higgsino like , depending on the relative values of @xmath8 and @xmath10 .
being the parameter @xmath10 in off - diagonal entries in the neutralino mass matrix , the transition between the two regimes is smoother than in the previous case , but the trend is the same , with higgsinos annihilating efficiently into gauge bosons and binos with suppressed annihilation rates .
in the computation of the relic density both the effects of mixing and coannihilations enter , and , for instance , a wiggle in the borderline of the excluded region corresponding to the @xmath22 threshold is clearly visible .
this is the scenario which has been considered up to now for estimates of the lsp relic abundance in split susy contexts @xcite .
the value of the higgsino parameter , hence of the higgsino mass , saturating the relic density constraint is smaller than for winos .
this is illustrated also in the left - panel of fig .
[ 3dlabels ] , where iso - level curves for @xmath23 are shown in the plane @xmath24 and for a few values of @xmath8 . in case @xmath8 is very heavy , we see the transition of the lsp from being a pure wino ( upper part of the figure ) to being a pure higgsino ( right - hand side ) .
the size of the saturating parameters differs because the coupling in the most important vertex , namely @xmath25 , reads @xmath26 for higgsinos and @xmath27 for winos , but also because the number of degrees of freedom in the coannihilation process in the two cases is different : as far as higgsinos are concerned , the coannihilating particles are two neutralinos and one chargino , all nearly degenerate in mass , while for winos only one neutralino and one chargino enter ( and this matters , because pair annihilation rates of charginos are larger than those for neutralinos ) .
we find that a pure higgsino of mass equal to @xmath28 gev and a pure wino of mass @xmath29 gev have a relic abundance equal to @xmath21 , the current best fit value from the wmap data .
lower values of @xmath8 imply smaller upper limits on the neutralino mass , corresponding , as sketched before , to bino lsps coannihilating either with winos or higgsinos , with the transition regimes getting progressively wider as @xmath8 decreases .
a 3-d representation of the hypersurface at @xmath30 in the @xmath31 space is given in the right - panel of fig .
[ 3dlabels ] ; the allowed low relic density region stands _ behind _ the plotted surface .
the value of @xmath11 and @xmath5 have not played much of a role in our discussion , since we expect very minor changes in the cosmologically allowed portion of the parameter space if these are varied ( and we have indeed verified numerically that this is the case ) .
such choice of parameters is actually crucial only for direct detection rates , as we will stress below .
finally , the bounds we have derived are valid under the assumption of a standard cosmological setup , and including only thermal sources of dark matter .
the bounds get tighter , or , from another perspectives , models that give here subdominant cdm candidates may become fully consistent with observations , if one introduces particle physics models with non - thermal sources of cdm , such as neutralino productions from gravitinos or moduli decays @xcite , or considers cosmological scenarios with faster expansion rates at the time of decoupling : sample cases are cosmologies with a quintessence energy density term dominating at the lsp freeze - out temperature @xcite , anisotropic universes with effective shear density terms @xcite or in scalar - tensor theories @xcite . relaxing
the upper bound is also possible , though maybe slightly more contrived , as it would involve wiping out ( totally , or in part ) the thermal relic component with an entropy injection : a mechanism of reheating at low temperature , _
i.e. _ lower than the lsp freeze - out temperature which is in the range
@xmath32 @xmath33 , would give this effect @xcite .
however , excluding this latter possibility , it is interesting to notice that , in this generic split susy setup , the cosmological bound is indeed the only issue forcing one of the sectors of the theory to be light , _
i.e. _ at a scale lower than 2.2 tev .
+ the issue of wimp dark matter detection has been studied at length ( for a review , see , _
e.g. _ , @xcite ; see also the more recent works , e.g. refs .
we will show here that this is especially relevant in the split susy scenario by systematically discussing all wimp detection techniques that currently provide , or that will provide in the future , constraints on the model .
rates are computed with the darksusy package @xcite ; the set of underlining assumptions will be briefly described here . in this section
we will present sample rates for dark matter candidates in the three extreme regimes described above , and focusing on models with thermal relic densities in the wmap preferred range @xcite , in case of a standard cosmological setup . in the next two sections ,
these results will be generalized , including also low relic density models , either by relaxing the requirement of thermal production of the whole cdm content of the universe or by referring to non - standard cosmological scenarios ; at that stage we will also put more emphasis on the comparison between different detection techniques . in the last decade
there has been a considerable experimental effort to detect wimps directly , measuring the recoil energy from wimps elastic scattering on nuclei @xcite .
our theoretical predictions are derived with the usual effective lagrangian approach , except that no contributions mediated by squarks are present , and taking a standard set of parameters @xcite for nucleonic matrix elements ( note , e.g. , the strange content here is slightly smaller than the values implemented in other analyses , see @xcite for details ) . in the left panel of fig .
[ curr_dirdet ] we plot the spin - independent ( si ) neutralino - proton scattering cross section @xmath34 , as a function of the neutralino mass .
the thick lines label models on the @xmath30 isolevel curves in fig .
[ m1m2 ] and [ 3dlabels ] , left , i.e. the regimes with one fermion mass parameter heavier than the other two . from the upper left side of figure , along a dashed line , we plot @xmath34 for a light lsp with significant bino - higgsino mixing turning into a heavier pure higgsino state ; along the dashed - dotted curve the higgsino state makes a transition into a pure wino , going through models with large higgsino - wino mixing , and finally on the dotted line the @xmath10 parameter is large and there is just a very sharp transition from pure winos to pure binos when moving again to smaller masses .
since the si scattering cross section is mediated by the light higgs in a t - channel , and the @xmath35 vertex scales with the higgsino - gaugino mixing , whenever this is small the cross section is small as well , as it can be clearly seen in the figure .
the gray - shaded region corresponds to the projection of all points lying onto the hypersurface at @xmath36 in the @xmath31 space ( the hypersurface plotted in fig . [
3dlabels ] , right ) . not surprisingly , this region is close to being the zone delimited by the curves labeling the three extreme regimes just described ( and the same holds for all other detection techniques we discuss in the present section , hence , for the sake of clarity , we omit the corresponding shadings ) .
the black dotted line indicates the current bound on @xmath34 from the cdms collaboration @xcite , while the green dot - dashed line stands for the projected sensitivity of next - generation ( sometimes dubbed _ stage 3 _ ) detectors ; for definiteness , we take as a benchmark experimental setup the xenon facility @xcite .
as it can be seen , the present sensitivity does not allow to set bounds on these models , but future projects will test a large portion of the parameter space . in the right panel of fig .
[ curr_dirdet ] we show the spin - dependent neutralino - proton cross sections , along the same parameter space slices as in the left panel . we also indicate the current @xcite and projected @xcite experimental sensitivities in this detection channel . as already pointed out @xcite , spin - dependent direct neutralino searches in the general mssm appear to be largely disfavored with respect to spin - independent ones : future facilities will reach a sensitivity which is expected to be various orders of magnitude below the theoretical expectations for the models under consideration .
+ the search for neutrinos produced by the annihilation of neutralinos trapped in the core of the gravitational wells of the sun or of the earth has been since long recognized as a promising indirect detection technique . in the present framework ,
fluxes from the earth are actually very low and will not be considered further ; to estimate neutrino fluxes from the sun we implement the standard procedure described in refs .
@xcite , except for a more careful treatment of neutralino capture rates @xcite . in the left panel of fig .
[ higgsinofrac ] , we present results in terms of muon fluxes , above the threshold of 1 gev , and compare them to the current best limits from the super - kamiokande collaboration @xcite and with the future projected sensitivity of the icecube experiment @xcite ( the mismatch in the energy threshold of icecube and the threshold considered here has been taken into account ) .
the color coding on the @xmath37 isolevel curves is the same as in the left panel , making it transparent that the emerging pattern is perfectly specular . the neutrino flux in the sun scales with a capture rate which is dominated by the spin dependent ( sd ) neutralino - proton coupling , and , as for the si term
, this is suppressed in models with small higgsino - gaugino mixing ( and the relevant vertex here is @xmath38 ) . while the sensitivity of current direct dark matter searches is more than one order of magnitude above the largest signals we obtain in the present framework , in the lowest mass range ( @xmath39 gev ) neutrino telescopes already rule out a few models .
however , future direct detection experiments will probe a much wider portion of the parameter space , as compared to what icecube is expected to achieve . the results shown in fig .
[ curr_dirdet ] and [ higgsinofrac ] are derived assuming , as in the previous section , that @xmath15 and @xmath16 gev . @xmath34
scales with @xmath40 , while the dependence on @xmath5 of the neutrino fluxes from the sun is negligible .
@xmath11 does not enter critically in any of the relevant couplings , however it slightly affects the higgsino - gaugino mixing .
we sketch a sample effect of this kind in the right panel of fig .
[ higgsinofrac ] , where the main point is however to illustrate the reason why the rates in the cases considered so far change sharply at @xmath41 tev , in the bino - wino transition along the @xmath37 isolevel curve at large @xmath10 ( red dotted lines ) .
the figure shows that going from the regime @xmath42 to the regime @xmath43 , there is not only a sharp switch in the gaugino content , but also a step - like fall in the ( already small ) higgsino fraction . as we show , a slight dependence on @xmath11 ( which enters in the neutralino mass matrix ) is present .
we will further discuss this effect in the next section .
other indirect detection techniques rely on the fact that there is a finite probability for dark matter wimps to annihilate in pairs in galactic halos , and in particular in the halo of the milky way , possibly giving rise to exotic sources of radiation and cosmic rays . in this analysis
we will consider the production of gamma - rays , antiprotons , positrons and antideuterons within our own galaxy .
the yield per annihilation in all these channels is read out of simulations with the @xcite 6.154 monte carlo code as included in the darksusypackage , except for @xmath44 , where we use the prescription suggested in ref .
since the source strength scales with the number density of neutralino pairs locally in space , i.e. , in terms of the dark matter density profile , with @xmath45 , the predictions for these signals are very sensitive to which @xmath46 is considered for the milky way . in fact , the latter is unfortunately not well - known , and one is forced to make some assumptions . in this analysis , we will mainly focus on two extreme models for the milky way dark matter halo .
they are both inspired to the current picture emerging from n - body simulations of hierarchical clustering in @xmath47cdm cosmologies , respecting , e.g. , the relation between halo mass and concentration parameter suggested therein .
the two models diverge however in the way the effect of the baryon infall in the dark matter potential well is sketched .
one extreme it to suppose that this process happened with a large transfer of angular momentum between the baryons and the dark matter . in this case the system of cold particles , sitting at the center of dark matter halo and forming the cusp seen in the simulations ( which describe the evolution of the collision - less dark halo component only ) , are significantly heated and removed from the central part of the halo .
one such extreme scenario was proposed in ref .
@xcite , suggesting that the resulting dark halo profile can be modeled by the so - called burkert profile @xcite : @xmath48 this profile has been shown to provide fair fits in case of a large sample of the rotation curves for spiral galaxies @xcite , and , from our perspective , having a very large core radius ( the parameter @xmath49 ) , is very conservative .
we will consider a sample choice of the free parameters in eq .
[ eq : burk ] , fixing the length scale parameter @xmath50 kpc and the local halo density @xmath51 gev @xmath52 . the second model we consider stands on the other extreme , and it is derived by supposing that the settling in of the baryons happened with no net transfer of angular momentum between baryons and dark matter .
the baryon infall happened as a slow and smooth process which gradually drove the potential well at the center of the galaxy to become deeper and deeper , accreting more and more dark matter particles in this central region .
this is the _
adiabatic contraction limit _ , which we implement according to the prescription of blumenthal et al .
@xcite ( circular orbit approximation ) .
we start from the cdm profile of the non - singular form extrapolated in ref .
@xcite ( which we label as n03 profile ) , assume as sample virial mass and concentration parameter , respectively , @xmath53 and @xmath54 , derive the enhancement on it induced by the stellar bulge and the disc components , and just cut the profile in the region which is dynamically dominated by the central black hole ( i.e. we assume complete relaxation due to the black hole formation ) .
the resulting spherical profile has a local dark matter density equal to @xmath55 gev @xmath52 .
the choice of parameters in both scenarios is justified by testing the dark matter model against available dynamical constraints .
finally , a self - consistent velocity distribution is derived as well , and the result for the burkert profile has already been exploited above when showing sensitivity curves for direct detection and computing capture rates in the sun ( and both of these change are essentially unchanged in case the adiabatically - contracted n03 profile is chosen ) .
further details on the definition of the halo models can be found in refs .
@xcite .
+ antimatter searches do not play a main role for the models we have considered so far , since we find that the predicted fluxes are well below both current constraints and the sensitivity of next - generation experiments ; at the same time , they look as the best chance for detection in other regions of the parameter space , hence we postpone the relative discussion . on the other hand , data on the gamma - ray flux in the galactic center ( gc ) direction
are already relevant to set combined constraints on our particle physics framework and the dark halo profile .
gamma - ray maps of the gc have been derived both from space , by the egret experiment on the compton gamma - ray observatory , below 10 gev @xcite , and , more recently , from ground by three different air cherenkov telescopes ( acts ) , with an overall energy coverage between 300 gev and 8 tev , and apparently discrepant results concerning spectral shapes of the source and normalization of the flux @xcite .
an interpretation of one of these signals in term of wimp pair annihilations is possible , see , e.g. , the modeling of the egret source in ref .
@xcite , while most probably it is not possible to fit with such a source all of them simultaneously .
we take here a conservative attitude and just assume that the gc gamma - ray flux produced in our model should not exceed the measured one ; of course , in case it were demonstrated that these fluxes were associated to another source , possibly not even located in the gc , as it has already been argued for the egret source @xcite , the constraints we derive here would get stronger . a clearer statement on
this respect should come with the next gamma - ray mission in space , the glast satellite @xcite . in fig .
[ curr_egret ] , left panel , we plot , with a black dotted horizontal line , the most stringent @xmath56 bound from the egret data set @xcite ( which turned out to be the bin at the largest energy @xmath57 ) , compared to the fluxes in this energy bin and computed for the cuspy , adiabatically contracted n03 model ( upper lines ) and cored burkert profile ( lower lines ) .
the fluxes are averaged over the egret angular resolution @xmath58 .
going to larger masses the expected flux decreases both because of the @xmath59 scaling in the number density of neutralino pairs and because we are looking at a rather low energy bin ; the trend is partially smoothed , especially for winos , by the transition along the @xmath37 isolevel curves into states that annihilate more and more efficiently into w bosons ( for pure winos this is maximal ) .
note also , even in this case , the sharp drop of the signal for pure binos , which is now due to the fact that they can annihilate just into fermions and with very suppressed rates ( of course coannihilation effects make a compensation in the effective thermally averaged annihilation cross section setting the relic abundance in the early universe , but not in the zero temperature cross section for annihilation for lightest state pairs , which is the relevant quantity here ) . at higher energies
we consider the constraint from the recent measurement by the h.e.s.s .
telescope , which has the largest statistics among recently reported results @xcite . since the the flux detected by h.e.s.s .
is surprisingly hard , the data point setting the strongest @xmath56 bound is the one corresponding to the _ smallest _ energy bin , @xmath60 gev . the predicted fluxes at this energy are shown in fig .
[ curr_egret ] , right panel ; in this case they are averaged over a 5.8@xmath61 cone around the gc , corresponding to the h.e.s.s .
angular resolution for this source . obviously the fluxes are zero for neutralino masses below the energy we are considering ; above that mass , the @xmath59 scaling in the number density of neutralino pairs is compensated ( except for the pure wino branch in the heavy @xmath10 case ) by the fact that we are considering a high energy bin , in which the pile up of the produced photons get efficient just for heavy masses , and again by the fact that the maximal pair annihilation rate involves pure winos .
note also the huge span in the predictions between the two halo models we are considering , even larger than in the case of the egret data , since the angular acceptance here is smaller .
we are ready to examine in more detail the role of dark matter searches in the split susy framework .
the approach we follow here is to choose a direction in the parameter space , to postulate that these selected models provide the lsp as a dark matter candidate ( i.e. relaxing the hypothesis that the thermal relic abundance of @xmath62s in a standard cosmological setup should match the wmap preferred range ) and to check their detectability . in this section
we concentrate , if not differently specified , on a case study where we take the soft breaking mass of the hypercharge gaugino @xmath63 gev , @xmath15 and @xmath64 gev .
this corresponds to a foliation of the parameter space along @xmath65 planes ; we consider a scan on these two parameters varying them in the range from 100 gev to 10 tev ( we remind again that no gut relation is assumed , and that @xmath8 , @xmath9 and @xmath10 are considered as independent parameters ) . this sample case is particularly illuminating , because , within a simple and readable setting of the relevant low - energy parameters in the neutralino sector , it allows to investigate the full set of neutralino compositions .
the pattern of mixing and composition for the lsp is schematically illustrated in fig .
[ sketch ] .
we give here a detailed , and , to our knowledge novel , analysis of the physics occurring when the wino and higgsino mixing is large ( violet strip in fig .
[ sketch ] ) , as well as an analysis of the wino - bino edge ( green strip , in fig .
[ sketch ] ) .
we present our results by resorting to _ visibility ratios _ ( vrs ) , i.e. ratios of expected signals to projected ( or current ) sensitivities .
a model featuring a vr@xmath66 in a given search channel will therefore be within the projected reach of ( or excluded by ) the corresponding experimental facility .
we start with the existing bounds from measurements of the gamma - ray flux towards the gc .
[ egret ] shows the exclusion curves one draws from the egret results , under the hypothesis of subdominant background ( see sec .
[ sec : rates ] ) .
the vr is defined as : @xmath67~{\rm gev}}\ , .
\label{eq : egret}\ ] ] here and in the analogous figures below , on the left we show 3d plots with values of vr on the @xmath24 plane , on the right the corresponding isolevel curves .
the cuspy adiabatically contracted n03 profile is assumed in the plots ; fluxes and vrs get a factor of @xmath68 smaller in case of the cored burkert halo .
the sharp decrease in the low masses regions , at @xmath69 gev , is due to the gauge boson threshold , _
i.e. _ when @xmath70 , in which case the gamma rays production cross section suddenly decreases .
we recall that this region is , however , already ruled out by chargino searches at lep , see fig . [
3dlabels ] , right .
the largest production of gamma rays occurs in the light @xmath9 region , and it does not depend either on @xmath8 or @xmath10 . in the region of bino like neutralino ,
the vrs sensibly decrease , along with the decrease in the higgsino fraction , and the expected fluxes are always well below the measured signal
. the egret data also constrain pure higgsinos , up to masses around 350 gev .
the vr from the h.e.s.s .
measurement is defined analogously , except that now in eq .
[ eq : egret ] the reference energy bin is the lowest in the h.e.s.s .
dataset , i.e the one at 281 gev ; this sets the threshold in vr at approximately 300 gev .
once again , the largest signal is obtained for wino like lsp s .
relaxing the @xmath63 gev assumption , we find that sufficiently pure winos with masses up to around 2 tev are excluded by h.e.s.s .
data , for the cuspy halo model we have considered ( but are perfectly viable in case of the cored profile , since fluxes are suppressed , in that case , by more than six orders of magnitude ) .
vrs for higgsinos are smaller , but still many configurations are excluded , while again binos give very small fluxes . turning to spin - independent direct detection , we take here , as a benchmark of next generation search experiments , the future reach of the xenon-1ton facility @xcite .
the corresponding projected sensitivity curve in the @xmath71 is shown in the left frame of fig .
[ curr_dirdet ] ; the vr is defined here as the ratio between the expected cross section and the projected sensitivity , at the corresponding neutralino mass . fig .
[ dirdet1 ] highlights again the role of gaugino - higgsino mixing in direct detections searches : @xmath72 , with @xmath73 and @xmath74 , respectively , the lightest neutralino higgsino and gaugino fractions ; their product is maximized on the ridge along the @xmath75 line ( explaining the shape of the vr=100 isolevel curve ) and has a line of local maxima around @xmath76 gev .
switching from a bino like to a wino like lightest neutralino , _
e.g. _ on lines at constant @xmath77 gev , the scattering cross section is enhanced by one order of magnitude or even more .
this can be retraced to the sudden increase in the higgsino fraction when the neutralino becomes wino like we have shown in the right panel of fig .
[ higgsinofrac ] , and slightly depends on @xmath11 .
the region at @xmath78 gev features a 500 gev bino like neutralino , whose scattering cross section off a proton keeps decreasing as the higgsino fraction , with increasing @xmath10 .
the isolevel curves indicate that , in this case study , neutralinos as heavy as 700 - 800 gev will be within reach of future detection facilities ; the critical parameter is found to be the value of @xmath10 , while the @xmath9 dependence is quite mild .
we show in fig .
[ dirdet2 ] the vr for a larger value of @xmath79 tev , and for both signs of @xmath10 . in this case , the reach of direct searches is clearly correlated to the degree of wino - higgsino mixing . strikingly enough , we find that , for both signs of @xmath10 , neutralinos as heavy as 5 to 6 tev will give a detectable signal at next generation direct detection searches , provided @xmath80 .
we emphasize that this situation is not theoretically unrealistic , as it is naturally realized along the hyperbolic - branch / focus - point ( hb / fp ) of msugra ( see ref .
* , where the phenomenology of bino - higgsino neutralino dark matter has also been extensively addressed ) ; these is the slice in the parameter space close to the region where there is no viable electroweak symmetry breaking ) of models featuring a wino like lsp , as in the mamsb model @xcite .
interestingly enough , we find cancellations in @xmath81 for @xmath82 ( see fig
. [ dirdet2 ] , right ) .
while in the past it was realized that interferences of heavy and light higgs bosons exchanges could induce low values for the scattering cross sections , we point out here that what we find is of a different nature . in split
susy , in fact , the heavy higgs contribution vanishes , and the @xmath81 is driven to low values by the neutralino - neutralino - higgs coupling itself .
the latter is in fact @xmath83 , hence a suitable interplay of @xmath9 and @xmath10 may entail , for a given value of @xmath11 , @xmath84 . apart from this novel feature
, we remark that the sign of @xmath10 does not affect much the vr in this channel .
the same conclusion holds for all other detection channels : this is why we always restrict to the case @xmath85 . to examine the prospects for indirect detection with neutrino telescopes , we consider as a projected future sensitivity the icecube expected limit on the muon flux from the sun @xcite , _
i.e. _ the green curve of fig . [ curr_dirdet ] , right .
the intensity of the signal we find highlights again the role of wino - higgsino mixing : the line at @xmath75 produces in fact the largest muon fluxes , as it combines large capture rates in the sun , due to large spin - dependent neutralino - matter couplings , and large annihilation rates . as the lsp turns into a bino , the capture - annihilation equilibrium ceases , and the vr suddenly collapses . in the wino like regions , at @xmath86 , we observe the role of the higgsino fraction at fixed neutralino mass ( _ i.e. _ along a given constant @xmath9 ) , an effect which is fully accounted for by the decrease in the capture rate of neutralinos inside the sun .
the same effect is visible , though less evident , in the higgsino like region .
turning to antiprotons and positrons searches , we address the perspectives of discrimination of the signal from neutralino pair annihilations against the background from the interaction of primary cosmic ray with the interstellar medium , following the approach outlined in ref .
we define the quantity : @xmath87 ^ 2 } { \phi_b(e)}{\rm d}e\,,\ ] ] where @xmath88 and @xmath89 refer , respectively , to the signal and the background flux , and the integral extends over the energy range in which the integrand is non - negligible .
this quantity stems from the continuum limit , up to an overall factor accounting for the exposure times effective area of a given future experiment , of a @xmath90 statistical variable defined assuming that the background is known and the signal is subdominant with respect to the background .
antiproton and positron backgrounds are obtained with the galprop @xcite propagation code , in a scheme with diffusion and convection tuned to give a fair estimate of ratios of primary to secondary cosmic ray nuclei .
signals are obtained within the same propagation framework .
vrs are defined as ratios between @xmath91 and the critical value in this quantity as estimated for the pamela experiment @xcite after three years of data taking , i.e. @xmath92 , see ref .
@xcite for details . in figs .
[ fig : pbar ] and [ fig : eplus ] we plot visibility ratios in case of the adiabatically contracted n03 halo model ; contrary to gamma rays , since fluxes are dominated by nearby sources , the suppression for the cored profile is not dramatic and discrimination pattern are comparable as we will discuss in sec .
[ sec : discussion ] . not surprisingly , we find a large dependence on the neutralino mass , and on the higgsino fraction in the bino like neutralino region at @xmath93 .
while antiprotons fluxes will be sensitive to winos and higgsinos up to masses as large as 500 gev , the sensitivity in the positrons fluxes is somewhat less stringent , extending up to higgsinos as heavy as 400 gev .
the fact that annihilating neutralinos give larger signal to background ratios for antiprotons than for positrons , has been already pointed out @xcite , and we give here a further confirmation to this point .
the case for antideuterons is different because here , restricting to a low energy window , the background flux is expected to be negligible @xcite , and even detection of 1 event would imply discovery of an exotic component . regarding the detection prospects
, we will consider , as the ultimate reach for an experiment in the future , that of the gaseous antiparticle spectrometer ( gaps ) @xcite .
this is a proposal for an instrument looking for antideuterons in the energy interval 0.1 - 0.4 gev per nucleon , with estimated sensitivity level of @xmath94 , to be placed either on a satellite orbiting around the earth or on a probe to be sent into deep space .
we compute the antideuteron flux induced by neutralino annihilations in this energy bin , and define the vr as the ratio between this and the mentioned expected sensitivity of gaps @xcite .
results for vrs in fig .
[ antideuteron ] show that a critical threshold in the signal is that of top quarks ; once again , the signal is maximized along the largest wino - higgsino mixing line .
the bino - higgsino mixing effect is also clearly distinguishable in the line of maxima in vr around @xmath95 ; the higgsino fraction is found to enhance the antideuteron production ( region @xmath96 ) . as discussed in @xcite , binos are somewhat less disfavored in antideuterons rather than in antiprotons or positrons searches , due to the hadronization of bino s dominant decay products , such as @xmath97 .
the critical detection line ( vr=1 ) extends here well beyond that of future antiprotons and positrons searches , as envisaged in ref .
@xcite , entering the region of bino like neutralinos with masses as large as 700 gev .
the last process we consider is the production of neutralinos and charginos at the cern large hadron collider ; we computed , with the prospino2.0 package @xcite , the total production cross section of charginos and neutralinos at a center of mass energy of 14 tev , _
i.e. _ : @xmath98 a thorough analysis of the expected lhc background and of the suitable cuts to optimize searches for light neutralinos and charginos goes beyond the scope of the present work .
we outline , however , that , in two extreme cases , dedicated studies have faced this very same problem : pure winos in mamsb with large common scalar masses @xmath99 , as discussed in @xcite and more recently in @xcite ; and higgsinos in the hb / fp region of msugra , which have been addressed in @xcite .
we find that in both cases the critical total production cross section tentatively amounts to values around 500fb , which we reproduce for illustrative purposes in the iso - level curves . with the aim of comparing different future neutralino dark matter detection strategies in the case study under scrutiny
, we collect in fig .
[ summary1 ] and [ summary2 ] the iso - level curves corresponding to the critical vr=1 lines .
we also include the bound stemming from the primordial production of @xmath0li @xcite induced by the residual pair annihilations of neutralinos , after freeze - out and during the period of synthesis of light elements ( the production of of @xmath0li in the standard bing bang nucleosynthesis scenario is very strongly suppressed ) .
the limit we implemented is extrapolated from ref .
@xcite for the cases of interest here ( namely those of neutralino annihilations with gauge bosons final states ) ; we shade the corresponding excluded regions in fig .
[ summary1 ] , where we show also the reach contours of direct spin - independent detection and of the muon - flux from the sun .
[ summary2 ] shows our results concerning halo - dependent quantities : results for the cuspy adiabatically contracted n03 profile and for the burkert cored profile are respectively given in the left and right panels .
+ interestingly enough , we find that the region excluded by the @xmath0li bound rules out a wide portion of parameter space where direct searches are not effective , namely that of pure winos .
we also remark that the whole region covered by the detection of a neutralino induced muon flux from the sun is completely contained in that of direct searches : a positive sign from icecube would imply , within this framework , a visible signal at next generation direct spin - independent facilities .
+ fig .
[ summary2 ] shows the excluded regions from current data on antimatter fluxes @xcite : remarkably , in both halo models a large portion of pure winos and higgsinos is already ruled out by available data .
constraints from gamma rays are effective only in the case of a cuspy profile , and exclude an even larger portion of models .
future antimatter searches will cover essentially the whole parameter space in the large @xmath10 region ; in the higgsino - bino mixing region , on the other hand , a clear hierarchy among antideuterons , antiprotons and positrons is visible ; with a cuspy profile , the region will be in all cases thoroughly probed by the pamela experiment . + turning to a conservative cored profile ,
we find again a large antideuterons reach along the higgsino - bino large mixing region ; in the pure wino region , instead , the lion s share is given by antiprotons . as a result
, the potentially undetectable regions shrink to very narrow corners of massive pure winos or higgsinos .
we remark , comparing fig .
[ summary1 ] and [ summary2 ] , a noticeable complementarity among neutralino dark matter searches : within this scenario , in fact , the higgsino like part of the parameter space would be thoroughly searched for by direct detection experiments , while the pure wino region , already constrained by current data on antimatter fluxes , will be largely accessible to next generation space based antimatter search facilities .
the bottom line is therefore that , except for a thin slice at the edge of the wino - bino transition line , _ in the case study we considered here , the whole parameter space will be accessible to future dark matter detection experiments _ ; this result holds quite independently of the halo model under consideration : a remarkable _ complementarity among direct searches and antimatter searches _ ensures in fact most corners of the parameter space to be within future reach .
a naive estimate of the lhc reach , instead , indicates that _ these models mostly lie beyond the reach of future super - colliders_.
in order to summarize the neutralino detection prospects at future dark matter search facilities , we now go back to the parameter space slices of split susy introduced in sec . [ sec : paramspace ] .
we will show on those planes our results concerning current and future reaches .
we recall that relic abundances inside the wmap range correspond to the blue contours , and that the regions shaded in red are ruled out by the lep2 searches for charginos @xcite .
we shade in yellow the region which is not consistent with the @xmath0li abundance . in order to be as conservative as possible
, we resort here to the cored burkert profile : 95% c.l .
exclusions limits from current positrons and antiprotons data are also indicated in the figures .
+ fig .
[ m1summary ] details on the @xmath24 plane at large @xmath100 tev . as a first remark , we point out once again the role of wino - higgsino mixing , which dramatically enters in direct detection searches : in particular , when the mixing is maximal , masses as large as 10 tev will be probed at future planned experiments !
antideuterons fluxes , as well as the flux of muons from the sun , are also sensitive to the degree of mixing between the wino and higgsino components in the lightest neutralino : in this setup , icecube may detect a signal for neutralino masses as heavy as 1 tev . as regards the case of pure neutralinos , either wino or higgsino like , we remark that direct detection techniques are still effective when @xmath101 1 tev ; on the other hand , if the mixing is tiny , then only antimatter searches are supposed to be effective .
we notice , moreover , an interesting and non - trivial complementarity between antiprotons and antideuterons searches : while the first will be more effective in the pure wino region , the latter are expected to do better in the pure higgsino case . in general , however , we find that the reach of antimatter searches is limited to around 500 gev in the wino like case and to 250 gev in the higgsino like one .
let us now turn to the @xmath18 and @xmath17 planes , fig .
[ m2musummary ] . in the well - known bino - higgsino plane
, a large region of sufficiently pure higgsinos turns out to be excluded by current antiprotons data .
as expected , direct detection will probe the region at large mixing , _
i.e. _ when @xmath102 , even at very large masses .
as the higgsino fraction drops ( upper part of the figure ) , all detection methods are not effective , in this plane .
the case of pure higgsinos reflects what we pointed out above : the region may be best searched for at antimatter detection experiments , up to masses around 250 gev .
the @xmath18 plane may be regarded , to some extent , as a low - energy blow - up of the so called _ hyperbolic branch / focus point region _
( hb / fp ) of minimal supergravity @xcite . in this respect
, we can directly compare the results we show in the left panel of fig .
[ m2musummary ] with those recent studies addressing the detection of neutralino dark matter in the hb / fp region @xcite .
we remark that , as far as those quantities that weakly depend on assumptions on the halo models are concerned , we mostly agree with those works . in particular
, we confirm the importance of the role of neutrino telescopes in the hb / fp region , already pointed out in baer et al .
, in ref .
@xcite : the timeline of the future experimental facilities might give the opportunity to icecube to provide the first indirect evidences for neutralino dark matter in this region of the msugra parameter space ; the would - be signal might later be confirmed by stage-3 direct detection experiments , whose sensitivity extends well beyond that of the future antartic neutrino telescope . in the present analysis we extended previous studies of the bino - higgsino mixing region of msugra with a detailed and thorough investigation of indirect neutralino detection through antimatter searches . with a conservative dark halo profile
, we find that positrons and antiprotons searches will not be able to probe models , in the hb / fp region , featuring a thermal neutralino relic abundance within the wmap range .
nevertheless , we point out , as a novel and somewhat exciting result , that antideuterons searches on a gaps - like experiment could be extremely comptetitive in the hb / fp region , with a reach comparable to that of neutrino telescopes . the @xmath17 plane , which reproduces a conceivable scenario in which the gaugino masses are light and the @xmath10 parameter is large , is essentially split in a pure wino and pure bino region , respectively below and above the diagonal , dashed line .
interestingly enough , in this plane direct detection and muon fluxes feature extremely low rates , well below future projected sensitivities .
the bino like region , in which most of the points producing a thermal relic abundance in the wmap range lie , is completely off - limits for any detection technique .
the wino like region , which is already constrained by current data up to @xmath103 masses as large as 300 gev , is going to be partly covered by antimatter searches , particularly in the antiprotons channel .
so far we have assumed that , at each point in the parameter space , the lightest neutralino has a relic density matching the cdm component , regardless of its thermal relic density in the standard cosmological framework
. we also mentioned that relaxing the overproduction bound is possible but contrived , while there is a plethora of possibilities to enhance a low thermal relic abundance into the cdm range . in a more conservative approach one can assume that only a fraction of the cdm is accounted for by supersymmetric models whose thermal relic abundance lies below the wmap preferred range ( @xmath104 ) . under this assumption ,
the simplest possible way to compute the detection rates is to rescale the neutralino dark matter density distribution according to the prescription @xmath105 direct detection and capture rates in the sun linearly scale with the local neutralino density @xmath106 ; on the other hand , gamma - ray and antimatter rates involve the squared of the neutralino density distributions .
we find that the only technique which could give a signal , in this approach , and in the parameter space we analyze here , is the spin - independent direct detection . in fig .
[ summaryrescale ] we present the _
rescaled _ direct detection reach in the @xmath24 plane , at @xmath100 tev .
accelerator searches would naturally not be involved in the above outlined rescaling : we show in the figure the iso-500 fb total production cross section at the lhc , and the kinematic reach of a future next linear collider ( nlc ) with a center - of - mass energy of 1 tev ( _ i.e. _ the iso-@xmath107 gev line ) .
+ for completeness , we also mention the detection prospects connected to another well - known feature of the split susy scenario , i.e. the fact that gluinos have very long lifetimes .
the phenomenological implications of a quasi - stable gluino have been recently studied in ref .
@xcite ; they mildly depend on the details of the spectrum , one of the crucial quantities being the gluino mass .
the maximal lhc reach is around @xmath108.3 tev @xcite . in the general setup we have considered
, the gluino mass is a free parameter . in order to give a flavor of what could be the lhc reach in the long - lived gluino detection channel
, one has to resort to a particular framework to relate the gaugino masses .
for instance , in the @xmath24 plane it is worthwhile inspecting the consequences of assuming an anomaly - mediated - inspired relation , where @xmath109 , the spread being given by rg effects on the @xmath110 and @xmath111 coupling which depend on the value of the gravitino mass parameter @xmath112 @xcite . in fig .
[ summaryrescale ] we plot , with a solid violet line , the putative reach of the lhc in the long - lived gluino detection channel assuming an anomaly mediated relation between @xmath9 and @xmath113 , which corresponds to @xmath114 gev .
the latter value is obtained solving for the equation : @xmath115^{\rm amsb}\cdot m^{0}_{\tilde g } , \quad m^{0}_{\tilde g}=2.3\ { \rm tev}.\ ] ] in summary , fig .
[ summaryrescale ] shows that in case the rescaling prescription is applied , models which could be discovered with the direct detection technique have non - negligible wino - higgsino mixing ; since the relic abundance of winos is smaller than that of higgsinos , at a given mass , we also notice that in this scheme the higgsino like region is somewhat favored
. the pure wino and pure higgsino regions could only be explored at accelerators : the complementarity between accelerator searches and the quest for dark matter therefore emerges even in the present setup of pure electroweak production of supersymmetric particles at colliders . as a concluding remark
, we want to stress that the results we presented here hold for more general supersymmetric models , where the scalar sector need not necessarily be largely `` _ split _ '' .
we can conservatively state that postulating a scalar sector lying even only one order of magnitude above the lsp mass scale would leave most of our results essentially unaffected .
even lighter scalars may , however , give rise to some possible caveats , particularly for the neutralino relic density , for instance as far as sfermion coannihilations , resonant annihilations with the heavy higgses , and @xmath19- and @xmath20-channels sfermions exchange diagrams are concerned .
we studied in full generality the neutralino dark matter phenomenology of models in which the scalar sector is heavy , _
i.e. _ in split supersymmetry scenarios .
the relevant parameters are the entries of the neutralino mass matrix , namely the soft breaking gaugino mass parameters @xmath8 and @xmath9 and the higgs mixing mass term @xmath10 . requiring that the thermal relic abundance of neutralinos does not exceed the observed amount of cold dark matter defines an hypersurface in the three - dimensional space of parameters :
the cosmological bound is indeed the only issue forcing one of the sectors of the theory to be light , _
i.e. _ at a scale lower than at least 2.2 tev ; the bound can be violated only invoking mechanisms for entropy production at low energy , a rather contrived setup .
models defined by parameters lying below the hypersurface are viable , and suitable mechanisms of relic density enhancement may drive low relic density models to produce the required amount of cold dark matter . we studied in detail the physics of wino - higgsino mixing , and pointed out that future experiments may be able to probe extremely large neutralino masses when the mixing is maximal : 10 tev and 1 tev neutralinos may give detectable signals respectively in next generation spin - independent searches and at icecube .
interestingly enough , we find that even though only one feynman diagram contributes to neutralino - proton scattering , cancellations among the various neutralino interaction - eigenstates components may conspire and suppress the relevant couplings . resorting to two extreme cases for the dark matter distribution in our galaxy , we showed that large parameter space portions are already ruled out by currently available data on antiprotons and positrons fluxes , as well as by the primordial @xmath116li abundance induced by neutralino residual annihilations . depending on the structure of the dark halo in the gc , measurements of the gamma - ray in the galactic center direction from the egret and the h.e.s.s .
telescopes have also been shown to set strong constraints in the parameter space .
future prospects for antimatter searches look promising , particularly in the pure wino and pure higgsino cases ; a remarkable complementarity between antideuterons and antiprotons detection has also been outlined .
we worked out an explicit analysis of a case study , in which we fixed the bino mass term @xmath1170.5 tev .
we showed that while most of the resulting parameter space slice will not be within the reach of the lhc , an interplay among direct detection and antimatter searches will allow future dark matter detection experiments to _ thoroughly explore _ the split supersymmetry parameter space . restricting to a standard cosmological scenario and taking into account for thermal components only
, a rescaling procedure should be implemented for subdominant supersymmetric dark matter candidates , suppressing detection rates and leaving a chance only for direct detection in the large wino - higgsino mixing region . in this scenario ,
the complementarity among collider searches and direct detection experiments explicitly emerges , although here superpartners can be produced only through electroweak processes . as a last remark
, we point out that a lighter scalar sector would in general yield , modulo cancellations , larger dark matter detection signals : in this respect our results may be regarded as _ lower bounds _ to neutralino searches in more general supersymmetric setups .
the work of a.m. and p.u . was supported by the italian infn under the project `` fisica astroparticellare '' and the miur prin `` fisica astroparticellare '' .
the work of s.p . was supported in part by the u.s .
department of energy under contract number de - fg02 - 97er41022 .
s.p . would like to thank h. baer , k. matchev and a. birkedal for useful discussions and remarks .
+ h. goldberg , phys .
lett . * 50 * ( 1983 ) 1419 ; j. r. ellis , j. s. hagelin , d. v. nanopoulos , k. a. olive and m. srednicki , nucl .
b * 238 * ( 1984 ) 453 ; l. e. ibanez , phys .
b * 137 * ( 1984 ) 160 .
d. n. spergel _ et al . _
[ wmap collaboration ] , astrophys .
j. suppl .
* 148 * ( 2003 ) 175 [ arxiv : astro - ph/0302209 ] .
n. arkani - hamed and s. dimopoulos , arxiv : hep - th/0405159 . g. f. giudice and a. romanino , nucl . phys .
b * 699 * ( 2004 ) 65 [ arxiv : hep - ph/0406088 ] . n. arkani - hamed , s. dimopoulos , g. f. giudice and a. romanino , arxiv : hep - ph/0409232 . i. antoniadis and s. dimopoulos , arxiv : hep - th/0411032 . b. kors and p. nath , arxiv : hep - th/0411201 . c. kokorelis , arxiv : hep - th/0406258 .
kumar and j. d. lykken , jhep * 0407 * ( 2004 ) 001 [ arxiv : hep - ph/0401140 ] . j. l. hewett , b. lillie , m. masip and t. g. rizzo , jhep * 0409 * , 070 ( 2004 ) [ arxiv : hep - ph/0408248 ] .
d. pierce and a. papadopoulos , phys .
d * 50 * ( 1994 ) 565 [ arxiv : hep - ph/9312248 ] ; d. pierce and a. papadopoulos , nucl .
b * 430 * ( 1994 ) 278 [ arxiv : hep - ph/9403240 ] ; a. b. lahanas , k. tamvakis and n. d. tracas , phys .
b * 324 * ( 1994 ) 387 [ arxiv : hep - ph/9312251 ] ; m. drees , m. m. nojiri , d. p. roy and y. yamada , phys .
d * 56 * ( 1997 ) 276 [ erratum - ibid .
d * 64 * ( 2001 ) 039901 ] [ arxiv : hep - ph/9701219 ] ; p. binetruy , g. girardi and p. salati , nucl .
b * 237 * ( 1984 ) 285 ; k. griest and d. seckel , phys .
d * 43 * ( 1991 ) 3191 ; j. edsjo and p. gondolo , phys .
d * 56 * ( 1997 ) 1879 [ arxiv : hep - ph/9704361 ] .
a. pierce , phys .
d * 70 * ( 2004 ) 075006 [ arxiv : hep - ph/0406144 ] ; b. murakami and j. d. wells , phys .
d * 64 * , 015001 ( 2001 ) ; t. moroi and l. randall , nucl .
b * 570 * , 455 ( 2000 ) ; m. fujii and k. hamaguchi , phys .
b * 525 * , 143 ( 2002 ) ; m. fujii and k. hamaguchi , phys .
d * 66 * , 083501 ( 2002 ) ; r. jeannerot , x. zhang and r. h. brandenberger , jhep * 9912 * , 003 ( 1999 ) ; w. b. lin , d. h. huang , x. zhang and r. h. brandenberger , phys .
lett . * 86 * , 954 ( 2001 ) .
p. salati , [ arxiv : astro - ph/0207396 ] ; f. rosati , phys .
b * 570 * ( 2003 ) 5 [ arxiv : hep - ph/0302159 ] .
s. profumo and p. ullio , jcap * 0311 * , 006 ( 2003 ) [ arxiv : hep - ph/0309220 ] .
s. profumo and p. ullio , arxiv : astro - ph/0404390 . m. kamionkowski and m. s. turner , phys .
d * 42 * ( 1990 ) 3310 .
r. catena , n. fornengo , a. masiero , m. pietroni and f. rosati , arxiv : astro - ph/0403614 .
n. fornengo , a. riotto and s. scopel , phys .
d * 67 * ( 2003 ) 023514 [ arxiv : hep - ph/0208072 ] .
m. kamionkowski , k. griest , g. jungman and b. sadoulet , phys .
lett . * 74 * ( 1995 ) 5174 [ arxiv : hep - ph/9412213 ] .
l. bergstrom , j. edsjo and p. gondolo , phys .
d * 58 * ( 1998 ) 103519 [ arxiv : hep - ph/9806293 ] .
j. l. feng , k. t. matchev and f. wilczek , phys .
d * 63 * ( 2001 ) 045024 [ arxiv : astro - ph/0008115 ] ; j. r. ellis , j. l. feng , a. ferstl , k. t. matchev and k. a. olive , eur .
j. c * 24 * ( 2002 ) 311 [ arxiv : astro - ph/0110225 ] ; j. r. ellis , k. a. olive and y. santoso , new j. phys .
* 4 * ( 2002 ) 32 [ arxiv : hep - ph/0202110 ] ; r. dermisek , s. raby , l. roszkowski and r. ruiz de austri , jhep * 0304 * ( 2003 ) 037 [ arxiv : hep - ph/0304101 ] ; a. bottino , f. donato , n. fornengo and s. scopel , phys . rev .
d * 70 * ( 2004 ) 015005 [ arxiv : hep - ph/0401186 ] ; d. hooper and j. silk , arxiv : hep - ph/0409104 . j. l. feng , k. t. matchev and f. wilczek , phys .
b * 482 * ( 2000 ) 388 [ arxiv : hep - ph/0004043 ] ; u. chattopadhyay , a. corsetti and p. nath , phys . rev .
d * 68 * ( 2003 ) 035005 [ arxiv : hep - ph/0303201 ] ; h. baer , a. belyaev , t. krupovnickas and j. ofarrill , jcap * 0408 * ( 2004 ) 005 [ arxiv : hep - ph/0405210 ] ; d. s. akerib _ et al . _
[ cdms collaboration ] , arxiv : astro - ph/0405033 .
e. aprile _ et al .
_ , arxiv : astro - ph/0207670 .
r. bernabai et al . ,
* b389 * ( 1996 ) 35 ; m. altmann et al .
, arxiv : astro - ph/0106314 ; see also http://dmtools.berkeley.edu/limitplots/ , by r. gaitskell and v. mandic .
f. donato , n. fornengo and p. salati , phys . rev .
* d62 * ( 2000 ) 043003 ; f. donato , n. fornengo , d. maurin , p. salati and r. taillet , phys .
* d69 * ( 2004 ) 063501 .
a. el - zant , i. shlosman and y. hoffman , astrophys .
j. * 560 * ( 2001 ) 336 .
s. profumo and p. ullio , jcap * 0407 * ( 2004 ) 006 [ arxiv : hep - ph/0406018 ] .
p. ullio , proceedings of the third international conference on frontier science `` physics and astrophysics in space '' , villa mondragone ( rome ) , italy , published on frascati physics series .
l. randall and r. sundrum , nucl .
b * 557 * ( 1999 ) 79 [ arxiv : hep - th/9810155 ] ; g. f. giudice , m. a. luty , h. murayama and r. rattazzi , jhep * 9812 * ( 1998 ) 027 [ arxiv : hep - ph/9810442 ] ; t. gherghetta , g. f. giudice and j. d. wells , nucl .
b * 559 * ( 1999 ) 27 [ arxiv : hep - ph/9904378 ] ; j. l. feng and t. moroi , phys .
d * 61 * ( 2000 ) 095004 [ arxiv : hep - ph/9907319 ] .
galprop numerical package , http://www.mpe.mpg.de/ aws / propagate.html k. mori , c. j. hailey , e. a. baltz , w. w. craig , m. kamionkowski , w. t. serber and p. ullio , astrophys . j. * 566 * , 604 ( 2002 ) [ arxiv : astro - ph/0109463 ] . w. beenakker , r. hopker and m. spira , arxiv : hep - ph/9611232 .
h. baer , j. k. mizukoshi and x. tata , phys .
b * 488 * ( 2000 ) 367 [ arxiv : hep - ph/0007073 ] .
a. j. barr , c. g. lester , m. a. parker , b. c. allanach and p. richardson , jhep * 0303 * , 045 ( 2003 ) [ arxiv : hep - ph/0208214 ] .
h. baer , t. krupovnickas and x. tata , jhep * 0307 * , 020 ( 2003 ) [ arxiv : hep - ph/0305325 ] .
h. baer , c. balazs , a. belyaev , t. krupovnickas and x. tata , jhep * 0306 * , 054 ( 2003 ) [ arxiv : hep - ph/0304303 ] .
k. jedamzik , phys .
d. * 70 * ( 2004 ) 063524 ; phys .
d. * 70 * ( 2004 ) 083510 . | we study the phenomenology of neutralino dark matter within generic supersymmetric scenarios where the gaugino and higgsino masses are much lighter than the scalar soft breaking masses ( split supersymmetry ) .
we consider a low - energy model - independent approach and show that the guidelines in the definition of this general framework come from cosmology , which forces the lightest neutralino to have a mass smaller than 2.2 tev .
the testability of the framework is addressed by discussing all viable dark matter detection techniques .
current data on cosmic rays antimatter , gamma - rays and on the abundance of primordial @xmath0li already set significant constraints on the parameter space .
complementarity among future direct detection experiments , indirect searches for antimatter and with neutrino telescopes , and tests of the theory at future accelerators , such as the lhc and a nlc , is highlighted .
in particular , we study in detail the regimes of wino - higgsino mixing and bino - wino transition , which have been most often neglected in the past .
we emphasize that our analysis may apply to more general supersymmetric models where scalar exchanges do not provide the dominant contribution to annihilation rates .
sissa 95/2004/ep + fsu
hep041122 + * a. masiero@xmath1 , s. profumo@xmath2 and p. ullio@xmath3 * + _ @xmath1 dipartimento di fisica ` g. galilei ' , universit di padova , and istituto nazionale di fisica nucleare , sezione di padova , via marzolo 8 , i-35131 , padova , italy _ + _ @xmath4 department of physics , florida state university + 505 keen building , fl 32306 - 4350 , u.s.a . _ + _ @xmath3 scuola internazionale superiore di studi avanzati , via beirut 2 - 4 , i-34014 trieste , italy and istituto nazionale di fisica nucleare , sezione di trieste , i-34014 trieste , italy _
+ _ e - mail : _
[email protected] , [email protected] , [email protected] |
quasar ( qso ) absorption line systems provide a unique means to study the intergalactic medium , as well as the interstellar medium ( ism ) of galaxies at all redshifts .
resonance absorption lines from metallic ions in the ism of distant galaxies along the line of sight to a qso provide a powerful method for the determination of chemical abundances at high redshifts , and also reveal the properties of low - density gas that is difficult or even impossible to detect with any other method , even in the nearby universe .
as this method is independent of the luminosity of the intervening galaxy , it is a relatively unbiased tracer of the chemical evolution of the universe as a whole .
the h i @xmath19 1215 absorption line in damped lyman-@xmath0 systems ( dlas , log n(h i)@xmath2020.3 ) and sub - dlas ( 19.0@xmath21log n(h i)@xmath2120.3 ) can be fit for accurate column density measurements due to the extended damping wings in the profile .
dlas and sub - dla systems have been shown to contain the majority of the neutral gas in the universe @xcite .
the dlas are of particular importance for galactic chemical evolution studies . with such high column densities ,
the absorption lines from metal atoms are easily visible .
also , systems with high column densities are expected to remain mainly neutral due to self shielding , alleviating the need for uncertain ionization corrections to the abundances .
there is still much debate as to the nature of the galaxies hosting dlas and sub - dlas . even at low redshift ,
finding a faint galaxy near the bright point source of the qso is challenging . at higher redshift ,
the cosmological dimming of surface brightness ( @xmath22 ) makes galaxy detections even more difficult . nonetheless , these systems are a unique laboratory to study the ism of galaxies over a wide range of redshifts . at @xmath23 ,
the lyman-@xmath0 line shifts above the atmospheric cutoff at @xmath24 , and ground based surveys can efficiently search for dla systems .
several such surveys have determined the @xmath25 neutral gas density @xmath1@xcite . at lower redshifts ,
space - based spectra are necessary to measure the neutral hydrogen column densities in the uv . as dlas ( and qsos
that are bright enough in the far ultraviolet ( fuv ) to be accessible to previous uv spectrographs ) are relatively rare , the number of such systems currently known at @xmath26 is small compared to available samples at high redshift .
the cosmic origins spectrograph ( cos ) is a new instrument package @xcite installed on the hubble space telescope that enables us to study these absorbers at low redshift with unprecedented efficiency . as @xmath27 spans @xmath28 40 percent of the age of the universe , the low redshift absorbers accessible with cos are crucial for understanding cosmic chemical evolution and the cosmological gas mass density and for linking their properties to their higher redshift counterparts . in this paper
, we report on the first dla systems observed with cos .
we present a detailed analysis of one such system , a dla in the line of sight to the qso sdss j1009 + 0713 , to illustrate the scientific potential of the observations .
the structure of the paper is as follows . in @xmath292
we describe in general the hst program and data reduction methods we have used for the cos spectra . in @xmath29 3 we describe the observations of the field of sdss j1009 + 0713 , including the ground based keck / hires spectra of the qso , the cos uv spectra , and imaging of the field with the wide field camera 3 ( wfc3 ) . in @xmath29 4
we derive chemical abundances for this system , discuss the physical state of the gas as determined by the ci and c ii * lines , and discuss the properties of the galaxies in the field as seen in the wfc3 images . in @xmath29 5
we combine the measurements of n(h i ) from other absorbers in this program to measure the cosmological mass density of neutral gas , @xmath1 , at @xmath3 in a blind survey with @xmath30 .
conclusions are summarized in @xmath29 6 . throughout this paper ,
we adopt a cosmological model with @xmath31 , @xmath32 , and h@xmath33=70 km s@xmath17 mpc@xmath17 .
the targets presented here were observed as part of program go11598 , a program focused on studying multiphase baryons in the halos of @xmath34 galaxies at 15 - 150 kpc impact parameters .
target qsos for this program were selected based on sufficient fuv flux and the presence of a galaxy seen in the sloan digital sky survey ( sdss , @xcite ) with impact parameter @xmath35 kpc and a spectroscopic or photometric redshift @xmath36 .
sightlines with strong mg ii systems seen in the sdss spectra were avoided due to the possibility of a lyman - limit system ( lls ) at @xmath370.35 which would block the fuv flux . in total ,
39 sightlines will ultimately be observed in this program .
followup ground - based spectra with the low resolution imaging spectrometer ( lris ) on keck will provide for redshifts , star formation rates and metallicities of the target galaxies as well as galaxies in the field .
these cos data were processed with calcos version 2.12 during retrieval from the archive . the individual reduced and wavelength calibrated x1d files were then coadded using custom software . in short , the gross source and background counts in each wavelength bin are summed after the bins are aligned in wavelength space by using a cross correlation analysis to allow for small shifts due to inaccuracies in the wavelength calibration provided by calcos .
simultaneously , an effective integration time at each wavelength bin is created which the total counts are ultimately divided by , returning the source count rate in photon s@xmath17 . during coaddition ,
the spectra are background subtracted and also corrected for the fixed pattern noise caused by grid wires above the micro - channel plate detectors using a reference file provided by the cos instrument team .
cos has extremely low backgrounds , and consequently , in some cases there are very few counts in the cores of strong absorption lines .
we determine uncertainties directly from the accumulated counts , using poisson statistics ( see , e.g. , @xcite ) when the count numbers are low .
the spectra were binned by 3 pixels , as the raw cos data are oversampled with a @xmath38 pixel wide resolution element .
all subsequent measurements and analysis were performed on the binned spectra .
the binned spectra have a resolution of @xmath2815 km s@xmath17per resolution element .
the s / n of these cos spectra range from @xmath39 per resolution element .
the dla and sub - dla systems that we have discovered in this program are shown in figures [ fig : other_dlas ] and [ fig : dla ] .
the lyman alpha lines from the systems were fitted using a single component with the continuum and profile fit simultaneously .
the redshifts of the systems were determined from the metal lines , which with narrow features provide for a more accurate redshift determination . in total , we find three dla systems and four sub - dla systems in these cos data .
absorption redshifts and neutral hydrogen column densities for the systems are given in table [ tab : other_dlas ] .
errors on n@xmath40given in table [ tab : other_dlas ] have been estimated by eye and are dominated by the continuum placement uncertainty .
in order to illustrate the capabilities of cos for future observations of dla systems , we present a detailed analysis of one system observed in the sample here .
similar analyses for the complete dla and sub - dla sample will be given in a forthcoming paper .
sdss j1009 + 0713 ( @xmath0=10:09:02.0 , @xmath41=+07:13:43.8 , @xmath42 ) , is an optically bright ( m@xmath43=17.08 , m@xmath44 ) qso observed in the sdss .
this qso was chosen for the survey of galaxy halos due to the presence of an @xmath45 galaxy at an impact parameter of 60 kpc at redshift @xmath46 .
the dla described here was serendipitously discovered in the cos spectrum of the qso .
no absorption lines from this system are present in the sdss spectrum of the qso , as the mg ii @xmath472798 , 2802 lines which are typically strong enough to be seen in the sdss spectra fall at 3115 - 3125 and are below the wavelength range that the sdss spectra cover . in this section
, we present a detailed analysis of this system , including the cos and keck high resolution spectrometer ( hires ) spectra as well as followup imaging with the wide field camera 3 ( wfc3 ) .
these cos data were acquired on 29 - 30 march 2010 in three orbits as visit 13 in program go11598 .
two exposures were obtained with the g130 m grating at central wavelengths of 1291 and 1309 with exposure times of 1497 and 2191 s. similarly , two exposures were taken with the g160 m grating at central wavelengths of 1577 and 1600 with exposure times of 2002 and 2007 s. the dla line of this system was fit with a single component centered at @xmath9=0.1140 with a best solution of log n(h i)[email protected] .
the fit to the dla line is shown in figure [ fig : dla ] .
profiles with column densities 1@xmath48 above and below the best fit value are also shown in figure [ fig : dla ] .
absorption lines from multiple species are shown in the left panel of figure [ fig : spec ] .
we detect lines from n i @xmath471134 ,
1199 , 1200 , 1201 ; o i @xmath19 1302 ; si ii @xmath471190 , 1304 , 1526 ; s ii @xmath471250 , 1253 , 1259 ; fe ii @xmath471142 , 1143 , 1144 ; ni ii @xmath19 1317 . a lyman limit system ( lls ) at @xmath9=0.3556 blocks the continuum blueward of @xmath281240 ( see @xcite for an analysis of the lls ) or @xmath28 1115 @xmath49 in the rest frame of the dla .
there are only two h@xmath50 lines covered in the cos spectrum due to the presence of the lls .
the lyman band ( @xmath51 ) transitions from the ( 1 - 0 ) band which have rest frame wavelengths @xmath52 are all blocked by the lls . only the h@xmath50 @xmath471120 , 1125 lines resulting from the ( 0 - 0 ) transitions are covered in the cos data .
these lines are however blended with strong high lyman series absorption lines from the lls ; consequently no direct indicator of molecular gas exists for this dla .
the cos spectra cover the lines of c iv @xmath471548,1550 as well as the si iv @xmath471393 , 1402 transitions .
the c iv lines lie at the edge of the detector where the sensitivity is low and only the c iv @xmath19 1548 line is minimally detected at s / n@xmath283 .
the si iv @xmath19 1393 line is blended with the lyman-@xmath0absorption line from a @xmath9=0.2774 system , and there is no detection of the si iv @xmath19 1402 with s / n@xmath285 in the region .
optical spectra of sdss j1009 + 0713 were obtained with the keck - ii / hiresb spectrograph @xcite on the night of 26 march 2010 .
the s / n of the data ranges from @xmath535 per 1.8 km s@xmath17pixel at 3200 to @xmath2820 at 4500 .
these data provide coverage from 2750 - 4000 in the rest frame of the @xmath9=0.1140 dla , and hence include the lines of mg ii @xmath472796 , 2803 , mg i @xmath19 2852 , ti ii @xmath473242 , 3384 , and ca ii @xmath473934 , 3969 .
the mg i @xmath19 2852 line is partially blended with the fe ii @xmath19 2344 line from the @xmath9=0.3558 lls , as can be seen in figure [ fig : spec ] .
in addition , with higher resolution ( r=44,000 ) than the cos spectra ( r@xmath2820,000 ) , these data provide valuable constraints on the kinematics and doppler parameters of the lines that are blended at the resolution of cos .
these data were processed with the keck / hires reduction code hiresredux in idl .
the right panel of figure [ fig : spec ] shows the hires spectra of several near - uv ( nuv ) and optical absorption lines that were observed in this dla .
the profile was fit with only two components having doppler widths of @xmath54 and 12.9 km s@xmath17and centroid velocities of @xmath55 and @xmath56 km s@xmath17respectively .
the majority of the low ionization gas appears to be in the @xmath57 km s@xmath17component , as this component is strongest in all of the non - saturated absorption lines detected here .
.properties of the @xmath58 dlas and sub - dlas in this survey .
[ tab : other_dlas ] . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] we have used the model of the depletion patterns in the ism of the milky way from @xcite to analyze the depletion in this dla system . in this model a single value ( f@xmath59 ) indicates the overall level of depletion in a sightline , with larger values of f@xmath59 indicating higher levels of dust depletion .
we find that the line of sight depletion factor f@xmath60 for this dla , implying an intrinsic metallicity of [ z / h]=@xmath61 , consistent with the metallicity determined from the s ii lines .
such negative values of f@xmath59 are typically seen along sightlines with extremely low average densities in the halo of our galaxy @xcite implying that the sightline passes through the outer edge of a dwarf galaxy .
kinematically , the component structure of this dla is extremely simple ( see figure [ fig : spec ] ) .
only two components are required to fit the profiles of the absorption lines . the kinematic width @xmath16(defined as the region where 90 percent of the apparent optical depth in the line occurs )
has often been used as an indicator of the kinematics of qso absorption line systems @xcite .
we find a kinematic width for this system of @xmath16=52 km s@xmath17based on the ti ii @xmath19 3384 line .
the nucleosynthetic origins of n seem to include both primary ( production of n from newly synthesized c via the cno cycle ) and secondary production ( n produced again in the cno cycle , but from previously created c and o from previous star formation ) processes in intermediate mass stars @xcite . of the @xmath2830 determinations of n abundances in dla systems , all are at high @xmath12 where the n i lines are redshifted into wavelengths accessible with ground based telescopes @xcite .
some constraints have been placed on nitrogen in low - z sub - dlas where , like the high - z dlas , nitrogen is found to be underabundant ( e.g.,@xcite ) .
abundances from n can be readily obtained from optical spectra of h ii regions in nearby star forming galaxies of a range of luminosities .
the low luminosity galaxies from @xcite show a flat plateau in the [ n / o ] ratio with respect to metallicity , while the higher metallicity h ii regions in @xcite show a sharp increase in [ n / o ] with increasing metallicity , indicative of secondary production of n. observations of n in extremely low metallcity dlas are important for the study of n production and galactic chemical evolution in general @xcite .
it is still yet to be seen if the [ n / o ] ratio in dlas displays the same upturn as metallicity increases as most dlas are metal poor and at high @xmath12 the n i lines are often blended with the lyman-@xmath0 forest .
the dla presented here is the highest metallicity dla with both n and @xmath0 abundance measurements . in figure
[ fig : nitrogen ] we show [ n/@xmath0 ] vs [ @xmath0/h ] for h ii regions in nearby galaxies from @xcite , as well as the dla sample from @xcite and this dla .
the dla from this work is shown with the red circular point , and lies below the `` knee '' of the diagram where n production transitions from primary to secondary production .
the mean [ n/@xmath0 ] ratio from the samples of @xcite and @xcite with [ @xmath0/h]@xmath21 - 0.3 is @xmath62[n/@xmath0]@xmath63 , 3@xmath48 higher than seen in this dla of [ n / s]=@xmath64 .
@xcite note this discrepancy in the sample of dlas shown in figure [ fig : nitrogen ] and propose that the lower values seen in high @xmath12 dlas could be due to the delayed enrichment of n with respect to o as the delay of @xmath28250 myr is a greater fraction of the time available for star formation at high redshift than in local galaxies , so at high @xmath12 this effect should be more pronounced .
however , in this low-@xmath12 dla we see the same decrement in [ n/@xmath0 ] compared to the values from local h ii regions
. it would be interesting to see if other high metallcity dlas are also below the plateau , perhaps indicating a systematic offset in abundances determined via absorption and emission line diagnostics .
both c i and c ii@xmath66 are important diagnostic resonance lines in the study of the ism .
emission from the 158 @xmath67 emission line of singly ionized carbon in the ism of the milky way is the dominant coolant , and is expected to also be so in high-@xmath12 dlas as well .
this line originates from the transition between the @xmath68 and @xmath69 levels in the ground state of c@xmath70 . in the cold neutral medium ( cnm , t@xmath71k )
the excited state is populated largely by collisions with hydrogen atoms , where in the more diffuse warm neutral medium ( wnm , t@xmath72k ) both collisions with free electrons as well as photon pumping can populate the excited state @xcite . in the wnm
the electron density can be determined from observations of the c ii and c ii@xmath65 lines as @xmath73 \
\rm cm^{-3 } \label{eq : eden}\ ] ] where t@xmath74 is t/10@xmath75k @xcite . in the ism of the galaxy
, the cnm is characterized by having large depletions of the refractory elements such as fe and ti relative to the volatile elements of s and zn which do not condense onto dust grains .
as the depletions levels in this dla are modest , with [ ti / s]=@xmath76 , we assume that the sightline is passing through mainly a wnm .
depletion levels seen in the cnm of the milky way are typically ten times higher than what is observed in this dla @xcite .
the c ii@xmath65 @xmath19 1335 line was not detected in these data , with s / n@xmath2811 per resolution element in the region around the line .
we have used the procedure discussed above to determine a limiting column density of log n(c ii@xmath65)@xmath2113.7 .
we show a synthetic voigt profile with this total column density overplotted on the data in figure [ fig : ciistar ] . as was mentioned above the c ii @xmath19 1334
is strongly saturated and only a limiting column density of log n(c ii)@xmath2015.90 can be determined from this transition . however , in sightlines with low depletion such as this carbon is mildly depleted and we can use the s abundance to estimate a total c ii column density assuming [ c / s]@xmath280.0 .
the above analysis yields log n(c ii)=16.45 . using the above values in equation [ eq : eden ] with a temperature of 6000 k characteristic of the wnm the electron density is determined to be n(e)@xmath210.08 @xmath77 .
the cooling rate per h atom due to c ii ] @xmath19 158 @xmath67 emission can be estimated by @xmath78 \
\rm{erg\ s^{-1}\ ( h\ atom)^{-1 } } \label{eq : emis}\ ] ] @xcite .
using the values above , we determine log @xmath79 .
this is typical of what has been observed in higher-@xmath12 dla systems @xcite .
c i is a tracer of cold gas in the ism , and is usually seen in dla absorbers bearing h@xmath50 @xcite . as such , c i absorption lines are indicative of cold , dense material in dlas .
i @xmath19 1277 line provides the strongest constraint on the c@xmath80 column density in the cos spectrum as the s / n in the region around the stronger transition of c i @xmath19 1560 was much lower .
this line was not detected in the spectrum and we have determined a limiting column density of log n(c i)@xmath2113.6 using the prescription above . in the two phase ism model of @xcite , the ratio of n(c ii)/n(c i ) is indicative of the gas density .
the observed ratio of n(c ii)/n(c i)@xmath20700 in this system implies that n(h)@xmath2110 @xmath77 .
the sample of qso sightlines from this hst program provides an opportunity to determine the incidence of dla systems at low @xmath12 , where there is a paucity of data due to the necessity of space - based observations . ideally , a totally `` blind '' survey in which there is no prior information about the sightlines prior to observation provides for the least amount of systematic biases in the determination of cosmological gas mass density . in order to eliminate as much observational bias as possible ,
we have defined the redshift interval over which to search for dlas and sub - dlas with the following criteria . as these sightlines
were selected due the presence of a galaxy within impact parameters of @xmath35 kpc , we have excluded @xmath81 km s@xmath17around the redshift of the target galaxy in the search path to avoid biasing the sample artificially high due to the possible clustering of galaxies . at @xmath82 the mg ii @xmath472796 ,
2803 lines are redshifted into the wavelength coverage of the sdss .
this program intentionally excluded sightlines with mg ii absorbers at @xmath83 due to the possibility of a lls which would block the fuv flux , which in effect biases the sample against dla detection at @xmath83 . to avoid this bias
, we focus only on @xmath84 . in the cases where the emission redshift of the qso is less than 0.34 ,
we have excluded a 3000 km s@xmath17region blueward of the emission redshift .
the sub - dlas towards sdss j0925 + 4004 , sdss j0928 + 6025 , and sdss j1435 + 3604 have been excluded from the calculation as the absorption redshifts are within 1000 km s@xmath17of the redshift of the target galaxy , leaving 3 dlas and 1 sub - dla in the final sample .
the total path length for the 37 sightlines in this sample using the definition above is @xmath85=11.94 , giving d@xmath2/@xmath86 and d@xmath2/dz=0.08@xmath6 for dlas and sub - dlas respectively , and assuming poisson statistical errors from @xcite .
the redshift density determined here is larger than the value determined in @xcite of d@xmath2/[email protected] , however the large statistical errors of this work make this difference unlikely to be statistically significant . from a large number of archived hst spectra , @xcite find d@xmath2/[email protected] for all absorbers with n(h i)@xmath8717.0 .
the central value determined here for dlas alone is @xmath282/3 of the total redshift density of all optically thick systems , again suggesting that the redshift density of dlas is lower than what is determined here .
the comoving gas mass density can be determined as @xmath88 where @xmath89 is the mean molecular mass of the gas taking into account the contribution of he , @xmath90 is the critical density of the universe , @xmath91 is the hubble constant , and @xmath92 is the mass of the hydrogen atom @xcite .
the absorption distance dx is defined as @xmath93 and is summed over each sightline in the sample .
the total path length for the 37 qsos in this sample using the criteria above is @xmath94x=15.1 .
we find @xmath1=1.4@xmath5 and @xmath1=4.2@xmath7 for dlas and sub - dlas respectively .
the neutral gas density in dlas at @xmath95 has been determined by previous surveys to be @xmath1=@xmath96 @xcite .
the value determined here agrees within the errors with these previous determinations .
the low-@xmath12 mass density of neutral gas from sub - dla systems has not yet been thoroughly investigated . at @xmath97 ,
@xcite find @xmath1=@xmath98 . due to the small number statistics ,
this sample is dominated by statistical errors . upper and lower bounds on the gas density have been determined by using the upper and lower 1@xmath48 limits on the number of systems , @xmath99 , from @xcite assuming poisson statistics .
we estimate the 1@xmath48 range of @xmath100 n(h i ) in equation [ eq : omega ] as @xmath100 n(h i)=@xmath101n(h i)@xmath102 where @xmath62n(h i)@xmath102 is the mean _ observed _ column density in the sample .
in this work , we have reported on the first damped lyman-@xmath0 systems observed with the cosmic origins spectrograph .
we summarize the results of this work as follows .
+ 1 ) we have determined the redshift density and gas density of dlas and sub - dlas in this sample covering a total redshift path of @xmath8 at @xmath84 .
we find d@xmath2/@xmath86 and @xmath1=1.4@xmath5 for dlas , and for sub - dlas d@xmath2/dz=0.08@xmath6 and @xmath1=4.2@xmath7 .
the redshift density of dlas determined here is several times higher than previous measurements of this value at @xmath95 of d@xmath2/[email protected] @xcite .
the large statistical errors of our sample make any differences unlikely to be statistically significant .
no measure of d@xmath2/@xmath103 or @xmath1exists for sub - dlas at @xmath104 , however the value determined here is lower than d@xmath2/dz=0.34@xmath105 and @xmath1=@xmath98 seen at @xmath97 @xcite .
the neutral gas density in dlas measured here is slightly higher than what has been found in previous surveys at @xmath95 which find @xmath1=@xmath96 @xcite , but the allowed range is fully consistent with the previous measurements .
based on 21 cm emission maps of galaxies in the nearby universe , @xcite find @xmath1=@xmath106 , which is well below the value determined here .
\2 ) we have measured the abundances of several elements of one dla in the sample and shown that the system has a sub - solar total abundance with [ s / h]=@xmath11 .
this metallicity is significantly higher than typical high redshift dlas , which have average metallicities @xmath62[zn / h]@xmath107 . the kinematical structure of the dla is remarkably simple .
only two components are needed to fit the profiles and the velocity width is well below that of the average dla population with @xmath16=52 km s@xmath17 . remarkably , the kinematics are similar to those of the extremely metal poor dlas from @xcite , even though the metallicity of this dla is @xmath28100 times greater than the dlas in that sample
. + 3 ) followup imaging with wfc3 reveals several star forming and presumably gas rich galaxies within 10@xmath18 of the qso , as well as a large spiral galaxy at small impact parameter to the bright qso . at first glance ,
any of these candidates appear to be promising candidates for the host galaxy of the dla system .
followup spectroscopy of the galaxies reveal that none of them are at a similar redshift to the dla .
spectra of an h ii region in the spiral galaxy show strong narrow emission lines at the redshift of the qso .
this , combined with the close alignment of the nucleus and disk of the galaxy leads us to identify the spiral galaxy as the host of the qso . the qso host is luminous , with m@xmath108 based on a decomposition of the wfc3 image with galfit .
a small feature is seen @xmath280.7 arcseconds from the center of the qso nucleus after psf subtraction , adding another host galaxy candidate of the dla .
even with this supporting data , the origin of the dla host galaxy remains enigmatic and exemplifies some of the difficulties of discovering the host galaxies of such systems .
+ 4 ) independent of the morphological properties of the dla host galaxy , we conclude that the sightline likely passes through the warm neutral medium of the host as the metals are very mildly depleted with [ s / fe]=@xmath109 , [ s / ni]=@xmath110 and [ s / ti]=@xmath111 and both c ii@xmath66 and c i are not detected . the depletion levels and low cold gas content are characteristic of higher @xmath12 dla systems @xcite .
we find that the line of sight depletion factor f@xmath112 , characteristic of sightlines through the halo of the milky way , reaffirming the conclusion that the sightline is passing through warm neutral gas .
+ in a forthcoming paper we will report on the metallicities and depletions in the other dlas and sub - dlas in this sample , as well as the n(h i)-weighted mean metallicity at @xmath84
. few observations of s or zn , undepleted elements in the ism and accurate metallicity indicators , exist at @xmath27 which spans @xmath113 percent of the age of the universe .
no observations of zn or s in sub - dla systems exist in this redshift range . at low redshifts , where there has been more time for the successive buildup of metals in the ism of galaxies and the abundances of dlas should be higher than at high redshift , it will be possible to see if dlas show an increase in the [ n/@xmath0 ] ratio characteristic of secondary production of n. our data show that over its lifetime , the cosmic origins spectrograph will have the ability to measure the abundances and physical properties of qso absorbers at @xmath114 with nearly the same precision as what can be done at @xmath115 with large ground based observatories .
some of the data presented herein were obtained at the w.m .
keck observatory , which is operated as a scientific partnership among the california institute of technology , the university of california and the national aeronautics and space administration .
the observatory was made possible by the generous financial support of the w.m .
keck foundation .
the authors wish to extend special thanks to those of hawaiian ancestry on whose sacred mountain we are privileged to be guests . without their generous hospitality these observations would not have been possible .
financial support for this research was provided by nasa grants hst - go-11598.03-a and nnx08aj44 g . j.x.p .
also acknowledges support from nsf grant ( ast-0709235 ) .
99 bahcall j.n . ,
kirhakos s. , saxe d.h . ,
schneider d.p . , 1997 ,
apj , 479 , 642 becker r.h . ,
white r.l . , helfand d.j . , 1995 ,
apj , 450 , 559 blanton m.r . ,
hogg d.w . ,
bahcall n.a . , brinkman j. , britton m. , connolly a.j . , csabai i. , loveday j. , et al . , 2003 ,
apj , 592 , 819 borthakur s. , tripp t.m .
, yun m.s . , momijian e. , meiring j.d . ,
bowen d.v .
, york d.g . , 2010 ,
apj , 713 , 131 bowen d.v . ,
tripp t.m .
, jenkins e.b . , 2001 , aj , 121 1456 bowen d.v .
, jenkins e.b . , pettini m. , tripp t.m . , 2005 ,
apj , 635 , 880 centurin m. , molaro p. , vladilo g. , proux c. , levshakov s.a . ,
dodorico v. , 2003 , a@xmath116a , 403 , 55 elmegreen b.g .
, parravano a. , 1994 , apj , 435 , 121 fitzpatrick e. l. , spitzer , l. , 1994 , apj , 427 , 232 froning c.s . , green j.c . ,
ap@xmath116ss , 320 , 181 gehrels n. , 1986 , apj , 303 , 336 ghavamian , p. , et al .
2009 , preliminary characterization of the post - launch line spread function of cos , ` http://www.stsci.edu/hst/cos/documents/isrs/ ` goddard q.e .
, kennicutt r.c . ,
ryan - weber e.v . , 2010 , mnras , 405 , 2791 haardt f. , madau p. , 1996
, apj , 461 , 20 henry r. b. c. , prochaska j.x . , 2007 , pasp , 119 , 962 hamilton t.s . , casertano s. , turnshek d.a . , 2002 ,
apj , 576 , 61 jenkins e.b . ,
tripp t.m . , 2001 ,
apjs , 137 , 297 jenkins e.b . ,
tripp t.m . , 2006 ,
apj , 637 , 548 jenkins e.b . , 2009 , apj , 700 , 1299 kennicutt r.c . , 1998 ,
apj , 498 , 541 kulkarni v.p . ,
khare p. , som d. , proux c. , york d.g .
, meiring j.d . , lauroesch j.t . , 2010 , new astronomy , in press liszt h. , 2002 , a@xmath116a , 389 , 343 ledoux , c. petitjean , p. , moller , p. , fynbo , j. , srianand r. , 2006 , a@xmath116a , 457 , 71 lehner n. , wakker b.p .
, savage b.d . , 2004 ,
apj , 615 , 757 lodders , k. , 2003 , apj , 591 , 1220 lupton r. , blanton m.r . , fekete g. , hogg d.w . , omullane w. , szalay a. , wherry n. , 2004 , pasp , 116,133 meiring j.d . , kulkarni v.p . ,
khare p. , bechtold j. , york d.g .
, cui j. , lauroesch j.t .
, crotts a.p.s.,nakamura o. , 2006 , mnras,370 , 43 meiring j.d .
, kulkarni v.p .
, lauroesch j.t . , proux c. , khare p. , york d.g . , 2009 , mnras , 393 , 1513 meiring j.d . , lauroesch j.t
. , kulkarni v.p .
, proux c. , khare p. , york d.g . , 2009 , mnras , 397 , 2037 marigo p. , 2001
, a@xmath116a , 370 , 194 noterdaeme p. , ledoux c. , petitjean p. , srianand r. , 2009 , a@xmath116a , 327 , 336 noterdaeme p. , petitjean p. , ledoux c. , srianand r. , 2009 , a@xmath116a , 1087 , 1098 noterdaeme p. , srianand r. , mohan v. , 2010 , mnras , 403 , 906 peng c.y . ,
ho l.c . , impey c.d . , rix h - w . , 2010 , aj , 139 , 2097 pettini m. , zych b.j .
, steidel c.c . , chaffee f.h .
, 2008 , mnras , 385 , 2011 prochaska j. x. , wolfe a. m. , 1997 , apj , 487 , 73 prochaska j.x . , wolfe a.m. , 2002 , apj , 566 , 68 prochaska j.x .
, gawiser e. , wolfe a.m. , castro s. , djorgovski s. g. , 2003 , apj , 595 , l9 prochaska j.x . , howk j.c .
, omeara j.m . , tytler d. , wolfe a.m. , kirkman d. , lubin d. , suzuki n. , 2002 , apj , 571 , 693 prochaska j.x . , herbert - fort s. , wolfe a.m. , 2005 , apj , 635 , 123 prochaska j.x . , omeara j.m . , herbert - fort s. , burles s. , prochter g.e . , bernstein r.a . , 2006 ,
apjl , 648 , 97 proux c. , dessauges - zavadsky m. , dodorico s. , kim t.s .
, mcmahon r. , 2003 , mnras , 345 , 480 proux c. , mcmahon r.g . ,
storrie - lombardi l.j .
, irwin m.j
. , 2003 , mnras , 346 , 1103 proux c. , dessauges - zavadsky m. , dodorico s. , kim t.s . ,
mcmahon r.g . ,
mnras , 363 , 479 rao s.m . ,
turnshek d.a . ,
2000 , apj , 130 , 1 rao s.m .
, turnshek d.a .
, nestor d.b . , 2006 , apj , 636 , 610 ribaudo j. , lehner n. , howk c.j . , 2010 ,
apj , submitted savage b.d .
, sembach k.r . , 1996 ,
araa , 34 , 279 schaye j. , 2004 , apj , 609 , 667 schlegel d.j .
, finkbeiner d.p .
, davis m. , 1998 , apj , 500 , 525 sembach k.r . , savage b.d . , 1996 ,
apj , 457 , 211 srianand r. , petitjean p. , ledoux c. , ferland g. , shaw g. , 2005 , mnras , 362 , 549 storrie - lombardi l.j . , wolfe a.m. , 2000 , apj , 543 , 552 thom c. , werk j.e . ,
tumlinson j. , prochaska j.x . , meiring j.d . ,
tripp t.m . ,
sembach k.r . , 2011 ,
apj , submitted toomre a. , 1964 , apj , 139 , 1217 tremonti c.a . ,
heckman t.m . , kauffmann g. , brinchmann j. , charlot s. , white s.d.m .
, seibert m. , peng e.w . , et al . 2004 , apj , 613 , 898 tripp t.m . , jenkins e.b . , bowen d.v . , prochaska j.x . , aracil b. , ganguly r. , 2005 , apj , 619 , 714 tumlinson j.t . , werk j.k . , thom c.t . , meiring j.d . , prochaska j.x . , tripp t.m . , okrochkov m. , sembach k.r . , 2011 ,
submitted van zee l. , salzer j.j . , haynes m.p .
, odonoghue a.a . , 1998 ,
apj , 116 , 2805 van zee l. , haynes m.p . , 2006 , apj , 636 , 214 vogt et al . , 1994 ,
spie , 2198 , 362 morton d.c . , 2003 , apjs , 149 , 205 wakker b.p .
, 2006 , apj , 163 , 282 welty d. e. , lauroesch j.t . , blades j.c .
, hobbs l.m . , york , d.g . , 1997 ,
apj , 489 , 672 wolfe a.m. , lanzetta k.m . ,
foltz c.b . , chaffe f. , 1995 , apj , 454 , 698 wolfe a.m. , prochaska j.x . , gawiser e. , 2003 , apj , 593 , 215 wolfe a.m. , gawiser e. , prochaska j.x . , 2005 ,
astrophys . , 43 , 861 wright e.l . , 1991 ,
apj , 381 , 200 york d.g . , adelman j. , anderson j.e . , jr . ,
anderson s.f . , annis j. , bahcall n.a .
, bakken j. a. , barkhouser r. , et al .
2000 , aj , 120 , 1579 zwaan m. a. , meyer m.j . , staveley - smith l. , webster r.l . , 2005 , mnras , 359 , 30 | we report on the first cosmic origins spectrograph ( cos ) observations of damped and sub - damped lyman-@xmath0 ( dla ) systems discovered in a new survey of the gaseous halos of low - redshift galaxies . from observations of 37 sightlines
, we have discovered three dlas and four sub - dlas .
we measure the neutral gas density @xmath1 , and redshift density d@xmath2/dz , of dla and sub - dla systems at @xmath3 .
we find d@xmath2/dz=0.25@xmath4 and @xmath1=1.4@xmath5 for dlas , and d@xmath2/dz=0.08@xmath6 with @xmath1=4.2@xmath7 for sub - dlas over a redshift path @xmath8 .
to demonstrate the scientific potential of such systems , we present a detailed analysis of the dla at @xmath9=0.1140 in the spectrum of sdss j1009 + 0713 .
profile fits to the absorption lines determine log n(h i)=@xmath10 with a metallicity determined from the undepleted element sulfur of [ s / h]=@xmath11 .
the abundance pattern of this dla is similar to that of higher @xmath12 dlas , showing mild depletion of the refractory elements fe and ti with [ s / fe][email protected] and [ s / ti][email protected] .
nitrogen is underabundant in this system with [ n / h]=@xmath14 , placing this dla below the plateau of the [ n/@xmath0 ] measurements in the local universe at similar metallicities .
this dla has a simple kinematic structure with only two components required to fit the profiles and a kinematic width of @xmath16=52 km s@xmath17 .
imaging of the qso field with hst / wfc3 reveals a spiral galaxy at very small impact parameter to the qso and several galaxies within 10@xmath18 , or 20 comoving kpc at the redshift of the dla .
followup spectra with lris on the keck telescope reveal that none of the nearby galaxies are at the redshift of the dla . the spiral galaxy is identified as the host galaxy of the qso based on the near perfect alignment of the nucleus and disk of the galaxy as well as spectra of an h ii region showing emission lines at the qso redshift .
a small feature appears 0.70@xmath18 from the nucleus of the qso after psf subtraction , providing another candidate for the host galaxy of the dla system . even with these supporting data
, we are unable to unambiguously identify the host galaxy of the dla , exemplifying some of the difficulties in determining dla hosts even at low redshift . |
this work is supported in part by the national natural science foundation of china under grant no.10975067 , the specialized research fund for the doctoral program of higher education(srfdp ) ( no.200801650002 ) . c. k. jung et al .
, ann . _ rev .
nucl . part .
sci . _ * 51 * , 451(2001 ) ; q. r. ahmad et al.[sno collaboration ] , _ phys .
lett . _ * 89 * , 011301(2002 ) ; k. eguchi et al .
[ kamland collaboration ] , _ phys .
lett . _ * 90 * , 021802(2003 ) ; m. h. ahn et al.[k2k collaboration ] , _ phys .
lett . _ * 90 * , 041801(2003 ) .
v. barger , d. marfatia , and k. whisnant , _ int
phys . e_*12 * , 569(2003 ) ; m. c. gonzalez - garcia and m. maltoni , _ phys .
rept . _ * 460 * , 1(2008 ) ; r. n. mohapatra and a. y. smirnov , ann .
nucl . part .
* 56*,569(2006 ) ; a. strumia and f. vissani , arxiv : hep - ph/0606054 ; z. z. xing , _ int . j. mod .
a_*23 * , 4255(2008 ) .
w. konetschny and w. kummer , _ phys .
b_*70 * , 433(1977 ) ; t. p. cheng and l. f. li , _ phys . rev . d_*22 * , 2860(1980 ) ; m. magg and c. wetterich , _ phys .
b_*94 * , 61(1980 ) . j. schechter and j. w. f. valle , _ phys .
d_*22 * , 2227(1980 ) .
j. f. gunion , j. grifols , a. mendez , b. kayser and f. i. olness , _ phys .
rev . d_*40 * , 1546(1989 ) ; j. f. gunion , c. loomis and k. t. pitts , _ hep - ph_/*9610237 * ; k. huitu , j. maalampi , a. pietila and m. raidal , _ nucl .
b_*487 * , 27(1997 ) ; j. maalampi and n. romanenko , _ phys .
b_*532 * , 202(2002 ) ; m. muhlleitner and m. spira , _ phys .
rev . d_*68 * , 117701(2003 ) ; g. azuelos , k. benslama and j. ferland , _ j. phys .
g_*32 * , 73(2006 ) ; t. rommerskirchen and t. hebbeker , _ j. phys .
g_*34 * , n47(2007 ) ; t. han , b. mukhopadhyaya , z. si and k. wang , _ phys . rev . d_*76 * , 075013(2007 ) .
e. ma , m. raidal and u. sarkar , _ phys .
lett . _ * 85 * , 3769(2000 ) ; e. ma , m. raidal andu .
sarkar , _ nucl .
b_*615 * , 313(2001 ) ; e. j. chun , k. y. lee and s. c. park , _ phys .
b_*566 * , 142(2003 ) ; m. kakizaki , y. ogura and f. shima , _ phys .
b_*566 * , 210(2003 ) .
t. abe et al.[american linear collider group ] , _ hep - ex_/*0106057 * ; j. a. aguilar - saavedra et al .
[ ecfa / desy lc physics working group ] , _ hep - ph_/*0106315 * ; koh abe et al.[acfa linear collider working group ] , _ hep - ph_/*0109166 * ; ilc technical review committee , second report , 2003 , slac - r-606 , february 2003 .
g. aarons et al .
[ ilc collaboration ] , `` international linear collider reference design report volume 2 : physics at the ilc , '' _ arxiv : _ * 0709.1893 * [ hep - ph ] ; brau et al .
[ ilc collaboration ] , `` ilc reference design report volume 1 - executive summary , '' _ arxiv : _ * 0712.1950 * [ physics.acc-ph ] .
s. chakrabarti , d. choudhury , r. m. godbole , b. mukhopadhyaya , _ phys .
b _ * 434 * , 347(1998 ) ; e. m. gregores , a. gusso , s. f. novaes , _ _ phys
d__**64 * * , 015004(2001 ) ; s. godfrey , p. kalyniak , n. romanenko , _ _ phys . rev . d__**65 * * , 033009(2002 ) ; chong - xing yue , shuang zhao , wei ma , _ nucl . phys .
b _ * 784 * , 36(2007 ) . c. s. lim and t. inami , _ prog . theor .
_ * 67 * , 1569(1982 ) ; a. pich , a. santamaria andj .
bernabeu , _ phys .
b_*148 * , 229(1984 ) ; m. l. swartz , _ phys .
rev . _ * d40*,1521(1989 ) ; m. raidal and a. santamaria , _ phys .
b_*421 * , 250(1998 ) .
g. abbiendi et al .
[ opal collaboration ] , _ phys .
b_*526 * , 221(2002 ) ; p. achard et al .
[ l3 collaboration ] , _ phys .
b _ * 576 * , 18(2003 ) ; j. abdallah et al .
[ delphi collaboration ] , _ phys .
b_*552 * , 127(2003 ) . v. m. abazov et al .
[ d0 collaboration ] , _ phys .
lett . _ * 93 * , 141801(2004 ) ; d. e. acosta et al .
[ cdf collaboration ] , _ phys .
* 93 * , 221802(2004 ) ; v. m. abazov et al .
[ d0 collaboration ] , _ phys .
_ * 101 * , 071803(2008 ) ; t. aaltonen et al .
[ the cdf collaboration ] , _ phys .
lett . _ * 101 * , 121801(2008 ) .
s. chakrabarti , d. choudhury , r. m. godbole , b. mukhopadhyaya , _ phys .
b_*434 * , 347(1998 ) ; e. m. gregores , a. gusso , s. f. novaes , _ phys .
d_*64 * , 015004(2001 ) ; s. godfrey , p. kalyniak , n. romanenko , _ phys .
rev . d_*65 * , 033009(2002 ) ; m. frank , k. huitu , s. k. rai , _ phys .
rev . d_*77 * , 015006(2008 ) .
u. klein , talk given at the 2nd cern - ecfa - nupecc workshop on the lhec , divonnes , france , september 2009 ; m. kuze , talk given at the 2nd cern - ecfa - nupecc workshop on the lhec , divonnes , france , september 2009 ; t. han , b. mellado , _ phys .
rev . d_*82 * , 016009(2010 ) ; b. jager , _ phys .
rev . d_*81 * , 054018(2010 ) .
m. g. albrow et al .
( fp420 r and d collaboration ) , _ arxiv : _ * 0806.0302 * [ hep - ph ] . v. avati and k. osterberg , report no .
cern - totem - note-2005 - 002 , 2006 ; o. kepka and c. royon , _ phys . rev . d_*78 * , 073005(2008 ) . | the @xmath0 triplet model ( @xmath1 ) predicts the existence of a pair of doubly charged @xmath0 bosons @xmath2 .
single production of @xmath2 via @xmath3 collision at the next generation @xmath4 international linear collider ( @xmath5 ) and the large hadron electron collider ( @xmath6 ) is considered .
the numerical results show that the production cross sections are very sensitive to the neutrino oscillation parameters .
their values for the inverted hierarchy mass spectrum are larger than those for the normal hierarchy mass spectrum at these two kinds of collider experiments . with reasonable values of the relevant free parameters ,
the possible signals of the doubly charged @xmath0 bosons predicted by the @xmath1 might be detected in future @xmath5 experiments . *
pacs number * : 12.60.cn , 14.80.cp , 13.10.+q
* 1 .
introduction * during the past decade , neutrino oscillation experiments have provided us with very convincing evidence that neutrinos are massive particles mixing with each other [ 1 ] .
this exciting breakthrough gives a strong motivation for physics beyond the standard model ( @xmath7 ) , since the @xmath7 itself only contains three massless neutrinos [ 2 ] .
various ways to go beyond the @xmath7 have been proposed in order to accomodate this observation .
the @xmath0 triplet model ( @xmath1 ) [ 3 , 4 ] is a simple extension of the @xmath7 with an @xmath8-triplet @xmath0 boson of hypercharge @xmath9 whose vacuum expectation value ( @xmath10 ) @xmath11 provides majorana neutrino masses without introducing right - handed neutrinos .
the @xmath1 [ 3 , 4 ] generates a majorana neutrino mass via the product of a triplet @xmath12 coupling @xmath13 and a triplet @xmath10 @xmath11 .
this model predicts the existence of seven physical @xmath0 bosons which are two @xmath14-even neutral bosons @xmath15 and @xmath16 , a @xmath14-odd neutral one @xmath17 , a pair of single charged bosons @xmath18 , and a pair of doubly charged @xmath0 bosons @xmath2 .
these new particles can produce rich phenomenology at present and in future high energy collider experiments .
since charge conservation prevents doubly charged scalars from decaying to a pair of quarks , @xmath2 carry the lepton number violation and would give rise to a distinctive experimental signal , which has lower background .
this fact has lead to many studies involving the doubly charged @xmath0 bosons @xmath2 in the literature .
for example , it has been shown that such particles can be produced with sizable rates at hadron colliders via the processes @xmath19 [ 5 , 6 ] and @xmath20 [ 5 , 7 , 8 ] . using these production mechanisms ,
discovery of @xmath2 at the large hadron collider ( @xmath21 ) has been studied in refs . [ 9 , 10 ] .
there is a direct connection between the @xmath12 coupling @xmath13 and the neutrino mass matrix which gives rise to phenomenological prediction for processes which depend on @xmath13 [ 11 ] .
the branching ratios of the lepton flavor violating ( @xmath22 ) decay processes @xmath23 and @xmath24 contributed by the doubly charged @xmath0 bosons @xmath2 are strong depend on @xmath13 in the context of the @xmath1 .
recently , it has been shown [ 12 ] that the current upper experimental limits for some of these @xmath22 decay processes can give serve constraints on the parameter space of @xmath13 in the @xmath1 . taking into account these constraints , in this paper , we consider single production of @xmath2 via @xmath3 collision at the next generation @xmath4 international linear collider ( @xmath5 ) [ 13 , 14 ] and the large hadron electron collider ( @xmath6 ) [ 15 ] . although single production of the doubly charged @xmath0 bosons via @xmath3 collision has been studied in refs . [ 16 ] , in this paper , we will focus our attention on the effects of the neutrino oscillation parameters on its production cross section and see whether the neutrino mass hierarchy can be disentangled via this kind of production processes at the @xmath5 and @xmath6 .
our numerical results show that the production cross sections at these two kinds of high energy colliders are very sensitive to the neutrino oscillation parameters .
the production cross section of the process @xmath25 for the inverted hierarchy mass spectrum is significantly larger than that for the normal hierarchy mass spectrum . in the sizable range of the parameter space of the @xmath1
, their values can be large enough to be detected in the future @xmath5 experiments .
our paper is organized as follows .
the essential features of the @xmath1 are briefly reviewed in section 2 . in this section
, we also summarized the constraints on the @xmath1 from some @xmath22 decay processes .
the effective production cross sections for the subprocess @xmath26 at the @xmath5 and @xmath6 are calculated in sections 3 and 4 .
our conclusions and discussions are given in section 5 .
* 2 . the essential features of the @xmath27 * the @xmath0 triplet model ( @xmath1 ) [ 3 , 4 ] is one of the appealing scenarios for generating neutrino mosses without the introduction of a right - handed neutrino . in this model , a complex @xmath8 triplet scalar with the hypercharge @xmath9
is introduced to the @xmath7 , which can be parameterized by @xmath28 a nonzero @xmath10 value @xmath29 gives rise to the following mass matrix for neutrinos @xmath30 the symmetric complex matrix @xmath13 @xmath31 is the @xmath12 coupling strength .
the interaction of the triplet scalar with lepton doublet @xmath32 is given by @xmath33 + h.c.,\ ] ] here @xmath34 is the charge conjugation operator .
the neutrino mass matrix @xmath35 can be diagonalized by the maki - nakagawa - sakata ( @xmath36 ) matrix @xmath37 [ 17 ] .
then the couplings @xmath13 can be written as following @xmath38_{ij}\ ] ] with@xmath39 where @xmath40 , @xmath41 , @xmath42 are the absolute masses of the three neutrinos .
the phase @xmath43 is the dirac cp - violating phase , @xmath44 and @xmath45 are referred to as the majorana phases [ 4 , 18 ] and there are @xmath46 , @xmath47 .
@xmath48 , @xmath49 with @xmath50 , in which @xmath51 are the mixing angles . the explicit expression forms for @xmath13 have been given in refs .
[ 8 , 9 ] .
after the eletroweak symmetry breaking , there are seven physical massive @xmath0 bosons ( @xmath52 ) left in the spectrum for the @xmath1 .
the doubly charged @xmath0 bosons @xmath2 are entirely composed of the triplet scalars @xmath53 , while the remaining eigen states are mixtures of the doublet and triplet scalars .
such mixing is proportional to the triplet @xmath10 @xmath11 , which is generally small even if @xmath11 assumes its largest value of a few gev .
thus , @xmath18 , @xmath16 , @xmath17 are predominantly composed of the triplet scalars , while the @xmath7-like @xmath0 boson @xmath15 is mainly composed of the doublet scalars .
it is well known that the doubly charged @xmath0 bosons @xmath2 have contributions to the @xmath22 decays @xmath24 , @xmath54 [ 19 ] . in the @xmath1
, their branching ratios depend on the neutrino mass matrix parameters , the @xmath10 value @xmath55 , and the mass @xmath56 of the doubly charged higgs bosons [ 12 ] .
the current experimental upper limits for some of these @xmath22 decay processes can give severe constraints on the relevant free parameters . in the @xmath1 , the @xmath22 decay process @xmath57 proceed via exchange of @xmath2 and @xmath18 at one - loop . in the case of ignoring the electron mass in the final state and all lepton masses in the loop
, the branching ratio @xmath58 can be written as [ 12 ] @xmath59 where @xmath60 is the fine structure constant , @xmath61 is the fermi constant . in above equation , it has been assumed @xmath62 .
the current experimental upper limit for the @xmath22 process @xmath63 is @xmath64 [ 20 ] , then we have @xmath65 using eq.(4 ) , there is @xmath66_{ij } , \ ] ] where @xmath67 and @xmath68 .
since the sign of @xmath69 is undetermined at present , distinct patterns for the neutrino mass hierarchy are possible .
@xmath70 is referred to as normal hierarchy @xmath71 with @xmath72 and @xmath73 is defined as inverted hierarchy ( @xmath74 ) with @xmath75 . combing eq.(7 ) and eq.(8 )
, there is the following relation @xmath76_{e\mu}.\ ] ] in our following numerical estimation , we will consider this constraint on the triplet @xmath10 @xmath11 to calculate the cross sections for single production of the doubly charged @xmath0 boson @xmath77 via @xmath3 collision at the @xmath5 and @xmath6 . *
3 .
single production of the doubly charged @xmath78 boson @xmath77 at the @xmath79 * the @xmath21 can generate very massive new particles and will essentially enlarge the possibilities of testing for new physics effects , while the @xmath5 is also need to complement the probe of the new particles with detailed measurement . for
the @xmath5 , the center of mass ( @xmath80 ) energy @xmath81 and the typical integrated luminosity @xmath82 @xmath83 is currently being designed [ 13 , 14 ] .
an unique feature of the @xmath5 is that it can be transformed to @xmath84 or @xmath3 collision with the photon beam generated by laser - scatting method .
the effective luminosity and energy of the @xmath84 and @xmath3 collisions are expected to be comparable to those of the @xmath5 . in some scenarios , they are the best instrument for discovery of the new physics signatures . in this section , we will consider single production of the doubly charged @xmath0 boson @xmath77 predicted by the @xmath1 via @xmath3 collision . from discussions given in section 2
, we can see that the doubly charged @xmath0 boson @xmath77 can be produced via @xmath3 collision associated with a positive lepton .
the relevant @xmath85 diagrams are shown in fig.1 , in which @xmath86 denotes the lepton @xmath87 or @xmath88 .
for the process @xmath89 , the renormalization amplitude can be written as @xmath90 here @xmath91 is the right - handed projection operator and @xmath92 is the polarization vector of the photon @xmath93 .
the explicit expressions for the @xmath12 coupling constants @xmath94 , @xmath95 , and @xmath96 can be written as [ 8 , 9 ] @xmath97 @xmath98,\end{aligned}\ ] ] @xmath99.\end{aligned}\ ] ] , the production cross section @xmath100 for the subprocess @xmath25 as a function @xmath101 for different values of the @xmath102 in the @xmath103@xmath104 and @xmath74@xmath105 cases.,title="fig : " ] ( -122,-1 ) ( a)(90,-1 ) ( b ) , the production cross section @xmath100 for the subprocess @xmath25 as a function @xmath101 for different values of the @xmath102 in the @xmath103@xmath104 and @xmath74@xmath105 cases.,title="fig : " ] after calculating the cross section @xmath106 for the subprocess @xmath26 , the effective cross section @xmath107 at the @xmath5 with the @xmath108 energy @xmath109 can be obtained by folding the cross section @xmath106 with the photon distribution function @xmath110 [ 21 ] @xmath111 where @xmath112 , in which @xmath113 is the c.m .
energy of the subprocess @xmath114 . in above equation
, we have neglected the mass of the lepton @xmath86 due to @xmath115 .
.,title="fig : " ] ( -120,-1 ) ( a)(90,-1 ) ( b ) .,title="fig : " ] it is obvious that the production cross section @xmath100 is dependent on the neutrino oscillation parameters , the triplet @xmath10 @xmath11 , and the mass parameter @xmath116 . direct searches for @xmath2 have been performed at @xmath117 [ 22 ] , tevatron [ 23 ] , and @xmath118 [ 24 ] via the production mechanisms @xmath119 and @xmath120 .
both of these production mechanisms are depend on only one unknown parameter @xmath102 .
the lower mass limits for @xmath102 from the tevatron searches are @xmath121 [ 23 ] . in this paper , we take @xmath102 as a free parameter and assume that its value is in the range of @xmath122 .
the current experimental upper limit for the @xmath22 process @xmath63 has give severe constraint on the triplet @xmath10 @xmath11 , as shown in eq.(9 ) . in our numerical estimation , we take the minimal value given by eq.(9 ) . in this case , the production cross section @xmath100 is only dependent on @xmath102 and the neutrino oscillation parameters .
there are nine neutrino oscillation parameters : three neutrino masses @xmath123 ( or @xmath35 ) , three mixing angles @xmath124 , the dirac phase @xmath125 and two majorana phases @xmath44 and @xmath45 . according the current constraints on the neutrinos and mixing parameters from neutrino oscillation experiments [ 25 ] , we use the following values in this article @xmath126 @xmath127 @xmath128 @xmath129 information on the mass @xmath101 of the lightest neutrino , the dirac and majorana phases can not be obtained from neutrino oscillation experiments .
to simply our paper , we will take @xmath130 and @xmath131 to calculate the cross sections for single production of @xmath77 in the cases of @xmath103 @xmath132 and @xmath74 @xmath133 . for the free parameter @xmath101
, we will assume its value in the range of @xmath134 . .
, title="fig : " ] ( -120,-1 ) ( a)(90,-1 ) ( b ) .
, title="fig : " ] from above discussions we can see that the production cross section for the process @xmath26 depends on two free parameters @xmath101 and @xmath102 . in our numerical estimation , we take the cross section @xmath100 as a function of the lightest neutrino mass @xmath101 for three values of @xmath102 .
our numerical results are summarized in fig.2 , fig.3 , and fig.4 , which correspond the subprocesses @xmath25 , @xmath135 , and @xmath136 , respectively .
one can see from these figures that the value of the production cross section @xmath100 increases as @xmath102 increases , which is because the minimal value of the triplet @xmath10 @xmath11 is proportional to the factor @xmath137 . in both of the nh and ih cases , the production cross section @xmath138 for the subprocess
@xmath139 is approximately equal to that for the subprocess @xmath140 and their values are smaller than @xmath141 in most of the parameter space of the @xmath1 .
this is due to @xmath142 , which reflects the neutrino mixing patterns .
however , for the subprocess @xmath143 , it is not this case . for the nh spectrum ,
its cross section @xmath144 is larger or smaller than the cross section @xmath138 ( or @xmath145 ) , which depends on the mass @xmath101 of the lightest neutrino . for the ih spectrum ,
the cross section @xmath144 is always larger than the cross section @xmath138 ( or @xmath145 ) . for @xmath134 and @xmath146 ,
the values of the cross sections for production of @xmath77 associated with a positron @xmath147 via @xmath3 collision at the @xmath5 with @xmath109 are in the ranges of @xmath148 and @xmath149 corresponding to the @xmath103 and @xmath74 cases , respectively .
( a ) and @xmath150 ( b)events at the @xmath5 . , title="fig : " ] ( -120,-1 ) ( a)(90,-1 ) ( b ) ( a ) and @xmath150 ( b)events at the @xmath5 .
, title="fig : " ] ref .
[ 9 ] has shown that , for @xmath151 , the leptonic decays of @xmath2 are the dominant decay modes , while the decay @xmath152 is negligible . in the context of the @xmath1 , the leptonic decay channels @xmath153 have been extensively studied in refs . [ 11 , 26 ] and the direct connections between their branching ratios and the triplet @xmath12 couplings @xmath13 ( i.e. the neutrino oscillation parameters ) are discussed in these papers . for the leptonic decays of @xmath2 , the possible decay modes are @xmath154 , @xmath155 , @xmath156 , @xmath157 , @xmath158 and @xmath159 . in the case of @xmath160 and @xmath131 , @xmath155 and @xmath156
are the dominate decay modes , their branching ratios are roughly equal to each other for the @xmath103 spectrum and @xmath161 .
while for the @xmath74 mass spectrum and @xmath162 , the doubly charged @xmath0 boson @xmath163 mainly decays to @xmath154 and its branching ratio is @xmath164 .
thus , for the final states generated by the subprocess @xmath25 , we consider the three leptons @xmath165 and @xmath150 for the @xmath103 and @xmath74 cases , respectively .
the number of the @xmath165 and @xmath150 events at the @xmath5 with @xmath109 and the integrated luminosity @xmath166 are shown in fig.5 , where we have taken @xmath101 in the range of @xmath167 . from this figure
, one can see that there will be several hundreds and up to ten thousands three lepton events to be generated per year in future @xmath5 experiments .
it is obvious that this is not the realistic number of the observed events . to take into account detector acceptance
we should impose appropriate cuts .
for example , the observed leptons must carry a minimum energy and respect a suitable rapidity cut , the angles of the observed leptons relative to the beam , @xmath168 , @xmath169 , must be restricted to some ranges [ 27 ] .
detailed analysis is needed , which is beyond the scope of this paper .
it is well known that , at the @xmath5 , the @xmath22 processes can provide extremely clear signatures and are experimentally interesting .
the leptonic decays of the doubly charged @xmath0 bosons give rise to same - sign lepton pairs , which can generate distinct experimental signals .
the three lepton states @xmath165 and @xmath150 generated by the subprocess @xmath143 are almost free of the @xmath7 backgrounds at the @xmath5 , which have been investigated in ref . [ 27 ] .
thus , as long as the triplet @xmath10 @xmath11 is enough small , the doubly charged @xmath0 boson @xmath77 predicted by the @xmath1 might be detected in future @xmath5 experiments .
* 4 .
single production of the doubly charged @xmath78 boson @xmath77 at the @xmath170 * recently , the high - energy @xmath171 collision has been considered at the @xmath21 , which is called the @xmath6 [ 15 ] . for the @xmath6 , the energy @xmath172 of the incoming proton
is given by the @xmath21 beam , and the energy @xmath173 of the incoming electron is in the range of @xmath174 , corresponding to the @xmath80 energies of @xmath175 .
its anticipated integrated luminosity is at the order of @xmath176 depending on the energy of the incoming electron and the design .
several studies have been performed to discuss the possibility of detecting the @xmath0 boson at the @xmath6 [ 28 ] . in this section ,
we consider single production of the doubly charged @xmath0 boson @xmath77 via @xmath3 collision at the @xmath6 .
it is well known that the equivalent photon approximation ( @xmath177 ) can be successfully to describe most of the processes involving photon exchange [ 29 ] .
a significant fraction of @xmath178 collision at the @xmath21 will involve quasi - real photon interactions occurring at energies well beyond the electroweak energy scale .
the quasi - real photons emitted by the incoming protons have low virtuality @xmath179 and scattered with small angles from the beam pipe .
the photon spectrum depending on virtuality @xmath179 and energy @xmath180 can be described by the following relation [ 30 , 31 ] @xmath181\ ] ] with @xmath182 @xmath183 where @xmath172 is the energy of the incoming proton which is related to the photon energy by @xmath184 and @xmath185 is the mass of the proton .
thus , @xmath186 is the proton momentum fraction carried by the photon .
the magnetic moment of the proton is @xmath187 , @xmath188 and @xmath189 are functions of the electric and magnetic form factors .
electromagnetic form factors are steeply falling as a function of @xmath179 .
then , the @xmath179 integrated photon flux can be written as @xmath190 where @xmath191 .
since the contribution to the above integral formula is very small for @xmath192 , in our numerical estimation , we will take @xmath193 . , the production cross section @xmath194 for the subprocess @xmath25 as a function @xmath101 for different values of the @xmath102 in the @xmath103@xmath104 and @xmath74@xmath105 cases .
, title="fig : " ] ( -120,-1 ) ( a)(90,-1 ) ( b ) , the production cross section @xmath194 for the subprocess @xmath25 as a function @xmath101 for different values of the @xmath102 in the @xmath103@xmath104 and @xmath74@xmath105 cases . , title="fig : " ] .
, title="fig : " ] ( -120,-1 ) ( a)(90,-1 ) ( b ) .
, title="fig : " ] at the @xmath6 , the effective production cross section @xmath195 for the subprocess @xmath26 can be written as @xmath196 where @xmath197 with @xmath198 and @xmath199 . to detect the intact proton or its loss energy , the forward proton detectors
are needed . the @xmath200 forward physics ( @xmath201 ) collaboration will have forward detectors with an acceptance of @xmath202 [ 31 ] .
the acceptance of the @xmath203 forward detectors can reach @xmath204 [ 32 ] . in our numerical estimation , we will take @xmath205 and @xmath206 , which corresponds the @xmath108 energy of the subprocess @xmath26 , @xmath207 .
our numerical results are summarized in fig.6 and fig.7 , which plot the production cross sections for the subprocesses @xmath25 and @xmath139 at the @xmath6 as functions of the lightest neutrino mass @xmath101 for three values of the mass @xmath102 of the doubly charged @xmath0 boson . in these figures
we have shown our results for the @xmath103 and @xmath74 cases .
our calculation shows that , at the @xmath6 , the effective production cross section for the subprocess @xmath208 is roughly equal to that for the subprocess @xmath209 , which is similar with the conclusion at the @xmath5 .
thus , we have not given the contours for the subprocess @xmath209 . from these figures ,
one can see that the production cross section of the subprocess @xmath210 at the @xmath6 is smaller than that for the @xmath5 . for @xmath134 and @xmath146 ,
the values of the production cross section @xmath211 and @xmath212 are in the ranges of @xmath213 and @xmath214 , respectively , for the @xmath103 spectrum , while are in the ranges of @xmath215 and @xmath216 , respectively , for the @xmath74 spectrum . if we assume the integrated luminosity @xmath217 of the @xmath6 with @xmath218 , which corresponds @xmath198 and @xmath199 , then there will be several and up to hundreds @xmath219events to be generated per year .
the process @xmath220 with @xmath77 decaying to @xmath154 can give rise to number of signal events with same - sign electron pair and an isolated positron , which is almost free of the @xmath7 backgrounds .
thus , it is also possible to detect the signatures of the doubly charged @xmath0 boson @xmath77 via its production associated with a positron at the @xmath6 .
certainly , detailed confirmation of the observability of the signals generated by the @xmath1 at the @xmath6 experiments would require monte - carlo simulations of the signals and backgrounds , which is beyond the scope of this paper .
* 5 . conclusions and discussions * doubly charged @xmath0 bosons ( @xmath2 ) appear in some popular new physics models beyond the @xmath7 , which can accommodate neutrino masses .
this kind of new particles have distinct experimental signals through their decay to same - sign lepton pairs .
their observation in future high energy collider experiments would be a clear evidence of new physics .
thus , searching for @xmath2 is one of the main goals of current and future high energy collider experiments .
the @xmath1 is one of the attractive new physics models , in which a @xmath0 triplet is introduced and the neutrino masses can be obtained by the product of the triple @xmath10 @xmath221 and the triplet @xmath12 coupling @xmath13 .
this model predicts the existence of the doubly charged @xmath0 bosons ( @xmath2 ) . in this paper , we consider production of @xmath77 associated with a positive lepton @xmath86 via @xmath3 collision at the @xmath5 and @xmath6 .
the production cross sections are dependent on the neutrino oscillation parameters , the triplet @xmath10 @xmath11 , and the mass parameter @xmath116 . in our numerical estimations
, we use the lower bound on @xmath222 given by the @xmath22 process @xmath63 , assume @xmath130 and @xmath131 , and take the experimental measurement values for the parameters @xmath223 , @xmath224 , and @xmath35 . in this case , the production cross sections depend only on two free parameters , the mass parameter @xmath102 and the lightest neutrino mass @xmath101 . from our numerical results ,
we can obtain following conclusions .
\(1 ) the cross section for production of the doubly charged @xmath0 boson @xmath77 associated with a positive lepton @xmath86 at the @xmath5 is generally larger than that at the @xmath6 .
the production cross sections for the @xmath74 mass spectrum are larger than those for the @xmath103 mass spectrum at these two kinds of collider experiments .
\(2 ) because of @xmath225 , the production cross section of the subprocess @xmath135 is roughly equal to that of the subprocess @xmath209 at both the @xmath5 and @xmath6 experiments . for the @xmath74 mass spectrum ,
the cross section @xmath226 is always larger than @xmath227 , while for the @xmath103 mass spectrum , @xmath226 is larger or smaller than @xmath227 which depends on the value of the lightest neutrino mass @xmath101 .
\(3 ) for @xmath134 and @xmath228 , the values of the cross section @xmath226 at the @xmath5 with @xmath229 are in the ranges of @xmath230 and @xmath231 corresponding with the @xmath103 and @xmath74 cases , respectively . while their values at the @xmath6 with @xmath232 are in the ranges of @xmath233 and @xmath234 for the @xmath103 and @xmath74 cases , respectively .
\(4 ) for the leptonic decays of @xmath77 , the process @xmath26 can produce three lepton final state , which is almost free of the @xmath7 backgrounds . in the context of the @xmath1 , this process can only generate number of three lepton events at the @xmath6 , while , with reasonable values of the relevant free parameters , it can give rise to large number of three lepton events at the @xmath5 .
for instance , for @xmath235 and @xmath146 , the subprocess @xmath26 with @xmath77 decaying to @xmath154 can generate @xmath236 @xmath150 events at the @xmath5 experiment with @xmath237 and the integrated luminosity @xmath166 in the @xmath74 mass spectrum .
thus , the possible signals of @xmath77 predicted the @xmath1 might be detected via this process in the future @xmath5 experiments . in our numerical estimation
, we have taken the minimal value of the triplet @xmath10 @xmath221 demanded by the current upper experimental bound for the @xmath22 process @xmath63 . however , the constraint on the @xmath1 from the process @xmath238 is generally stronger than that from the process @xmath63 .
for example , ref .
[ 12 ] has shown that , for @xmath131 and @xmath239 , the processes @xmath240 and @xmath63 demand @xmath241 and @xmath242 , respectively .
it is obvious that the minimal value of @xmath221 demanded by @xmath57 is smaller than that for @xmath238 .
thus , if we use the constraints on the free parameters of the @xmath1 coming from the process @xmath238 , the production cross sections and the number of the signal events will be decreased at least by one order of magnitude . however , as long as @xmath243 , there will be several @xmath244 events to be generated at the @xmath5 .
it has been shown [ 26 ] that the majorana phases @xmath44 and @xmath45 have large effects on the branching ratios @xmath245 .
thus , @xmath44 and @xmath45 also have effects on the cross section for production of the doubly charged @xmath0 boson @xmath77 associated with a positive lepton @xmath86 .
if we vary the values of @xmath44 and @xmath45 , the numerical results for the cross sections @xmath226 , @xmath227 , and @xmath246 are also changed .
however , our physical conclusions are not changed .
certainly , the doubly charged @xmath0 boson @xmath247 can also be singly produced at the @xmath5 and the @xmath6 via the charge - conjugation processes of the corresponding processes for @xmath77 .
similar to above calculation , we can give the values of the production cross sections for @xmath247 .
thus , the conclusions for @xmath77 also apply to @xmath247 .
if the doubly charged @xmath0 bosons @xmath2 are indeed found at the @xmath21 , various consistency checks will have to be performed .
one of the important tasks is to exact calculate and precise measure the branching ratios @xmath248 and see whether their values are consistent with a neutrino mass matrix from oscillation data .
we expect that our work will be helpful to test the higgs triplet mechanism for neutrino masses and further to distinguish different new physics models . |
observers of interstellar sight lines are familiar with the strong @xmath10@xmath7@xmath11 permitted absorption bands of co below 1545 .
less familiar to observers are the intersystem bands of co , which involve the triplet states @xmath12 , @xmath13 , and @xmath14 , whose electric dipole transitions to the ground state @xmath11 are spin - forbidden .
these triplet states have potential energy curves that cross the curve of @xmath10see figure 1 of rostas et al .
( 2000)and via mutual perturbations acquire a small percentage of @xmath15 character at the expense of the singlet state .
the resulting triplet - singlet @xmath6-values are relatively small , thus turning the intersystem bands into very useful astrophysical probes along interstellar ( is ) sight lines for which the _ a@xmath7x _ bands are optically thick , see ( * ? ? ?
* hereafter mn ) , ( * ? ? ?
* hereafter f94 ) . to exploit such probes , accurate @xmath6-values of intersystem transitions are required .
* hereafter rostas et al . ) presented predictions from extended calculations of the triplet - singlet mixing coefficients , partly as an attempt to resolve discrepancies between previous theoretical ( mn ) and astrophysical @xmath6-values .
rostas et al . also conducted low - resolution laboratory measurements at room temperature to improve upon the experimental results of @xcite .
high - resolution laboratory measurements on two intersystem bands were also reported by @xcite . using the _ hubble space telescope _ ( @xmath16 ) , we obtained a high - resolution echelle spectrum of the o9.5v star x per ( hd 24534 ) with the space telescope imaging spectrograph ( stis ) . permitted _
a@xmath7x _ bands along this line of sight
were first modeled by @xcite , who based his analysis of @xmath0co on archival @xmath16 goddard high - resolution spectrograph ( ghrs ) observations . for our stis observations ,
the weakest permitted band available for analysis is the _ a@xmath7x _ ( 8@xmath70 ) band , which still has a large optical depth ( @xmath17 ) in the cores of the strongest lines , @xmath17(@xmath18(0 ) ) @xmath19 70 .
therefore , the column density ( @xmath20 ) of @xmath0co toward x per can be measured more reliably from the observed intersystem bands , which have a much smaller @xmath17 . in this letter
, we determine @xmath20(@xmath0co ) through a comparison with our previous observations of intersystem bands toward @xmath1 oph ( f94 ) .
then we derive is @xmath6-values toward x per for all seven `` old '' bands and for five intersystem bands seen in is spectra for the first time .
finally , we compare our @xmath6-values with the predictions of rostas et al .
regarding nomenclature , we make use in this paper of the short notation introduced by rostas et al . , where , e.g. , @xmath2112 stands for _ d@xmath7x _
( 12@xmath70 ) .
stis spectra of x per were acquired in 2001 february and march with grating e140h during 10,728 s over five orbits , yielding the data sets o64812010@xmath7030 and o64813010@xmath7020 .
the star was observed through the smallest aperture ( 0.1 @xmath2 0.03 ) , which was designed to provide the highest resolving power , @xmath18 = 200,000 @xcite .
elaborate reduction procedures had to be employed in order to correct for odd - even sub - pixel effects in the unbinned ( 2048-pixel ) exposures .
normal reduction routines were then followed by processing the spectra in the stsdas environment of iraf .
the five orbital exposures were co - added into a single spectrum of 42 orders , sporting a signal - to - noise ratio of 70 , or 100 for features detected and combined from two adjacent overlapping orders .
since the continuum is modulated by shallow photospheric features from x per , short spectral segments centered on features ( bands ) of interest were rectified and normalized before being compared to a fitted synthetic model of the pertinent feature .
flux and wavelength calibrations were performed by the space telescope science institute pipeline and supplied with the binned , 1024-pixel exposures .
a previous analysis of @xmath0co by @xcite inferred that this line of sight involves two unresolved velocity components .
however , the significantly higher resolution of our spectrum shows profile asymmetries due to underlying structure , which required four components for a good fit .
this is the same number of components found in preliminary fits of high - resolution spectra of x per ( d. e. welty , private communication ) , although we derived a different set of relative velocities , fractions , and component widths .
some 87% of the total @xmath20(@xmath0co ) resides in a main component with a doppler parameter ( @xmath22-value ) of 0.39 km s@xmath5 .
the other three components have between 2% and 9% shares of @xmath20 , with @xmath22-values between 0.17 and 2.09 km s@xmath5 , as will be fully described in a future paper .
the highly saturated _
a@xmath7x _ bands could be fitted only once the total @xmath20 was established via the intersystem bands .
our method of analysis employed a multi - parametric spectrum synthesis code based on voigt profile line transfer equations given in @xcite .
the computed absorption spectrum was convolved with a gaussian instrumental profile ( jenkins & tripp 2001 ) and then matched with the data via rms - minimization down to relative parameter steps of 10@xmath23 or less .
in order to determine @xmath6-values for intersystem bands , @xmath20(@xmath0co ) toward x per must be known . for optically thin features ,
the equivalent width ( @xmath24 ) is proportional to @xmath17 , i.e. , @xmath24 @xmath25 @xmath26 .
we can , therefore , utilize the published value of @xmath20 toward @xmath1 oph and the ratio of @xmath24s for optically thin bands toward the two stars in the determination of @xmath20 toward x per
. the result should be independent of the intersystem @xmath6-values and of the modeled cloud structure , provided we exclude the @xmath24 ratio for the optically thick @xmath2714 .
column 4 of table 1 lists the ratios of band @xmath24 toward x per over band @xmath24 toward @xmath1 oph . among the optically thin bands
, the highest @xmath24 ratio belongs to @xmath284 , which was the faintest intersystem band analyzed by f94 .
in addition , @xmath284 also sports the largest deviation of derived @xmath6-value in the low-@xmath17 sample of f94 .
it is clear that f94 underestimated the low - resolution @xmath24 for @xmath284 due to placing the continuum too low ( see their figure 1 ) .
dropping the @xmath284 band , and thus avoiding a 5% increase in the average , we obtain @xmath29(x per)/@xmath29(@xmath1 oph ) = 5.54 @xmath30 0.87 . under the necessary and fulfilled condition of optical thinness ,
this is also the ratio of @xmath20(x per ) over @xmath20(@xmath1 oph ) .
here we shall adopt the @xcite column density toward @xmath1 oph , @xmath20(@xmath0co ) = ( 2.54 @xmath30 0.16 ) @xmath2 10@xmath31 @xmath4 , which was also employed by f94 .
based on the measured @xmath20 ratio , we infer that @xmath20(@xmath0co ) toward x per is ( 1.41 @xmath30 0.22 ) @xmath2 10@xmath3 @xmath4 .
kaczmarczyk ( 2000b ) found ( 1.0 @xmath30 0.2 ) @xmath2 10@xmath3 @xmath4 from an analysis of permitted _ a@xmath7x _ bands ; the agreement between the two determinations is at the 2-@xmath32 level .
recall , however , that kaczmarczyk s model has fewer cloud components and larger @xmath22-values , readily accounting for its reduced column density .
the crossings of vibrational levels in a triplet state and vibrational levels in the @xmath10 state occur near certain rotational @xmath33 levels . therefore , mixing coefficients depend on @xmath33 and thus intersystem @xmath34-values generated by interaction with @xmath10 have a dependence on @xmath35 . whereas all previous calculations of mixing coefficients focused on the vibrational level of the a state ( @xmath36 ) closest to the perturbing triplet level , i.e. , a single-@xmath36 treatment , rostas et al
. were the first to use multi-@xmath36 computations that included all @xmath37 levels with @xmath38 = 0 to 12 .
they derived @xmath34-values that for some intersystem bands were in better agreement with their own experimental results ( @xmath35 = 300 k ) , and/or with the is results of f94 ( @xmath35 = 4.2 k ) .
still , the calculated @xmath6-values of rostas et al .
are `` a few percent '' off , because the changes in @xmath39@xmath40 of the `` parent '' _ a@xmath7x _ bands induced by mixing with other triplet levels were ignored .
in addition , differences of a few percent are expected due to the adopted sources for _ a@xmath7x _ @xmath6-values : mn and we used values from @xcite , while rostas et al.s preference lies with values published by @xcite . the mn line oscillator strength , @xmath41@xmath40 , which we use as input for the fits , includes @xmath39@xmath40 for the permitted _ a@xmath7x _ band and the usual h@xmath42nl - london rotational factor , as well as the single-@xmath36 strength of the perturbation . according to rostas et al .
, for triplet levels that cross the @xmath37 levels at low values of @xmath33 , single-@xmath36 and multi-@xmath36 calculations should not differ by a large factor since the interaction with the nearest `` parent '' @xmath36 level is predominant .
therefore , using @xmath41@xmath40-values from mn is a valid approach while searching for relatively small corrections due to multi-@xmath36 interactions .
this condition is appropriate for all bands studied here , except for @xmath217 , @xmath2110 , and to a lesser extent , @xmath284 . on the other hand , intersystem bands that are farther from their nearest parents
have `` @xmath43-independent '' @xmath34-values due to their high-@xmath33 crossing , especially under is conditions .
again , using @xmath41@xmath40-values from mn poses no difficulties for scaling the integrated band @xmath6-values thanks to negligible high-@xmath33@xmath44 level populations .
indeed , we made no attempt to fit rotational lines independently of each other ; our syntheses did not indicate a need to adjust individual @xmath41@xmath40-values . rather than varying the @xmath6-values ,
our variable parameters were @xmath20(@xmath0co ) , which scales the entire band , and @xmath35 , which determines its shape .
eleven @xmath0co bands exhibit similar excitation temperatures with averages of @xmath45 = 6.1 @xmath30 1.0 k and @xmath46 = 5.9 @xmath30 0.4 k. lower @xmath35 were suggested by the fit to @xmath2714 , presumably due to effects of higher @xmath17 . although @xmath34 depends on @xmath35 , the difference in @xmath6-values between @xmath35 = 4.2 k and @xmath35 = 6.0 k is @xmath47 1% , which we ignore here .
one additional variable fitting parameter was the heliocentric radial velocity .
for the main component , the average value from nine @xmath0co intersystem bands is 14.7 @xmath30 1.0 km s@xmath5 , where the 1 @xmath32 error is dominated by stis wavelength calibration uncertainties .
this radial velocity is in very good agreement with the fits mentioned in 2 .
table 1 lists results of synthetic fits for 12 intersystem bands detected in our spectrum of x per . for the reference
@xmath6-values we used the mn @xmath41@xmath40-values of the @xmath18(0 ) lines , which agree to @xmath30 1% of @xmath34-values calculated for @xmath35 = 4.2 k , except for @xmath2714 and @xmath2717 , which have low-@xmath33 crossings and for which the @xmath18(0 ) @xmath6-values are off by @xmath47 2% .
each ratio of a fitted @xmath34-value over the reference mn value was computed from the ratio of each band s fitted @xmath20(@xmath0co ) over the actual @xmath20 determined in 3 .
column 5 lists the resulting band @xmath6-values toward x per . for comparison , in columns
6@xmath78 we list ratios of @xmath34 from this paper , from f94 , and from theoretical predictions in rostas et al . over the values from mn .
the rostas et al .
values are based on multi-@xmath36 calculations while the mn values refer to single-@xmath36 results .
the theoretical ratios in column 8 lack formal error bars , although deviations due to the presence of other triplet levels and to absolute scales should amount to a few percent .
the last column directly compares @xmath6-values from x per to those from rostas et al .
the first seven bands listed in table 1 were previously detected toward @xmath1 oph by f94 .
as discovered by f94 , the measured @xmath6-value of @xmath217 toward x per is five times greater than the value listed by mn in their tables 5 and 14 .
however , mn list another @xmath6-value for @xmath217 in their table 7 , based on measurements by eidelsberg et al .
( 1992 ) ; the latter result is more in line with both is @xmath6-values and was confirmed in room temperatures experiments by rostas et al .
the more rigorous multi-@xmath36 calculation reproduces the astronomical @xmath6-values much better than a single-@xmath36 prediction because @xmath217 is relatively far from its nearest parent , @xmath372 .
our new @xmath6-value is 6% ( 0.4 @xmath32 ) lower than the prediction by rostas et al . , while the f94 result differs by 1% only .
consequently , we do not confirm the recent laboratory result by stark et al .
( 2002 ) that @xmath34(@xmath217 ) should be reduced by ( 29 @xmath30 15)% relative to rostas et al .
our synthesis of the optically thick @xmath2714 toward x per returns an @xmath6-value 6% smaller than that listed by mn , @xmath48 the f94 result of @xmath724% .
the agreement with the newer predictions is better : the f94 @xmath6-value is lower by 19% , whereas the x per result is identical to that of rostas et al .
according to them , single-@xmath36 and multi-@xmath36 @xmath6-values are also identical : 8.3 @xmath2 10@xmath23 .
thus , ironically , the strong @xmath2714 can not help to distinguish between the two methods because its ( single ) interaction with @xmath374 is extremely predominant .
all other five bands , @xmath2711 , @xmath2717 , @xmath2112 , @xmath284 , and @xmath285 , are also found to be in very good agreement with the multi-@xmath36 predictions of rostas et al .
, i.e. , to within 20% . of these , the f94 @xmath27 bands already showed an excellent agreement with mn ( @xmath9 6% ) , and are within 7% of rostas et al .
the last three f94 bands showed large differences ( down to @xmath749% ) from mn predictions .
all three show better agreement with rostas et al .
: @xmath2112 now differs by @xmath716% , @xmath285 differs by @xmath723% , while @xmath284 is off by @xmath737% ( 3.1 @xmath32 ) .
the f94 @xmath6-value for @xmath284 remains their most discrepant result ( after @xmath217 ) .
this weakest band of f94 was affected by an improper continuum placement , forcing its exclusion from our derivation of @xmath20(@xmath0co ) in 3 .
stark et al.s ( 2002 ) result for @xmath284 , that its @xmath6-value is marginally too small by ( 17 @xmath30 15)% relative to rostas et al . , is very similar to the x per measurement : ( @xmath711 @xmath30 15)% . out of the seven bands under discussion
, the largest x per difference relative to rostas et al.s predictions belongs to @xmath2717 ( + 20% = + 1.0 @xmath32 ) , a band for which a few weaker @xmath33@xmath44 = 2,3 lines are blended with another intersystem band , @xmath2112 . in our spectrum
there are first is detections of four @xmath0co intersystem bands : @xmath218 , @xmath219 , @xmath2110 , and @xmath2812 . of these ,
@xmath34 was predicted by rostas et al . for @xmath218 , while single-@xmath36 @xmath6-values were listed by mn for @xmath218 and @xmath2812 .
we identified @xmath219 and @xmath2110 as intersystem co features thanks to their characteristic shapes .
final confirmation was provided by a comparison with laboratory bandhead wavelengths in table 27 of @xcite , or through the use of molecular state constants from tilford & simmons to compute reasonably accurate line positions .
the newly detected bands are shown in fig .
our wavelength calculations with the constants for the @xmath49 state identify the new _
d@xmath7x _ bands as the @xmath50 ( @xmath51 = 1 ) branch of the @xmath13 state . according to the lists of mn , @xmath52 ( @xmath51 = 2 ) components are at least ten times weaker . whereas @xmath218 was already mentioned by mn , who adopted eidelsberg et al.s ( 1992 ) room temperature @xmath6-value , the newer experiments of rostas et al
. led to a refined value .
our determination clearly favors the new value of rostas et al .
calculated for an is temperature , since the x per @xmath6-value is @xmath724% ( @xmath71.7 @xmath32 ) away from their value , but is @xmath761% ( @xmath77.6 @xmath32 ) away from the value listed by mn .
the strongest of the new bands , @xmath219 , has a @xmath24 between those of @xmath217 and @xmath2112 .
its strong detection contradicts its non - inclusion in the tables of mn , who imposed a limiting @xmath6-value smaller than inferred here .
the third _ d@xmath7x _ band , @xmath2110 , has the smallest secure @xmath6-value found in this study , @xmath19 6 @xmath2 10@xmath53 .
the @xmath2110 band also lacks a previously published @xmath6-value . toward x per ,
it is blended with the red wing of the much wider @xmath541402 , which was successfully removed by profile rectification .
the new is detection of @xmath2812 close to @xmath378 confirms to within 7% the @xmath6-value given by mn , but it was not included in the study of rostas et al .
however , in the last column of table 1 we compare @xmath219 , @xmath2110 , and @xmath2812 @xmath6-values with unpublished multi-@xmath36 calculations ( m. eidelsberg , private communication ) .
two bands agree to @xmath30 12% with these predictions , while the @xmath6-value of our faintest band , @xmath2110 , is @xmath734% ( @xmath72.3 @xmath32 ) away from its predicted value .
the high @xmath20(co ) toward x per also provides us with a clean detection of a fully resolved intersystem band from the @xmath8co isotopomer , namely , @xmath2112 .
this band was suggested to exist by mn in the @xmath1 oph spectrum of @xmath376 analyzed by @xcite . for @xmath8co ,
the @xmath2112 band was predicted to have substantial mixing , borrowing 19% of the @xmath376 @xmath6-value according to mn , versus the mere 1% borrowed by @xmath2112 from @xmath376 in @xmath0co .
since the @xmath378 band of @xmath8co has @xmath55 1 , we could find @xmath20(@xmath8co ) toward x per directly .
using the four - component model from @xmath0co , we derived @xmath20(@xmath8co ) = ( 1.94 @xmath30 0.08 ) @xmath2 10@xmath56 @xmath4 , which yields a @xmath0co/@xmath8co ratio of 73 @xmath30 12 toward x per .
as can be seen in table 1 , the measured @xmath6-value of @xmath8co @xmath2112 is 13% ( 2
@xmath32 ) larger than the single-@xmath36 prediction in mn .
unfortunately , multi-@xmath36 calculations for @xmath8co have yet to be performed .
the agreement between the band @xmath6-values determined from our measurements of absorption along the is line of sight to x per and the calculations of rostas et al .
is very good .
rostas et al . suggested that multi-@xmath36 computations are superior for intersystem @xmath6-values whenever the spin - forbidden band is not fully overlapped by the permitted band .
we concur , because the 11-band average of column 9 in table 1 is 0.95 @xmath30 0.16 , i.e. , the @xmath75% difference can be accounted for by known @xmath35 approximations and _ a@xmath7x _ @xmath6-value uncertainties , the latter estimated to be @xmath57 10% by rostas et al . as a caveat , had we retained the @xmath284 band in the calculation of @xmath20(@xmath0co ) in 3 ,
the average would have changed to 0.90 @xmath30 0.15 due to a systematic adjustment of all @xmath6-values by @xmath75% .
individually , nine intersystem bands show very good agreement with their predictions : @xmath2711 , @xmath2714 , @xmath217 , and @xmath2112 differ by @xmath58 ( 7% ) , while @xmath2717 , @xmath284 , @xmath285 , @xmath219 , and @xmath2812 agree to within 1 @xmath32 ( 20% ) .
such small differences between empirical and predicted @xmath6-values may be primarily attributed to assigned observational uncertainties ( @xmath9 17% for the stis data ; 16% for @xmath20 toward x per ) rather than to the possibility that current quantum - mechanical treatments are inadequate .
the bands @xmath219 , @xmath2110 , and @xmath2812 were not included in rostas et al . , but a comparison with unpublished multi-@xmath36 calculations shows that two @xmath6-values agree to within 12% , while @xmath2110 , our weakest and deblended band , is @xmath734% off .
the agreement between empirical results and theoretical predictions assures that accurate column densities can be extracted from is spectra showing very strong absorption in the permitted bands of @xmath0co .
finally , there is a need for theoretical multi-@xmath36 calculations of triplet - singlet mixing coefficients in @xmath8co .
we thank dr .
m. eidelsberg and dr .
d. e. welty for providing results before publication , and an anonymous referee for helpful comments .
the research presented here was supported in part by a nasa grant for @xmath16 program go-08622 and nasa grant nag5 - 11440 to u. toledo .
black , j. , & van dishoeck , e. 1988 , , 331 , 986 chan , w. f. , cooper , g. , & brion , c. e. 1993 , chem .
phys . , 170 , 123 eidelsberg , m. , rostas , f. , breton , j. , & thieblemont , b. 1992 , , 96 , 5585 eidelsberg , m. , jolly , a. , lemaire , j. l. , tchang - brillet , w .- .
l. , breton , j. , & rostas , f. 1999 , , 346 , 705 federman , s. r. , cardelli , j. a. , sheffer , y. , lambert , d. l. , & morton , d. c. 1994 , , 432 , l139 jenkins , e. b. , & tripp , t. m. 2001 , , 137 , 297 kaczmarczyk , g. 2000a , , 312 , 794 2000b , , 316 , 875 lambert , d. l. , sheffer , y. , gilliland , r. l. , & federman , s. r. 1994 , , 420 , 756 morton , d. c. , & noreau , l. 1994 , , 95 , 301 rostas , f. , eidelsberg , m. , jolly , a. , lemaire , j. l. , le floch , a. , & rostas , j. 2000 , , 112 , 4591 sheffer , y. , federman , s. r. , lambert , d. l. , & cardelli , j. a. 1992 , , 397 , 482 stark , g. , smith , p. l. , yoshino , k. , esmond , j. r. , matsui , t. , & ito , k. , 2002 , , 138 , 305 tilford , s. g. , & simmons , j. d. 1972 , j. phys .
chem . ref .
data , 1 , 147 .
llrcccccccc @xmath0co & @xmath2711 [ @xmath372 ] & [email protected] & [email protected] & [email protected](@xmath75 ) & [email protected] & [email protected] & 1.14 & [email protected] + & @xmath2714 [ @xmath374]&[email protected] & [email protected] & [email protected](@xmath74 ) & [email protected] & [email protected] & 0.94 & [email protected] + & @xmath2717 [ @xmath376 ] & [email protected] & [email protected] & [email protected](@xmath75 ) & [email protected] & [email protected] & 1.00 & [email protected] + & @xmath217 [ @xmath372 ] & [email protected] & [email protected] & [email protected](@xmath75 ) & [email protected] & [email protected] & 5.07 & [email protected] + & @xmath2112 [ @xmath376 ] & [email protected] & [email protected] & [email protected](@xmath75 ) & [email protected] & [email protected] & 0.69 & [email protected] + & @xmath284 [ @xmath372 ] & [email protected] & [email protected] & [email protected](@xmath75 ) & [email protected] & [email protected] & 0.81 & [email protected] + & @xmath285 [ @xmath373 ] & [email protected] & [email protected] & [email protected](@xmath75 ) & [email protected] & [email protected] & 0.69 & [email protected] + & @xmath218 [ @xmath373 ] & [email protected] & & [email protected](@xmath76 ) & [email protected] & & 0.52 & [email protected] + & @xmath219 [ @xmath374 ] & [email protected] & & [email protected](@xmath75 ) & & & & ( [email protected] ) + & @xmath2110 [ @xmath375 ] & [email protected] & & [email protected](@xmath76 ) & & & & ( [email protected] ) + & @xmath2812 [ @xmath378 ] & [email protected] & & [email protected](@xmath75 ) & [email protected] & & ( 0.96 ) & ( [email protected] ) + @xmath8co & @xmath2112 [ @xmath376 ] & [email protected] & & [email protected](@xmath73 ) & [email protected] & & & + | in an echelle spectrum of acquired with the space telescope imaging spectrograph we have identified individual rotational lines of 11 triplet - singlet ( intersystem ) absorption bands of @xmath0co .
four bands provide first detections for interstellar clouds . from a comparison with the @xmath1 oph sight line we find that x per is obscured by a higher @xmath0co column density of 1.4 @xmath2 10@xmath3 @xmath4 .
together with the high spectral resolution of 1.3 km s@xmath5 , this allows ( i ) an improved measurement of previously published @xmath6-values for seven bands , and ( ii ) an extraction of the first astrophysical oscillator strengths for _ d@xmath7x _ ( 8@xmath70 ) , ( 9@xmath70 ) , and ( 10@xmath70 ) , as well as for _ e@xmath7x _ ( 12@xmath70 ) .
the @xmath8co _ d@xmath7x _ ( 12@xmath70 ) band , previously suspected to exist toward @xmath1 oph , is now readily resolved and modeled .
our derived intersystem @xmath6-values for @xmath0co include a few mild ( @xmath9 34% ) disagreements with recent predictions from a perturbation analysis calculated for the interstellar excitation temperature .
overall , the comparison confirms the superiority of employing multiple singlet levels in the calculations of mixing coefficients over previous single - level predictions . |
a graph @xmath12 is a _
pairwise compatibility graph _ ( pcg )
if there exists a tree @xmath1 , an edge - weight function @xmath13 that assigns positive values to the edges of @xmath1 and two non - negative real numbers @xmath2 and @xmath3 , with @xmath14 , such that each vertex @xmath5 is uniquely associated to a leaf @xmath4 of @xmath1 and there is an edge @xmath6 if and only if @xmath7 where @xmath8 is the sum of the weights of the edges on the unique path from @xmath4 to @xmath9 in @xmath1 .
in such a case , we say that @xmath0 is a pcg of @xmath1 for @xmath2 and @xmath3 ; in symbols , @xmath15 .
it is clear that if a tree @xmath1 , an edge - weight function @xmath13 and two values @xmath2 and @xmath3 are given , the construction of a pcg is a trivial problem .
we focus on the reverse of this problem , i.e. , given a graph @xmath0 we have to find out a tree @xmath1 , an edge - weight function @xmath13 and suitable values , @xmath2 and @xmath3 , such that @xmath16 .
such a problem is called the _ pairwise compatibility tree construction problem_. the concept of pairwise compatibility was introduced in @xcite in a computational biology context and the weight function @xmath13 has positive values , as it represents a not null distance .
there are several known specific graph classes of pairwise compatibility graphs , e.g. , cliques and disjoint union of cliques @xcite , chordless cycles and single chord cycles @xcite , some particular subclasses of bipartite graphs @xcite , some particular subclasses of split matrogenic graphs @xcite .
furthermore a lot of work has been done concerning some particular subclasses of pcg as leaf power graphs @xcite , exact leaf power graphs @xcite and lately a new subclass has been introduced , namly the min - leaf power graphs @xcite .
initially , the authors of @xcite conjectured that every graph is a pcg , but this conjecture has been confuted in @xcite , where a particular bipartite graph with 15 nodes has been proved not to be a pcg .
this latter result has given rise to this research as it is natural to ask for the smallest graph that is not a pcg .
a _ caterpillar _ @xmath17 is an @xmath18-leaf tree for which any leaf is at a distance exactly one from a central path called _
spine_. a _ centipede _ is an @xmath18-leaf caterpillar , in which the edges incident to the leaves produce a perfect matching . deleting from an @xmath18-leaf centipede
the degree two vertices and merging the two edges incident to each of these vertices into a unique edge , results in a new caterpillar that we will call _ reduced centipede _ and denote by @xmath19 ( as an example , @xmath20 is depicted at the top left of fig .
[ fig.5nodes ] ) .
caterpillars are interesting trees in the context of pcgs , as in most of the cases , the pairwise compatibility tree construction problem admits as solution a tree that is in fact a caterpillar . for this reason , we focus on this special kind of tree . in this note
, we prove that all the graphs with at most seven vertices are pcgs .
more precisely , we demonstrate the following results : * if @xmath21 , then there always exist a new edge - weight function @xmath22 , and a new value @xmath23 such that it also holds : @xmath24 . *
it is well known that graphs with five vertices or less are all pcgs and the witness trees not all caterpillars are shown in @xcite .
for each one of these graphs we prove that it is pcg of a reduced centipede , providing accordingly , an edge - weight function @xmath13 and the two values @xmath2 and @xmath3 .
* all the graphs with six and seven vertices , except for the wheel @xmath11 ( i.e. the graph formed by connecting a single vertex to all vertices of a cycle of length six see figure [ fig.wheel].a ) , are pcgs of a reduced centipede and , for each of them , we provide the edge - weight function @xmath13 and the two values @xmath2 and @xmath3 such that it is @xmath25 , @xmath26 .
* for what concerns the wheel @xmath11 , it is known @xcite that @xmath11 is not pcg of the reduced centipede @xmath27 ( and hence it is not pcg of a caterpillar ) . we show that @xmath11 is pcg of a tree different from a caterpillar .
in this section we list some results that will turn out to be useful in the rest of the paper .
let @xmath1 be a tree such that there exist an edge - weight function @xmath13 and two non - negative values @xmath2 and @xmath3 such that @xmath16 . observe that if @xmath1 has at least @xmath28 vertices and contains a vertex @xmath29 of degree @xmath30 , then we can construct a new tree @xmath31 in which @xmath29 is eliminated , the two edges @xmath32 and @xmath33 incident to @xmath29 are merged into a unique edge @xmath34 and a new function @xmath22 is defined from @xmath13 only modifying the weight of the new edge , that is set equal to the sum of the weights of the old edges : @xmath35 .
it is easy to see that @xmath36 .
for this reason , from now on , we will assume that all the trees we handle do not contain vertices of degree two .
@xcite [ prop.integer ] let @xmath37 , where @xmath38 and the weight @xmath39 of each edge @xmath40 of @xmath1 are positive real numbers .
then it is possible to choose @xmath41 such that for any @xmath40 , the quantities @xmath42 and @xmath43 are natural numbers and @xmath44 .
we prove here the following useful lemma : [ lemma.1 ] let @xmath37 .
it is possible to choose @xmath41 such that @xmath45 , where the minimum is computed on all the edges of @xmath1 , and @xmath44 . according to proposition [ prop.integer ]
, we can assume that the edge weight @xmath13 and the two values @xmath46 are integers .
let @xmath47 be the edges of @xmath1 incident to the leaves .
without loss of generality , we can assume @xmath48 .
we define @xmath49 as follows : @xmath50 and for each @xmath51 define @xmath52 .
clearly , the function @xmath49 is well defined as all its values are positive . as the weight of any edge incident to a leaf
has been decreased by exactly @xmath53 and the rest of the weights remained unchanged , then for of any two leaves @xmath54 it holds that @xmath55 .
let @xmath56 and @xmath57 .
it is easy to see that @xmath44 indeed , if @xmath58 then it means that there was no path weight below @xmath2 , with respect to @xmath13 .
the previous results imply that it is not restrictive to assume that the weights and @xmath2 and @xmath3 are integers and that the smallest weight is @xmath59 .
thus , in the rest of the paper we will use these assumptions .
in this section we will prove that we can get rid of different kinds of caterpillar structures and restrict to consider only reduced centipedes .
[ th.caterpillar ] let @xmath0 be an @xmath18 vertex graph , @xmath17 and @xmath19 be an @xmath18-leaf caterpillar without degree 2 vertices and an @xmath18-leaf reduced centipede , respectively .
let @xmath21 .
it is possible to choose @xmath22 and @xmath23 such that @xmath24 . in order not to overburden the exposition ,
let @xmath60 and @xmath61 .
if @xmath62 is a reduced centipede , the claim is trivially proved , so assume it is not .
we lead the proof into two steps .
first we define a non - negative edge - weight function @xmath63 proving that @xmath62 weighted by @xmath13 and @xmath64 weighted by @xmath63 generate the same pcg @xmath0 for the same values @xmath2 and @xmath3
. then we modify @xmath63 into a positive weight function @xmath22 and introduce two new values @xmath65 and @xmath23 proving that @xmath0 is also @xmath66 .
draw @xmath62 so that : i ) the spine lies on a horizontal line , ii ) all the leaves lie on a parallel line and iii ) the edges between the spine and the leaves are represented as non - crossing line segments ; number the leaves and the vertices of the spine from left to right @xmath67 and @xmath68 , @xmath69 , respectively . by drawing the reduced centipede @xmath64 in a similar way , we number the leaves and the vertices of the spine from left to right by @xmath70 and @xmath71 . we define the edge - weight function @xmath63 as follows : * let @xmath72 the unique adjacent vertex of @xmath73 in @xmath62 ; for each @xmath74 , define @xmath75 ; * define @xmath76 and @xmath77 ; * for each @xmath78 , define @xmath79 if and only if @xmath80 in @xmath62 ; * for each @xmath78 , define @xmath81 if and only if @xmath82 in @xmath62 . observe that @xmath63 is well defined , as @xmath62 has no degree 2 vertices .
it is quite easy to convince oneself that for each pair of leaves in @xmath62 , @xmath73 and @xmath83 , @xmath84 is exactly the same as @xmath85 and that @xmath2 and @xmath3 remain unchanged , so @xmath86 .
it remains to show that we can reassign the edge - weights of @xmath64 in a way that any edge gets a positive weight and @xmath64 is the pairwise compatibility tree of @xmath0 . to this purpose
, we denote by @xmath87 the edge set of any graph @xmath88 , and we introduce the following two quantities : @xmath89 @xmath90 is the smallest distance between the quantities @xmath46 and the weighted distances on the tree of the paths corresponding to non - edges of @xmath0 ; @xmath91 is the number of edges of @xmath64 of weight @xmath92 .
observe that , unless @xmath0 coincides with the clique @xmath93 ( which trivially is pcg of the reduced centipede ) , there always exists a pair of leaves such that their distance on @xmath64 falls out of the interval @xmath94 $ ] and hence @xmath95 .
furthermore , as any edge incident to a leaf in @xmath64 is strictly greater than @xmath92 and in view of the hypothesis that the caterpillar @xmath62 is not a reduced caterpillar , it holds @xmath96 ( the bound @xmath97 is reached when @xmath62 is a star ) .
so , the value @xmath98 is well defined . now define a new weight function @xmath22 on @xmath64 by assigning the weight @xmath99 to any edge of weight @xmath92 . more formally , @xmath100 if @xmath101 and @xmath102 otherwise . in this way the distance between any two leaves in @xmath64 can result increased by a value upper bounded by @xmath103 .
set the new value @xmath104 .
the following three observations conclude the proof : * any distance between leaves in @xmath64 that was strictly smaller than @xmath2 with respect to the weight function @xmath63 remains so after this transformation in view of the fact that @xmath103 ; * any distance that was strictly greater than @xmath3 with respect to the weight function @xmath63 is strictly greater than @xmath23 due to the definition of @xmath90 ; * any distance that was in the interval @xmath94 $ ] with respect to the weight function @xmath63 is now in the interval @xmath105 $ ] .
observe that the previous statement suggests not to consider all kinds of caterpillars , but to restrict to reduced centipedes , only .
in the next section we exploit this result .
in this section we show that all graphs with at most seven vertices , except for the wheel @xmath11 , are pcgs of a reduced centipede .
analogously to what we did in the proof of theorem [ th.caterpillar ] , name the leaves of @xmath19 from left to right with @xmath67 and the vertices of the spine from left to right with @xmath106 .
as , for any @xmath18 , there exists a unique unlabeled reduced centipede with @xmath18 leaves @xmath19 , in the following we consider the edges of @xmath19 as ordered in the following way : @xmath107 ; @xmath108 for each @xmath109 ; @xmath110 ; finally , @xmath111 for each @xmath112 . in figure [ fig.5nodes ]
all the 18 connected non isomorphic cyclic graphs with 5 vertices are depicted , together with the vector @xmath114 and the values of @xmath2 and @xmath3 that witness that all of them are pcgs of @xmath20 .
observe that the connected non isomorphic graphs on 5 vertices are 21 , we have omitted the 3 graphs that are trees , which are trivially pcgs .
we remind that it is already proved in @xcite that all the graphs with at most five vertices are pcg , but the provided trees were all different and not all caterpillars . for what concerns graphs with 6 and 7 vertices , except for the wheel @xmath11 , we get a similar result . for the sake of brevity we do not depict all these graphs ( there are 112 connected non isomorphic graphs with 6 vertices and 853 with 7 vertices ) , but the values of @xmath114 , @xmath2 and @xmath3 we got with the help of an enumerative c program can be found at the web page https://sites.google.com/site/pcg6and7vertices/ .
+ thus , we obtain the following result : calamoneri , t . , petreschi , r. , sinaimeri , b. on relaxing the constraints in pairwise compatibility graphs , in : md . s. rahman and s .-
nakano ( eds . ) , walcom 2012 , lecture notes in computer science , vol . 7157 , pp .
124135 , springer , berlin ( 2012 ) .
kearney , j. i. munro and d. phillips : efficient generation of uniform samples from phylogenetic trees . in : benson , g. ,
page , r.d.m .
lecture notes in computer science , vol . 2812 , pp .
springer , berlin ( 2003 ) . | a graph @xmath0 is called a pairwise compatibility graph ( pcg ) if there exists an edge - weighted tree @xmath1 and two non - negative real numbers @xmath2 and @xmath3 such that each leaf @xmath4 of @xmath1 corresponds to a vertex @xmath5 and there is an edge @xmath6 if and only if @xmath7 where @xmath8 is the sum of the weights of the edges on the unique path from @xmath4 to @xmath9 in @xmath1 . in this note , we show that all the graphs with at most seven vertices are pcgs . in particular all these graphs exept for the wheel on @xmath10 vertices
@xmath11 are pcgs of a particular structure of a tree : a centipede .
pairwise comparability graphs , caterpillar , centipede , wheel . |
the sunspot oscillations , as measured in the photosphere , show broad range of frequencies that peak at 3 mhz ( period of approx . 5 min ) .
they are generally coherent over a large fraction of the sunspot umbra and five minute p - modes may propagate completely across the entire sunspot . moreover , there are indications that the oscillations proceed in 20 40 min long wave trains the origin of which is not well understood .
oscillations of the sunspot umbra as measured chromospheric lines show major peak around 5 mhz ( period of 3 min ) and amplitudes of several kilometers per second . they are thought to represent various magneto - hydrodynamic waves propagating within the sunspot and thought to be generated by sunspot oscillations ( e.g. , * ? ? ?
* ; * ? ? ?
* and references therein ) .
there is a small time lag in the oscillatory patterns measured at different altitudes , which is consistent with the general picture of waves propagating up from the photosphere to the base of the corona ( e.g. , * ? ? ?
although the energy flux associated with with 3 min oscillations appears be too low to maintain the observed temperature of coronal loops @xcite , and later showed that when the entire spectrum of propagating slow magneto - acoustic waves is considered then they could provide energy flux sufficient to account for heating of active region loops .
for comprehensive review of sunspot oscillations reader is referred to papers by @xcite and @xcite .
it was found that the intensity of coronal loops associated with plage areas display periods of the order of 321 @xmath1 74 s , whereas sunspot loops oscillate with periods of the order of 172 @xmath1 32 s , which indicates that photospheric oscillations above a sunspot have the ability to propagate through the transition region ( tr ) and into the corona .
@xcite suggested that the propagating waves may modulate magnetic reconnections between the loops thus driving high - speed outflows along the loops .
velocities of disturbances propagating along the sunspot loops appear to be temperature dependent indicating the slow magneto - acoustic nature of the propagating waves . in the corona
the oscillations are limited to small regions that often coincide with the endpoints of sunspot coronal loops @xcite and show no signs of the resonances @xcite predicted by the chromospheric resonator theories @xcite .
sunspot oscillations also manifest themselves via umbral flashes ( ufs , * ? ? ?
* ) , which are emissions in the core of chromospheric lines caused by upward propagating waves steepening into shocks @xcite .
the temperature in ufs may exceed the surrounding values by up to 1000k and fluctuations of umbral fields are possible on sub - arcsecond scales , although the field inclination and the overall magnetic configurations does not seem to change .
@xcite recently reported on observations of very small scale ( height @xmath2 1 mm , width @xmath2 0.1 mm ) jet - like structures in the sunspot umbra that thought to be driven by sunspot oscillations .
@xcite suggested may be a sunspot version of short dynamic fibrils observed in plage areas . observed density and field oscillations
further support the idea of possible slow mode wave propagation in the atmosphere above sunspot s umbra . in this study
we analyse data from the space based interface region imaging spectrograph ( iris ) and the new solar telescope ( nst ) operating the big bear solar observatory ( bbso ) to better understand the fine structure of chromospheric umbral oscillations , the structure and evolution of umbral flashes and shocks .
we analysed nst @xcite and iris @xcite data acquired during joint iris - bbso - dst observations on 2 september 2013 .
this was a medium sit - and - stare iris data set with the slit crossing the center of the umbra of the main sunspot in noaa ar 11836 .
the iris observing run began at 16:39 ut and lasted for about 100 min .
the pixel size of the data is [email protected] and the time cadence was 3 s and 12 s for spectral and slit - jaw data , correspondingly .
in this study we used calibrated level 2 data available online .
the same iris data set was used for studying shocks in the upper chromosphere and the tr above the sunspot by @xcite .
we utilized spectra and slit - jaw images obtained using mg ii 2796.35 and si iv 1393.76 spectral lines , which form in the chromosphere ( formation temperature 10@xmath4 k ) and middle of the transition region ( tr , 10@xmath5 k ) , respectively .
this data set also includes series of images from atmospheric imaging assembly on board of solar dynamic observatory ( sdo / aia ) @xcite .
the nst data were acquired with the aid of the 308 sub - aperture adaptive optics ( ao-308 ) and span the same time interval as iris observations .
time series of broadband ( 10 ) images of the photosphere were taken with a tio filter ( 7057 , time cadence 15 s , pixel scale [email protected] ) .
this absorption line ( the head of the tio @xmath6-system ) is only formed at low temperatures below 4000 k , and is well suitable for observing sunspot umbra , where the temperatures are significantly lower .
we also utilized available series of narrow - band h@xmath0 ( 6563 ) images taken at @xmath1 0.8 , @xmath1 0.4 , and 0.0 from the line center acquired with nst s visible imaging spectrometer ( vis , pixel size of [email protected] ) , which combines a 5 interference filter with a fabry - prot etalon to produce a resulting bandpass of 0.07 over a 70@xmath3 @xmath7 70@xmath3 field of view .
the difference in the acquisition time at two sequential line positions ( e.g , + 0.4 and -0.4 ) was about 2 s. the five point line scan was recorded every 23 s. at each line position we acquired a burst of 25 images with the exposure times ranging from 12 ms at @xmath10.8 to 25 ms at the line center .
the bursts were then speckle reconstructed @xcite .
in @xcite we have estimated that the errors introduced by local misalignment and residual seeing into measurements at each line position ranges from 2% to 9% with the average of 5% .
we found that vis h@xmath0 + 0.8 and iris mg ii 2796.35 images are best suitable for co - aligning the nst and iris data sets .
we estimate the maximum error of co - alignment to be about [email protected] ( 3 iris pixels ) .
the co - alignment was performed manually using prominent features in the umbra and several brightening events outside the sunspot .
iris images and spectra were co - aligned using the fiducial line on the slit . [ cols="^ " , ] finally , in figure [ nst_iris_ufs ] we compare an h@xmath0 doppler map ( left ) generated by subtracting an h@xmath0 + 0.4 image from the corresponding blue wing image to the nearly co - temporal iris mg ii k slit - jaw image ( right ) .
the purpose of this comparison is to make a connection between the ufs seen in h@xmath0 and iris data .
the arrows in both panels indicate the location of the uf lane discussed in figure [ ha_ufs ] , while the two broken lines mark the position of diffuse mg ii k ufs . although these mg ii k ufs are not high contrast features they were clearly identified from a movie of the slit - jaw images .
it turns out that the mg ii k flashes are co - spatial with locations of red - shifted phase of h@xmath0 shock waves , i.e. , times when the shocked plasma was falling down and the h@xmath0 line was red - shifted .
note , that the iris ufs are somewhat delayed relative to the corresponding h@xmath0 features , which can be explained by the time difference between the two images . in the next iris image taken at 17:25:29 ut the ufs are already further away from the sunspot center and better coincide with the h@xmath0 features , however , their contrast is too low to be reproduced in a figure .
we note that the two red - shifted h@xmath0 arches and the mg ii k ufs marked by the broken lines are co - spatial with the r2 arch in figure [ ufs ] and the umbra - penumbra boundary , respectively .
moreover , similarly to h@xmath0 ufs , the mg ii k flashes , when observed away from the light bridges , appear much weaker and harder to track , thus further confirming our earlier suggestion about possible importance of the lbs in formation and/or enhancement of shocks .
iris and nst spectral data further confirm the idea of upward propagating magneto acoustic waves , which has been reported earlier by a number of observers .
the average time lag between the arrival times of the shocks at the mg ii k and si iv line formation levels , determined from a cross correlation analysis of 100 min long time profiles of blue wing intensities , was found to be about 15 s. we find , however , that this lag is highly dependent on the selection of the correlated segments and peak to peak delay may range from 40 s to zero . analyzing a 100 min long time profile
we find that the intensity of chromospheric shocks above the sunspot s umbra may vary on large time scales of order of 20 min .
however , the available data set was obviously too short to make more confident conclusions .
these long period waves may have internal or external ( relative to a sunspot ) origin .
one may speculate that the modulation may arise from a hysteresis effect , where the current physical conditions of the shocked plasma are defined not only by the currently propagating shock but also by the previous history of plasma heated by passing shocks .
as shocks constantly push the chromosphere up , not all the plasma may timely fall back to the initial state by the time the next one arrives , which may lead to a temporarily slightly over - inflated chromosphere .
when the critical mass that can no longer be supported by shock waves is accumulated , it falls back suppressing the upwardly moving shocks ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
an alternative external cause for the irregular beat may be the sub - photospheric interaction of the upward propagating fast magneto - acoustic waves and those reflected at the photosphere back into the convective zone as discussed by @xcite .
these effects may have a global influence on the sunspot causing 18 - 24 min oscillations in evershed flows @xcite , 18 min periodicities at the chromospheric layers that were concentrated at the outer penumbra , chromospheric flows above sunspot ( e.g. , * ? ? ?
* ) , as well as upward wave propagation .
nst s h@xmath0 data showed that bright ufs prefer to be located near lbs and clusters of uds .
this is a new and significant finding , indicating that although magneto - acoustic waves are present everywhere inside the umbra , the effectiveness of uf production may depend on the non - uniformity in the umbra s magneto - convection properties and magnetic fields .
this inference agrees with conclusion that there are umbral areas with slightly different physical and/or magnetic field conditions .
this preference for ufs to be linked to lbs and ud clusters may be explained as follows .
the uds and lbs are known to have weaker magnetic fields as compared to the dark umbra dotted only with small , dim and barely discernible uds .
therefore they are thought to be surrounded by casp - like fields with a horizontal field gradient @xcite .
@xcite found that the shock formation depends on the vertical gradient of alfven speed , @xmath8 , in the medium .
in particular , if @xmath8 , decreases with height then shocks form more effectively .
such conditions are met either in rapidly expanding vertical flux tubes and/or near edges of flux tubes , where fields may become mostly horizontal . in sunspot umbra , such conditions exist near lbs and ud clusters , which are considered field - free structures . in this case
adjacent umbral fields expand into the atmosphere above these features thus creating a narrow lane of negative vertical gradient of @xmath8 with favourable conditions for formation of chromospheric shocks .
the observations show that ufs tend to appear on the sunspot - center side of lbs .
such dynamics suggests that they may be linked to a presumed horizontal trans - umbral wave that causes ufs to sequentially appear at progressively distant light bridges . according to @xcite simulations of magneto - acoustic waves in sunspots using a 3d nonlinear mhd code
, we speculate that we might be dealing with a compact sub - photospheric driver of oscillations as opposed to photospheric oscillations being coherent on the scales of the entire umbra .
@xcite showed that when a fast ( acoustic ) wave moves horizontally across the umbral field lines , up and away from the subphotospheric source located near the axis of a sunspot , it is partially converted into a slow mode wave when it reaches the layer where the alfven and sound speeds are equal .
the slow mode waves then steepen into shocks thus creating a pattern of shocks ( ufs ) propagating away from the sunspot center .
this behavior inferred from simulations is very similar to the dynamics of ufs observed here .
most of the sunspot umbra , observed in the tr si iv spectral line , appears bright and filled with oscillations and shocks .
however , there are regions that appear very dark at the chromospheric and tr level and do not exhibiting detectable oscillations . while other authors have pointed out the connection between the footpoints of coronal loops and the bright tr umbra ( e.g. , * ? ? ?
* ) , we emphasize a possible connection between the bright magneto - convection features ( i.e. , lbs , clusters of uds ) and bright uv umbra and coronal loop footpoints . to the best of our knowledge ,
the role of lbs in coronal heating has not been widely discussed in the science literature .
we find that the bright si iv umbra is situated above the underlying chromospheric lbs and other bright umbral structures as well as with the footpoints of the overlying bright coronal loops .
the intensity of shocks is not uniform over the umbra and the most intense shock structures were detected in the parts of the umbra associated with one of the lbs .
we suggest that the non - homogeneities in the umbral magnetic field may play an important role in heating coronal loops rooted in the sunspots by facilitating generation of shocks .
the non - uniformities may also increase effectiveness of fast to slow mode conversion , however this question has to be answered by future mhd simulations . finally , as we have already noted in this study
, the second quadrant of the umbra appears dark in si iv and aia 171 ( as well as other ) spectral lines .
careful inspection of the relevant aia and bsbo h@xmath0 images showed that the dark part of the umbra was associated with a system of what appears to be dim horizontal threads possibly representing a filaments channel .
@xcite used spectropolarimetric data acquired at the fe i 15643 - 15658 to conclude that in the presence of moderate and strong magnetic fields ( 300 - 1800 g ) the photospheric waves tend to run upward , while in the areas with low closed loops oscillations exhibited a lack of power and standing waves dominate .
while these observations offer reasonable explanations for existence and spatial distribution of hot and bright uv loops and dim umbra , we still need to understand the mechanisms of wave energy transfer and conversion , heating of umbral loops and the role of light bridges in this process .
this work was conducted as part of the effort of nasa s living with a star focused science team `` jets '' .
we thank bbso observing and engineering staff for support and observations .
this research was supported by nasa lws nnx11ao73 g and nsf ags-1146896 grants .
vyu acknowledges support from korea astronomy and space science institute during his stay there , where a part of the work was performed .
authors thank prof .
nakaryakov for insightful comments .
we also thank anonymous referee for valuable criticism . , k. s. , pevtsov , a. a. , & olmschenk , s. 2008 , in astronomical society of the pacific conference series , vol .
383 , subsurface and atmospheric influences on solar activity , ed .
r. howe , r. w. komm , k. s. balasubramaniam , & g. j. d. petrie , 279 | we analyse sunspot oscillations using interface region imaging spectrograph ( iris ) slit - jaw and spectral data and narrow - band chromospheric images from the new solar telescope ( nst ) for the main sunspot in noaa ar 11836 .
we report that the difference between the shock arrival times as measured the mg ii k 2796.35 and si iv 1393.76 line formation levels changes during the observed period and peak - to - peak delays may range from 40 s to zero .
the intensity of chromospheric shocks also displays a long term ( about 20 min ) variations .
nst s high spatial resolution h@xmath0 data allowed us to conclude that in this sunspot umbral flashes ( ufs ) appeared in the form of narrow bright lanes stretched along the light bridges and around clusters of umbral bright points .
time series also suggested that ufs preferred to appear on the sunspot - center side of light bridges , which may indicate the existence of a compact sub - photospheric driver of sunspot oscillations .
the sunspot s umbra as seen in the iris chromospheric and transition region data appears bright above the locations of light bridges and the areas where the dark umbra is dotted with clusters of umbral dots .
co - spatial and co - temporal data from the atmospheric imaging assembly on board solar dynamics observatory showed that the same locations were associated with bright footpoints of coronal loops suggesting that the light bridges may play an important role in heating the coronal sunspot loops .
finally , the power spectra analysis showed that the intensity of chromospheric and transition region oscillations significantly vary across the umbra and with height , suggesting that umbral non - uniformities and the structure of sunspot magnetic fields may play a role in wave propagation and heating of umbral loops . |
early in the 90th swift and hohenberg formulated an amplitude approach to describe systems , where the stable states are periodic , as e.g. the case for rayleigh - bnard convection @xcite .
more recently this idea has been taken up by the materials science community to model crystals at the atomic scale .
elder et al .
proposed a functional for a scalar dimensionless field @xmath0 of form @xmath1\phi + \frac{1}{4}\phi^4 \right)\,dr \,,\ ] ] with two phenomenological parameters @xmath2 and @xmath3 and a corresponding dynamical equation @xmath4 for this purpose @xcite . since its introduction , this phase - field crystal ( pfc ) method
@xcite has emerged as a computationally efficient alternative to molecular dynamics ( md ) simulations for problems where the atomic and the continuum scale are tightly coupled .
the reason is that it operates for atomic length scales and diffusive time scales .
thus for a simple application such as diffusion in gold or copper it runs @xmath5-@xmath6 times faster than the corresponding md calculation @xcite . in that sense it provides from point of view of multiscale materials modeling an interesting link between the traditional phase - field method and md .
moreover , a connection between classical density functional theory of freezing and phase - field crystal modeling could be identified in @xcite .
thereby a second theoretical foundation besides the swift - hohenberg amplitude equation approach could be established .
essentially it motivates the application of pfc models also for spatially non - uniform _ non - periodic _ states .
recently the phase - field crystal method has been applied to a variety of different growth phenomena .
one of its interesting features is that other than the phase - field method , in which elasticity explicitly needs to be integrated in the functional to be taken into account @xcite , it includes elastic effects inherently .
thus it allows to simulate for example features of crack propagation @xcite and plasticity @xcite from the atomic to the micro - scale .
to model the elastic behavior of different kinds of materials , the parameters of the phase - field crystal model equation can be adjusted to match the elastic moduli of a given experimental system . however , in its most simplistic form , in which it is a reformulation of the swift - hohenberg equation @xcite with a conserved dynamics as introduced by elder _
_ @xcite the poisson ratios which can be modeled , are restricted to @xmath7 ( in the one mode approximation )
. moreover , since in the simplistic pfc model the material is defined by only three parameters , it is restricted with respect to the crystal lattice structures which it can describe as well .
these are triangular symmetries in two dimensions and @xmath8 symmetry in three dimensions @xcite .
another crystal symmetry applying to protein crystals in a membrane could be obtained by including higher order correlation functions @xcite . moreover ,
liquid crystals have been simulated by combining the original phase - field crystal equation with an orientational field @xcite . in @xcite
we followed the above direction to extend the phase - field to apply to a larger class of condensed matter systems taking a different route : we derived a generalized pfc model for isotropic as well as anisotropic crystal lattice systems of arbitrary poisson ratios as well as condensed matter systems built - up from non - spherical units such as for example anisotropic colloids and liquid crystals . to this end
we extended the simplest pfc model proposed by elder @xcite , to a conservative , anisotropic langevin equation and applied it to the heterogeneous nucleation of colloids at a wall . whereas our previous work @xcite was devoted to a mere introduction of our model and its application to heterogenous nucleation at a wall , here
we show for the first time how its parameters can be derived from dynamical density functional theory ( ddft ) ( see ) .
further we report for the first time in detail on its morphology diagram .
to do so , we proceed as follows : first we give a thorough derivation of our anisotropic phase - field model based on the ddft in .
we then - after a brief summary of the model in - study the influence of anisotropy and undercooling on the morphology of the final states .
we analyze these morphologies and summarize our results in terms of a morphology diagram in .
finally we conclude with a summary and an outlook of our study in .
by now phase - field crystal ( pfc ) modeling is widely used to predict crystal nucleation growth and to model microstructural pattern formation during different physical phenomena such as solidification . as usual , the pfc model used in this studies is based on a free energy fuctional @xmath9 $ ] of phase - field @xmath10 and a dynamical equation which represents the time evolution of the phase - field . in this pfc model ,
periodic nature of a crystal lattice is incorporated by using a free energy functional which is minimized by periodic density field .
the equation of motion used in this model was introduced in @xcite for the case of simplest phase - field crystal ( spfc ) model , and is given by @xmath11,\ ] ] here , @xmath2 and @xmath12 are constants . in order to simplify the model , a dimensionless parameter @xmath13 is introduced which is defined as @xmath14 in our model @xmath13 represents undercooling in the same manner as @xmath15 is defined in @xcite , therefore , @xmath13 can be written as @xmath16 the anisotropic version of phase - field crystal model originally introduced in @xcite is used in this study .
this pfc model is basically an extension of the spfc model which is derived in @xcite .
the apfc model is capable of simulating isotropic and anisotropic crystal lattice system of any arbitrary poisson ratio as well as condensed matter systems such as colloids and liquid crystals .
the free energy functional used in this model is given by @xmath17\phi + \frac{1}{4 } c \phi^4 \big ) dr,\end{aligned}\ ] ] where @xmath18 is a symmetric matrix and @xmath19 is a fourth rank tensor with the symmetry of an elastic tensor : @xmath20 . from the free energy functional defined in , the corresponding langevin differential equation of motion for anisotropic lattice system
can be written as follows : @xmath21\phi + c \phi^3 \big).\end{aligned}\ ] ] the coefficients occuring in the anisotropic phase - field crystal model proposed in ref .
@xcite can be derived from microscopic density functional theory @xcite .
here we follow a similar line as proposed recently by van teeffelen et al @xcite for radially symmetric interactions .
we generalize this route here to anisotropic interactions .
we assume that the anisotropic colloids are completely aligned in space .
cartesian coordinates @xmath22 will be used in the following , @xmath23 denoting the spatial dimension .
the interaction pair potential between two aligned particles is @xmath24@xcite .
the latter function is anisotropic , in general , i.e. it does not only depend on @xmath25 .
other examples for these anisotropic interactions with fixed orientations are oriented hard spherocylinders @xcite and charged rods @xcite , anisotropic gaussian potentials @xcite , board - like colloidal particles @xcite , colloidal molecules @xcite , as well as patchy colloids @xcite and proteins @xcite . henceforth _
symmetry is assumed @xmath26 dynamical density functional theory for anisotropic situations @xcite is now generalized from the isotropic case as follows .
the dynamical evolution of the time - dependent one - particle density field @xmath27 is : @xmath28}{\delta \rho({\bf r},t)}\right]\,.\end{aligned}\ ] ] here @xmath29 is the thermal energy , and @xmath30 is the @xmath23-dimensional gradient .
@xmath31 denotes the diagonalized diffusion tensor with the anisotropic short - time translational diffusivities of the anisotropic particle . for a given ( hydrodynamic ) shape of the particle ,
explicit expressions for @xmath32 are available @xcite .
furthermore , in , @xmath33 $ ] is the equilibrium density functional which can be split as @xmath34= f_{\rm{id}}\left[\rho({\bf r})\right ] + f_{\rm{ex}}\left[\rho({\bf r})\right]+f_{\rm{ext}}\left[\rho({\bf r})\right]\,.\ ] ] where @xmath35= k_b t \int\,{\mathrm d } { \bf r } \rho({\bf r})\left\{\ln\left[\rho({\bf r})\lambda^d\right]-1\right\}\,,\ ] ] with @xmath36 denoting the thermal de broglie wavelength .
the external part involves an external one - body potential @xmath37 and is given by @xmath38=\int\,{\mathrm d } { \bf r}\rho({\bf r})v({\bf r},t)\,.\ ] ] finally , the excess part @xmath39 $ ] , embodies the nontrivial correlations between the particles and must be further approximated .
henceforth we assume small deviations of the inhomogeneous density profile around a homogeneous reference density @xmath40 . in this limit
, the leading approximation for @xmath41 $ ] is given by the ramakrishnan and yussouff @xcite expression : @xmath42\simeq f_{\rm{ex}}(\rho ) -\frac{k_bt}{2}\int\int{\mathrm d } { \bf r}{\mathrm d } { \bf r}^\prime \delta\rho({\bf r})\delta\rho({\bf r}^\prime)c_0^{(2)}({\bf r}-{\bf r}^\prime;\rho)\,.\end{aligned}\ ] ] where @xmath43 is the anisotropic direct correlation function of the fluid at density @xmath40 which possesses the same symmetry as the underlying pair potential @xmath24 .
in particular , it is inversion - symmetric @xmath44 moreover , @xmath45 . in fourier space reads @xmath46&=&f_{\rm ex}(\rho ) - \frac{k_{\rm b}t ( 2\pi)^d}{2 } \int { \rm d { \bf k } } { \delta \tilde { \rho}}({\bf k } ) { \delta { \tilde \rho}}({\bf -k})\tilde c_0^{(2)}({\bf k},\rho)\end{aligned}\ ] ] with @xmath47 denoting a fourier transform .
we now expand the direct correlation function @xmath48 in terms of @xmath49 around @xmath50 .
( alternatively fitting procedures can be used , e.g. around the first peak of @xmath48 . )
this leads to the taylor expansion in fourier space @xmath51 corresponding to a gradient expansion in real - space .
inversion symmetry enforces all odd orders to vanish .
possible additional symmetries in the shape of the particles will lead to corresponding restrictions on the tensorial coefficients @xmath18 and @xmath19 as discussed below . inserting this expansion into
, one gets @xmath52\,.\end{aligned}\ ] ] if one further uses the constant mobility approximation , @xmath53 in front of the functional derivative in and if one approximates @xmath54\approx
k_b t\rho \int\,{\mathrm d } { \bf r } \big\{\frac{1}{2}\phi({\bf r},t)^2-\frac{1}{6}\phi({\bf r},t)^3 + \frac{1}{12}\phi({\bf r},t)^4-{\rm const.}\big\}\end{aligned}\ ] ] with @xmath55 , one arrives at : @xmath56 \,.\end{aligned}\ ] ] this exactly reduces to the anisotropic phase - field model of ref .
@xcite for the special case @xmath57 , @xmath58 , and a neglected cubic term in the ideal gas functional expansion in . as a remark
the latter was retained in other variants of the pfc model @xcite . concluding this section , the anisotropic phase - field crystal model as used in @xcite can be derived and justified from dynamical density functional theory .
the derivation points , however , to more realistic approximations for anisotropic diffusivities .
furthermore , if is used , some approximations can be avoided but these were not found to change the results significantly for spherical interactions @xcite .
we finally present phenomenological symmetry arguments for the expansion coefficients @xmath18 and @xmath19 of the anisotropic pfc model .
first we assume that the orientation of the fixed particles is set by a single unit vector @xmath59 only which is invariant under space inversion ( @xmath60 ) .
this is the case for @xmath57 and for rotationally symmetric particles in @xmath61 .
then , any gradient term in the scalar free energy functional must involve an even number of gradients due to space inversion symmetry .
rotational symmetry of space then requires that only combinations of @xmath62 and @xmath63 are nonvanishing in the functional .
therefore the only possibility for physically relevant gradient terms is @xmath64 and @xmath65 where @xmath66 , @xmath67 , @xmath68 , @xmath69 , and @xmath70 are scalar prefactors .
this reduces the number of independent degrees of freedom in @xmath18 and @xmath19 down to 5 . in case
there are different fixed vectors , say @xmath59 and @xmath71 , there are corresspondingly more terms allowing for more freedom in @xmath18 and @xmath19 .
this is realized , e.g. , for biaxial colloidal particles in two crossed external fields along @xmath59 and @xmath71 .
as initial condition , a square domain is defined with a sphere in the center to initialize the nucleus .
periodic boundary conditions in all directions in the square box are used .
the values for the parameters are the same as defined in @xcite , namely @xmath72 ( which sets the time scale ) , @xmath73 , @xmath74 , @xmath75 and @xmath76 , respectively @xcite .
however , a typical set of values for @xmath13 and @xmath77 is used for each simulation since our basic objective is to study the dependence of the stable state phase on these parameters .
a simple explicit numerical scheme is used to obtain a reasonably well approximated solution .
a forward euler scheme is used for the time derivative with a sufficiently small time step of @xmath780.00075 to ensure the stability of the scheme .
the laplace operator is approximated by using second order difference scheme given by @xmath79 for the following simulations @xmath80 is chosen as @xmath81 .
convergence of our results was ensured via convergence studies .
the morphologies depicted in , and show simulations of 256 times 256 numerical grid units .
however , morphologies in show simulation of 128 times 256 numerical grid units .
in this section we present the simulation results obtained from our studies .
these simulation results demonstrated the following issues concerning nucleation and successive microstructure formation : + ( i ) the effect of undercooling on crystal growth .
+ ( ii ) the dependence of anisotropy and undercooling on stable state phase .
+ ( iii ) the effect of anisotropy and undercooling on distance between the neighbouring stripes . in order to quantify the anisotropy of the material at the atomic scale ,
a dimensionless parameter @xmath77 is introduced and is given by @xmath82 the effect of this dimensionless parameter is studied by performing numerical simulations with @xmath83 and @xmath84 .
the initial and boundary conditions as well as the values for all other simulation parameters used in these simulations as same as given in .
the results obtained for both cases after 30,000 time steps are shown in .
[ htbf ] when a liquid is supercooled just below the melting temperature the crystal starts growing and the crystal growth is directly related to the undercooling . depending on the formal undercooling , which quantifies the distance from the phase - equilibrium line in the phase diagram , in the system the final state will have different morphologies .
these are categorised and analysed here in detail to get an overview of the state phase of the apfc model given by the variables @xmath77 and @xmath13 . in this section
the apfc model is used to examine the rate of crystal growth from a supercooled liquid state . as explained above
, @xmath13 represents the undercooling in our model .
a number of simulations with different values of @xmath13 are performed for a specific @xmath77 value .
we used the same initial condition i.e. single solid nucleus ( nucleation site ) for all simulations .
the results showed that the rate of crystal growth increases with an increase in the value of @xmath13 .
as an example the simulation results with @xmath83 and for two typical values of @xmath13 i.e. with @xmath85 and @xmath86 respectively at 40,000 time steps are depicted in .
[ htbf ] in this section we studied the dependence of a stable state phase on anisotropy and undercooling .
more specifically , we show how a stable state can be composed of different phases such as triangular phase , stripe phase and co - existance of stripes and triangular phase , depending on the values of @xmath77 and @xmath13 i.e. anisotropy and undercooling respectively . to study these patterns we performed a number of simulations with different values of @xmath77 and @xmath13 in each simulation .
however , the other parameters for all these simulations are same as described in the previous section .
the results in the form of a diagram of stable state phase are demonstrated in .
[ htbf ] these simulation results demonstrate that the stable state phase always consists of stripes if @xmath87 irrespective of the @xmath13 value .
the co - existence of stripes and triangular phases is found only in case of @xmath84 and @xmath88 , while for other values of @xmath77 and @xmath13 , the stable state consist of a triangular phase .
typical picture of the triangular phase , stripes , and a co - existence of stripes and triangular phase are given in .
[ h ] the precise shape of the grain boundary between the two solid phases depends on the system size due to @xcite . the two - phase coexistence as such
, however , is independent of the size of the system . to analyse the stripe morphology further , we studied the effect of anisotropy and undercooling on the distance between the neighbouring stripes . as discussed in the previous section ,
the stable state consists of stripes when @xmath87 irrespective of the @xmath13 value , as shown in .
it is observed that the stripe phase obtained for different values of @xmath77 are different from each other in terms of the spacing between the neighbouring stripes in the stripe phases .
however , simulations with different values of @xmath13 result in similar stripe phases .
our results reveal that the spacing between neighbouring stripes decreases with an increase in the value of @xmath77 .
thus the stripe phase obtained with @xmath89 are much finer compared to the ones obtained with @xmath90 .
further , to investigate the effect of the undercooling @xmath13 on the stripe phase , we performed several simulations by fixing a specific value for @xmath77 and varying @xmath13 .
the results show that for a specific value of @xmath77 , similar stripe morphologies are obtained with different value of @xmath13 .
demonstrates this interesting finding for different values of @xmath77 and @xmath13 .
[ h ] from the above discussion , we can conclude that the distance between neighbouring stripes decreases as we increase the value of @xmath77 .
however , the undercooling has no significant effect on the stable stripe phase , i.e. the stripe phases for different values of @xmath13 are similar . in this section
we discuss the influence of the anisotropic factor @xmath77 on the final triangular phase obtained with @xmath91 and @xmath92 .
more specifically , we analysed the form of triangles in each of the final triangular phase .
the form of a triangle is determined in terms of three internal angles and ratio of length of the longest and shortest sides of the triangle .
it is observed that the triangular phase obtained in case of @xmath83 i.e. without any anisotropy , consist of triangles with same three angles of @xmath93 each .
however , the triangular phase obtained with non - zero values of @xmath77 contains triangles with dissimilar sides .
the details of the angles calculated for each case are shown in .
[ h ] these results demonstrate that for isotropic case i.e. @xmath83 , the final triangular phase consist of equilateral triangles .
however , for anisotropic case i.e. non zero values of @xmath77 , the final triangular phase contain scalene triangles i.e. no two sides are similar .
the ratio of length of the longest and shortest sides of the triangles , calculated for each case is presented in .
this underlines the capability of our apfc model to give rise to truly anisotropic morphologies .
in this article we presented a ddft based derivation of the apfc model proposed by two of the authors ( d. l. and h. e. ) in @xcite previously .
further we investigated the state phase of this model given by variation of @xmath13 and @xmath77 to demonstrate its capacity to model structures beyond those captured by the spfc equations originally introduced by elder et al .
@xcite .
in particular we studied the influence of anisotropy and undercooling on final states using numerical techniques to minimize the free energy functional in our model .
more specifically , a number of numerical simulations are performed by using different sets of values for our dimensionless parameters @xmath77 and @xmath13 which represent anisotropy and undercooling , respectively .
the results obtained from these numerical simulations are analysed .
our studies reveal that : + ( i ) the rate of crystal growth increases with increase of @xmath13 i.e. undercooling ; + ( ii ) the stable state phase consists of a stripe phase if @xmath87 irrespective of @xmath13
. however , the stable state is a co - existence of stripes and triangular phase when @xmath84 and @xmath88 , while for other values of @xmath77 and @xmath13 , the stable state consists of a triangular phase ; + ( iii ) for @xmath87 , the undercooling @xmath13 has no effect on the resulting stripe phases , however , the distance between the neighbouring stripes decreases with an increase of @xmath77 . + ( iv ) triangular crystals with cells that are neither equilateral nor rhombic are possible for our anisotropic model . in the future we plan to extend the approach to reactive systems @xcite to simulate morpholgenesis in such systems from the atomic to microscale .
we thank s. van teeffelen and t. ala - nissila for helpful discussions . this work has been supported by the dfg through the dfg priority program spp 1296 .
daming li is also supported by national sciences foundation of china ( grant no .
10701056 ) .
8 j. swift and p. c. hohenberg , phys . rev .
a * 15 * , 319 ( 1977 ) .
note that the resulting model equations of our apfc model are not rotationally invariant just like the ones of the famous wbk model @xcite .
this might cause problems when studying for example several grains and issues of their orientation with respect to each other might play an important role . | the anisotropic phase - field crystal model recently proposed and used by prieler et al .
[ j. phys . : condens .
matter * 21 * , 464110 ( 2009 ) ] is derived from microscopic density functional theory for anisotropic particles with fixed orientation . further its morphology diagram is explored .
in particular we investigated the influence of anisotropy and undercooling on the process of nucleation and microstructure formation from atomic to the microscale . to that end numerical simulations were performed varying those dimensionless parameters which represent anisotropy and undercooling in our anisotropic phase - field crystal ( apfc ) model .
the results from these numerical simulations are summarized in terms of a morphology diagram of the stable state phase .
these stable phases are also investigated with respect to their kinetics and characteristic morphological features . |
supersymmetry , if it is realized in nature , must be spontaneously broken on a ground state that is at least metastable , in such a way to comply with the negative experimental searches performed so far . in the context of supergravity theories ,
this has motivated several systematic studies of the circumstances under which metastable non - supersymmetric vacua may arise .
finding such vacua turns out to be surprisingly difficult , even when allowing the freedom of adjusting all the parameters of the theory , because the scalar potential has a restricted functional form in supersymmetric theories .
this difficulty moreover increases with the number of supersymmetries . one general strategy that can be pursued to uncover possible obstructions against metastable supersymmetry breaking in generic supergravity theories consists of studying the general structure of the masses of the sgoldstini , which represent an endemic danger of instability .
for n=1 theories , this study has been performed exhaustively in @xcite ( see also @xcite ) .
the outcome is that metastable de sitter vacua may exist only if the scalar manifold has a suitable curvature . for n=2 theories ,
the same study has only been carried out in certain special classes of theories . in particular
, it has been shown that metastable de sitter vacua are ruled out in theories with only hypermultiplets @xcite or only abelian vector multiplets @xcite . on the other hand , for the more general classes of theories with charged hypermultiplets or non - abelian vector
multiplets a few examples of metastable de sitter vacua are known @xcite , but no constraint on the existence of such vacua has been worked out so far . for n=4 and n=8 theories ,
a similar study has been done in @xcite , where it was proven that metastable de sitter vacua could only arise in a very constrained region of parameter space .
no genuine examples of such vacua are known in this context , but their existence has not yet been excluded ( examples of unstable de sitter vacua @xcite and marginally stable minkowski vacua possessing flat directions @xcite are known ) . finally , the same kind of study has also been applied to theories defined through a truncation that reduces the amount of supersymmetry , where an unstable de sitter vacuum of the original theory may lead to a metastable de sitter vacuum in the truncated theory .
the simplest case of n=2 to n=1 truncations with only scalar multiplets has been studied in @xcite , and several non - trivial examples in the context of n=8 to n=4 and n=4 to n=2 were described in @xcite .
the same issue of vacuum metastability when supersymmetry is spontaneously broken already arises in the simpler context of theories with global supersymmetry .
one may then also study the problem in this simpler context , without losing any of its essential features , at least for n=1 and n=2 theories where , contrary to n=4 and n=8 theories , one can have a generic scalar geometry even in the rigid limit . moreover , for simplicity one may discard constant fayet - iliopoulos terms , since these are essentially an accidental feature of rigid supersymmetry and are only compatible with gravity under quite restrictive circumstances
. this rigid version of the metastability problem was studied in some detail in @xcite , again using the strategy of looking at the masses of the sgoldstini .
the rigid limit of all the results known from the supergravity analyses for generic n=1 and special n=2 theories were recovered in a simpler and more transparent way .
furthermore , it was argued that in n=2 theories with non - abelian vector multiplets there is no obstruction against metastability from the sgoldstini . on the other hand ,
no constraint has been derived so far with this approach for the general case of n=2 theories with charged hypermultiplets . in the context of rigid supersymmetry
, one can also use other types of analyses that better exploit the control that one has over the off - shell theory .
for instance , assuming that metastable supersymmetry breaking is possible and that the only massless states are the goldstini , one can try to explicitly construct their low - energy effective theory in which supersymmetry is realized non - linearly .
any obstruction showing up in this attempt can then be interpreted as signaling the impossibility of realizing the assumed metastable supersymmetry breaking vacuum .
proceeding along these lines , it was shown in @xcite that n=2 theories possessing an @xmath9 global symmetry can not admit a genuinely metastable supersymmetry breaking vacuum , at least under the further , more technical assumption that they admit a well - defined supercurrent superfield satisfying a conservation law that involves at most a superconformal linear anomaly multiplet .
this result strongly suggests that a crucial requirement for a generic n=2 theory to admit a viable metastable non - supersymmetric vacuum is that it should not possess any global @xmath9 symmetry .
remarkably , the presence of such an @xmath9 symmetry rests on radically different features in the hypermultiplet and vector multiplet sectors , and this introduces an important distinction between them . in the vector multiplet sector ,
an @xmath9 symmetry automatically emerges whenever constant fayet - iliopoulos terms are absent , even if the scalar manifold is arbitrarily curved . in the hypermultiplet sector , on the other hand , no @xmath9 symmetry can arise if one considers a sufficiently generic curved scalar manifold . as a consequence ,
the simplest candidates for n=2 theories admitting metastable non - supersymmetric vacua are those involving charged hypermultiplets .
following this expectation , it has been shown in @xcite that the simplest class of such theories based on just one hypermultiplet and one abelian vector multiplet with suitably curved scalar manifolds does indeed admit metastable non - supersymmetric vacua .
the aim of this work is to construct a simple , minimal class of n=2 supergravity theories that admit metastable de sitter vacua , without involving either constant fayet - iliopoulos terms or non - abelian gaugings , by generalizing the construction presented in @xcite from rigid to local supersymmetry .
the setup involves one hypermultiplet and one vector multiplet spanning suitably curved quaternionic - khler and special - khler geometries , with an abelian gauging based on a single triholomorphic isometry .
our aim is to construct the most general model of this type and study the circumstances under which this admits a metastable supersymmetry breaking vacuum .
we shall see how this can be achieved by fixing a point in the scalar manifold and then tuning the geometry in the neighborhood of that point such that it corresponds to a metastable de sitter vacuum . as already argued in @xcite , this possibility could a priori depend on the value of the cosmological constant @xmath0 relative to the gravitino mass squared @xmath1 in planck units , and the relevant dimensionless parameter is therefore expected to be given by the following expression , which we restrict for simplicity to be positive : = .
[ epsilon ] more precisely , when @xmath10 the influence of gravitational effects on scalar masses is negligible .
therefore , by virtue of the results derived in @xcite for the rigid limit , it should be possible to achieve a viable metastable vacuum by adjusting the form of the scalar manifolds . on the other hand , when @xmath11 the influence of gravitational effects on scalar masses is a priori significant and the possibility to achieve a viable metastable vacuum by adjusting the form of the scalar manifolds must be reexamined within supergravity .
the rest of this paper is structured as follows . in section 2
we briefly review the main features of n=2 gauged supergravity theories that will be relevant for our purposes .
in particular , we describe useful explicit parametrizations of the most general quaternionic - khler and special - khler manifolds of minimal dimensions four and two in terms of toda and laplace potentials , respectively , which are the main building blocks of our model . in sections 3 and 4
we then review the known properties of the models with only one hypermultiplet and only one vector multiplet that can be constructed with these spaces , and recall specifically the structure of their scalar mass matrices and the sum rules forbidding the existence of metastable vacua . in section 5 we then explicitly construct the class of models involving one hypermultiplet and one vector multiplet that we are interested in . next we compute the structure of the scalar mass matrix and study the constraints that can be put on its eigenvalues .
finally , we derive a sharp upper bound on the smallest scalar mass as a function of the parameter @xmath12 , and deduce from this the range of values for @xmath12 for which metastable vacua are allowed . in section 6 , we briefly discuss the rigid limit of our supergravity analysis and show how it matches the rigid supersymmetry analysis of @xcite . in section 7 , we then present a class of explicit examples of models admitting metastable de sitter vacua and compute for these the full spectrum of scalar masses . in section 8
we present our conclusions .
let us briefly review the general structure of n=2 gauged supergravity theories , restricting to abelian symmetries and using planck units . in general , there can be @xmath13 hypermultiplets and @xmath14 vector multiplets . the @xmath15 real scalars @xmath16 from the hypermultiplets span a quaternionic - khler manifold with metric @xmath17 while the @xmath14 complex scalars
@xmath18 from the vector multiplets span a special - khler manifold with metric @xmath19 .
the @xmath14 vectors @xmath20 from the vector multiplets and the graviphoton @xmath21 , denoted altogether by @xmath22 , have kinetic metric @xmath23 and topological angles @xmath24 in terms of the so - called period matrix @xmath25 associated with the special - khler manifold .
they can be used to gauge a maximum of @xmath26 of isometries , which are described by triholomorphic killing vectors @xmath27 on the quaternionic - khler manifold .
the scalar and vector kinetic energy is given by @xcite : t = - _ ab f^a _
f^b + _ ab f^a _ f^b - g_uv d_q^u d^q^v - g_i | _ z^i ^|z^| .
[ t ] in this expression @xmath28 , @xmath29 and @xmath30 .
the scalar potential is instead given by v = 2 g_uv k_a^u k_b^v l^a |l^b + g^i | f_i^a |f_|^b p_a^x p_b^x - 3 p_a^x p_b^x l^a |l^b .
[ v ] here @xmath31 denotes the covariantly holomorphic symplectic section of the special - khler manifold and @xmath32 , while @xmath33 denote the three killing potentials admitted by each of the triholomorphic killing vector @xmath27 .
finally , let us also recall that the gravitino mass is given by m^2_3/2 = p_a^x p_b^x l^a |l^b . in the following , we will concentrate on the simplest cases where @xmath13 is @xmath34 or @xmath35 and @xmath14 is @xmath34 or @xmath35 . to construct the most general theories of this type , we will need to consider the most general quaternionic - khler manifold of dimension four with at least one triholomorphic isometry and the most general special - khler manifold of dimension two .
fortunately , there exist general local parametrizations for these two kinds of manifolds , in terms of potentials @xmath36 and @xmath37 of three and two real variables satisfying the toda and the laplace equations , respectively .
it is then possible to construct general theories based on generic choices for these two functions . in the hypermultiplet sector
, we will consider a generic four - dimensional quaternionic - khler space admitting at least one triholomorphic isometry . with the canonical normalization adopted in ( [ t ] ) for the scalar kinetic term ( corresponding to @xmath38 in the notation of @xcite )
, the scalar curvature must be @xmath39 .
it turns out that the line element of such a manifold can always be locally brought into the following przanowski - tod form @xcite , using four real coordinates @xmath40 : d s^2 = g_uv dq^u dq^v = 12 ^2 ( f d^2 + f e^h ( d^2 + d^2 ) + f^1 ( d+ ) ^2 ) .
this depends on a single function @xmath36 of the three variables @xmath41 , which must satisfy the three - dimensional toda equation : [ toda - h ] ( e^h ) _ + h _
the function @xmath42 is then related to the function @xmath36 by f = 2 - h_. the @xmath35-form @xmath43 is instead determined , modulo an irrelevant exact form , by the following equation , whose integrability is guaranteed by the toda equation : d = ( f _ d- f _ d ) d+ ( f e^h ) _ dd .
further details about the geometric properties of this space can be found for example in @xcite .
the @xmath44 connection @xmath45 is found to be : ^1 = - ^1 e^h/2 d , + ^2 = - ^1 e^h/2 d , + ^3 = - ^1 ( d+ ) + h_d- h_d .
the three hyper - khler forms @xmath46 , which satisfy @xmath47 as a consequence of the equation defining @xmath43 and are thus only covariantly closed , are given by : j^1 = ^2 e^h/2d(d+ ) - ^2 f e^h/2 dd , + j^2 = - ^2 e^h/2 d(d+ ) - ^2 f e^h/2 dd , + j^3 = ^2 d(d+ ) + ^2 f e^h dd . finally , the manifest isometry amounts to a shift in the variable @xmath48 and the corresponding killing vector takes the following expression , involving an arbitrary parameter @xmath49 of dimension one : k = _ . in the above parametrization in terms of the four
real coordinates @xmath50 , the components of the metric are : g_uv = 12 ^2 ( f + f^1 _ ^2 & f^1 _ _ & f^1 _ _ & f^1 _ + f^1 _ _ & f e^h + f^1 _ ^2 & f^1 _ _ & f^1
_ + f^1 _ _ & f^1 _ _ & fe^h + f^1 _ ^2 & f^1 _ + f^1 _ & f^1 _ & f^1 _ & f^1 + ) .
positivity of this metric requires @xmath51 .
a simple choice for the vielbein @xmath52 that can be used to locally trivialize this metric as @xmath53 is given by : e_u^p = 1 ( f^1/2 & 0 & 0 & f^1/2 _ + 0 & f^1/2 e^h/2 & 0 & f^1/2 _ + 0 & 0 & f^1/2 e^h/2 & f^1/2 _ + 0 & 0 & 0 & f^1/2 ) . the components of the three hyper - khler forms are also easily worked out and seen to satisfy the algebra @xmath54 .
one can then easily verify that the killing vector with components @xmath55 is triholomorphic and compute the corresponding killing potentials @xmath56 from their defining relation @xmath57 .
the result is : k^u = ( 0 + 0 + 0 + ) , |p| = ^1 .
[ killingpt ] it will be useful for the analysis of the forthcoming sections to define the following three - dimensional matrix of dimensionless parameters associated to the second derivatives of the function @xmath42 with respect to its variables @xmath41 : _ u v = ( e^1)_u^p f_p q ( e^1t)^q_v . [
a ] a simple computation shows that the entries of this matrix are given by _
= f^1 ^2 f _ , + _
= f^1 ^2 f _ e^h/2 , _ = f^1 ^2 f _ e^h/2 , + _
= f^1 ^2 f _ e^h , _ = f^1 ^2 f _ e^h , _ = f^1 ^2
f _ e^h . the definitions and properties of the functions @xmath36 and
@xmath42 imply an important constraint on the second derivatives of @xmath42 , and thus on the parameter @xmath58 .
more precisely , using the relation between @xmath42 and @xmath36 and the toda equation ( [ toda - h ] ) satisfied by @xmath36 , one easily shows that @xmath59 . as a consequence of this property ,
it follows that the trace of the matrix @xmath58 involving the second derivatives of @xmath42 is completely fixed in terms of the first derivatives of @xmath42 and the function @xmath42 itself , and one finds : ^u v _ u v = - ( f-1)(f-2 ) + ( 3 f - 4 ) f^1 f_. [ constrainta ] in the vector multiplet sector , we will consider a general two - dimensional special - khler manifold of the local type .
this can be locally described by a metric of the following form , using the coordinates @xmath60 : ds^2 = g_| dq^d q^| = 2 g_z |z dz d |z = 2l |d z|^2 .
the line element depends on a single real function @xmath37 of @xmath61 and @xmath62 .
this function is restricted by the fact that it can be expressed in terms of a holomorphic prepotential .
the resulting restriction on the curvature of the space can be taken into account efficiently in terms of the symplectic section @xmath63 , where @xmath64 . in the local case ,
this section is defined to be covariantly holomorphic and thus satisfies : _
|z l^a = 0 .
moreover , the constrained form of the geometry implies two special properties for the two possible kinds of second covariant derivatives of the section .
the first is that @xmath65 l^a = g_{z \bar z } l^a$ ] .
this implies that _ |z _ z l^a = l l^a .
the second is that @xmath66 , where @xmath67 is covariantly holomorphic and thus satisfies @xmath68 .
acting with a further derivative on this relation we then deduce that @xmath69 . multiplying this equation with @xmath70 and using the previous relation ,
one finally finds that _ |z _
z^2 l^a _ |z |l^b = _
z^2 l^b _ |z^2 |l^a . in the above parametrization in terms of the coordinates @xmath71 , the components of the metric are given by : g_| = ( l & 0 + 0 & l + ) .
an obvious choice for the vielbein @xmath72 that allows the metric to be locally trivialized as @xmath73 is given by : e_^= ( l^1/2 & 0 + 0 & l^1/2 + )
. it will be convenient for the analysis of the next sections to introduce the following dimensionless parameter related to the third derivatives of the section : = .
[ b ] it will also be useful to define the following quantity : = - .
in models with just one hypermultiplet , the only possible source of potential comes from a gauging of the isometry with the graviphoton .
this corresponds to taking @xmath74 and @xmath75 , where @xmath55 and @xmath56 are given by eq .
( [ killingpt ] ) .
furthermore , we can choose @xmath76 .
the resulting potential is @xmath77 and only depends on the three variables @xmath41 .
its explicit form reads v = ^2 ^2 ( f^1 - ) .
the first derivatives @xmath78 are given by @xmath79 and v_u = ^2 ^2 .
the second derivatives @xmath80 are instead given by @xmath81 , @xmath82 and v_u v = ^2 ^2 .
the gravitino mass reads : m_3/2 ^ 2 = ^2 ^2 .
it follows that the parameter ( [ epsilon ] ) takes the value = - 3 + 4 f^1 .
[ epsilonh ] we see from these formulae that the strength of supersymmetry breaking is controlled by the value of @xmath83 . in order to study the possible vacua and their properties , we must first find the critical points of the scalar potential .
the value of the cosmological constant @xmath0 fixes the value of @xmath42 to be given by : f = 43 + .
the stationarity conditions @xmath84 fix instead the values of the first derivatives @xmath85 to be given by @xmath86 , and since @xmath87 this gives : f_u = - ^1 _ u .
[ stath ] using the above relations , one can then compute the hessian matrix @xmath88 and the the physical scalar mass matrix @xmath89 in terms of the values of the second derivatives @xmath90 , which are related to the parameters @xmath58 by the definition ( [ a ] ) . this
is found to be given by @xmath91 , @xmath92 and m^2_u v = m_3/2 ^ 2 .
the parameters @xmath58 can be adjusted by suitably choosing the function @xmath42 and therefore the values of its second derivatives at the point under consideration .
the only constraint is that the corresponding function @xmath36 should solve the toda equation .
this then results in the constraint ( [ constrainta ] ) which implies the following sum rule , after using the stationarity condition ( [ stath ] ) : ^u v _ u v = . [ constrainth ] to check whether the scalar masses can all be positive , we may now compute the average of the three eigenvalues of the physical mass matrix : m^2 ^u v m^2_u v .
it turns out that the value of this average mass is entirely fixed by the constraint ( [ constrainth ] ) and is found to be : m^2 = m^2_3/2 .
[ m2onlyhyper ] by construction @xmath93 represents an upper bound on the smallest mass eigenvalue and also a lower bound on the largest mass eigenvalue .
moreover , both of these bounds can be saturated by tuning the parameters @xmath58 .
we then conclude that : \{m^2_i } - m_3/2 ^ 2 - v .
this result implies that minkowski and de sitter vacua are necessarily unstable .
it reproduces the result derived in @xcite , which looked at the average sgoldstino mass for theories with an arbitrary number of hypermultiplets .
in models with just one vector multiplet , the only possible source of potential comes from a fayet - iliopoulos term associated to constant killing potentials for an arbitrary linear combination of the graviphoton and the matter vector . without loss of generality , we can choose @xmath94 and @xmath95 , where @xmath49 is a constant and @xmath96 is an arbitrary unit vector .
the potential is then independent of @xmath97 and only involves @xmath98 .
it is given by @xmath99 and depends on the two variables @xmath100 .
its explicit form reads v = ^2 .
the first derivatives @xmath101 are found to be : v_z = ^2 .
the second derivatives @xmath102 are instead given by v_z |z = ^2 , + v_z z = ^2 .
the gravitino mass reads : m_3/2 ^ 2 = ^2 |l|^2 .
it follows that the parameter ( [ epsilon ] ) takes the value = - 3 + l^1 ||^2 .
we see from these formulae that the strength of supersymmetry breaking is controlled by the value of @xmath103 . in order to study the possible vacua and their properties
, we must again first find the critical points of the scalar potential .
the value of the cosmological constant @xmath0 fixes the value of @xmath104 to be given by _
z l = l^1/2 e^- i l , where @xmath105 was defined in ( [ gamma ] ) . the stationarity conditions @xmath106 fix instead the value of @xmath107 to by given by _
z^2 l = 2le^2
i l . using the above relations , one
can then compute the hessian matrix @xmath108 and the physical scalar mass matrix @xmath109 in terms of the value of @xmath110 , which is related to the parameter @xmath111 by the definition ( [ b ] ) .
this gives : m^2_z m^2_z z = m^2_3/2 e^2 i .
the parameter @xmath111 can be adjusted to any desired value by changing the function @xmath37 and therefore the value of its second derivative at the point under consideration , without any constraint . to check whether the scalar masses can all be positive
, we can now compute the average of the two eigenvalues of the physical mass matrix : m^2 ^| m^2_| .
we see that this is completely fixed , independently of the parameter @xmath111 , and reads m^2 = m^2_3/2 .
[ m2onlyvec ] by construction @xmath93 represents an upper bound on the smallest mass eigenvalue and also a lower bound on the largest mass eigenvalue .
moreover , both of these bounds can be saturated by tuning the parameter @xmath111 . from this
we deduce that \{m^2_i } -2 v .
this result implies that de sitter vacua are necessarily unstable .
it reproduces the result derived in @xcite , which looked at the average sgoldstino mass for theories involving an arbitrary number of abelian vector multiplets .
in models with one hypermultiplet and one vector multiplet , the source of potential comes from a gauging of the hypermultiplet isometry with a linear combination of the graviphoton and the matter vector . without loss of generality , we can choose @xmath112 , @xmath113 and @xmath94 , @xmath114 , where @xmath55 and @xmath56 are given by eq .
( [ killingpt ] ) .
once again , the potential involves @xmath98 but not @xmath97 .
it is given by @xmath115 and depends only on the five variables @xmath116 .
its explicit expression is v = ^2 ^2 .
the first derivatives @xmath117 are given by @xmath79 and v_u = ^2 ^2 , + v_z = ^2 ^2 .
the second derivatives @xmath118 , defined to be ordinary and covariant in the hyper and vector sectors , respectively , are instead given by @xmath81 , @xmath119 and v_u v = ^2 ^2 , + v_z |z = ^2 ^2 , + v_z z = ^2 ^2 , + v_u z = ^2
the gravitino mass reads : m_3/2 ^ 2 = ^2 ^2 |l|^2 .
it follows that the parameter ( [ epsilon ] ) is given by = - 3 + 4 f^1 + l^1 | breaking originating from the hyper and vector sectors is controlled by the expectation values of @xmath83 and @xmath103 , respectively .
the relative importance of these two contributions is most conveniently parametrized by an angle @xmath120 , defined as : = || . in order to study the possible vacua and their properties
, we must again first find the critical points of the scalar potential .
the values of the cosmological constant @xmath0 and the angle @xmath120 fix the values of @xmath42 and @xmath104 to : f = 43 + ^2 [ cos ] , + _ z l = l^1/2 e^- i l [ sin ] , where we recall again that @xmath105 was defined in ( [ gamma ] ) . the stationarity conditions @xmath121 fix instead the values of @xmath85 and @xmath107 to by given by @xmath122 and @xmath123 , and since @xmath124 this gives : f_u = - ^4 ^1 _ u [ stathv1 ] , + _
z^2 l = - l ( ( 3 + ) ^2 - 2)e^2 i l [ stathv2 ] . using the above relations , one
can then compute the hessian matrix @xmath125 and the physical scalar mass matrix @xmath126 in terms of the values of @xmath90 and @xmath110 , which are related to the parameters @xmath58 and @xmath111 by the definitions ( [ a ] ) and ( [ b ] ) .
this is found to be given by @xmath91 , @xmath92 , @xmath127 and m^2_u v = m^2_3/2 , + m^2_z |z = m^2_3/2 , + m^2_z z = m^2_3/2
e^2 i , + m^2_u z = m^2_3/2 e^i . the parameters @xmath58 can be adjusted by suitably choosing the function @xmath42 and therefore the values of its second derivatives at the point under consideration .
the only constraint comes from the toda equation for @xmath36 , which can be rewritten as in ( [ constrainta ] ) . after using the stationarity condition ( [ stathv1 ] ) this in turn
implies the following sum rule : ^u v _ u v = .
[ constrainthv ] the parameter @xmath111 can instead be adjusted by suitably choosing the function @xmath37 and the associated section , and is completely arbitrary .
finally , the angle @xmath120 can be adjusted by suitably choosing the relative overall sizes of @xmath42 and @xmath37 . to check whether the scalar masses can all be positive
, we can now study the two - dimensional matrix obtained by averaging over the three non - trivial directions in the hyper sector and the two directions in the vector sector .
this real symmetric matrix takes the form m^2 ( m^2_hh & m^2_hv + m^2_hv & m^2_vv + ) , [ m2avr ] and its three independent elements are defined by m^2_hh ^u v m^2_u v , m^2_vv ^| m^2_| , m^2_hv .
it turns out that the values of these average entries of the mass matrix are entirely fixed , as in the previous cases , and are found to be m^2_hh = m^2_3/2 [ m2hh ] , + m^2_vv = m^2_3/2 [ m2vv ] , + m^2_hv = m^2_3/2 [ m2hv ] .
note that in the hyper sector one correctly recovers @xmath128 when @xmath129 , matching the case with only one hypermultiplet ( [ m2onlyhyper ] ) , and one finds instead @xmath130 when @xmath131 .
similarly , in the vector sector one correctly recovers @xmath132 when @xmath131 , matching the case with only one vector multiplet ( [ m2onlyvec ] ) , and one finds instead @xmath133 when @xmath129 .
finally , for the mixing between the two sectors , one finds @xmath134 when either @xmath129 or @xmath131 . to get the sharpest possible bounds on the mass eigenvalues for the case we consider here , it is natural to consider the two eigenvalues of the two - dimensional averaged mass matrix ( [ m2avr ] ) .
these are given by m^2 _
= ( m^2_hh + m^2_vv ) . using the results ( [ m2hh ] ) , ( [ m2vv ] ) and ( [ m2hv ] ) ,
these two eigenvalues are found to be of the form m^2 _
= m^2_3/2 , [ m2 + - ] where : x = ( 3 + ) ^2 ^4 - ( 3 + ) ( 6 + ) ^2 + ( 4 + 3 + 2 ^2 ) , [ x ] + y = ( 3 + ) ^4 ^8 + ( 3 + ) ^3 ^6 - ( 3 + ) ^2 ( 4 + 9 - 3 ^ 2 ) ^4 + - ( 3 + ) ( 16 + 27 + 5 ^2 ) ^2 + ( 4 + 9 + 2 ^2)^2 .
[ y ] we can now study the behavior of @xmath135 as functions of @xmath120 for given value of @xmath12 . for @xmath136 ,
one finds a local maximum for some finite value of @xmath120 , with a magnitude that is negative or positive depending on whether @xmath12 is smaller or larger than a certain critical value , as illustrated in figure 1 .
more precisely , one finds the following results .
for @xmath137 , @xmath136 is everywhere negative , except at @xmath138 where it vanishes .
this implies that minkowski vacua ( @xmath139 ) can be at most loosely metastable . for @xmath140 , @xmath136 is everywhere negative , and de sitter vacua with @xmath141 are thus necessarily unstable . for @xmath142 , @xmath136 is everywhere negative except at @xmath143 where it vanishes , and de sitter vacua with @xmath144 can thus be loosely metastable .
finally for @xmath145 , @xmath136 becomes positive for a finite range of values of @xmath120 , centered around a value between @xmath146 and @xmath147 where the maximum occurs .
therefore de sitter vacua with @xmath148 can be genuinely metastable . for @xmath149
, one finds a local minimum for some finite value of @xmath120 , with a magnitude that is always positive . as a function of @xmath120 for various values of @xmath12 , in units of @xmath1.,scaledwidth=60.0% ] by construction @xmath136 represents an upper bound on the smallest mass eigenvalue and @xmath149 a lower bound on the largest mass eigenvalue , for given values of @xmath120 and @xmath12 . from these quantities
one can then derive two other bounds that only depend on @xmath12 by suitably extremizing @xmath136 and @xmath149 over @xmath120 , for fixed @xmath1 and @xmath0 .
more precisely , we can compute : m^2_up _ \{m^2_- } , + m^2_low _ \{m^2_+ } .
one can then see that @xmath150 and @xmath151 .
moreover , any of these bounds can be saturated by tuning the parameters @xmath58 , @xmath111 and @xmath120 .
we can now study the behavior of @xmath152 as functions of @xmath12 . for @xmath153 ,
one finds a non - monotonic function that first decreases and is negative and then increases and becomes positive , as shown in figure 2 . for @xmath154 ,
one finds instead a monotonically increasing function that is always positive . as a function of @xmath12 , in units of @xmath1.,scaledwidth=60.0% ] to summarize , we see that when @xmath0 is positive and small compared to @xmath1 , as required for particle phenomenology , the vacuum is necessarily unstable , but when @xmath0 is positive and large compared to @xmath1 , as could be desirable for inflation , the vacuum can be metastable . in these two limits of small and large cosmological constant , one can actually derive some simpler result for the bound on the smallest mass eigenvalue
. this will also allow us to describe the general case of intermediate cosmological constant in a simpler , qualitatively way .
in the limit where @xmath12 is small , which corresponds to a situation where the cosmological constant is made small through a tuning of parameters , the entries of the two - dimensional averaged mass matrix are given by : m^2_hh m^2_3/2 , [ mhhsmall ] + m^2_vv m^2_3/2 , [ mvvsmall ] + m^2_hv m^2_3/2 .
[ mhvsmall ] when @xmath155 , the two eigenvalues @xmath135 of ( [ m2avr ] ) take degenerate and vanishing extremal values @xmath152 for the same angle @xmath156 given by @xmath157 . when @xmath158 , the two eigenvalues @xmath135 of ( [ m2avr ] ) instead take non - degenerate and non - vanishing extremal values @xmath152 for two slightly different angles
@xmath159 . to compute the approximate values of @xmath152 for small @xmath12 , one can then further expand the entries ( [ mhhsmall ] ) , ( [ mvvsmall ] ) and ( [ mhvsmall ] ) at @xmath159 up to first order in @xmath160 , then compute the eigenvalues and finally extremize them with respect to the parameter @xmath161 .
proceeding in this way one finds : m^2_up - m^2_3/2 , + m^2_low m^2_3/2 .
it follows that in this regime : \{m^2_i } - v .
this quantifies the extent to which de sitter vacua with a small cosmological constant are unstable in this class of models . in the limit where @xmath12 is large , which includes the situation where the main contribution to the vacuum energy is non - gravitational , the entries of the two - dimensional averaged mass matrix are instead given by : m^2_hh ^2 m^2_3/2 , + m^2_vv ^2 m^2_3/2 , + m^2_hv ^2 m^2_3/2 .
after computing @xmath162 , one can check that @xmath136 develops a maximum for @xmath163 and @xmath149 develops a minimum for @xmath164 , with the values m^2_up
^2 m_3/2 ^ 2 , + m^2_low 1.05 ^2 m_3/2 ^ 2 .
it follows that in this regime : \{m^2_i } .
this quantifies the extent to which de sitter vacua with a large cosmological constant are stable in this class of models . in more general situations where @xmath12 is neither very large nor very small
, one has in principle to use the exact entries of the two - dimensional averaged mass matrix .
the extrema @xmath153 and @xmath154 must then be computed by extremizing the corresponding complicated expressions ( [ m2 + - ] ) for @xmath136 and @xmath149 .
however , it turns out that the exact results for these bounds are reasonably well approximated for any value of @xmath12 by just adding up the behaviors that we derived for small and large values of @xmath12 .
one can in fact write : m^2_up
< ( - + ^2 ) m^2_3/2 , + m^2_low > ( + 1.05 ^2)m^2_3/2 .
it follows in particular that : \{m^2_i } < - v + .
this quantifies in a simpler but weaker way the situation for de sitter vacua with an intermediate cosmological constant in this class of models .
note that this alternative bound can no longer be strictly saturated , and in relation to this the transition between positive and negative values on the right - hand side occurs at @xmath165 rather than @xmath142 .
let us briefly study what happens in the rigid limit , when gravitational effects are negligible , in order to compare our results with those of @xcite . in this decoupling limit @xmath166 while @xmath0 stays finite but
is dominated by non - gravitational effects .
one is then in the situation where @xmath167 , and the entries of the mass matrix simplify .
moreover , all these entries take the form of a finite coefficient depending on the angle @xmath120 times the squared energy scale @xmath168 , which stays finite in the limit .
more precisely , one finds m^2_u v = m^2_a , [ muvrigid ] + m^2_z |z = m^2_a , [ mzzbrigid ] + m^2_z z = m^2_a e^2 i , [ mzzrigid ] + m^2_u z = m^2_a e^ i [ muzrigid ] , where @xmath169 , which using the relations ( [ cos ] ) and ( [ sin ] ) is recognized to correspond to the vector mass in the rigid limit , namely m^2_a = ^2 ^2 f^1 l^1 |_z l|^2 . notice also that in the rigid limit the constraint that the parameters @xmath58 must satisfy simplifies to ^u v _
u v = 6 + 8 ^2 .
the averages of the above blocks are then found to be : m^2_hh = ( 1 + ^2 ) m^2_a , [ m2hhrigidlimit ] + m^2_vv = ( 1 + 2 ^2 ) m^2_a , [ m2vvrigidlimit ] + m^2_hv = ( + ) m^2_a .
[ m2hvrigidlimit ] notice finally that in the rigid limit it is more meaningful to study the extrema of @xmath135 with respect to @xmath120 at fixed @xmath170 . by doing so
, one finds that the extrema of @xmath135 both occur at @xmath171 , with values @xmath170 and @xmath172 .
this implies in particular that \{m^2_i } m^2_a . at this point
, we can compare the above formulae with the results derived in @xcite in rigid supersymmetry , where the przanowski - tod quaternionic - khler manifold reduces to a gibbons - hawking hyper - khler manifold @xcite and the local special - khler manifold reduces to a global special - khler manifold .
we see that the averaged masses ( [ m2hhrigidlimit ] ) , ( [ m2vvrigidlimit ] ) and ( [ m2hvrigidlimit ] ) exactly match the corresponding results in @xcite .
this is expected , since these quantities are , by definition , independent of any parameter and any coordinate choice . on the other hand ,
the non - averaged masses ( [ muvrigid ] ) , ( [ mzzbrigid ] ) , ( [ mzzrigid ] ) and ( [ muzrigid ] ) can be compared with those in @xcite only after taking into account the differences in the coordinates used here and in @xcite to describe the scalar manifold and the associated parameters . in the hypermultiplet sector
, the parameters @xmath58 defined here must map to the parameters @xmath173 defined in @xcite .
however , the coordinates @xmath174 used here are related in a non - trivial way to the coordinates @xmath175 used in @xcite . by comparing the form of the metric , the killing vector and the killing potentials for the two spaces in the ricci - flat limit where they should coincide , one can see that @xmath176 is related to @xmath177 , while @xmath178 and @xmath179 are related to the two angular variables @xmath180 describing the orientation of the vector @xmath181 , and finally @xmath182 is related to @xmath48 .
moreover , the function @xmath42 in the przanowski - tod metric is directly related to the function @xmath42 in the gibbons - hawking metric by a simple rescaling involving two powers of the radial coordinate .
the precise relation between the complete @xmath58 and the complete @xmath173 is then not totally straightforward to determine , due to this non - trivial change of coordinates involving the planck scale ( see @xcite for a related discussion in a specific example ) .
for this reason , we will not attempt to compare more explicitly the non - averaged mass matrix in this sector with the results of @xcite . in the vector multiplet sector ,
the parameter @xmath111 defined here must map to the parameter @xmath183 defined in @xcite .
moreover , the coordinate @xmath61 used here maps to the coordinate @xmath61 used in @xcite .
similarly , the function @xmath37 used here maps to the function @xmath37 in @xcite .
the precise relation between @xmath111 and @xmath183 is thus straightforward to determine .
using special coordinates and taking the rigid limit ( in which the khler connection drops out from the covariant derivatives ) the symplectic section simply reads @xmath184 , and one finds that @xmath185 .
taking into account this relation , the structure of the non - averaged mass matrix in this sector then precisely matches the results of @xcite in the rigid limit , as it should .
in order to illustrate the general statements of the previous sections , let us now study a class of more explicit examples and compute the full spectrum of their mass eigenvalues . for concreteness and simplicity , we shall focus on a family of solutions where
the eigenvalues of the mass matrix can be computed analytically .
more precisely , let us keep @xmath120 and @xmath12 arbitrary , but take all the real parameters @xmath58 to be correlated and controlled by a single real parameter @xmath186 and similarly choose the complex parameter @xmath111 to be controlled by a single real parameter @xmath187 : _
= , + _ = x , _ = x , _ _ = 0 , _ = 0 , _ = 0 , + = y . for simplicity we shall also set @xmath188 , as this does not affect the eigenvalues . with this choice of parameters
the mass matrix takes the form m^2_i
|j = _ i |j m^2_3/2 , where the non - trivial entries of the matrix @xmath189 are given by _ =
( 3 + ) ^2 ( 1 + x ) ^4 - ( 3 + ) ( 6 + 5 ) ^2 + 4 ( 1 + ) ( 2 + ) , + _ = - ( 3 + ) ^2 x ^4 , _ = - ( 3 + ) ^2 x ^4 , + _ = 0 , _ = 0 , _ = 0 , + _ z |z = ( ( 3 + ) ^2 - 2 ) ( ( 3 + ) ^2 + ) , + _ z z = ( 3 + ) ^2 ^4 y - ( ( 3 + ) ^2 - 1 ) ( ( 3 + ) ^2 - 2 ) , + _ z = ( 3 + ) , _ z = 0 , _ z = 0 .
the five non - trivial masses are then given in terms of the eigenvalues @xmath190 of the non - trivial block @xmath191 of this matrix : m^2_i = _
i m^2_3/2 . in this simple four - parameter special family of vacua , the eigenvalues @xmath190
can be computed analytically as functions of the parameters @xmath120 , @xmath12 , @xmath186 and @xmath187 characterizing the scalar geometry in the vicinity of the vacuum point .
they are found to be _
1,2 = - ( 3 + ) ^2 x ^4 , + _
3 = - ( 3 + ) ^2 y ^4 + ( ( 3 + ) ^2 -2 ) ( 2 ( 3 + ) ^2 -1 + ) , + _
4,5 = p , where p = ( 3 + ) ^2 y ^4 + ( 3 + ) ^2 ( 1 + x ) ^4 + - ( 3 + ) ( 5 + 4 ) ^2 + ( 1 + ) ( 3 + 2 ) , + q = ( 3 + ) ^4 y^2 ^8 - ( 3 + ) ^4 y ( 1 + x ) ^4 ^4 + + ( 3 + ) ^3 y ( 7 + 6 ) ^4 ^2 - ( 3 + ) ^2 y ( 1 + ) ( 5 + 2 ) ^4 + + ( 3 + ) ^4 ( 1 + x)^2 ^8 - ( 3 + ) ^3 ( 1 + x ) ( 7 + 6 ) ^6 + + ( 3 + ) ^2 ( 69 + 112 + 28 ^2 + 4 x ( 1 + ) ( 5 + 2 ) ) ^4 + - ( 3 + ) ( 35 + 79 + 44 ^2 + 8 ^3 ) ^2 + ( 1 + ) ^2 ( 5 + 2 ) ^2 . using the same kind of notation , we also denote m^2_up = _ up m^2_3/2 , + m^2_low = _ low m^2_3/2 . in this class of models , it is straightforward to verify all the statements of the previous section concerning the range that the mass eigenvalues are allowed to take . for any given value of @xmath12 , one may make the vacuum as stable as possible by first adjusting @xmath120 to the optimal value that allows the upper bound @xmath192 on the smallest eigenvalue to be maximized , and then adjusting @xmath186 and @xmath187 to saturate this value .
it turns out that this best situation occurs when @xmath193 , while @xmath194 is always bigger . in table 1
we list some sample models illustrating this point .
we see that in this simple class of models where only two real parameters are retained among @xmath58 and @xmath111 , there are thus always four of the five eigenvalues that become degenerate when the parameters are adjusted in such a way to saturate the bound defined by @xmath192 . in more general models where additional independent parameters are retained among @xmath58 and @xmath111 , this feature is expected to disappear .
in this work we have shown that metastable de sitter vacua may arise in rather simple n=2 supergravity theories with a single hypermultiplet and a single vector multiplet , without fayet - iliopoulos terms or non - abelian gauge symmetries , provided that the scalar manifold is suitably curved . a first , crucial requirement for this to be possible
is that both sectors should be involved in the process of supersymmetry breaking , as it is known that in theories with only one hypermultiplet or only one vector multiplet no metastable vacua can ever arise . a second , important
limitation is that the cosmological constant @xmath0 should be sufficiently large compared to the gravitino mass squared @xmath1 in planck units .
more precisely , for positive @xmath195 we showed that metastable de sitter vacua are only possible if @xmath145 .
as a consequence , the simple de sitter vacua that we constructed can realize slow - roll inflation with a large hubble parameter corresponding to @xmath10 but not a late - time vacuum with small cosmological constant corresponding to @xmath196 . in the simple class of models that we have studied , the quaternionic - khler manifold describing the hypermultiplet sector is a generic four - dimensional przanowski - tod space possessing at least one triholomorphic isometry .
the special - khler manifold describing the vector multiplet sector is a totally general two - dimensional special space and the potential is generated by gauging the isometry of the hypermultiplet geometry .
for this class of models , we were able to derive a simple upper bound on the mass of the lightest scalar , which depends only on the angle @xmath120 defining the spread of the supersymmetry breaking direction between the two sectors and the parameter @xmath12 related to the cosmological constant . in the particular cases where only the hypermultiplet or only the vector multiplet is involved in supersymmetry breaking , corresponding to @xmath129 and @xmath131 respectively , one recovers for any @xmath12 the results of @xcite and @xcite , namely the saturable bounds : \{m^2_i } - m^2_3/2 - v , [ hyp ] + \{m^2_i } - 2 v .
[ vec ] in the more general case where both the hyper- and vector multiplets are involved in supersymmetry breaking , there exists for any given @xmath12 an optimal choice for @xmath120 that maximizes the smallest mass .
this allows us to define a bound depending only on @xmath12 for the situation where the two sectors are optimally mixed .
the exact result for the saturable upper bound on @xmath197 as a function of @xmath12 is plotted in figure 2 but can not be expressed in any simple , analytic form .
however , we showed that the following simpler bound , which is slightly weaker and non - saturable , is also true and gives a good approximation to the exact bound : \{m^2_i } < - v + .
[ hypvec ] notice that on the right - hand side of this bound , the transition between positive and negative values arises at @xmath198 , but as already said the precise critical point for which metastable vacua really become possible is @xmath142 .
it is straightforward to generalize the analysis presented in this paper to study the possibility of getting metastable supersymmetry breaking anti de sitter vacua with a negative cosmological constant @xmath199 in the same setting .
in fact , the equations ( [ m2 + -])([y ] ) defining the bounds @xmath135 on the scalar masses hold true unchanged also in the case of negative @xmath12 , and one can repeat the same analysis as for positive @xmath12 . the maximal value @xmath153 that @xmath136 can take is found to be positive for @xmath200\!-\!3,-\frac 32 [ \,\ , \cup \,\ , ] \!-\!1,0[\,$ ] and negative for @xmath200\!-\!\frac 32,1[$ ] , but always larger than the breitenlohner
freedman bound @xmath201 for stability @xcite .
this implies that metastable supersymmetry breaking anti de sitter vacua may exist for any @xmath202\!-\!3 , 0[\,$ ]
. it would be interesting to investigate whether the results found in this paper for models with a single hypermultiplet and a single vector multiplet can be extended to theories involving several of these multiplets , at least in the simplest case of abelian gaugings without constant fayet - iliopoulos terms .
a natural way to try to do this is to look at the mass matrix in the subspace of scalar fields defined by the various sgoldstini , along the lines of @xcite .
indeed , this approach allows one to extend the results ( [ hyp ] ) and ( [ vec ] ) to theories with an arbitrary number of hypermultiplets and an arbitrary number of vector multiplets , respectively . in this way , one might also hope to derive a result similar to ( [ hypvec ] ) for theories with several hyper- and several vector multiplets .
a first hint in favor of this comes from the fact that the known examples of metastable de sitter vacua in such theories @xcite do indeed satisfy this type of bound , albeit in a rather trivial way , thanks to the fact that they all lead to @xmath203 .
we leave a thorough analysis of this general problem for future work .
* acknowledgements * we would like to thank g. dallagata and m. trigiante for useful comments and discussions . the research of c. s. and p. s. is supported by the swiss national science foundation ( snsf ) under the grant pp00p2 - 135164 , and that of f. c. by the german science foundation ( dfg ) under the collaborative research center ( sfb ) 676 `` particles , strings and the early universe '' . m. gomez - reino and c. a. scrucca , _ locally stable non - supersymmetric minkowski vacua in supergravity _ , jhep * 0605 * ( 2006 ) 015 [ hep - th/0602246 ] .
m. gomez - reino and c. a. scrucca , _ constraints for the existence of flat and stable non - supersymmetric vacua in supergravity _ , jhep * 0609 * ( 2006 ) 008 [ hep - th/0606273 ] . m. gomez - reino and c. a. scrucca , _ metastable supergravity vacua with f and d supersymmetry breaking _ ,
jhep * 0708 * ( 2007 ) 091 [ arxiv:0706.2785 ] .
f. denef and m. r. douglas , _ distributions of nonsupersymmetric flux vacua _ , jhep * 0503 * ( 2005 ) 061 [ hep - th/0411183 ] .
l. covi , m. gomez - reino , c. gross , j. louis , g. a. palma and c. a. scrucca , _ de sitter vacua in no - scale supergravities and calabi - yau string models _ ,
jhep * 0806 * ( 2008 ) 057 [ arxiv:0804.1073 ] .
l. covi , m. gomez - reino , c. gross , j. louis , g. a. palma and c. a. scrucca , _ constraints on modular inflation in supergravity and string theory _ ,
jhep * 0808 * ( 2008 ) 055 [ arxiv:0805.3290 ] .
p. fr , m. trigiante and a. van proeyen , _ stable de sitter vacua from n = 2 supergravity _ , class .
quant . grav .
* 19 * ( 2002 ) 4167 [ arxiv : hep - th/0205119 ] .
o. ogetbil , _ stable de sitter vacua in 4 dimensional supergravity originating from 5 dimensions _ , phys
. rev .
d * 78 * ( 2008 ) 105001 [ arxiv:0809.0544 [ hep - th ] ] .
m. gomez - reino , j. louis and c. a. scrucca , _ no metastable de sitter vacua in n=2 supergravity with only hypermultiplets _
, jhep * 0902 * ( 2009 ) 003 [ arxiv:0812.0884 ] .
e. cremmer , c. kounnas , a. van proeyen , j. p. derendinger , s. ferrara , b. de wit and l. girardello , _ vector multiplets coupled to n=2 supergravity : superhiggs effect , flat potentials and geometric structure _ ,
b * 250 * ( 1985 ) 385 .
a. borghese and d. roest , _ metastable supersymmetry breaking in extended supergravity _ , jhep * 1105 * ( 2011 ) 102 [ arxiv:1012.3736 ] .
a. borghese , r. linares and d. roest , _ minimal stability in maximal supergravity _ , [ arxiv:1112.3939 ] .
m. de roo , d. b. westra and s. panda , _ de sitter solutions in n=4 matter coupled supergravity _
, jhep * 0302 * ( 2003 ) 003 [ hep - th/0212216 ] .
m. de roo , d. b. westra , s. panda and m. trigiante , _ potential and mass matrix in gauged n=4 supergravity _ , jhep * 0311 * ( 2003 ) 022 [ hep - th/0310187 ] . c. m. hull and n. p. warner ,
_ the potentials of the gauged n=8 supergravity theories _ , nucl .
b * 253 * ( 1985 ) 675 .
r. kallosh , a. d. linde , s. prokushkin and m. shmakova , _ gauged supergravities , de sitter space and cosmology _ , phys .
d * 65 * ( 2002 ) 105016 [ hep - th/0110089 ] .
g. dallagata and g. inverso , _ de sitter vacua in n = 8 supergravity and slow - roll conditions _ ,
b * 718 * ( 2013 ) 1132 [ arxiv:1211.3414 ] .
g. dallagata and g. inverso , _ on the vacua of n = 8 gauged supergravity in 4 dimensions _ , nucl .
b * 859 * ( 2012 ) 70 [ arxiv:1112.3345 ] .
f. catino , c. a. scrucca and p. smyth , _ metastable de sitter vacua in n=2 to n=1 truncated supergravity _ , jhep * 1210 * ( 2012 ) 124 [ arxiv:1209.0912 ] .
d. roest and j. rosseel , _ de sitter in extended supergravity _ , phys .
b * 685 * ( 2010 ) 201 [ arxiv:0912.4440 ] .
jacot and c. a. scrucca , _ metastable supersymmetry breaking in n=2 non - linear sigma - models _ , nucl .
b * 840 * ( 2010 ) 67 [ arxiv:1005.2523 ] . l. brizi and c. a. scrucca , _ the lightest scalar in theories with broken supersymmetry _
, jhep * 1111 * ( 2011 ) 013 [ arxiv:1107.1596 ] .
i. antoniadis and m. buican , _ goldstinos , supercurrents and metastable susy breaking in n=2 supersymmetric gauge theories _ ,
jhep * 1104 * ( 2011 ) 101 [ arxiv:1005.3012 ]
. b. lgeret , c. a. scrucca and p. smyth , _ metastable spontaneous breaking of n=2 supersymmetry _ , [ arxiv:1211.7364 ] .
l. alvarez - gaum and d. z. freedman , _ geometrical structure and ultraviolet finiteness in the supersymmetric sigma model _
, commun .
phys . * 80 * ( 1981 ) 443 .
l. alvarez - gaum and d. z. freedman , _
potentials for the supersymmetric nonlinear sigma model _
, commun . math .
* 91 * ( 1983 ) 87 .
j. bagger and e. witten , _ matter couplings in n=2 supergravity _ , nucl .
b * 222 * ( 1983 ) 1 .
b. de wit and a. van proeyen , _ potentials and symmetries of general gauged n=2 supergravity : yang - mills models _ , nucl .
b * 245 * ( 1984 ) 89 .
b. de wit , p. g. lauwers , and a. van proeyen , _ lagrangians of n=2 supergravity - matter systems _ , nucl .
b * 255 * ( 1985 ) 569 . c. m. hull , a. karlhede , u. lindstrom and m. rocek , _ nonlinear sigma models and their gauging in and out of superspace _ , nucl .
b * 266 * ( 1986 ) 1 .
a. strominger , _ special geometry _ ,
phys . * 133 * ( 1990 ) 163 .
r. dauria , s. ferrara and p. fr , _ special and quaternionic isometries : general couplings in n=2 supergravity and the scalar potential _ ,
b * 359 * ( 1991 ) 705 .
l. andrianopoli , m. bertolini , a. ceresole , r. dauria , s. ferrara , p. fr and t. magri , _
n = 2 supergravity and n = 2 super yang - mills theory on general scalar manifolds : symplectic covariance , gaugings and the momentum map _
, j. geom .
* 23 * ( 1997 ) 111 [ hep - th/9605032 ] .
m. przanowski , _ killing vector fields in selfdual , euclidean einstein spaces with @xmath204 _ , j. math
. phys .
* 32 * ( 1991 ) 1004 .
k. p. tod , _ the su(@xmath205)-toda field equation and special four - dimensional metrics _ ,
notes in pure and appl . math .
* 184 * ( 1997 ) 307 .
m. davidse , f. saueressig , u. theis and s. vandoren , _ membrane instantons and de sitter vacua _ ,
jhep * 0509 * ( 2005 ) 065 [ hep - th/0506097 ] .
s. w. hawking , _ gravitational instantons _
a * 60 * ( 1977 ) 81 .
g. w. gibbons and s. w. hawking , _ gravitational multi - instantons _ , phys .
b * 78 * ( 1978 ) 430 .
n. j. hitchin , a. karlhede , u. lindstrom and m. rocek , _ hyperkahler metrics and supersymmetry _ , commun .
* 108 * ( 1987 ) 535 .
n. ambrosetti , i. antoniadis , j. -p .
derendinger and p. tziveloglou , _ the hypermultiplet with heisenberg isometry in n=2 global and local supersymmetry _ , jhep * 1106 * ( 2011 ) 139 [ arxiv:1005.0323 ] .
p. breitenlohner and d. z. freedman , _ stability in gauged extended supergravity _ , annals phys .
* 144 * ( 1982 ) 249 . | we construct a simple class of n=2 gauged supergravity theories that admit metastable de sitter vacua , generalizing the recent work done in the context of rigid supersymmetry .
the setup involves one hypermultiplet and one vector multiplet spanning suitably curved quaternionic - khler and special - khler geometries , with an abelian gauging based on a single triholomorphic isometry , but neither fayet - iliopoulos terms nor non - abelian gauge symmetries .
we construct the most general model of this type and show that in such a situation the possibility of achieving metastable supersymmetry breaking vacua crucially depends on the value of the cosmological constant @xmath0 relative to the gravitino mass squared @xmath1 in planck units . in particular , focusing on de sitter vacua with positive @xmath0 , we show that metastability is only possible when @xmath2 .
we also derive an upper bound on the lightest scalar mass in this kind of model relative to the gravitino mass @xmath3 as a function of the cosmological constant @xmath0 , and discuss its physical implications . @xmath4 0.5 cm francesca catino@xmath5 , claudio a. scrucca@xmath6 and paul smyth@xmath6 0.5 cm @xmath7_ii .
institut fr theoretische physik , universitt hamburg , + d-22761 hamburg , germany _
+ @xmath8_zentrum fr mathematische physik , universitt hamburg , + bundesstrasse 55 , d-20146 hamburg , germany _ + @xmath6_institut de
thorie des phnomnes physiques , epfl , + + _ 1.5 cm |
current investigations of the strongly correlated electron systems often deal with collective contributions to the electrical conductivity ( and to the other transport coefficients ) .
this includes in particular the conductivity in the charge - density - wave ( cdw ) or spin - density - wave ( sdw ) ordered states . in most of those studies the electrical conductivity is determined , theoretically and experimentally , relative either to the conductivity in the metallic state or with respect to the conductivity in the semiconducting state of the pinned cdw / sdw .
obviously , the prerequisite to such procedures is the accurate evaluation of the conductivity in the metallic state or in the state of the pinned cdw / sdw , and this simple , yet not completely solved , problem is in the focus of our interest here . actually , using the microscopic transverse response theory , lee , rice and anderson have found that the single - particle optical conductivity in the ordered cdw state with the negligible small number of scattering centers is given in terms of the semiconducting current - current correlation function which describes excitations across the gap @xcite .
however , it is also shown that for the typical value of the cdw gap and of the zero - frequency damping energy ( arising from the impurity scattering processes ) their result matches up neither the result of the longitudinal response theory @xcite nor the experimental observation @xcite . yet , the longitudinal and transverse response have to coincide for fast enough ( quasi)homogeneous longitudinal fields , as implicit for example in the maxwell equations of the medium , which employ only one dielectric function . in his textbook
@xcite , mahan has further shown that an alternative , so - called force - force correlation function method gives a good description of the ( high - frequency ) optical processes in both metallic and semiconducting systems , including various excitations within the conduction band and across the ( pseudo)gaps , but also fails to reach correctly the @xmath1 limit , requiring a specific @xmath2 field - theory approach @xcite . also important
is the observation that most of the transport - coefficient analyses are based on the boltzmann equations applied to the nearly - free electron models @xcite , completely neglecting band periodicity in the reciprocal space .
most of these issues can be settled down using the longitudinal response theory with a particular care devoted to the continuity equation @xcite . in the present article
, it will be shown that this can be done alternatively ( but in a somewhat less strict way ) using the gauge - invariant form of the transverse approach .
for this purpose , we consider a quasi - one - dimensional ( q1d ) two - band model with the impurity scattering taken into consideration .
the two bands are taken to result from the site - energy dimerization in the highly conducting direction , and , consequently , the bloch functions and all relevant vertex functions can be determined analytically ( sec .
iii ) . using the equation of motion approach ( which is found to be the generalization of the force - force correlation function approach ) ,
the intra- and interband optical conductivity are calculated ( sec .
iv ) . in the intraband channel of the transverse correlation functions ,
the most singular processes in powers of @xmath0 are collected , resulting in the optical conductivity which matches up the dc conductivity obtained by the boltzmann equations @xcite or by the landau response theory @xcite . the resulting interband conductivity in the cdw ordered state is found to be consistent with the experimental observation .
theorywise , the semiconducting current - current correlation function @xcite is replaced by a slightly modified function containing an additional factor which comes from the gauge invariant treatment of the diamagnetic current contributions .
finally ( sec .
v ) , the optical and dc conductivity are determined for a few typical q1d cases .
the optical conductivity tensor @xmath3 is a measure of the absorption rate for the ( transverse ) electromagnetic waves traveling across the crystal , and the measured spectra , together with the dc conductivity data and other transport coefficients , are an extremely valuable source of the information about the electronic subsystem .
although some aspects of the microscopic response theory can be found in the textbooks @xcite , there is no systematic microscopic solution to the multiband optical conductivity problem .
actually , it is easy to determine macroscopic symmetry features of @xmath3 , even in a general case . in this respect , we shall combine the macroscopic symmetry features with the microscopic description of the electron - photon coupling functions ( determined for a simple , exactly solvable q1d electronic model ) to develop a consistent microscopic multiband response theory .
the optical conductivity analysis starts with the hamiltonian @xcite @xmath4 which comprises the bare photon term @xmath5 , the bare electronic hamiltonian @xmath6 and the first - order and the second - order electron - photon coupling term , @xmath7 and @xmath8 .
the bare photon contribution is @xmath9 .
\label{eq2}\end{aligned}\ ] ] here @xmath10 and @xmath11 are the wave vector and the polarization of the photon field @xmath12 , @xmath13 is the field conjugate to @xmath12 , and @xmath14 is the bare photon dispersion . the structure of a typical q1d tight - binding electronic hamiltonian @xmath15 is determined below . however , notice that general symmetry properties of @xmath3 discussed here do not depend on details of @xmath16 . to obtain @xmath3 , the retarded photon green function
is required @xcite @xmath17 \rangle \\
\nonumber & = & { \rm e}^ { -\eta t } \frac{1}{2 \pi } \int_{-\infty}^{\infty } { \rm d } \omega \ , { \rm e}^{{\rm i } \omega t } \langle \langle q_{{\bf q } \alpha } ; q_{-{\bf q } \alpha } \rangle \rangle_{\omega } , \label{eq3}\end{aligned}\ ] ] with @xmath18 and @xmath19 representing the photon fields in the heisenberg picture at time @xmath20 and @xmath21 , respectively . using the equation of motion formalism
, we get @xmath22 \langle \langle q_{{\bf q } \alpha } ; q_{-{\bf q } \alpha } \rangle \rangle_{\omega } & = & \hbar , \label{eq4}\end{aligned}\ ] ] with the adiabatic term @xmath23 .
the renormalized photon frequency @xmath24 is given by @xcite @xmath25 where @xmath26 and @xmath27 are , respectively , the diamagnetic and current - current contributions to the photon self - energy , shown in fig .
1 ( for the q1d model under consideration , the explicit form of @xmath28 and @xmath26 is given in secs .
ivb , c and ref .
@xcite , respectively ) . combining the maxwell equations with eq .
( [ eq4 ] ) , it can be shown that the transverse dielectric function @xmath29 satisfies the relation @xmath30
\langle \langle q_{{\bf q } \alpha } ; q_{-{\bf q } \alpha } \rangle \rangle_{\omega } & = & \hbar , \label{eq6}\end{aligned}\ ] ] with @xmath31 the optical conductivity defined by eq .
( [ eq7 ] ) is @xmath32 , \label{eq8}\end{aligned}\ ] ] irrespective of whether the electromagnetic field is treated as a classical or a quantum field .
this is a quite general and in many respects very useful result . in the general case , with several valence bands or with several scattering channels in a single band
, @xmath26 includes one or more diamagnetic contributions ( depending on the number of bands intersecting the fermi level ) , and @xmath28 represents all intra- and interband current - current correlation functions .
the following general properties of the expression ( [ eq8 ] ) are important for interpreting the measured spectra .
first , there are at least two distinct structures in the optical conductivity spectrum @xmath33 , the first one is a delta function at @xmath2 , related to the diamagnetic current , and the second one represents various contributions , including the exciton contributions if the short - range dipole - dipole interactions are present @xcite .
this is easily seen from @xmath34 \delta ( \omega ) \nonumber \\ & & -\frac{1}{\omega } { \rm i m } \ { \pi_{\alpha \alpha } ( \omega ) \}. \label{eq9}\end{aligned}\ ] ] obviously , in the normal metallic or insulating state the delta - function term vanishes @xcite .
therefore , any consistent treatment of the electron - photon coupling functions has to fulfill the relation @xmath35 the total optical conductivity in the normal metallic or insulating state can be written then in the form @xmath36 . \label{eq11}\end{aligned}\ ] ] noteworthy , in the superconducting state the sum @xmath37
\label{eq12}\end{aligned}\ ] ] measures the weight of the missing area in optical conductivity spectra @xcite , while the single - particle contributions are still given by eq .
( [ eq11 ] ) .
finally , notice that in absence of local dipolar excitations ( the case of the site - energy dimerization discussed in this article ) the total current - current correlation function is the sum of only two contributions describing , respectively , the creation of the `` free '' intra- and interband electron - hole pairs , while the processes associated with excitons ( i.e. the quasiparticles representing the bound electron - hole pairs ) are not present . the optical conductivity determined by eq .
( [ eq11 ] ) , with the current - current correlation function calculated by the equation of motion approach , is in the focus of the present analysis .
the causality requirement @xcite ( i.e. the kramers kronig relations ) , the effective mass theorem @xcite and the gauge - invariance requirement @xcite will be used to test the obtained results .
a particular care will be devoted to the non - physical singularity at @xmath2 related to the prefactor of @xmath38 in eq .
( [ eq11 ] ) , and to the construction of the optical conductivity model with the correct behaviour in the @xmath1 limit , giving rise to a unified description of the optical and transport phenomena .
although our model is q1d , the present response theory is quite general and the electronic hamiltonian could represent an arbitrary multiband model .
we assume that in addition to the bare electronic hamiltonian ( denoted below by @xmath39 ) there is a static single - particle potential ( @xmath40 ) characterized by a commensurate wave vector , which describes , for example , the scattering of electrons on the site - energy dimerization potential .
the other single - electron scattering processes ( on impurities , phonons , etc . ) are represented by @xmath41 . for most of the questions discussed here the main effects of the two - electron interactions
are taken satisfactorily into account through the effective mean fields included in @xmath40 or @xmath41 , @xcite resulting finally in @xmath42 in eq .
( [ eq1 ] ) .
@xmath43 couples the valence electrons to transverse electromagnetic fields .
the resulting total electronic hamiltonian is @xmath44 we start by diagonalizing the hamiltonian @xmath45 . as mentioned above , we consider the simplest q1d model , where @xmath40 represents the site - energy dimerization in the highly conducting direction and in @xmath41 only the impurity scattering is taken into account .
next , we determine the related electron - photon coupling functions . at the end of this section , the multiband current - current correlation function is introduced .
the single - particle properties of the q1d site - energy - dimerization model come from the exact diagonalization of the hamiltonian @xcite @xmath46 .
\label{eq14}\end{aligned}\ ] ] the bare electron dispersions of two subbands , artificially dimerized along the highly conducting direction @xmath47 , are @xmath48 with @xmath49 , and with the wave vector * k * restricted to the new ( reduced ) brillouin zone , @xmath50 , @xmath51
. @xmath52 and @xmath53 ( @xmath54 ) are the bond energies in the direction @xmath47 and the perpendicular direction @xmath55 , respectively , and @xmath56 is the magnitude of the dimerization potential in the direction @xmath47 .
note that such potential corresponds to imperfect nesting in eq .
( [ eq15 ] ) in contrast to dimerization in all directions ; the former nesting is chosen because it is more interesting in context of the conductivity studies . the transformations between the unperturbed states , the band index @xmath57 , and the bloch states , the band index @xmath58 , of the form @xmath59 lead to @xmath60 with the dispersions @xmath61 \pm \sqrt{\frac{1}{4}\varepsilon_{\underline{c}c}^2 ( { \bf k } ) + \delta ^2 } , \nonumber \\ \varepsilon_{\underline{c}c } ( { \bf k } ) & = & \varepsilon_{\underline{c } } ( { \bf k } ) - \varepsilon_c ( { \bf k } ) .
\label{eq18}\end{aligned}\ ] ] the transformation - matrix elements are given by @xmath62 where @xmath63 , @xmath64 , with the auxiliary phase @xmath65 defined in the usual way @xmath66 the bands are shown in fig .
hereafter , the lower ( conduction ) band is assumed to be partially filled and the upper ( valence ) band is empty .
for the lower band completely filled and @xmath67 , the band structure corresponds to the commensurate cdw system , otherwise we have the metallic behaviour . in the reduced ( a ) and extended ( b ) zone representation , for @xmath68 ev and @xmath69 ev . in fig .
( b ) , the solid and dashed lines correspond , respectively , to @xmath70 and @xmath71 in the relations @xmath72 , @xmath73 .
the dot - dashed line is the fermi level @xmath74 in a typical metallic case .
, title="fig : " ] in the reduced ( a ) and extended ( b ) zone representation , for @xmath68 ev and @xmath69 ev . in fig .
( b ) , the solid and dashed lines correspond , respectively , to @xmath70 and @xmath71 in the relations @xmath72 , @xmath73 .
the dot - dashed line is the fermi level @xmath74 in a typical metallic case .
, title="fig : " ] using the generalized minimal substitution method for the tight - binding electrons @xcite , we obtain that the conduction electrons described by the hamiltonian ( [ eq14 ] ) are coupled to the external electromagnetic fields through @xmath75 where , to the second order in the vector potential @xmath76 , @xmath77 the photon annihilation operator @xmath78 enters in eq .
( [ eq22 ] ) through @xmath79 where @xmath80 \label{eq23}\end{aligned}\ ] ] is the electromagnetic displacement field of eq .
( [ eq2 ] ) @xcite . finally , in the bloch representation the coupling hamiltonian becomes @xmath81 with @xmath82 representing , respectively , the current density and raman @xcite density operators .
the structure of the related vertex functions @xmath83 and @xmath84 is given in appendix a. ) , with the current vertices given by eqs .
( [ eqa2 ] ) and ( a2@xmath85 ) .
the shaded square is the electron - hole propagator defined by eq .
( [ eq28 ] ) .
( b ) the intraband current - current correlation function at high - frequencies ( the leading term in powers of @xmath86 ) .
( c ) the structure of the high - frequencies self - energy term .
the direct processes are associated with the creation of an electron - hole pair with wave vectors * k * and @xmath87 ( @xmath10 is the wave vector of the external field ) , while the indirect processes involve the impurity - assisted electron - hole pairs with wave vectors @xmath88 and @xmath89 , or @xmath90 and @xmath91 ( @xmath92 is the momentum relaxed on impurities , and the crosses represent the impurity scattering ) . ]
we consider here only the intraband impurity scattering processes in the perturbation @xmath93 for which one usually assumes @xmath94 @xcite .
this form of @xmath95 is consistent with the regime in which @xmath96 ( @xmath97 is the relaxation time defined below ) .
the generalization is straightforward . in the equation of motion approach
@xcite used in the next section , the starting point is the current - current correlation function of eqs .
( [ eq8])([eq12 ] ) shown in fig .
it is defined by @xcite @xmath98 \rangle .
\label{eq27}\end{aligned}\ ] ] here the current operator @xmath99 includes all intra- and interband current density fluctuations , as seen from eq .
( [ eq25 ] ) .
according to fig .
3 , @xmath100 comprises two intraband and two interband contributions , and the problem is reduced , as will be seen immediately below , to the self - consistent calculation of intra- and interband electron - hole propagator in presence of the perturbation @xmath95 . in eq .
( [ eq27 ] ) , as well as in eqs .
( [ eq28 ] ) , ( [ eq30 ] ) , ( [ eq33 ] ) and ( [ eq34 ] ) , the usual notation for the retarded correlation functions is used : @xmath101 \rangle$ ] , with @xmath102 being the operator @xmath103 in the heisenberg picture .
the retarded electron - hole propagator @xmath104 \rangle \nonumber \\ & & \hspace{5 mm } = { \rm e}^ { -\eta t } \frac{1}{2 \pi } \int_{-\infty}^{\infty } { \rm d } \omega \ , { \rm e}^{{\rm i } \omega t } { \cal d}_1^{ll ' } ( { \bf k } , { \bf k}_+',{\bf k}_+ , { \bf k } ' , \omega ) \label{eq28}\end{aligned}\ ] ] ( @xmath105 is the abbreviation for @xmath106 ) is the central quantity to all long - wavelength correlation functions , as can be seen from fig . 4 , or from the expression @xmath107 which represents a generalized long - wavelength correlation function , with @xmath108 and @xmath109 being the charge , current , raman or some other vertex functions . in eq .
( [ eq28 ] ) the abbreviation @xmath110 is used .
the symmetry of the @xmath108 and @xmath109 vertices , together with the nature of the singularity of the leading term in the perturbation series , determines the summation rule for the related feynman diagrams .
the simplest way to collect the most singular diagrams is to consider the equations of motion connecting @xmath111 with the correlation function @xmath112 defined as @xmath113 ; { l'}_{{\bf k } ' + { \bf q } \sigma}^{\dagger } l_{{\bf k } ' \sigma } \rangle \rangle_{t}. \end{aligned}\ ] ] the direct calculation gives the exact relation @xmath114^{-1 } { \cal d}_1^{ll ' } ( { \bf k } , { \bf k}_+ , { \bf k}_+ ' , { \bf k } ' , \omega ) \label{eq31 } \\ \nonumber & & \hspace{5 mm } = \hbar \delta_{{\bf k } , { \bf k } ' } \big[f_l ( { \bf k } ) - f_{l ' } ( { \bf k}_+ ) \big ] + { \cal d}_2^{ll ' } ( { \bf k } , { \bf k}_+ , { \bf k}_+ ' , { \bf k } ' , \omega).\end{aligned}\ ] ] here @xmath115^{-1 } = \hbar ( \omega + { \rm i } \eta ) + e_l ( { \bf k } ) - e_{l ' } ( { \bf k}_+ ) \nonumber \\ \label{eq32}\end{aligned}\ ] ] is a useful abbreviation .
@xmath116 is the fourier transform of @xmath117 , and @xmath118 \ }
\big]^{-1}$ ] is the fermi
dirac function , with @xmath119 .
the way to evaluate @xmath120 depends on the choice of representation of this function .
there are two alternative ways , @xmath121 _ t , \big ( { l'}_{{\bf k } ' + { \bf q } \sigma}^{\dagger } l_{{\bf k } ' \sigma } \big)_0 ] \rangle\end{aligned}\ ] ] or @xmath122_0 , \big ( { l'}_{{\bf k } ' + { \bf q } \sigma}^{\dagger } l_{{\bf k } ' \sigma}\big)_{-t } ] \rangle,\end{aligned}\ ] ] leading to two different self - consistent schemes ( one described below and another encountered in the longitudinal response theory @xcite ) , both giving , as will be argued below , the same result .
again , @xmath123_t$ ] is the abbreviation for @xmath124 $ ] .
it is important to realize at the outset that the first term on the right - hand side of eq .
( [ eq31 ] ) is significant for all interband correlation functions @xmath125 in eq .
( [ eq29 ] ) , irrespective of the vertex symmetries .
the probability for the direct creation of the interband electron - hole pair is proportional here to @xmath126 , so that all occupied states in the conduction band(s ) are equally important . for vertices @xmath108 and @xmath109 taken to represent the intraband current vertices , the correlation function @xmath125 in eq .
( [ eq29 ] ) becomes the intraband current - current correlation function @xmath127 . according to the discussion of sec.iia
, the related intraband optical conductivity is ruled by the prefactor @xmath38 in eq .
( [ eq11 ] ) .
the direct processes in eq .
( [ eq31 ] ) ( see the first term on the right - hand side of fig
. 3(c ) ) related to @xmath128 are insignificant , and @xmath129 can be adequately described by the second , indirect term in this equation ( the second term in fig .
3(c ) ; see also fig . 5 ) , with @xmath130 given by eq .
( [ eq34 ] ) . ) , indirect intraband ( @xmath131 ) , direct interband ( @xmath132 ) and indirect interband ( @xmath133 ) optical excitations . in this article
the second and third processes are considered , the indirect intraband , using the exact summation of the impurity scattering processes , and the direct interband , using the phenomenological treatment of the impurity scattering .
the dotted line is the photon dispersion , and the dashed line corresponds to the impurity scattering . ] before turning to the evaluation of @xmath127 it is interesting to contrast this conclusion with its analog in the longitudinal response theory . in the longitudinal approach ,
the long - range coulomb forces are activated , and the first term in eq .
( [ eq31 ] ) dominates the intraband charge - charge correlation function , even in the dynamic limit , since the small probability for the intraband electron - hole pair creation , proportional to @xmath134 , cancels out the @xmath135 singularity of the long - range forces , which is the well - known rpa result .
the indirect scattering processes , on the other hand , are proportional to the effective intraband charge vertex , analogous to the effective intraband current vertex introduced below , eq .
( [ eq41 ] ) . since the bare long - wavelength intraband charge vertex satisfies @xmath136 , where @xmath137 is the bare electron charge , the effective intraband charge vertex for the indirect scattering processes vanishes , because @xmath138 .
since the longitudinal and transverse approaches are to be equivalent , this means that the direct longitudinal processes are to be equivalent to the indirect transverse processes in the intraband channel , while the contributions of both the indirect longitudinal and the direct transverse scattering processes are to be negligible .
this issue will be further discussed in sec.ivb5 .
the above transverse approach can now be applied to the current - current correlation function of the two band model , rewritten in the form @xmath139 here the vertices @xmath108 and @xmath109 in eq .
( [ eq29 ] ) are replaced by @xmath140 and @xmath141 , respectively .
optical processes relevant to the two - band model ( including the indirect interband ones not considered in the present analysis ) are illustrated in fig .
5 . in order to make presentation of the results more transparent
, we shall first determine the intraband contributions , and then give the analysis of the interband optical excitations .
for the sake of brevity , in the next section the ( intra)band index @xmath142 will be omitted ( @xmath143 , @xmath144 , , with @xmath145 ) .
it should be noticed that the results obtained below for the intraband optical conductivity are quite general , i.e. they cover various physically different regimes . as explained in sec.ivb2 , when the electron filling of the dimerized band varies between 0 and 1 , the electronic system transforms from an electron - like semiconducting , through a metallic , into a hole - like semiconducting regime . a more detailed discussion of this issue is given in sec.v , in context of the total optical conductivity .
according to the aforementioned arguments , the intraband optical processes are described by the equations @xmath146 , { c}_{{\bf k } ' + { \bf q } \sigma}^{\dagger } c_{{\bf k } ' \sigma } ] \rangle,\end{aligned}\ ] ] where @xmath147 is the intraband term in eq .
( [ eq32 ] ) and @xmath148 is the fourier transform of the force - force correlation function @xcite @xmath149_0 , \big [ c_{{\bf k } ' + { \bf q } \sigma}^{\dagger } c_{{\bf k } ' \sigma},\ , h ' \big]_{-t } ] \rangle.\end{aligned}\ ] ] the second term in eq .
( [ eq37 ] ) is the ground - state average of the four - operator product at @xmath21 .
it is off - diagonal in the bloch representation , and its value can be obtained by putting @xmath150 in the left - hand side of eq .
( [ eq37 ] ) and then taking the formal limit @xmath1 .
the result is @xmath151 , { c}_{{\bf k } ' + { \bf q } \sigma}^{\dagger } c_{{\bf k } ' \sigma } ] \rangle \nonumber \\ & & \hspace{20 mm } \approx { \cal d}_3({\bf k } , { \bf k}_+ , { \bf k}_+ ' , { \bf k } ' , \omega = 0 ) .
\label{eq39}\end{aligned}\ ] ] for the impurity scattering , we have the cancellation of this constant term with its counterpart obtained by the electron @xmath152 hole replacement ( see eq .
( [ eq47 ] ) , for example ) , but for a time - dependent perturbation @xmath95 , eq .
( [ eq39 ] ) achieves an interesting structure , as pointed out in refs .
@xcite due to the large velocity of light , the energy and wave vector transfers in the external points of the intraband current - current correlation function fulfill @xmath153 , and the factors @xmath154 and @xmath155 in eqs .
( [ eq36 ] ) and ( [ eq37 ] ) , representing the propagator of the virtual electron - hole pairs ( related to the process @xmath156 in fig . 5 ) , can be replaced by @xmath157 .
the resulting intraband correlation function becomes @xmath158 \nonumber \\ & & \hspace{10mm}\times \big [ j_{\alpha } ( { \bf k}''_1 ) - j_{\alpha } ( { \bf k}_1 ) \big ] \bigg < v({\bf k } - { \bf k } '' ) v({\bf k}''_1 - { \bf k}_1 ) \nonumber \\ & & \times \big [ { \cal d}_1({\bf k } , { \bf k } '' , { \bf k}_1 '' , { \bf k}_1 , \omega ) - { \cal d}_1({\bf k } , { \bf k } '' , { \bf k}_1 '' , { \bf k}_1 , 0 ) \big ] \bigg > . \label{eq40 } \end{aligned}\ ] ] @xmath159 is here and subsequently the average over the impurity sites @xcite . and @xmath160 are encircled into the effective vertex .
the dot is the bare intraband current vertex . ]
the diagrammatic representation of @xmath161 is given in fig .
6(a ) , with the structure of the leading term , proportional to @xmath86 , shown explicitly in fig .
the expression @xmath162 \label{eq41}\end{aligned}\ ] ] can be recognized as an effective current vertex for the indirect intraband photon absorption / emission processes , and is a sum of two terms encircled in fig .
notice also that @xmath163 is proportional to @xmath164 , canceling out the factor of @xmath38 in one of the effective vertices .
it is important to notice that in the leading term ( fig .
6(b ) ) the expression ( [ eq40 ] ) is identical to the result of the force - force correlation function method @xcite . while the latter method is usually limited to the examination of this leading term , or to the summation of irrelevant higher order diagrams which results in the well - known hopfield formula @xcite , the present approach is focused instead on the exact summation of the most singular contributions in powers of @xmath86 and should be regarded as a generalization of the standard force - force correlation function approach . after determining the structure of the external points ( i.e. the effective vertices ) in the diagram shown in fig .
6(a ) , we have to find the internal structure of the diagram .
the latter is represented by the indirect ( impurity - assisted ) electron - hole propagator @xmath165 , characterized by the momentum transfer @xmath166 , rather than by the negligibly small external momentum transfer @xmath167 of the direct processes in eq .
( [ eq31 ] ) . its internal structure is determined here by the self - consistent solution of the exact equation @xmath168 + { \cal d}_2 ( { \bf k } , { \bf k } '' , { \bf k}_1 '' , { \bf k}_1 , \omega).\end{aligned}\ ] ] @xmath169 represents the electron - hole pair created by the indirect photon absorption of fig . 5 , which means that the energy transfer @xmath170 is close to the electron - hole pair energy @xmath171 .
the wave vectors * k * and @xmath172 ( @xmath173 and @xmath174 , as well ) are independent of each other , and therefore the first term in eq .
( [ eq42 ] ) dominates the low - energy physics , preferring the optical processes between the states @xmath175 and @xmath176 . ) .
( b ) typical contributions omitted in the self - consistent scheme .
( c ) the self - consistent equation of the kernel @xmath177 and ( d ) of the electron - hole propagator @xmath169 .
the diamond represents the electron - hole self - energy consisting of the single - particle self - energy and vertex - correction contributions . ] in combining eq .
( [ eq42 ] ) with the equation of motion for @xmath178 ( see eq .
( [ eq53 ] ) for more detail ) it is sufficient to collect only the terms in the summation which are relevant to the self - consistent equation for the kernel @xmath179 in the current - current correlation function ( see figs .
7(c , d ) , and the criterion of the validity shown in figs .
7(a , b ) ) @xmath180 \nonumber \\ & & \hspace{15 mm } \times \big [ { \cal f}_{\alpha}({\bf k}_1 '' , { \bf k}_1 , \omega ) - { \cal f}_{\alpha}({\bf k}_1 '' , { \bf k}_1 , 0 ) \big]\bigg > , \label{eq43 } \end{aligned}\ ] ] where @xmath181 { \cal d}_1 ( { \bf k } , { \bf k } '' , { \bf k}_1 '' , { \bf k}_1 , \omega).\end{aligned}\ ] ] the idea of the present approach is the self - consistent treatment of eqs .
( [ eq42 ] ) and ( [ eq44 ] ) , which represent two equations connecting the electron - hole propagator @xmath182 with the kernel @xmath183 .
the solution is based on the expansion @xmath184 where @xmath185 includes only the self - consistent terms proportional to @xmath186 . as easily seen in the longitudinal analysis , the self - consistent expression for @xmath187 , with @xmath188 ,
plays the leading role .
accordingly , to compare both response theories , one needs the self - consistent scheme for @xmath183 and the recurrence relations for @xmath185 ( see sec.ivb3 ) . for pedagogical reasons , it is convenient first to consider the zeroth order contribution to ( [ eq43 ] ) and define the effective number of conduction electrons and the related electron - hole damping energy .
the direct calculation gives the first term in the expansion ( [ eq45 ] ) @xmath189 { \cal d}_0 ( { \bf k},{\bf k } '' , \omega).\end{aligned}\ ] ] the related contribution to the current - current correlation function ( which is a good approximation for @xmath190 at high frequencies ) is given by @xmath191 \langle | v({\bf
k } - { \bf k } '' ) | ^2 \rangle \nonumber \\ & & \hspace{20mm}\times \frac{1}{\hbar } \big [ { \cal d}_0 ( { \bf k } , { \bf k } '' , \omega ) - { \cal d}_0 ( { \bf k } '' , { \bf k } , \omega ) \big ] \label{eq47 } \\ & & \hspace{10 mm } \approx \frac{1 } { \big(\hbar \omega \big)^2 } \frac{1}{v } \sum_{{\bf k } { \bf k } '' \sigma } j_{\alpha } ( { \bf k } ) \big ( j_{\alpha } ( { \bf k } ) - j_{\alpha } ( { \bf k } '' ) \big ) \nonumber \\ & & \hspace{15mm}\times \big [ f ( e({\bf k } ) ) - f ( e({\bf k } ) + \hbar \omega ) \big ] \langle | v({\bf k } - { \bf k } '' ) | ^2 \rangle \nonumber \\ & & \hspace{20mm}\times \frac{1}{\hbar } \big [ { \cal d}_0 ( { \bf k } , { \bf k } '' , \omega ) + { \cal d}_0 ( { \bf k } '' , { \bf k } , \omega ) \big ] . \hspace{5mm}(47 ' ) \nonumber \end{aligned}\ ] ] for the impurity scattering processes ,
the real part of this function is negligible , while the imaginary part is given by @xmath192 \frac{\hbar}{\tau ( { \bf k } , \omega ) } , \nonumber\end{aligned}\ ] ] where the electron - hole damping energy is @xmath193 furthermore , in this case , the frequency dependent part in @xmath194 is negligibly small , and we can put the average over the fermi surface @xmath195 in eq . ( [ eq48 ] ) instead of @xmath194 .
we finally get @xcite @xmath196 with @xmath197 being the effective number of conduction electrons .
the behaviour of @xmath198 with band filling @xmath199 is shown in fig . 8 for a few typical values of the ratio @xmath200 , and compared to the prediction of the free - electron and free - hole approximation .
notice that for @xmath201 one obtains the well - known result @xmath202 ( @xmath203 is the primitive - cell parameter of the dimerized lattice ) . ) shown in the dimensionless form , for @xmath68 ev , @xmath204 , @xmath205 k and @xmath206 ( dotted line ) , 10 mev ( solid line ) .
the dot - dashed line is the dimensionless concentration of conduction electrons @xmath207 , i.e. the prediction of the free - electron approximation .
@xmath208 is the primitive cell volume and @xmath209 is the mass scale parameter .
the doping regions with the electron - like ( @xmath210 ) and hole - like ( @xmath211 ) conductivity are also indicated .
inset : the effective number @xmath212 for @xmath68 ev , @xmath204 , @xmath205 k and @xmath213 ev .
the dot - dashed line is the prediction of the free - hole approximation , @xmath214 . ] beyond this approximation the expression ( [ eq47 ] ) requires the evaluation of two coupled integrations over * k * and @xmath172 . in the low - dimensional electronic systems , where the van hove singularities in the band structure may play important role , this is not a trivial task .
the most outstanding advantage of the present approach is the fact that the effective current vertex @xmath215 consists of two terms , which , together with two terms in @xmath216 , give four contributions to @xmath217 , two of them representing the so - called self - energy contributions , and the other two the vertex corrections ( see fig .
6(b ) ) @xcite .
we shall show next that in the @xmath218th order term in ( [ eq45 ] ) the single - particle vertex corrections and the single - particle self - energy contributions are also treated on equal footing , even if the relaxation processes on impurities are treated beyond the approximation @xmath219 .
the effects of the impurity scattering on the correlation function ( [ eq37 ] ) can be exhibited in the following way : @xmath220 .
\label{eq52 } \end{aligned}\ ] ] when combined with this expression , eqs .
( [ eq42 ] ) and ( [ eq46 ] ) lead to @xmath221 .
\nonumber \end{aligned}\ ] ] on the right - hand side of this equation , only the self - consistent terms are taken into account .
the first term in the brackets represents the single - particle self - energy contributions and the second one the single - particle vertex corrections .
it is important to remember again that these corrections are the largest for @xmath222 , so that @xmath223 can be approximated by @xmath38 . as a consequence
, the solution of this equation can be sought in powers of @xmath224 .
however , due to the dependence on @xmath87 and @xmath225 of the vertex - corrections term , we have to turn back to the kernel ( [ eq44 ] ) and apply several changes to its dummy variables to obtain the self - consistent description of @xmath183 and the desired recurrence relations between the contributions @xmath226 .
the kernel is described by @xmath227 \nonumber \\ & & \hspace{15 mm } \times { \cal d}_1 ( { \bf k } , { \bf k } '' , { \bf k}_1 '' , { \bf k}_1 , \omega ) \sigma({\bf k } , { \bf k } '' , \omega ) , \label{eq54 } \end{aligned}\ ] ] where @xmath228 is the electron - hole self - energy and @xmath229 is the kernel corresponding to the replacement of @xmath165 in eq .
( [ eq44 ] ) by @xmath230 .
the only approximation made in the derivation of eq .
( [ eq54 ] ) is @xmath231 \frac{j_{\alpha } ( { \bf q } ) } { j_{\alpha } ( { \bf k } '' ) } , \nonumber \\ \label{eq56}\end{aligned}\ ] ] which allows a simple description of the vertex corrections in @xmath183 , and which treats correctly the disappearance of the forward scattering contributions ( @xmath232 ) to both @xmath233 and @xmath183 . as briefly discussed at the end of sec.ivb5 , this approximation is directly related to the restrictions enforced by the the continuity equation for the charge density .
similarly , the definition ( [ eq45 ] ) , together with the self - consistent relations ( [ eq53 ] ) and ( [ eq54 ] ) , gives rise to the recurrence relations illustrated in fig . 7(d ) : @xmath234 the simplest way to solve eq . ( [ eq54 ] ) is to replace @xmath235 by its imaginary part averaged over the fermi surface , @xmath236 . as mentioned above , for the impurity scattering processes , the real part of @xmath235
can be ignored .
even if @xmath237 is not small , we can turn back to the electronic hamiltonian and tray to include the @xmath237 effects into the effective single - particle hamiltonian , and to diagonalize it , as we did here with the scattering processes on the dimerization potential @xmath238 .
the real part of the new self - energy @xmath235 is minimized in this way , with only the imaginary part playing an important role in eq .
( [ eq57 ] ) . in this case
, we obtain the expression @xmath239 which leads to the well - known results of the memory - function approximation @xcite @xmath240 with @xmath241 being the intraband memory ( relaxation ) function .
this result is consistent with the causality requirement , @xmath242 provided that @xmath243 . if @xmath244 is frequency dependent , the corresponding @xmath237 is non - zero , but the result ( [ eq59 ] ) is still acceptable . for @xmath237
not too large we can introduce the effects of @xmath237 through the mass redefinition @xmath245 through the kramers kronig relations .
the result is the generalized drude formula for the intraband optical conductivity .
here we show two important results .
first , the memory - function approximation , which in the traditional form has not been found to be transparent , can be understood as a simple replacement @xmath246 of the exact - summation result given below .
second , the memory - function results are acceptable even in the cases with the pronounced optical excitations across a gap ( or pseudogap ) ( where @xmath247 ; see , for example , the hole - like semiconducting regime in fig . 8 at @xmath248 ) , provided that the electron group velocity @xmath249 in eq .
( [ eq51 ] ) is determined using the relation ( [ eqa5 ] ) . the intraband conductivity spectrum obtained in this way
is related with the interband conductivity spectrum through the well - controlled conductivity sum rule @xcite . in the case where the wave vector dependence of the imaginary part of @xmath235 is significant and
the real part of @xmath235 is not too large the full recurrence relations for @xmath185 can be used to obtain the intraband current - current correlation function .
the result is @xmath250 ^ 2 \frac { \langle | v({\bf
k } - { \bf k } '' ) | ^2 \rangle } { \hbar \omega } \nonumber \\ & & \hspace{15 mm } \times \big[f ( { \bf k } ) - f ( { \bf k } '' ) \big ] \frac{1}{\hbar } \frac { { \cal d}_0 ( { \bf k},{\bf k } '' , \omega ) } { \hbar\omega + \hbar \sigma({\bf k } , { \bf k } '' , \omega ) } , \label{eq61 } \\ & & \hspace{10mm}\approx \frac{1}{v } \sum_{{\bf k } { \bf k}''\sigma } j^2_{\alpha } ( { \bf k } ) \frac{\partial f ( { \bf k})}{\partial e({\bf k } ) } \frac { 1 } { \hbar\omega + \hbar \sigma({\bf k } , { \bf k } '' , \omega ) } \nonumber \\ & & \hspace{20 mm } \times \langle | v({\bf k } - { \bf k } '' ) | ^2 \rangle \bigg ( 1 - \frac{j_{\alpha } ( { \bf k } '' ) } { j_{\alpha } ( { \bf k } ) } \bigg ) \nonumber \\ & & \hspace{20mm}\times \frac{1}{\hbar } \big [ { \cal d}_0 ( { \bf k } , { \bf k } '' , \omega ) + { \cal d}_0 ( { \bf k } '' , { \bf k } , \omega ) \big ] . \hspace{5mm}(61 ' ) \nonumber \end{aligned}\ ] ] physically the most important case corresponds to the approximation @xmath251 . in this case , we obtain @xmath252 the final form of the optical conductivity comes from eqs .
( @xmath253 ) and ( 11 ) @xmath254 which is the result identical to the result of the longitudinal response theory .
the longitudinal response theory gives a simpler way to obtain the same @xmath255 .
the difference between the two approaches is in the way how the continuity equation connecting the charge density and current density fluctuations is treated .
the consideration of the direct processes of the wave vector * q * in the longitudinal approach allows a more precise treatment of the continuity equation , in the way analogous to the landau response theory @xcite .
but here only an approximate solution is possible , since the theory is formulated in terms of the indirect intraband processes ( of the wave vector @xmath256 ) .
first significant consequence of the present exact summation method is the behaviour of the intraband optical conductivity in the zero - frequency limit . when the real part of @xmath257 is small enough
, we can write @xmath258 resulting in the dc conductivity which is equal to the well - known boltzmann result @xcite @xmath259 here @xmath260 is the effective number of conduction electrons shown in the dimensionless form and @xmath261 is the temperature dependent @xmath262 relaxation time . the relation ( [ eq62 ] ) , together with eq .
( [ eq66 ] ) , gives the complete description of the optical conductivity in a general multiband model , with the symmetry of the intra- and interband current vertices playing an essential role .
thus , eq . ( [ eq64 ] ) provides the direct link between the low - frequency optical conductivity and various transport coefficients not only in the single - band but also in the multiband models .
the approximation in which the @xmath263 term in eq . ( [ eq31 ] ) is omitted leads to the ideal interband current - current correlation function and to the ideal interband conductivity characterized by a sharp threshold at the energy @xmath264 @xcite .
the former one is given by @xmath265 with @xmath266 .
the latter one comes from eqs .
( [ eq11 ] ) and ( [ eq65 ] ) @xmath267 } { \hbar \omega + { \rm i } \hbar \eta - e^2_{\underline{c}c } ( { \bf k})/\big ( \hbar \omega \big ) } , \label{eq66 } \end{aligned}\ ] ] with @xmath268 put in the numerator and @xmath23 in the denominator .
the leading term in the interband electron - hole self - energy @xmath269 \label{eq67 } \end{aligned}\ ] ] comes from the single - particle self - energy contributions .
a reasonable generalization for the interband optical conductivity is given by eq .
( [ eq66 ] ) with the replacement @xmath270 by @xmath271 .
, @xmath272 mev , @xmath273 k , @xmath274 , @xmath68 ev , @xmath204 and @xmath275 5 mev .
@xmath276 labels the dc conductivity at @xmath206 .
] the interplay between the self - energy and vertex - corrections terms in @xmath277 , the correct treatment of the indirect interband optical excitations , as well as the role of the effective mass theorem in resolving all these issues will be explained elsewhere @xcite .
it should be noticed here that the landau - like function ( [ eq66 ] ) gives the in - gap optical conductivity slightly different from the corresponding lindhard - like function , as easily seen by comparing fig . 9 with fig . 4 reported in ref .
the understanding of the difference between these two results is of general importance as well , and will be discussed in detail in ref .
@xcite . : @xmath272 mev , @xmath273 k , @xmath274 , @xmath68 ev , @xmath204 and @xmath275 5 mev . notice that this figure also represents the temperature behaviour of the normalized dc conductivity , since @xmath278 .
in the simples limit , @xmath279 and @xmath280 , the optical conductivity of the present two - band model , @xmath281 is a function of the fermi wave vector @xmath282 , three band parameters , @xmath283 , @xmath284 and @xmath285 , and two damping energies , @xmath286 and @xmath287 . for @xmath288 ,
the model illustrates optical properties of various q1d imperfectly nested cdw systems ( including both the ordered cdw state and the pseudogap effects at temperatures above the critical temperature @xmath289 ) . in this section
, we shall briefly discuss a few qualitative results important to the q1d cdw systems .
first , the temperature dependence of the dc and optical conductivity in the ordered cdw state is discussed for the strictly 1d case ( @xmath204 ) .
then we contrast the interband conductivity found here to the oversimplified semiconducting optical conductivity usually encountered in the textbooks @xcite .
the temperature dependence in @xmath290 , @xmath291 , @xmath292 and @xmath293 is responsible for the transfer with increasing / decreasing temperature of the optical conductivity spectra between the intraband and interband channels .
this effect is particularly large in the vicinity of the metal - to - insulator phase transition . it should be noticed that , in the approximation in which @xmath294 is set in eq .
( @xmath295 ) , the temperature dependence of @xmath296 is given by the product of the effective number of conduction electrons @xmath297 and the relaxation time @xmath298 .
we have two adjustable parameters at any temperature , and the analysis of the dc conductivity data is thus possible only by the combination with the optical conductivity measurements .
the latter method allows also the determination of the magnitude of cdw order parameter @xmath299 and the critical exponent @xmath300 in @xmath301 , as well as the bond energy @xmath283 and the damping energy @xmath302 . in figs .
9 and 10 the optical conductivity normalized to the dc conductivity at @xmath206 and the temperature dependence of the dc conductivity are shown for typical values of the parameters .
the typical result for the interband conductivity in the ordered cdw state is shown in fig .
11 ( solid and dotted curves ) , and compared to the data measured in the blue bronze k@xmath303moo@xmath304 ( diamonds ) @xcite .
notice that the gauge - invariance factor @xmath305 in eq .
( [ eq66 ] ) ensures the disappearance of @xmath306 at @xmath307 , independently of the value of the phenomenologically introduced damping energy @xmath287 .
the dashed curve is the prediction of the usual optical model for semiconductors @xcite , with the factor @xmath308 absent , which is characterized by a significant ( but non - physical ) contribution to @xmath309 at @xmath307 . )
( @xmath310 ) , for @xmath311 ev , @xmath204 , @xmath312 ev , @xmath274 , @xmath205 k and @xmath313 0.3 ev ( solid curve ) and 1 mev ( dotted curve ) .
the dashed curve is the prediction of the usual ( gauge - non - invariant ) optical conductivity for semiconductors , calculated using the same band parameters , with @xmath313 0.15 ev .
@xmath314 is the dc conductivity of the @xmath206 , @xmath315 0.15 ev case .
the diamonds are experimental data measured in k@xmath303moo@xmath304 at @xmath205 k @xcite . ]
in this article , the response of the conduction electrons in a q1d multiband model to the transverse electromagnetic fields has been studied in presence of two scattering mechanisms .
in contrast to the coherent scattering on the site - energy - dimerization potential , the impurity scattering processes give negligibly small contribution to the real part of the electron self - energy , but dominate in the relaxation processes in both the intra- and interband optical excitations .
we determine the current - current correlation function , i.e. the optical conductivity of the related two - band model as a function of band filling .
the transverse equation of motion approach has been used to collect the most singular contributions to the optical conductivity .
it is shown that the present multiband optical analysis represents a generalization of the usual force - force correlation function method , and that in the dc limit it approaches correctly the results of boltzmann equations , due to its gauge - invariant form .
the present optical conductivity model gives the frequency and temperature dependences of the single - particle contributions to the optical conductivity spectra in the ordered cdw state which are consistent with both the experimental observation and the prediction of the longitudinal response theory .
it is explained that , while the response to the longitudinal fields is associated with the direct electron - hole pair excitations , the response to the transverse electromagnetic fields can be understood in terms of the indirect ( impurity - assisted ) electron - hole pair excitations .
this research was supported by the croatian ministry of science and technology under the project 0119 - 256 .
one of the authors ( i.k . ) would like to acknowledge the hospitality of the institute of physics of complex matter , lausanne , where parts of this work were completed .
the vertex functions in the expression ( [ eq25 ] ) depend on the unperturbed vertices @xmath316 and @xmath317 in the way @xmath318 for the @xmath319 polarization of the electromagnetic fields , the result is @xmath320 j^{cc}_{a } ( { \bf k } ) \nonumber \\ & = & \mp \cos \varphi ( { \bf k } ) j^{cc}_{a } ( { \bf k } ) , \label{eqa2 } \\ j^{\underline{c } c}_{a } ( { \bf k } ) & = & 2u ( { \bf k } ) v ( { \bf k } ) j^{cc}_{a } ( { \bf k } ) \nonumber \\ & = & \sin \varphi ( { \bf k } ) j^{cc}_{a } ( { \bf k } ) , \hspace{30 mm } ( { \rm a2 } ' ) \nonumber \\ \gamma^{\underline{c } \underline{c } , cc}_{a a } ( { \bf k};2 ) & = & \mp \big [ u^2 ( { \bf k } ) - v^2 ( { \bf k } ) \big ] \gamma^{cc}_{a a } ( { \bf k};2 ) \nonumber \\ & = & \mp \cos \varphi ( { \bf k } ) \gamma^{cc}_{aa } ( { \bf k};2 ) .
\label{eqa3}\end{aligned}\ ] ] similarly , for @xmath321 @xmath322 finally , using eqs .
( [ eq18 ] ) , ( [ eqa2 ] ) and ( [ eqa4 ] ) we can check the ward identity @xcite which connects the intraband current vertex @xmath323 with the electron group velocity @xmath324 @xmath325 p. a. lee , t. m. rice and p. w. anderson , solid state commun .
* 14 * ( 1974 ) 703 .
f. wooten , _ optical properties of solids _ , academic press , 1972 ; n. w. ashcroft and k. sturm , phys .
b * 3 * ( 1971 ) 1898 .
g. grner , _ density waves in solids _ , addison - wesley , new york , 1994 .
i. kupi , physica b * 322 * ( 2002 ) 154 .
l. degiorgi , st .
thieme , b. alavi , g. grner , r. h. mckenzie , k. kim and f. levy , _ physics and chemistry of low - dimensional inorganic conductors _ , eds .
c. schlenker at al . , plenum press , new york , 1996 , p. 337 .
k. kim , r. h. mckenzie and j. w. wilkins , phys .
* 71 * ( 1993 ) 4015 .
g. d. mahan , _ many - particle physics _ , plenum press , new york , 1990 .
a. a. abrikosov , l. p. gorkov and i. e. dzyaloshinski , _ methods of quantum field theory in statistical physics _ ,
dover publications , new york , 1975 .
s. doniach and e. h. sondheimer , _
green s functions for solid state physicists _ , w. a. benjamin , london , 1974 . j. m. ziman , _ electrons and phonons _ , oxford university press , london , 1972 .
d. pines and p. nozires , _ the theory of quantum liquids i _ , addison - wesley , new york , 1989 .
i. kupi and s. barii , to be published . c. kittel , _ introduction to solid state physics _ ,
john wiley and sons , new york , 1996 .
notice the redundant adiabatic factor @xmath326 in the fourier transformation , which is necessary in the differential treatment of the retarded correlation functions .
i. kupi , physica b * 344 * ( 2004 ) 27 .
s. l. adler , phys . rev .
* 126 * ( 1962 ) 413 .
n. wieser , phys . rev .
* 129 * ( 1962 ) 63 .
p. upanovi , a. bjeli and s. barii , z. phys .
b * 97 * ( 1995 ) 113 .
j. r. schrieffer , _ theory of superconductivity _
, w. a. benjamin , new york , 1964 .
d. c. mattis and j. bardeen , phys . rev . * 111 * ( 1958 ) 412 .
m. tinkham , _ introduction to superconductivity _ , krieger , 1980 .
a. a. abrikosov and v. m. genkin , zh .
* 65 * ( 1973 ) 842 ( sov .
jetp * 38 * ( 1974 ) 417 ) . n. w. ashcroft and n. d. mermin , _ solid state physics _ , saunders college publishing , 1976 .
in contrast to that , in the longitudinal approach , based on the charge - charge correlation functions , the long - range coulomb interactions play the crucial role .
w. gtze and p. wlfle , phys .
b * 6 * ( 1972 ) 1226 . t. giamarchi , phys . rev .
b * 44 * ( 1991 ) 2904 . | the electrodynamic features of the multiband model are examined using the transverse equation of motion approach in order to give the explanation of several long - standing problems .
it turns out that the exact summation of the most singular terms in powers of @xmath0 leads to the total optical conductivity which , in the zero - frequency limit , reduces to the results of the boltzmann equation , for both the metallic and semiconducting two - band regime .
the detailed calculations are carried out for the quasi - one - dimensional ( q1d ) two - band model corresponding to imperfect charge - density - wave ( cdw ) nesting .
it is also shown that the present treatment of the impurity scattering processes gives the dc conductivity of the ordered cdw state in agreement with the experimental observation .
finally , the dc and optical conductivity are calculated numerically for a few typical q1d cases . |
it is now generally accepted that long - duration grbs are associated with bright supernova - like explosions .
one of the primary differences between these grb - associated `` supernovae '' and normal supernovae is that grb - associated supernovae have larger ( by more than a factor of 4 in some cases ) @xmath0ni yields .
how is this extra @xmath0ni produced .
it is known that the grb jets themselves do not produce large amounts of @xmath0ni and researchers focused on two different paradigms for producing large amounts of @xmath0ni : the @xmath0ni is produced in a disk around the black hole ( see surman et al .
2006 and references therein ) , or the @xmath0ni is produced in the explosion of the star that accompanies the grb jet ( e.g. nagataki et al .
2003 ; mazzali et al . 2006 ) .
calculations assuming an accretion - disk site for @xmath0ni production tend to approximate the disk using the structure from advection dominated accretion flow solutions ( e.g. popham et al . 1999 ) and use a neutrino - driven wind solution to derive particle trajectories . depending upon the exact conditions in the disk ( in particular , the electron fraction ) , a large amount of @xmath0ni can be produced .
although surman et al .
( 2006 ) have shown that the amount of @xmath0ni produced does depend upon the conditions in the disk , one can reasonably assume that this mechanism always produces large amounts of @xmath0ni .
the explosive nucleosynthesis site uses strong shocks in the stellar explosion ( beyond the narrow jet region ) to heat material and cause it to burn to @xmath0ni .
it depends sensitively on the structure of the star and the velocity of the shock beyond the jet itself .
nomoto and collaborators @xcite have argued that a significant portion of the star in a grb is still ejected in faster - than - normal velocities , allowing the production of enhanced amounts of @xmath0ni .
the problem with this site of @xmath0ni production is that grbs may be produced both as direct collapse black holes or through weak supernovae that do not impart enough energy onto the star to prevent the fallback of matter onto the proto - neutron star , causing it to collapse to a black hole .
fryer et al . ( 2006 ) found that if the delay between this weak supernova explosion is more than a few seconds , the amount of @xmath0ni produced in explosive nucleosynthesis is very small indeed . and with very little @xmath0ni produced in the explosion
, these grbs should produce dim supernovae .
where are these grbs with faint associated supernovae ?
all observations suggested that bursts were associated with bright , high @xmath0ni - yield , high - velocity supernovae ( also known as `` hypernovae '' - nomoto et al .
could it be that fallback black holes do not produce grbs ?
the same month the fryer et al .
( 2006 ) paper was accepted , the first of two very peculiar long - duration bursts exploded .
these two bursts were well - localized , but their late - time emission showed no presence of an associated supernova ( fynbo et al .
2006 , gal - yam et al .
2006 ) . on the surface
, these bursts seem to be exactly the dim - supernova grbs predicted by theory . but gal - yam et al .
( 2006 ) argue that grb060614 is not a typical long burst and may require a different engine altogether .
grb060605 s duration ( @xmath1s ) lies on the border between short and long bursts and ofek et al .
( 2007 ) argue that it is a short burst .
so the observational evidence of low @xmath0ni - yield grb - supernovae is far from conclusive .
in addition , we must be wary of making any strong claims between @xmath0ni yields and supernova light - curves .
the process by which the energy from the decay of @xmath0ni and its daughter product @xmath0co is converted into optical photons is not linear .
the energy from decay is released in gamma - rays which then scatter and ultimately absorb in the star .
the star thermalizes this energy and re - emits it in optical photons .
but the star is also hot because it just exploded and the shock energy has heated it . in supernovae , the peak of the supernova light - curve can be powered either by this shock energy or , for the most compact stars , the decay of radioactive material . in this paper
, we provide the first radiation - hydrodynamics explosion calculations of these delayed - collapse hypernovae . in 2 ,
we describe both our initial stellar models and a description of our light - curve code .
we conclude with a discussion of our results and comparison of these results to the current data .
for initial conditions , we use the two 40m@xmath2 fallback grbs from @xcite with two different delays between the weak supernova explosion and the grb outburst : a 1s delay producing 0.33m@xmath2 of @xmath0ni , and a 6.8s delay producing 0.016m@xmath2 .
we will focus on the longer delay explosion in our efforts to produce a dim supernova explosion . at early times
, the radiation is completely trapped , so we grab the models from these calculations after the shock has moved a few times @xmath3 cm .
@xcite calculated the detailed yields from these models , but for our opacities we have reduced these yields to 30 base elements from hydrogen up to zinc .
figure [ fig : init ] shows the initial velocity structure used in our models .
note that these explosions are very fast with shock velocities @xmath4 in excess of @xmath5 , and the stars are rather compact ( @xmath6 cm ) so the shock will quickly break out of the star . on top of this basic stellar structure
, we have added a wind structure using the mass loss rates from the last @xmath710,000 - 100,000y in the life of our wolf - rayet stars based on models presented by @xcite : @xmath8 with velocities in the 700@xmath9 range .
note that the wind velocities could be as high as 3,000@xmath9 and it is likely that grbs arise from lower metallicity objects where the mass - loss rate is lower than our chosen values .
the primary effect of both of these is a lowering of the density in the wind .
we have included a run where the density is scaled down by a factor of 100 to study this effect .
the initial density profile of all our models are shown in fig .
[ fig : init ] . for our calculations , we use the multidimensional radiation - hydrodynamics code rage ( radiation adaptive grid eulerian ) , which was designed to model a variety of multimaterial flows @xcite .
the conservative equations for mass , momentum , and total energy are solved through a second - order , direct - eulerian godunov method on a finite - volume mesh @xcite .
it includes a flux - limited diffusion scheme to model the transport of photons using the levermore - pomraning flux limiter @xcite .
rage has been extensively tested on a range of verification problems @xcite and applied to ( and tested on ) a range of astrophysics problems @xcite , including the strong velocity gradients that exist in supernova explosions @xcite .
the rage code can be used in 1,2 , and 3 dimensions with spherical , cylindrical and planar geometries in 1-dimension , cylindrical and planar geometries in 2-dimensions , and planar geomeotries in 3-dimensions . for this paper , we focus on 1-dimensional , spherical calculations .
rage uses an adaptive mesh refinement technique , allowing us to focus the resolution on the shock and follow the shock as it progresses through the star .
even so , we were forced to regrid in our calculations to ensure that we resolved the shock at early times but still allow us to model the shock progression out to 100d ( the shock moves from @xmath3 cm out to @xmath10 cm in the course of a simulation ) . for our current set of calculations ,
the energy released from the decay @xmath0ni and @xmath0co is deposited directly at the location of the @xmath0ni using the following formula : @xmath11\ ] ] where @xmath12mev and @xmath13mev are the mean energies released per atom for the decay of @xmath0ni and @xmath0co , respectively , and @xmath14s , @xmath15s . especially at late times , this energy is not deposited into the matter surrounding it , but rather escapes the star .
we have included a simulation where after the shock has reached @xmath16 cm , we assume 99% of the energy escapes , allowing a lower bound on our models .
this provides us with 4 models in total : 1s delay , 6.8s delay , 6.8s delay with large escape fraction of gamma - rays , and a 6.8s delay with both a large escape fraction and a lowered wind . for opacities ,
we use the sesame opacities produced at lanl @xcite .
these opacities have been extensively used in astrophysics including many problems in supernovae ( e.g. fryer et al . 2000 ; deng et al .
2005 ; mazzali et al .
. this opacity base is being continually updated , and we use the latest updates on these opacities .
the assumption with these opacities is that the atoms are in local thermodynamic equilibrium with the matter and , hence , the opacity can be determined assuming a single temperature in each cell .
only a few atoms deviate significantly from their local thermodynamic solution and this affect should be minor for our calculations of bolometric light curves .
if there still is considerable internal energy in the shock when it reaches the edge of the star , the shock can accelerate even further when it reaches the wind . from figure
[ fig : init ] , we see this is the case for our short delay model , but not for our long delay models . at this time , roughly @xmath17 ( @xmath18 for our short delay model ) the shock is just becoming optically thin .
this occurs between @xmath19 when the shock is well beyond the stellar surface and out in the wolf - rayet wind . for our short - delay model ,
the shock has taken on a homologous velocity ( @xmath20 ) structure , but the long - delay models take more than @xmath21 to develop the canonical homologous velocity structure .
more important , note the density structure in these models . in the long - delay models ,
the weak supernova shock combined with the accretion of fallback material produces an evacuated region around the compact remnant .
when the strong explosions sweeps up the weak supernova shock , it produces a shell of fast moving material .
this ring of material causes the shock to remain optically thick longer than one would expect from more standard shock profilesni decay gamma - rays are identical . ] . clearly , the nature of the explosion will have a strong affect on our light - curves .
simple , homologous - velocity profiles , with power - law density profiles will not be able to model the broad range of explosion scenarios .
the short - delay explosion experiences both a shock break - out bump at 3days followed by a @xmath0ni - powered peak at roughly 25days after the launch of the explosion .
the only surprise in this model is that shock break - out for these strong explosions propagating through dense wind material can produce strong emission , in excess of @xmath22 !
this emission becomes the dominant emission in models where the @xmath0ni is small .
the light - curves produced by our long - delay ( @xmath23 low @xmath0ni yield ) simulations are shown in figure [ fig : lc ] .
these light - curves only get a large peak at shock break - out .
this sharp peak drops within 8 days of the launch of the explosion . by allowing the gamma - rays to escape
, we actually get a brighter peak luminosity ( the shock is slightly more compact and is hotter ) , but by 20 days , the model with 100% gamma - ray deposition is brighter and by 100d it has a luminosity that is an order of magnitude higher than the model with only 1% gamma - ray deposition .
these features can be explained if the bulk of the light - curve is powered by shock energy , not gamma - ray deposition .
the peak in the light curve is powered by shock break - out .
indeed , it appears that shock energy dominates out beyond 20 - 30days .
the fact that the supernova explosion is strong ( @xmath24erg ) and the @xmath0ni production is low ( @xmath25 ) , this is not unexpected .
in such a system , one would expect that the amount of mass in the wind is more important .
our simulation with a low wind is an order of magnitude dimmer than our strong wind candidates .
the strong shock drives the matter temperature ( and the emitted planck - averaged radiation temperature ) well beyond the optical bands .
but as the shock moves out of the envelope , the averaged radiation energy in the shock decreases ( fig .
it is the temperature in the shocks that drive most of the emission and this temperature is characteristic of the radiation temperature .
6 days into the explosion , when the luminosity peaks , the planck - averaged radiation temperature lies between 1.5 - 4ev for all models , with most of the photons emerging in the ultraviolet .
let s compare these models with the observational limits placed by fynbo et al.(2006 ) and gal - yam et al .
( 2006 ) . at 10 days for a grb at 388mpc ( roughly equivalent to a redshift of 0.089 corresponding to grb060505 ) ,
the bolometric luminosities for these three models are 23.5 , 24.7 , and 26.8 magnitudes for the 1% gamma - ray deposition , 100% gamma - ray deposition , and low wind models respectively . for the more distant grb060614 ,
the expected late - time flux is even lower .
it is likely that these models would all be below the observed limits at this time .
however , at shock break - out , these spherically - symmetric models predict a very bright luminosity ( above 20th magnitude at 6d ) . why do nt we see the burst at this time ?
although the afterglow is still strong at this time , our predicted luminosities could well dominate . but remember that the peak flux is in the ultraviolet , not the optical .
the fact that nothing is observed could place constraints on the model .
either there is considerable dust extinction ( a few magnitudes in the uv ) , the explosion is weaker than our spherically symmetric @xmath26erg explosion has assumed , or this grb truly is different than our standard long - duration burst classification as gehrels et al .
( 2006 ) and gal - yam et al .
( 2006 ) argue . we have learned a number of features of delayed black - hole forming grbs .
first , the emission of these supernovae are likely to peak at shock break - out .
the luminosity is dominated by the shock energy , and not the decay of radioactive elements .
as such , any observation places more constraints on the surrounding wind than it does on the @xmath0ni yield . also , the structure of the shock may well be very different than the canonical density and velocity structures seen in more normal supernovae and we must pay particular attention to the explosions in calculating these light - curves .
the models in these papers are examples of a larger set of possibilities .
finally , bear in mind that we are assuming spherically symmetric models run in the gray approximation and asymmetric , multi - group models may have very different light - curves .
this work was carried out in part under the auspices of the national nuclear security administration of the u.s .
department of energy at los alamos national laboratory and supported by contract no .
de - ac52 - 06na25396 . | the currently - favored model for long - duration gamma - ray bursts ( grbs ) invokes explosions from the collapse of a massive star down to a black hole : either directly or through fallback .
those grbs forming via fallback will produce much less radioactive nickel , and hence it has been argued ( without any real calculation ) that these systems produce dim supernovae .
these fallback black - hole grbs have been recently been argued as possible progenitors of a newly discovered set of grbs lacking any associated supernovae .
here we present the first ever radiation - hydrodynamics calculations of the light - curves produced in the hypernova explosion by a delayed - fallback gamma - ray burst .
we find that the bolometric light - curve is dominated by shock - deposited energy , not the decay of radioactive elements .
as such , observations of such bursts actually probe the density in the progenitor wind more than it does the production of radioactive nickel . |
the main goal of the _ chandra _ multiwavelength plane survey ( _ champlane _ ) is to identify and study the populations of low - luminosity ( @xmath1ergs@xmath2 ) accretion - powered binaries in the galaxy @xcite .
_ champlane _ includes a deep survey of the galactic bulge consisting of ( near-)simultaneous _ chandra _ and _ hubble space telescope _
( _ hst _ ) pointings of three low - extinction regions . with minimal
obscuration by dust , for the bulge , and positions progressively closer to the galactic center ( gc ) , these `` windows '' are chosen to trace the x - ray point - source population towards the gc .
the bulge population of interacting binaries is also of interest for comparison with those in globular clusters , where stellar encounters have enhanced the numbers of close binaries , and with x - ray populations in other galaxies .
here we report initial results of observations of baade s window ( _ bw _ , ( @xmath3,@xmath4)=(106,@xmath5383 ) ) and stanek s window ( _ sw _ , ( @xmath3,@xmath4)=(025,@xmath5215 ) ; * ? ? ?
* ) where we have found a significant number of sources with candidate m - giant counterparts . in @xmath6[sec_obs ] we describe the optical identification , and x - ray and optical / near - infrared spectra . as x - ray emission from single m - giants
is rarely detected @xcite , we have investigated whether these sources could be symbiotic binaries in which a cool giant typically transfers mass to a white dwarf via a wind ( @xmath6[sec_dis ] ) .
more results , including analysis of our `` limiting window '' at ( @xmath3,@xmath4)=(010,@xmath5143 ) will be presented in future papers .
we obtained _ chandra_/acis - i observations of _ bw _ ( obsid3780 ; 98 ks ) and _ sw _ on 2003 july 9 and 2004 february 14/15 , respectively .
observation was split in two exposures of 80 ks ( obsid4547 ) and 19 ks ( obsid5303 ) , with a gap of 12 , to keep the same roll angle ; these were stacked for the remainder of the analysis .
the _ champlane _
x - ray data reduction pipeline @xcite was used to create source lists from images in the 0.38 kev band and 95% confidence radii on positions @xmath7 .
a total of 365 ( _ bw _ ) and 389 ( _ sw _ ) acis - i sources were detected .
other pipeline products include net source counts between 0.58 kev , and energy quantiles @xmath8 corresponding to the energies below which @xmath9% of the counts are detected .
we searched the optical gravitational lensing experiment ii ( _ ogle - ii _ ) star catalog ( * ? ? ?
* u02 hereafter ) for counterparts . due to the high probability for spurious matches , we used only variable stars @xcite to measure the _ chandra _ boresight correction ( @xmath103@xmath1110@xmath12 stars in the acis - i areas , versus @xmath101@xmath1110@xmath13 ( _ bw _ ) and @xmath104@xmath1110@xmath13 ( _ sw _ ) variables ) .
as the u02 and wozniak et al.catalogs are not cross - linked , we first identified the variables in u02 using match criteria based on difference in position ( @xmath142 pixels or @xmath10082 ) and @xmath15 magnitude .
we used the boresight method described in @xcite adopting a 1@xmath16 error on optical positions @xmath172(u02 ) . the resulting _
chandra_@xmath5_ogle ii _ offsets ( @xmath18 , @xmath19 ) are ( @xmath20005(4),@xmath5005(5 ) ) and ( @xmath20022(3),@xmath5040(3 ) ) , based on 26 and 49 matches in _
bw _ and _ sw _ , respectively .
the light curves suggest many of these are interacting binaries for which we indeed expect enhanced x - rays . after correcting for boresight , we searched for counterparts in 2@xmath16 radii , with @xmath16 the quadratic sum of @xmath21 , @xmath7 ( normalized to 1@xmath16 ) and the boresight error . to reduce spurious matches we only consider acis - i sources with @xmath22 , yielding 351 and 384 _ bw _ and _ sw _
sources , for which 547 and 684 matches are found for 289 and 313 sources@xmath5@xmath15 color information ] , respectively . given several initial identifications with m - giants ,
bulge giants were selected to be at least as red ( @xmath23@xmath5@xmath24 for solar metallicity ; * ? ? ? * ) and at least as bright ( @xmath25 from 10gyr , @xmath26 isochrones by @xcite ) as m0iii giants at the gc ( 8kpc ) .
the maps by @xcite based on photometry of bulge clump giants were used to correct for reddening and extinction@xmath23@xmath5@xmath27 and @xmath28 are overestimated .
foreground m - giants with @xmath15@xmath291112 would however be saturated in u02 . ] .
four ( _ bw _ ) and 9 ( _ sw _ ) candidate counterparts satisfy the criteria ( table [ tab_prop ] , fig .
[ fig_cmd ] ) . repeating the matching for
randomly offset x - ray positions ( @xmath30 and @xmath31 varied with 45 increments on a 49@xmath1151 grid ) shows that in _ bw _ the average number of m - giant counterparts found by chance @xmath14@xmath32@xmath33 , is [email protected] , and that @xmath35 with 4.4% probability ; in _ sw _ @xmath14@xmath32@[email protected] , and @xmath36 with 0.64% probability .
star counts by @xcite predict @xmath290.4 m - dwarfs in the acis - i area that can contaminate our sample .
the inner 6@xmath116 of _ bw _ and _ sw _ were observed with the advanced camera for surveys on _ hst_. optical identification using these data is the topic of future papers .
we note that for the sources in table [ tab_prop ] in the _ hst _ pointings ( bw1 and 2 ) we find that all additional objects in the 2@xmath16 match circles have colors consistent with them being late - type main - sequence stars at @xmath37 kpc , which would not normally be luminous enough for detection ( @xmath38 ergs s@xmath2 , * ? ? ?
* ) in our _ chandra _ data .
( @xmath23@xmath5@xmath15 ) , @xmath28)@xmath39(0.84,0.81 ) in _ bw _ and ( 1.32,1.27 ) in _
sw_. matches from table [ tab_prop ] are labeled ; an open symbol indicates there are multiple u02-stars in the 2@xmath16 match circle .
the curves are the first and asymptotic giant branches of a @xmath40=0.019 , 10-gyr isochrone @xcite at 8 kpc ; the spread of stars around the isochrone can be due to distance or non - solar compositions .
expected colors for m - giants are marked @xcite .
[ fig_cmd ] ] llllllll i d & counts & @xmath41/ & @xmath42 & @xmath43@xmath44 & spt & _ ogle - ii _
i d + & & @xmath45 & & ( 8kpc ) & m@xmath46 & + bw1 & 10(5 ) & 0.7 & 2.3 & 31.3(2 ) & 3 & 45/113191/1472 + bw2 & 15(5 ) & 0.7 & 2.6 & 31.5(2 ) & @xmath46 & 01/640074 + bw3 & 12(5 ) & -1.2 & 2.2 & 31.7(2 ) & 4@xmath55 & 45/256236/1224 + bw4 & 30(7 ) & 0.0 & 2.7 & 32.2(1 ) & @xmath46 & 45/244411 + sw1 & 521(26 ) & 1.9 & 4.3 & 33.04(2 ) & @xmath46 & 39/126098 + sw2 & 30(8 ) & 2.6 & 4.5 & 31.8(1 ) & @xmath46 & 04/684358/4865 + sw3 & 54(9 ) & 1.5@xmath47 & 60@xmath47 & 32.88(7 ) & 5@xmath56 & 04/293339/4625 + sw4 & 19(6 ) & 1.1 & 3.8 & 31.6(1 ) & 3 & 04/293323/4310 + sw5 & 16(8 ) & 1.5 & 4.1 & 31.5(2 ) & @xmath46 & 04/065565/3277 + sw6 & 18(9 ) & 1.2 & 4.2 & 31.6(2 ) & 4@xmath55 & 04/065539/3113 + sw7 & 124(13 ) & 0.6 & 4.1 & 32.48(4 ) & 3@xmath55 & 04/266999/2810 + sw8 & 10(6 ) & 1.0 & 4.1 & 31.3(3 ) & 4@xmath56 & 04/648022/2915 + sw9 & 23(10 ) & 1.0 & 4.5 & 31.8(2 ) & 4@xmath56 & 04/051718/2234 + since most of the selected sources have only 1030 counts , we use quantile analysis @xcite to derive spectral information .
their median energies @xmath48 are relatively high with all _ bw _ and 5 _ sw _ sources having @xmath49 kev .
of all the detected sources in _ bw _ and _ sw _ , only 14% ( _ bw _ ) and 20% ( _ sw _ ) have @xmath49 kev .
spectra for about half of the sources can be described with a bremsstrahlung model ( @xmath50 kev ; fig .
the remaining sources are too hard ; power - law models constrain their photon indices to @xmath0 ( fig .
2b ) . adding a broad 6.4 kev emission line brings these points even closer to the grid and increases @xmath41 ( fig .
2c ) ; this suggest the presence of a fluorescent fek line to avoid the unphysical @xmath51 in the hardest sources .
the column density @xmath52 is consistent with the optically - derived value for bulge stars in _ bw _ and _ sw_. an exception is sw3 for which the inferred @xmath52 exceeds the field value by a factor of @xmath1015 , suggesting the source is intrinsically absorbed .
table [ tab_prop ] summarizes quantile results for power - law models . only for sw3 ,
for which @xmath41 is not well - constrained , is a bremsstrahlung model used to estimate @xmath52 and @xmath45 .
for the two sources with @xmath53100 counts , source spectra were grouped to have @xmath54 counts bin@xmath2 and were corrected for background .
sherpa _ to fit power - law and thermal - bremsstrahlung models , and account for absorption by neutral hydrogen .
fixing @xmath52 to the appropriate values from @xcite gave acceptable results for both sources , in agreement with quantile analysis ; fitting for @xmath52 does not yield significant improvements ( table [ tab_specres ] ) .
lllllll i d & @xmath52 & @xmath41 & @xmath55/dof & @xmath52 & @xmath45 & @xmath55/dof + & & & & & ( kev ) + sw1 & ( 4.3 ) & [email protected] & 0.52/22 & ( 4.3 ) & 3.4@xmath56 & 0.60/22 + & 2.6@xmath57 & [email protected] & 0.51/21 & 2.2@xmath57 & 6.0@xmath58 & 0.52/21 + sw7 & ( 4.1 ) & 0.44@xmath59 & 0.014/4 & ( 4.1 ) & @xmath14335 & 0.27/4 + & 8.4@xmath60 & 0.72@xmath61 & 0.012/3 & 17@xmath62 & @xmath14159 & 0.05/3 + optical spectra were obtained with the flwo-1.5 m fast spectrograph in 2004 may and june ( bw3 ; sw3 , 4 , 6 , 7 , 9 ) , and the ctio-4 m hydra multi - object spectrograph ( bw1 , 3 ; sw3 , 6 , 8 , 9 ) in 2004 june and 2005 june .
fast spectra cover 38007500 with 3 resolution while hydra spectra cover 40006800 with 4.9 resolution .
stars were classified using templates in @xcite and tio - band strengths @xcite , see table [ tab_prop ] .
h@xmath30 , and h@xmath30 and h@xmath63 are in emission in sw7 and 9 , respectively , but no strong high - ionization emission lines ( often present in spectra of symbiotics , see @xmath6 [ sec_dis ] ) are observed .
no variability is seen for stars observed in multiple runs .
near - infrared spectra of the hei 1.083@xmath64 line were obtained for bw3 and sw7 on 2005 may 19 using keck - ii / nirspec in high - resolution cross - dispersed echelle mode .
each observation consisted of a single ab nod pair of exposures with 900s integrations at both slit positions .
strong p - cygni profiles are seen , indicative of wind velocities of 230 ( bw3 ) and 180kms@xmath2 ( sw7 ; fig.[fig_nirspec ] ) .
similar profiles in ultraviolet lines are found in symbiotic star systems , for example agdra @xcite .
single m - giants do not show hei 1.083@xmath64 with the emission strength and absorption as in bw3 and sw7 @xcite . ]
we explore possible explanations for the x - ray emission from these giants .
single m - giants are rarely detected in x - rays but @xcite point out a few candidates that are matched securely to x - ray sources ( @xmath65 erg s@xmath2 for a 10@xmath66 k coronal plasma , but @xmath67 is @xmath1015 times larger for a @xmath68 power law , for @xmath69 @xmath70 ) .
their x - rays are yet unexplained , but the absence of optical emission lines is used to argue against a symbiotic nature . in rscvn binaries ,
activity of a ( sub)giant is increased by tidal coupling of the stellar rotation to the orbit , but no rscvns with m - giants are known .
tidal forces in a detached binary wide enough to fit an m - giant are too weak to spin up a main - sequence companion
. moreover , neither single stars nor active binaries are in general as hard as some of our sources . of the @xmath10200 symbiotic binaries currently known ,
@xmath1080% contain an m36 red giant @xcite .
mass transfer occurs mostly via a wind as opposed to roche - lobe overflow .
@xcite show in a rosat study that symbiotics are typically soft ; nuclear burning on the white dwarf produces very soft emission ( @xmath71 kev ) , while somewhat harder x - rays are associated with hot ( @xmath72 kev ) shocked gas in the colliding winds of the giant and hot star .
note that our sample contains many hard sources ( @xmath73% , @xmath6[sec_xspec ] ) that may be similar to the hardest sources in mrset et al .
emission above a few kev is so far poorly studied for symbiotics ; possible explanations include accretion ( chcyg ; * ? ? ? * ) or shocked gas ( zand ; * ? ? ?
recently , symbiotics have been detected also in very hard x - rays by _ integral _ and _ swift _
( e.g.rtcru , * ? ? ?
x - ray luminosities of symbiotics range from 10@xmath74 to 10@xmath75 erg s@xmath2 , up to @xmath1010@xmath76 erg s@xmath2 for the neutron - star accretor gx1 + 4 .
we estimate the upper limit on the temperature of the shock - heated gas in the colliding - wind model by requiring that the thermal energy per gas particle does not exceed the kinetic energy .
a wind velocity @xmath77 km s@xmath2 gives @xmath78 kev , lower than implied for any of our sources by quantile analysis .
thus , if the hei 1.083@xmath64 lines in bw3 and sw7 ( two of the hardest sources ) trace the conditions in the shocked winds , this model is an unlikely explanation . to reach @xmath79 kev as in sw3 , requires @xmath80 km s@xmath2 .
a mass - loss rate @xmath81 @xmath82 yr@xmath2 gives a total kinetic luminosity 1/2@xmath83 erg s@xmath2 , sufficient to power @xmath84=7.6@xmath11(@xmath85/8kpc)@xmath8610@xmath87 erg s@xmath2 for sw3 .
looking for such fast outflows would be one test whether shocked winds are indeed a plausible explanation for our soft sources . to check whether accretion is a viable source of x - rays , we use eq . 3 from @xcite for the total accretion luminosity @xmath88
we choose masses for the giant and white dwarf of 1 and 0.6 @xmath82 , a white - dwarf radius of 10@xmath89 cm ,
@xmath81 @xmath82 yr@xmath2 , and @xmath77 km s@xmath2 ; the orbital speed of the white dwarf is assumed to be large compared to the local sound speed .
the range of known orbital periods @xmath90 for symbiotics ( 2005700 d ) implies 6@xmath1110@xmath91 erg s@xmath2 , scaling linearly with @xmath92 . thus it is likely that accretion onto a white dwarf can power the x - rays : @xmath93 = ( 2.0110)@xmath11(@xmath85/8 kpc)@xmath94 10@xmath95 erg s@xmath2 ( table [ tab_prop ] ) .
accretion onto a neutron star is less probable : for a 1.4 @xmath96 , @xmath97 km companion , @xmath98 erg s@xmath2 for the assumed @xmath92 .
assuming we observe the white dwarf through the giant s wind , we estimate @xmath52 for the simplified case of a circular orbit and a spherically symmetric neutral - hydrogen wind of constant velocity : @xmath99/(16@xmath100@xmath101@xmath102 ) @xmath103 , with @xmath102 the semi - major axis , @xmath104 the orbital inclination , and @xmath105 the orbital phase . for the parameters and range in @xmath90 as
above , 9@xmath1110@xmath106 2@xmath1110@xmath107 @xmath70 for @xmath108 , below or similar to the galactic @xmath52 for _ bw _ and _
sw_. @xcite have found a mass - loss rate @xmath10910@xmath110 @xmath82 yr@xmath2 for red - giant symbiotics which allows for an enhanced @xmath52 , but values as high as for sw3 also require a larger @xmath104 .
quantile analysis suggests 6.4 kev fek emission in our hardest sources , which may be another manifestation of the giant s cool wind via interaction with hard x - rays ; fek lines are observed in e.g. chcyg @xcite and rtcru ( j. sokoloski 2006 , private communication ) .
how do other properties of our sources compare to those of m - giants and symbiotics ?
variability and mass loss , both associated with stellar pulsations , are typical properties of single late m - giants .
@xcite found that many m5 and all m6 and later giants in baade s window are variables .
of the 10 matches included in @xcite , the matches to bw1 , sw7 , 9 , and possibly sw3 , 4 , 5 and 8 show semi - regular variability as commonly observed in late m - giants ( amplitudes @xmath111 mag , time scales 20200 d ) .
@xcite note that bw3 however shows more regular variability .
the period they find in _ macho _ data ( 135.5d ) is the same as what we find in _ ogle - ii _ data ( [email protected] ) .
if this is due to ellipsoidal variations then @xmath112 d. variability of sw2 , 6 and possibly 8 is irregular with @xmath113 in 4 seasons . from mid - infrared photometry @xcite derive @xmath114 and @xmath115 @xmath96 yr@xmath2 for the matches to bw1 and 4 , respectively .
outflow velocities for bw3 and sw7 are high for single red giants ( @xmath11630 km s@xmath2 ) but a similar velocity has been derived from hei 1.083@xmath64 for gx1 + 4 @xcite .
optical spectra of our matches lack strong high - ionization nebular emission lines .
the presence of such lines is one of the original defining properties of symbiotics .
examples of `` weakly symbiotic '' ( i.e. without strong nebular emission ) systems that are hard in x - rays are 4u1700 + 24 and 4u1954 + 319 ( both suspected neutron - star systems that look like normal m - giants in the optical , * ? ? ?
* ) and the aforementioned _ integral _ and _ swift _ sources , for which h@xmath30 and h@xmath63 are the strongest optical emission lines . if these are similar to our sources , the large fraction of such systems in our sample then suggests these properties are characteristic for a bulge sub - population and/or not identified previously due to sensitivity limits or limited astrometric precision
. further study of the _ bw _ and _ sw _ sources but also those of @xcite is therefore of interest for estimates of the total number of galactic symbiotics and their progeny , which may include snia systems .
the evolution of symbiotics depends on the orbital period .
long - period systems become wide double white dwarfs . for @xmath117
d as for bw3 , and 1 and 0.6 @xmath82 stars , the roche - lobe radius is likely @xmath118 . in the models by @xcite , a 1@xmath82 star reaches @xmath119 at the tip of the red - giant branch and expands even further as an asymptotic giant .
if bw3 is a binary , it may start unstable mass transfer leaving a close double white dwarf , which could evolve to a snia .
we use the results by @xcite to constrain the number of symbiotics among the initial sample of 1453 still unclassified faint , hard x - ray sources in the 10 @xmath11 10 region around the gc @xcite . for @xmath120 and @xmath121 , @xmath122 for m - giants at the gc .
laycock et al .
find that @xmath123% ( 90% confidence ) of the 110 hard sources within 1 from the gc can have counterparts with @xmath124 , while this is true for @xmath125% of the 1343 hard sources at larger offsets .
these likely include high - mass x - ray binaries with b0 v primaries but also symbiotics .
we thank j. sokoloski for discussions and comments on the manuscript .
this work is supported by chandra grants go3 - 4033a and ar6 - 7010x .
we are grateful for the use of the keck observatory on mauna kea . | we have searched the _ ogle - ii _ archive for candidate counterparts of x - ray sources detected in two low - extinction windows included in our galactic bulge _
chandra / hst _ survey .
we find that a significant number i.e . in excess of the expected level of random associations
can be matched with probable m - giants .
their x - ray properties can be understood if these sources are symbiotic binaries where the x - rays are typically , either directly or indirectly , the result of a white dwarf accreting from the wind of a cool giant .
optical and near - infrared properties of selected sources are consistent with a symbiotic nature , although none of the spectra collected for 8 out of 13 candidate counterparts show the high - ionization nebular emission lines observed for many symbiotics .
the hard x - ray emission for several sources ( power - law photon indices @xmath0 ) suggests our sample includes systems similar to the symbiotics recently detected with _
integral _ and _ |
the existence of supermassive black holes ( bhs ) in normal galaxy nuclei was predicted forty years ago by @xcite , but until fifteen years ago it was still considered an interesting possibility which had to be demonstrated .
nowadays bhs are regarded as a crucial ingredient for our understanding of how galaxies form .
key to this paradigm shift was the launch in 1990 of the hubble space telescope ( hst ) .
it all started with the realisation that the mass of the bh is correlated to other global characteristics of the host galaxy as a whole . initially a correlation @xmath8 was found between the mass of the bh and the luminosity of the host - galaxy stellar spheroid @xcite . in 1997
the installation of the stis long - slit spectrograph on hst allowed the spatially - resolved kinematical observations to probe inside the radius of the subarcsecond bh sphere of influence @xmath9 in nearby galaxies ( @xmath10 being the velocity dispersion of the stars in the galaxy ) .
the increased accuracy in the determinations contributed to the discovery of the tighter @xmath11 correlation @xcite .
as @xmath10 is a good predictor of galaxy properties , it is perhaps not surprising that similar correlations were found between and respectively the galaxy concentration @xcite , the dark - halo mass @xcite , the bulge mass @xcite and the stars gravitational binding energy @xcite .
the existence of these correlations is broadly consistent with a scenario in which the bh regulates the galaxy formation , during the hierarchical galaxy merging , by shutting off the conversion of gas into stars via a feedback mechanism due to its powerful outflows @xcite . even though the current scenario explains a number of observed facts , our interpretation of the role of bhs in galaxy formation is far from secure .
one of the problems lies in the fact that the models have few observables to compare with , mainly the galaxy mass ( or @xmath10 ) and .
moreover , even after a decade of hst spectroscopy and models , only about 40 secure bh determinations exist ( compilations are given e.g. in * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . with few exceptions
the bh determinations have been performed with a single dynamical tracer ( either gas or stars ) , so that no independent test of the two measurements methods could be made .
very little is known about the orbital distribution near the bhs , which is expected to contain key information on the bh accretion process @xcite .
the advent of integral - field spectroscopy on all the 810-m class telescopes , combined with adaptive optics ( ao ) to reduce the blurring effect of the earth s atmosphere is opening a new epoch for bh studies .
the large mirrors allow for shorter exposure times and higher signal - to - noise ratios ( @xmath12 ) of the observations , than what was possible with the 2.4-m mirror and the stis long - slit spectrograph of hst .
observations at near - infrared wavelengths allow dust absorption effects to be virtually eliminated .
the integral - field observations , due to the tight constraint on the orbital distribution , dramatically improve the accuracy of determination , for a given spatial resolution and @xmath12 @xcite .
integral - field data are also needed for a unique recovery of the orbital distribution from the observations @xcite .
this can be understood from dimensional arguments , considering that most orbits in a stationary potential conserve three isolating integrals of motion .
this three - dimensional orbital distribution can not be recovered without the knowledge of at least another three - dimensional observed quantity .
motivated by these arguments , determinations from ao - assisted integral - field observations in the near - infrared are starting to appear in the literature @xcite .
the elliptical galaxy ngc 5128 ( centaurus a ) is a prime candidate for ao - assisted integral - field observations .
at least nine independent distance @xmath13 determinations for cen a , based on different methods , are available in the literature ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the median value is @xmath14 mpc , with extreme ranges of 3.4 mpc and 4.4 mpc respectively .
here we adopt a value does not affect our results but sets the scale of our models in physical units .
specifically , lengths and masses scale as @xmath13 , while mass - to - light ratios scale as @xmath15 . ]
@xmath16 mpc to be consistent with all the earlier papers on determination on this galaxy @xcite .
at this close distance cen
a is the nearest elliptical galaxy and one arcsec corresponds to 17 pc .
cen a is among the only @xmath17 galaxies on the whole sky with an observed @xmath18 ( see * ? ? ?
* for a partial list ) .
moreover cen a possesses a nuclear gaseous disk in regular rotation from which a number of independent determinations of have been performed , with ever increasing accuracy @xcite .
this allows for an accurate comparison between the bh mass derived with gaseous or stellar kinematics .
the @xmath0-band central surface brightness of cen a is quite bright at @xmath19 mag @xcite , which allows a high @xmath12 in the stellar spectra to be achieved in reasonable exposure times .
a bright star , sufficiently close to the nucleus , can be used as reference for the ao correction .
all these facts make cen a a unique observational benchmark for determinations in the near - infrared . in section 2
we describe our data and the extraction of the stellar kinematics . in section 3
we present the stellar dynamical models .
we discuss our results in section 4 and we finally summarize them in section 5 .
for the dynamical modeling of the stellar kinematics we used integral - field spectroscopy obtained with sinfoni on the ut4 ( yepun ) of the very large telescope of the european southern observatory on the cerro paranal .
sinfoni consists of the cryogenic near - ir integral field spectrometer spiffi @xcite coupled to the visible curvature ao system macao @xcite .
we observed the nucleus of cen a with two different spatial scales : 0250@xmath200125 ( 250mas scale ) with a field of view ( fov ) of 8@xmath208and 010@xmath20005 ( 100mas scale ) with a fov of 32@xmath2032 .
the 100mas observations and data reduction were already described in @xcite .
the observations were performed in excellent seeing conditions of 05 fwhm , as measured by the seeing monitor in the @xmath21-band . with the 250mas spatial scale
the observations were taken in natural seeing , while with the 100mas scale the sinfoni ao module was locked onto an r@xmath2214 mag star about 36southwest of the nucleus .
although the reference star is relatively distant from the galaxy nucleus , the good seeing allowed us to achieve a nearly diffraction - limited correction in the @xmath0-band .
both data sets were obtained in the @xmath0-band , which covers the wavelength range 1.932.47 .
the spectral resolution of the observations was @xmath23 in both scales , and corresponds to an instrumental dispersion of @xmath24 .
the 250mas observations followed an object - sky - object sequence with equal exposure time of 300 s each , for a total on - source exposure time of @xmath25 s. the sky exposure were taken at 200from the nucleus to make sure the spectra were not contaminated by the large galaxy .
the 100mas observations followed a similar sequence with individual exposure times of 900 s each and a total on - source exposure time of @xmath26 s. the different exposures were dithered with shifts of 02 to allow for the removal of detector defects and cosmic rays .
the data were reduced using the sinfoni data reduction pipeline provided by eso as described in section 2 of @xcite .
the spatial point - spread - function ( psf ) of the 100mas observation was determined in section 2.1 of @xcite from a fit to the non - thermal nucleus , which is unresolved down to a 006 ( fwhm ) spatial resolution @xcite .
the psf can be approximated by two gaussians with fwhm of 012 and 030 respectively , with the smallest gaussian containing 17% of the total flux .
a two - dimensional fit with a single gaussian to the nucleus in the reconstructed image from the data cube gives a fwhm of 017 .
for the stellar dynamical modeling the integral - field high - spatial resolution sinfoni observations are essential to tightly constrain the bh mass and stellar orbital distribution .
however they are not sufficient as they sample only a small fraction of the half - light radius of cen a ( @xmath27 in @xmath0-band ; @xcite ) .
@xcite showed that one has to sample with the kinematics a significant fraction of to accurately constrain bh masses .
for this reason in this paper we also use the @xmath0-band kinematics obtained with the gemini near infrared spectrograph ( gnirs ) at gemini south by @xcite .
the long - slit observations extend to a distance of @xmath28 from the galaxy nucleus and were obtained at two position angles : ( i ) centered on the nucleus , along the galaxy major axis at large radii ( @xmath29 from north through east ) , and ( ii ) along the galaxy minor axis ( @xmath30 ) , but offset from the nucleus by 085 . in this paper
we increased the gnirs @xmath10 by 5% to match our sinfoni kinematics .
differences at this level between the two datasets are almost unavoidable and are likely due to low - level systematic calibration errors , or differences in the adopted stellar templates .
for this reason we can not state which of the two datasets has the proper absolute @xmath10 calibration . as none of the results in this paper depends on whether we apply the small shift to the gnirs or sinfoni data , we arbitrarily decided to increase the gnirs data to match our sinfoni ones , which we adopt as reference .
a 5% uncertainty in @xmath10 approximately translates into a 10% uncertainty in the @xmath31 .
the individual sinfoni exposures were registered using reconstructed images and merged into a single data cube . in the process
they were resampled to a spatial scale of 0125@xmath200125 ( 250mas scale ) and 005@xmath20005 ( 100mas scale ) .
the stellar kinematics requires a high @xmath12 for an unbiased extraction from the spectra . for typical observational setups , values on the order of @xmath32 per spectral pixel ,
are generally required to detect deviations of the line - of - slight velocity distribution from a simple gaussian shape ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
for this reason long - slit data are invariably adaptively binned along the spatial direction , before the kinematic extraction ( e.g. * ? ? ?
the goal of binning is to maintain the maximum spatial resolution , given a constraint on the minimum @xmath12 .
four our integral - field data we use the voronoi binning technique , mxc / idl/ ] which constitutes an optimal solution to this problem in two - dimension @xcite . given that the observations are dominated by the photon noise
, we spatially binned the data by requiring an equal number of counts in each spatial bin . in this way
we did not have to rely on the , often inaccurate , noise propagation by the reduction pipeline , to estimate the @xmath12 .
this constant flux per bin is achieved in the voronoi binning algorithm by setting the input noise @xmath33 , where @xmath34 is the total flux ( in arbitrary units ) in each unbinned spectrum . for the 100mas scale
the target @xmath12 for the binning was fixed by the requirement for the data to be unbinned ( pixel size 005@xmath20005 ) within a radius @xmath35 and start being binned at larger radii , where the surface brightness decreases .
this led to 622 voronoi bins , out of the original @xmath36 pixels .
similarly for the 250mas scale we required the data to be unbinned ( pixel size 0125@xmath200125 ) within @xmath37 , leading to 554 bins .
the actual minimum @xmath12 achieved by the adopted binning was determined afterwards , from the rms residuals of a spectral template fit ( section [ sec : ppxf ] ) , to be @xmath38 for both spatial scales .
the near - infrared @xmath0-band spectral region is dominated by the strong stellar absorption feature of the 2.30 ( 2 - 0 ) @xmath39co band head . at this wavelength
the galaxy spectrum is dominated by the light from cool and evolved giant stars .
the strength of the co absorption is very sensitive to the star surface gravity @xcite and this implies that , contrary to the optical region , in the @xmath0-band it is _ essential _ to include giant stars in the construction of an optimal stellar template for an unbiased extraction of the stellar kinematics @xcite .
the extraction of the stellar kinematics was performed using the penalized pixel - fitting method@xmath40 ( ppxf ; * ? ? ?
* ) , which fits the logarithmically - rebinned spectra with a template convolved with a line - of - sight velocity - distribution described by a gauss - hermite expansion @xcite .
the method allows the template to be carefully optimized during the kinematics fit , it permits emission lines to be easily masked , and it includes a penalty criterion to deal with low @xmath12 or insufficient resolution in the spectra . as library of stellar templates
we used a set of 11 dwarfs and giants stars ( luminosity class ii
v ) of late spectral types ( k m ) , observed with the same instrumental setup as for the cen a observations . the optimal positive linear combination of the 11 templates was determined only once , using a spectrum with very high @xmath41 , obtained by co - adding all the spectra in the 250mas sinfoni observations , while excluding the unresolved non - stellar nucleus . as expected the flux in the resulting optimal template fit
is dominated by a giant m5iii star ( 52% of the flux ) .
this star provides by itself an excellent fit to the central spectrum of cen a. however the contribution from a dwarf m0v star ( 30% of the flux ) is required to properly reproduce the depth of the na feature at 2.21 ( cf . * ? ? ?
the template provides a very accurate description of the spectrum from 2.22.4 , which contains all the significant stellar absorption features in the @xmath0-band ( fig .
[ fig : template_fit ] ) . 8 250mas field ( black solid line ) is compared to the broadened optimal template determined with ppxf ( red solid line ) .
the fit residuals , vertically shifted by an arbitrary amount , are shown as blue dots .
the gaps in the residuals correspond to some non well corrected telluric features and two h@xmath42 gas emission lines at 2.23 and 2.25 , which were excluded from the fit.,title="fig : " ] for maximum consistency the same combination of stellar templates was used to extract the kinematics at all spatial positions and for both the 250mas and 100mas observations .
additive polynomials of degree four were used to allow for possible variations in the stellar line - strength at different radii , to account for imperfections in the sky subtraction or spectral calibration , and to model the contribution of the non - thermal nucleus ( section [ sec : nonthermal ] ) .
consistent results were obtained using polynomials of degree 18 . from numerous tests on other galaxies we found that , even for the most extreme gradients in the stellar population , the additive polynomials are sufficient to approximate possible low - frequency variations in the line - strength , once the global galaxy template is accurately determined from a high-@xmath12 spectrum , to account for the high - frequency content in the spectrum .
we prefer this faster and more robust approach than trying to fit the template mix in every bin , from spectra of lower @xmath12 .
generally both approaches give nearly indistinguishable results , and we tested that this is true also in the present case .
we fitted with ppxf the mean velocity @xmath21 , the velocity dispersion @xmath10 and the first two gauss - hermite parameters @xmath43 and @xmath44 @xcite , using the same spectral range and masked regions as in the real data .
we adopted a penalty @xmath45 ( defined in equation 12 of @xcite ) .
the losvd is always very well sampled by our observations and the @xmath12 is high , so any @xmath46 gives essentially the same results and the measurements are virtually unbiased ( fig .
[ fig : ppxf_simulation ] ) .
the @xmath47 measurement errors were determined as the biweight dispersion @xcite of 100 monte carlo realizations . for the determination of realistic errors we used non - penalized fits with @xmath48 . at our minimal @xmath49 and for @xmath50 the typical random errors in @xmath21 , @xmath10 , @xmath43 and @xmath44 , are 4 , 5 , 0.02 and 0.03 respectively .
the kinematics extraction was performed while fitting precisely the same wavelength region , from 2.252.37 , in both the 250mas and 100mas sinfoni observations ( fig .
[ fig : spectrum_variation ] ) .
this spectral region includes the ( 20 ) @xmath39co band head and three other prominent band heads of the @xmath39co series .
emission from the highly ionized species of [ caviii ] at 2.32 and an h@xmath42 line at 2.35 were excluded from the fits ( see also * ? ? ?
* ) . of our data and an losvd with @xmath51 , @xmath52 and @xmath10 in the range between 30 and 240 .
we extracted the kinematics with ppxf and a penalty @xmath45 .
the lines in the top two panels show the 50th ( median , solid line ) , 16th and 84th percentiles ( 1@xmath10 errors , dotted lines ) of the differences between the measured values and the input values of the mean velocity @xmath53 and the velocity dispersion @xmath54 .
the bottom panels show the same lines for the recovered values of @xmath43 and @xmath44 , compared to the input values ( dashed line ) . in the range of our observations @xmath55
the recovered values are virtually unbiased.,title="fig : " ] the extracted sinfoni kinematicsmxc / cena09 . ] in the two scales is shown in fig . [ fig : kinematics250 ] and fig .
[ fig : kinematics100 ] .
for the first time our observations detect a low - level clear sense of stellar rotation in cen a. we used the method of appendix c of @xcite to determine the global kinematical major axis @xmath56 and the systemic velocity of @xmath57 .
a maximum velocity of just @xmath58 is reached along @xmath59 at about @xmath60 from the nucleus , at the edge of our 250mas sinfoni fov .
the @xmath10 field shows a gradual increase towards the center , before a sudden drop in the very center , where the non - thermal continuum prevents a proper kinematics extraction .
the @xmath43 field is symmetric about the center , indicating low template - mismatch , and anti - correlated with @xmath21 , as generally observed in early - type galaxies @xcite , and the @xmath44 field is generally close to zero over the whole field .
, the velocity dispersion @xmath10 and the two gauss - hermite moments @xmath43 and @xmath44 .
overlaid are the contours of the surface brightness derived from the reconstructed image , in 1 mag intervals .
north is up and east is to the left.,title="fig : " ] for the sinfoni 100mas observations .
the color levels are the same in this figure and in fig .
[ fig : kinematics250].,title="fig : " ] the nuclear stellar rotation is _ counter - rotating _
( by about 180@xmath61 ) with respect to the regular h@xmath42 nuclear gas rotation presented in ( * ? ? ?
* their fig .
the stars also rotate much slower than the gas , which reaches a maximum velocity @xmath62 at @xmath63 .
this indicates that the recent gas acquisition was not able to produce a significant fraction of stars near the nucleus .
this is consistent with the lack of evidence for any change in the nuclear stellar population of cen a. this is different from a similar case of gas versus stars counter - rotation observed in the nucleus of the spiral galaxy ngc 5953 @xcite . in that case
the nuclear gas rotation is associated to evidence of young stars which co - rotate with the gas and counter - rotate with respect to the outer part of the galaxy . in the case of cen
a adopting a fixed template becomes important very close to the center , due to the presence of the non - thermal nucleus @xcite . at radii where the nucleus dominates ,
the stellar absorptions are diluted , resulting in a strong decrease in the observed line - strength @xmath64 . due to the strong correlation between @xmath64 and the stellar velocity dispersion @xmath10 ( section 2.2 of @xcite ) , an inaccurate modeling of the non - thermal dilution can cause large errors in the measured @xmath10 . similarly the rise in the non - thermal continuum
could be incorrectly interpreted as a variation of the stellar population , requiring a change in the stellar template mix and also producing an error in @xmath10 .
as there is no evidence for a sudden change in the population in the nucleus of cen a the safest choice is to assume the stellar template is fixed and to model the non - thermal continuum via additive polynomials in ppxf .
additive polynomial still allow for low - order variations in the stellar line - strength and account for possible instrumental effects .
this approach allows the @xmath10 to be reliably extracted in the high - resolution 100mas observations down to @xmath65 , before the photon noise of the nucleus eliminates all stellar information from the spectra . as an illustration of the importance of taking the non - thermal continuum accurately into account in the kinematic extraction , in fig .
[ fig : spectrum_variation ] we show the best fit with ppxf , using a gaussian losvd , to the spectra extracted at different radii from the 100mas sinfoni data .
we also performed separate fits including the gauss - hermite parameters @xmath43 and @xmath44 . in the range 031 , where we can trust our extraction , the losvd is essentially consistent with a gaussian as @xmath66 , in agreement with fig .
[ fig : kinematics100 ] .
the high @xmath12 spectra in these plots were obtained by co - adding all the spectra within circular annuli of one pixel width ( 005 ) . the standard kinematics extraction approach for this paper
consists of modeling the observed spectrum as the sum of a convolved fixed optimal template plus a fourth degree additive polynomial . to test the sensitivity of the measured @xmath10 to the details of the extraction , especially in the critical continuum - dominated nuclear region
, we compared our standard approach with others using different options in the ppxf routine : ( i ) we use a first degree polynomial ; ( ii ) we fit the optimal template at each radius using our 11 stars ; ( iii ) we use as template a single m5iii star ; ( iv ) we include in the fit the additive contribution of a scaled version of the nuclear non - thermal spectrum ( as described in @xcite , to subtract the sky spectrum ) .
all these four approaches give dispersion profiles which agree within the statistical errors for @xmath67 .
the last approach is the one which produces the best fit and describes almost every details of the observations .
this is the one we show in fig .
[ fig : spectrum_variation ] .
we also give in the plots the line - strength @xmath64 , defined as the ratio between the mean flux within the fitted range in the observed cen a spectrum , and the flux contributed by the best fitting stellar template alone in the same range . at @xmath35
the non - thermal source dominates and the spectrum shows a nearly linear trend . at @xmath68
it becomes possible to estimate the velocity dispersion , however the stars still contribute only 19% of the flux in the observed spectrum , the statistical errors are large and the determination is still affected by large systematic uncertainties . at this radius the measured @xmath10 is sensitive to the details of the extraction . only at larger radii the kinematics can be reliably extracted and will be included in our models . and one pixel width ( 005 ) .
the best - fitting ppxf model ( red solid line ) consists of the stellar template ( fig .
[ fig : template_fit ] ) convolved with a gaussian losvd plus a fourth degree additive polynomial , plus a scaled copy of the non - thermal nuclear spectrum ( top panel ) .
the residuals are shown at the bottom of each panel with the blue dots .
the value @xmath69 at @xmath70 is likely due to a small difference in the line strength of the stellar spectrum , or in the sky subtraction , with respect to the adopted optimal template .
emission from the [ caviii ] line at 2.32 and the h@xmath42 line at 2.35 were excluded from the fits ( grey areas ) .
these lines are prominent also outside the nucleus and can not be properly subtracted by our model . _
right column : _ the convolved optimal template ( red solid line ) is compared to the observed spectrum after subtraction of the nuclear spectrum and polynomial contributions ( black solid line ) .
the blue dots show the residuals.,title="fig : " ] fig .
[ fig : spectrum_variation ] shows that the non - thermal continuum is still present in the spectra at the 25% level at @xmath63 .
the effect of the non stellar continuum is still clearly visible in the varying slope of the spectrum .
still at all radii the stellar spectrum is well approximated by the fixed convolved template , once the continuum is removed ( right column of fig .
[ fig : spectrum_variation ] ) . to better quantify the extent of the dilution due to the central non - thermal continuum fig .
[ fig : gamma_profile ] shows the surface brightness profile @xmath71 measured from the reconstructed image of the 100mas data cube , together with the radial profile of the the pure stellar contribution , estimated as @xmath72 from the result of the ppxf fit in the individual bins ( see @xcite for a similar analysis on the non - thermal nucleus of m 87 ) .
the plot shows that the underlying galaxy profile is smooth and well approximated by a shallow power - law @xmath73 , as expected in a ` core ' elliptical @xcite .
this confirms that the observed radial spectral variation is consistent with being entirely due to the psf effect and not to an intrinsic change in the stellar population .
the measurement of @xmath74 from the spectra allows for an extremely accurate determination of the halo of the psf , which would be difficult to disentangle from the underlying galaxy via photometry alone . as measured from the individual pixels in the sinfoni 100mas scale ( filled squares )
is compared to the estimated profile of the stellar light @xmath72 ( red open circles ) . as expected for a normal galaxy profile ,
the underlying stellar light profile is smooth and well approximated by a power - law @xmath73 ( solid line ) .
the dashed vertical line marks the radius where the psf has reached 50% of its peak value.,title="fig : " ]
cen a appears very round on the sky and shows little rotation within one .
@xcite mapped the full velocity field of cen a with 71 slit positions out to 100 from the nucleus .
they found a maximum rotation of around 40 roughly along the direction of the major axis at @xmath75 .
@xcite find a comparable maximum rotation of about 20 within 40 .
adopting a luminosity - weighted dispersion @xmath76 @xcite , the amount of rotation can be quantified by the parameter @xmath77 @xcite . from the empirical relation @xmath78 ( equation [ 23 ] of @xcite ) , we estimate for cen a @xmath79 @xcite .
the ellipticity of the galaxy within 1 , measured from 2mass @xmath0-band photometry , is very small @xmath80 @xcite .
these measured values seem to place cen a on the region of the face - on fast rotator early - type galaxies on the diagram ( figure 11 of @xcite ) .
although the ellipticity of cen a remains low @xmath81 out to about 3 , a well - defined photometric pa can still be determined . the major axis that we measure from the inertia ellipsoid of the surface brightness on the 2mass @xmath0-band photometry @xcite at large radii is @xmath82 ( east of north ) .
the axis of maximum rotation that we determine from our 250mas sinfoni data is quite different at @xmath56 ( section [ sec : ppxf ] ) .
the sense of stellar rotation that we measure is in support to the determination by @xcite , and our accurate @xmath59 is probably consistent with their determination , given their large uncertainties .
the strong kinematical misalignment of @xmath83 is an indication that the galaxy is certainly not an axisymmetric object .
this is consistent with the strong twist in the kinematical position angle of the planetary nebula system at large radii @xcite .
however the galaxy is unlikely to be well described by a stationary triaxial geometry either .
in fact at radii where cen a starts to appear more elongated , its surface brightness is dominated by stellar shells , which constitute the relics of a recent accretion event and are likely associated with the polar geometry of the strong dust lane which crosses the galaxy nucleus @xcite .
this shows that at those radii the galaxy has not reached an equilibrium configuration . the relatively low but ordered rotation near the nucleus , the presence of prominent stellar shells at large radii and the strong kinematical misalignment of cen a are reminiscent of the fast rotator s0 ngc 474 in @xcite .
however only much more extended integral - field observations in the near - infrared , sampling up to 1 , could conclusively reveal the dynamical status of the central regions of cen a. at different radii the galaxy is expected to reach an equilibrium configuration in very different time scales , starting from the center .
we estimated the circular velocity @xmath84 of cen
a from the multi - gaussian expansion ( mge ; * ? ? ?
* ) parametrization of its @xmath0-band surface brightness tabulated in @xcite and the best fitting mass - to - light ratio @xmath31 we derive in this paper .
it is nearly flat over two orders of magnitude in radius @xmath85 from 4400 .
this implies that at the edge of our sinfoni observations ( @xmath86 ) the characteristic orbital period is @xmath87 myr , while it is @xmath88 gyr at large radii where the stellar body starts to become more elongated and shells are still visible . the very short orbital period in the central regions , where the galaxy is nearly circular in projection , suggests that it has already reached an equilibrium configuration there .
given that the kinematics at a certain projected radius @xmath89 is mostly influenced by stellar orbits having that same characteristic radius @xmath89 @xcite , this justifies the use of stationary dynamical models to study the central dynamics of cen a , even though the galaxy is still dynamically evolving at larger radii . for this paper we will construct two types of models , to test the sensitivity of the estimate to the assumed geometry and dynamics .
we build ( i ) an axisymmetric orbit - based model , to reproduce in detail the kinematic observations , and ( ii ) a simple anisotropic spherical jeans model to qualitatively check the determination .
our axisymmetric three - integral dynamical model is based on @xcite numerical orbit - superposition method .
this has become the current standard for all determinations from the stellar kinematics , available in the literature from different groups ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the axisymmetric implementation we use , the orbital ` dithering ' and the setup we employ , are described in detail in @xcite .
the relatively limited spatial extension of the integral - field data , and the likely lack of equilibrium at large radii , does not justify the use of more general triaxial models @xcite .
our dynamical model assumes constant @xmath31 .
this is likely a good approximation in the central regions ( @xmath90 ) of cen a , where dark matter is expected to contribute only a small fraction of the mass @xcite as observed via dynamics or gravitational lensing in larger samples of early - type galaxies @xcite .
the model still allows for dark matter in the form of a constant shift in the global @xmath31 .
our model assumes an mge parametrization for the @xmath0-band surface brightness of cen a using the values tabulated in @xcite .
the models are fitted to our integral - field sinfoni 100mas ( with ao correction ) and 250mas kinematics in the central regions ( @xmath91 ) , and to the two - slits gnirs kinematics of @xcite at larger radii ( @xmath92 ) .
for @xmath93 we only used the sinfoni 100mas data in the fit , not to spoil the information contained in the high - resolution ao - assisted observations with the seeing - limited 250mas ones .
similarly we do not fit the gnirs kinematics for @xmath94 , where we have higher quality sinfoni data . as the models are bi - symmetric by construction , the sinfoni kinematics were bi - symmetrized before the fit along @xmath95 .
for all datasets we fit , and we compute the predictions , for the gauss - hermite moments up to @xmath43@xmath96 , where @xmath43 and @xmath44 are the measured values , while we set @xmath97 ( @xmath47 errors ) .
this is done to include in the model the additional observational constraint that extreme values for the high - order moments are never observed ( see two extreme examples in fig . 4 of @xcite ) .
we constructed a set of models with slightly different geometries to test the variation in the recovered : ( i ) an edge - on model ( @xmath98 ) ; ( ii ) an edge - on model with the major axis gnirs kinematics aligned with @xmath95 . in this case
we set @xmath99 for the gnirs minor axis data ; ( iii ) an edge - on model with ` circularized ' gnirs kinematics .
here we assumed the @xmath10 and @xmath44 field have circular symmetry , and a profile given by the average of the two slits .
we fitted the @xmath10 and @xmath44 gnirs kinematics along five polar sectors , covering the whole galaxy from the major to the minor axis , and not just the observed major / minor axes .
( iv ) same as model [ ii ] , but with @xmath100 , which is the average inclination for random orientations ; ( v ) same as model [ ii ] , but with @xmath101 .
all these models are expected to produce similar results given that the differences between the major and minor axis gnirs kinematics are almost at the level of the measurement errors , and given that cen a is so close to circular within the region where we have kinematics , that for large ranges of inclination the intrinsic density is essentially spherical .
we find that indeed all these models produce bh mass estimates well within the errors , and have similar uncertainties .
the orbital distribution of all the models is also nearly the same .
for this reason we will not present the individual results . instead for all our figures below we adopt as representative the model ( ii ) .
the two best - fitting parameters of the model , and @xmath31 , were determined as follows .
after an initial approximate search for the best fitting parameters using isotropic jeans models ( section [ sec : jeans ] ) , we computed a set of orbit libraries , each consisting of 444,528 individual orbits , at our first guess for @xmath31 . for each set
the was sampled linearly in @xmath102 , where @xmath103 is our first guess for .
we then scaled the velocities of the orbits to compute models at different @xmath31 @xcite .
in the fit we use a modest amount of regularization @xmath104 ( see * ? ? ?
* for a definition ) .
the contours of the resulting @xmath105 , which describe the agreement between the models and the data , are shown in fig .
[ fig : chi2_grid ] .
the best fitting bh has mass @xmath2 , where the @xmath3 error bars are marginalized over the @xmath31 and correspond to one degree - of - freedom ( @xmath106 ) .
the corresponding best fitting @xmath0-band @xmath107 . , which describes the agreement between the data and the model , are plotted as a function of and @xmath0-band @xmath31 .
the lowest levels show the @xmath108 which correspond to confidence levels of 1 , 2 and @xmath3 ( thick contour ) , for one parameter .
additional contours are separated by factors of two in @xmath109 .
the open square marks the best fitting derived from the gas kinematics by @xcite .
the open circle is the best fitting isotropic model from section [ sec : jeans ] .
the small dots are the locations of the models that were run , while the @xmath105 was interpolated at other locations.,title="fig : " ] the data - model comparison for our best fitting model are shown in fig .
[ fig : best_fit_kinematics ] .
all the significant features of the data are reproduced in detail by the model , for the two sinfoni datasets .
unsurprisingly the gnirs long - slit data are also reproduced in detail ( not shown ) .
although we only have observations along two slits , we still predict the model kinematics over a full quadrant , to verify that the model does not contain un - physical features or sharp discontinuities , which are never observed in real galaxies @xcite . .
the second row show the best fitting dynamical model predictions .
the central bins that were excluded from the fit are shown with the white diamonds .
_ bottom two panels : _ same as in the top two panels , for the 250mas sinfoni kinematics . for each quantity
the color scale is the same in the two instrumental configurations.,title="fig : " ] the orbital anisotropy of our best fitting model is shown in fig .
[ fig : anisotropy ] .
we plot the ratio @xmath110 between the radial velocity dispersion @xmath111 and the tangential one defined as @xmath112 . here
@xmath113 includes only random motion and not ordered rotation , so that an isotropic system has @xmath114 , and @xmath115 are the standard spherical coordinates .
we also plot for reference the radius at which the total luminous mass in the mge model is equal to , for our best fitting @xmath31 and values .
the derived value is @xmath116 , which is very close to the usual practical definition of the bh radius of influence @xmath117 , for @xmath118 ( see also * ? ? ?
also shown is the break radius @xmath119 of a nuker - law @xcite fit to the @xmath0-band hst photometry @xcite .
both radii are very well resolved by the observations , which then allow the nuclear orbital distribution to be robustly recovered .
the velocity ellipsoid appears to deviate from a sphere ( isotropy ) by just 15% in the whole range where we have kinematics .
this is just above the errors of 5% in this determination @xcite .
the main significant trend is a decrease in the @xmath110 ratio inside @xmath6 , but no sharp transition is seen near @xmath120 .
( see text for a definition ) measured at different polar angles , from the equatorial plane to the symmetry axis , in the galaxy meridional plane .
as the model is nearly spherical , the differences at a given radius , provide a rough indication of the model uncertainties .
also shown are the break radius ( or core radius ) @xmath120 , the bh radius of influence @xmath6 and the @xmath10 of the best fitting gaussian model of the sinfoni 100mas psf.,title="fig : " ] in the previous section we used general axisymmetric schwarzschild models to determine and to derive the anisotropy profile in the nucleus of cen a. in this section we use simpler anisotropic spherical jeans models to test the reliability of these modeling results . in the past 25 years
the dynamical models of galaxies have evolved from simple spherical models based on the jeans equations @xcite to quite general orbit- or particle - based numerical techniques with spherical @xcite , axisymmetric ( e.g. * ? ? ? * ) or triaxial geometry @xcite .
contrary to the models based on the jeans equations , which can generally be solved and tested to machine precision , the results of the more general models are sensitive to the numerical implementation details and more difficult to test thoroughly .
moreover the complexity of the general models often prevents a qualitative assessment of the results , which sometimes have to rely entirely on subtle details of the kinematics ( e.g. fig . 2 of @xcite or fig .
7 of @xcite ) .
one way to gain confidence in the determinations and the recovered anisotropy is by comparing them with the simpler jeans models @xcite .
these models are based on a completely different set of assumptions and do not suffer from possible incompleteness in the sampling of the orbit library or from limited kinematic coverage . to test our best fitting and @xmath31 and the corresponding recovered anisotropy profile of fig .
[ fig : anisotropy ] we use the jeans anisotropic mge ( jam ) package@xmath40 of @xcite .
the method allows one to compute the predicted second moments , projected onto the sky plane , for a model with a variable anisotropy profile , via a quick and accurate single numerical quadrature . given that our standard model of cen a is intrinsically close to spherical
, we use the spherical formalism of equation ( 50 ) of @xcite .
we construct a spherical model having the same mge surface brightness as our schwarzschild model , but we set the axial ratios of all the gaussians @xmath121 .
we assume the same anisotropy profile of our best fitting schwarzschild model of section [ sec:3i - model ] , with the same and @xmath31 .
the method allows for a different anisotropy @xmath122 to be assigned to different gaussian components of the mge model . in practice
we assign an anisotropy @xmath123 ( radial anisotropy ) to the gaussians with @xmath124 and an anisotropy @xmath125 ( tangential anisotropy ) to the remaining gaussians .
although @xmath126 changes in a discontinuous fashion for the different mge gaussians , the resulting @xmath126 profile is smooth .
in fact the gaussians overlap with each other and the @xmath126 at a certain spatial position is a luminosity - weighted sum of the @xmath126 of the individual components .
this anisotropic model , mimicking the best - fitting three - integral model , is shown with the solid line in the top panel of fig .
[ fig : jeans ] .
although the model is not by itself a fit to the data , it provides a remarkably accurate description of the observed and radially - averaged sinfoni and gnirs ( outside @xmath127 ) second velocity moments @xmath128 .
we ignore the @xmath44 gauss - hermite moment in the @xmath129 estimation , as it is essentially zero within the errors .
the purely gaussian losvd extraction of fig .
[ fig : spectrum_variation ] gives consistent @xmath129 values within the errors .
this test shows that the best - fitting discrete schwarzschild s numerical representation of cen a is an excellent approximation for the continuum limit represented by the jeans model .
it provides a strong confirmation for the recovered anisotropy profile and the need for tangential anisotropy in the nucleus of cen a. we also show in fig .
[ fig : jeans ] , with the two dashed lines , the jeans predictions for two corresponding to the @xmath3 upper and lower confidence limits of the schwarzschild model .
they show that , as expected , inside @xmath130 the data are very sensitive to a change in at the level of the quoted errors . for
the 100mas ( red open circles ) and 250mas scale ( blue open squares ) , biweight averaged over circular annuli , are plotted as a function of the radius from the center of cen a. this is the same kinematics presented in fig .
[ fig : kinematics250 ] [ fig : kinematics100 ] , which was used in the schwarzschild models .
we also show for comparison , as black open upward triangles , the @xmath10 measured in fig .
[ fig : spectrum_variation ] , assuming a purely gaussian losvd .
the corresponding gnirs @xmath129 observations of @xcite along the major axis ( filled downward triangle ) and the minor axis ( filled upward triangle ) , are also shown for @xmath131 ( see text for a discussion of the discrepancy between the gnirs and sinfoni data inside @xmath130 ) .
these have been arbitrarily scaled up by 5% to match our data at @xmath132 ( see text ) .
the solid line shows the prediction of a jam model , having the best fitting and derived from the schwarzschild model of fig .
[ fig : chi2_grid ] and the corresponding anisotropy profile of fig .
[ fig : anisotropy ] .
the model was convolved with the psf of the 100mas observations .
the two dashed lines have a corresponding to the upper and lower @xmath3 errors in the schwarzschild model .
the dotted line has @xmath133 , corresponding to the best fitting value of @xcite .
_ bottom panel : _ same as in the top panel , but assuming isotropy ( @xmath134 ) for all the jeans models . the has been decreased to match the data at large radii.,title="fig : " ] the dotted line shows the model prediction , still with the same @xmath126 profile as for the previous models , for @xmath135 .
this is the best fitting value determined by @xcite from the gnirs data .
this model has a much steeper @xmath129 profile near the center , which seems to qualitatively reproduce the steeper rise of the @xmath129 in the gnirs data within @xmath136 .
unfortunately we can not compare our jeans models with the more nuclear gnirs kinematics ( @xmath37 ) .
those measurements have rather extreme @xmath137 values . in this case
it becomes not possible to reliably translate the kinematics into a true @xmath129 value , as required by the jeans equations .
in fact when the gauss - hermite moments are large , the @xmath129 one can derive by formally integrating over the losvd is extremely sensitive to the wings of the losvd which are observationally not well constrained .
for this reason we limit our comparison to the remaining values , for which @xmath44 is consistent with zero within the @xmath47 errors .
our comparison suggests that the @xmath138 higher value for derived by those authors is not due to differences in the modeling , but more likely to the differences in the measured nuclear kinematics profile with gnirs .
a more conclusive evidence would require a full schwarzschild model of the nuclear gnirs data , which is outside the scope of this paper .
it is clear however that our sinfoni nuclear kinematics is not consistent with the gnirs one of @xcite when @xmath130 .
those authors measure larger @xmath10 values and especially strong @xmath44 values in the nucleus , while we find @xmath139 over the full field - of - view ( fig .
[ fig : kinematics100 ] ) .
the reasons for these differences are likely the following : ( i ) our sinfoni data represent a significant improvement over the gnirs one , both in @xmath12 and in spatial resolution ( compare fig . 6 of @xcite with the right column of our fig .
[ fig : spectrum_variation ] ) .
the gnirs data forced them to extract the kinematics inside @xmath93 by fitting only the first co band , while we could accurately fit all three co bands from the sinfoni data , down to the limit where they disappear due to the rise of the non - thermal nucleus ; ( ii ) we used different approaches for the subtraction of the non - thermal continuum .
as discussed in detail in section [ sec : nonthermal ] , a non accurate subtraction of the non - thermal contribution from the spectrum , leads directly to an error in @xmath10 .
we subtracted the non - thermal source with additive polynomials during the ppxf fit , so that both the kinematics and the continuum contribution are simultaneously optimized to fit the spectrum .
@xcite instead removed the continuum based on a measure of co line - strength and only subsequently fitted the kinematics on the continuum - free spectrum . for comparison , in the bottom panel of fig .
[ fig : jeans ] , we show four spherical jeans models with the same as in the top panel , but in this case assuming isotropy ( @xmath134 ) .
the @xmath31 was decreased to fit the data at large radii , where the bh contribution is minimal .
a reasonable fit to the @xmath129 is obtained in this case at the upper limit @xmath140 of our schwarzschild confidence interval .
this general agreement confirms previous comparisons , showing that although general three - integral models are needed for accurate determinations , much simpler isotropic models , applied to the same data , provide a useful sanity - check for the determination @xcite .
these simple tests would be especially useful to give confidence in the determinations in critical cases , where @xmath6 is barely resolved and a determination of seems at the limit of what is possible with current technologies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite compared the determinations between two - integral and three - integral models for 10 galaxies and found that , even using seeing - limited ground - based data , the two - integral of @xcite are systematically too high by just a factor @xmath141 , with respect to more recent three - integral models , based on high - resolution hst observations .
it would be interesting to perform a similar type of comparison , but using the same data for both anisotropic jeans and schwarzschild models .
the reason for the reasonable agreement between the simpler and the more general models is likely due to the fact that many real galaxies do not ` utilize ' all the freedom in the orbital distribution that would be required to produce more dramatic disagreements , and do not deviate very strongly from being isotropic in their centers ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this is certainly true for the results presented in this paper .
our determination , using the stellar kinematics , for the bh mass in the center of cen a @xmath2 agrees very well with the determination @xmath142 we performed using the gas kinematics in @xcite from the same sinfoni data .
this gas - stars comparison constitutes one of the most robust and accurate ones , due to a very well resolved bh sphere of influence ( @xmath5 compared to a psf fwhm@xmath143 ) and thanks to the use of high - resolution integral - field data for both kinematical tracers .
comparable and equally successful comparisons were done by @xcite and @xcite , while a less good agreement was found in @xcite , likely due to the disturbed gas kinematics .
in @xcite we made a detailed comparison with all the numerous previous determinations for cen a @xcite .
in summary the only significant disagreement in is with the previous stellar determination by @xcite . in section [ sec : jeans ] we showed that the disagreement is likely caused by a difference in the data quality and in the treatment of the contribution from the central non - thermal continuum in the kinematic extraction .
it is not due to differences in the details of the modeling methods . as shown in @xcite , with our new gas and stars determination cen a lies within the errors on the @xmath11 relation as given by either @xcite or @xcite .
the value @xmath144 is well consistent with what one would expect from a @xmath145 gyr luminosity - weighted age of the stellar population of cen a , almost independently of the assumed near - solar metallicity .
this adopts the @xcite _ normalization _ of the initial mass function ( imf ) and uses the population models of @xcite .
these population models include a proper treatment of the tp - agb stars which is essential for a reliable prediction of the near - infrared flux and @xmath146 .
the @xmath146 of the stellar population would decrease to 50% of the observed dynamical value if cen a had a mean age of @xmath147 gyr . in that case
dark - matter would be required to explain the observations .
however the younger age is at the lower extreme of the estimated age of 38 gyr , for the main body of cen a , from the analysis of its globular cluster system @xcite . adopting the normalization of the @xcite imf
, the mean age of cen a would have to be lower than @xmath147 gyr for the population @xmath146 not to over - predict the dynamical one .
however this imf normalization has been shown to be inconsistent with the observed dynamical of early - type galaxies as a class ( @xcite ; see @xcite for a review ) .
cen a possesses a clear core in the luminosity profile , with a break from a shallow central slope to a steeper outer one at a radius @xmath7 @xcite , which is extremely well resolved by our kinematics .
every galaxy spheroid seems to host a supermassive bh @xcite , and galaxies are expected to form by mergers .
these cores have been interpreted as due to the scouring of the galaxy profile due to the ejection of stars in radial orbits passing close to the bh binary which forms shortly after mergers @xcite .
the models predicts that the anisotropy should show a bias towards tangential orbits inside @xmath120 . to test this prediction , in this paper we present the recovered stellar anisotropy profile @xmath110 for cen a ( fig .
[ fig : anisotropy ] ) .
the profile is to first order flat around the value @xmath148 .
the maximum deviations of the velocity ellipsoid from a sphere ( isotropy ) are on the order of 15% .
the profile shows no sudden bias towards tangential orbits inside @xmath120 , while a tangential bias appears only inside @xmath6 .
this profile seems typical of the few nuclear anisotropy profiles that have been published so far using integral - field observations or spherical models ( the only cases for which the orbital distribution can be robustly recovered ) : m32 @xcite , m87 @xcite , ngc 3379 @xcite , ngc 1399 @xcite .
a similar anisotropy trend was observed with high - resolution hst long - slit data by @xcite , although in that case the profiles appear more noisy .
although the predicted tangential bias is observed , its size does not seem related to that of the core .
so these anisotropy determinations would not seem to support the scenario in which the cores have been produced by the core - scouring mechanism .
however a big caveat is that the current simulations are not entirely realistic : they only consider binary mergers of cusp galaxies which are initially isotropic . now that high - quality anisotropy measurements are starting to appear
, it would be valuable if the numerical simulations could revisit their predictions in a more cosmological motivated setting , considering sequences of multiple mergers . for the moment we must refrain from making firm statements on what our anisotropy profile implies for bh formation .
we present a determination of the mass of the supermassive bh in the nucleus of the nearby elliptical galaxy cen a , using high spatial resolution integral - field observations of the stellar kinematics , combined with large - scale long - slit kinematics .
our high - resolution observations were obtained in the near infrared @xmath0-band with sinfoni on the vlt , using adaptive optics to correct for the blurring effect of the atmosphere .
the psf of our data has a close to diffraction - limited resolution of 017 fwhm and very high @xmath1 per spectral pixel .
we discuss the extraction of the stellar kinematics , with particular emphasis on the treatment of the non - thermal continuum coming from the nucleus of cen a , which contaminates the central stellar spectra .
we use a standard three - integral axisymmetric numerical orbit - superposition model to determine the mass of the supermassive bh , which turns out to be @xmath2 .
this value is in good agreement with our previous accurate determination @xmath142 , obtained from the gas kinematics extracted from the same sinfoni data @xcite .
it is also in agreement with the recent analysis of the hst gas kinematics by @xcite , and with other similar , but with lower spatial resolution , determinations .
this mass is consistent with the prediction of the @xmath11 relation of either @xcite or @xcite .
we carefully test the and anisotropy recovery from our three - integral stellar modeling using simple anisotropic spherical mge jeans models .
we find very good consistency between the two different approaches .
we study the nuclear orbital distribution of cen a. in agreement with a few previous studies of other galaxies , we find a tangential bias in the anisotropy near the bh .
however the size of the tangential region do not seem to be associated with that of the core radius @xmath120 of the surface brightness profile , but has the size of the much smaller bh sphere of influence @xmath6 .
this does not add support to the scenario in which galaxy cores are scoured by binary bhs .
more realistic numerical simulations would be required for a comparison with the new accurate data and to draw conclusions from this observation .
mc acknowledge support from a stfc advanced fellowship ( pp / d005574/1 ) .
nn acknowledges support from the christiane - nsslein - volhard foundation .
jr acknowledges financial support from the academy of finland ( project 8121122 ) .
we are grateful to the referee karl gebhardt for a thoughtful report , which improved the presentation of our work . based on observations collected at the european southern observatory , paranal , chile , eso program 075.b-0490(a ) .
d. c. , mosteller f. , tukey j. w. , 1983 , understanding robust and exploratory data anlysis .
wiley series in probability and mathematical statistics , new york : wiley , 1983 , edited by hoaglin , david c. ; mosteller , frederick ; tukey , john w. | we present a determination of the mass of the supermassive black hole ( bh ) and the nuclear stellar orbital distribution of the elliptical galaxy centaurus a ( ngc 5128 ) using high - resolution integral - field observations of the stellar kinematics .
the observations were obtained with sinfoni at the eso very large telescope in the near - infrared ( @xmath0-band ) , using adaptive optics to correct for the blurring effect of the earth atmosphere .
the data have a spatial resolution of 017 fwhm and high @xmath1 per spectral pixel so that the shape of the stellar line - of - sight velocity - distribution can be reliably extracted .
we detect clear low - level stellar rotation , which is counter - rotating with respect to the gas .
we fit axisymmetric three - integral dynamical models to the data to determine the best fitting values for the bh mass @xmath2 ( @xmath3 errors ) and @xmath4 in solar units .
these values are in excellent agreement with previous determinations from the gas kinematics , and in particular with our own published values , extracted from the same data .
this provides one of the cleanest gas versus stars comparisons of determination , due to the use of integral - field data for both dynamical tracers and due to a very well resolved bh sphere of influence @xmath5 .
we derive an accurate profile of the orbital anisotropy and we carefully test its reliability using spherical jeans models with radially varying anisotropy .
we find an increase in the tangential anisotropy close to the bh , but the spatial extent of this effect seems restricted to the size of @xmath6 instead of that @xmath7 of the core in the surface brightness profile , contrary to detailed predictions of current simulations of the binary bh scouring mechanism .
more realistic simulations would be required to draw conclusions from this observation .
[ firstpage ] black hole physics
galaxies : individual ( ngc 5128 ) galaxies : elliptical and lenticular , cd galaxies : kinematics and dynamics instrumentation : adaptive optics |
deep inelastic scattering ( dis ) off protons has provided decisive information on the parton distribution functions ( pdfs ) of the proton .
inclusive measurements of the cross section for the reaction @xmath2 as a function of the virtuality of the exchanged boson ( @xmath3 ) and of the bjorken scaling variable ( @xmath1 ) have been used to determine the proton structure function @xmath4 .
perturbative qcd ( pqcd ) in the next - to - leading - order ( nlo ) approximation has been widely used to extract the proton pdfs from such measurements and to test the validity of pqcd . in the standard approach ( dglap @xcite ) , the evolution equations sum up all leading double logarithms in @xmath5 along with single logarithms in @xmath6 and are expected to be valid for @xmath1 not too small . at low @xmath1 , a better approximation is expected to be provided by the bfkl formalism @xcite in which the evolution equations sum up all leading double logarithms along with single logarithms in @xmath7 .
the dglap evolution equations have been tested extensively at hera and were found to describe , in general , the data . in particular , the striking rise of the measured @xmath4 at hera with decreasing @xmath1
can be accomodated in the dglap approach .
nevertheless , the inclusive character of @xmath4 may obscure the underlying dynamics at low @xmath1 , and more exclusive final states like forward axis pointing in the proton beam direction , referred to as the `` forward direction '' .
] jets @xcite need to be studied .
bfkl evolution predicts a larger fraction of small-@xmath1 events containing high-@xmath8 forward jets than predicted by dglap @xcite .
parton dynamics at low @xmath1 is particularly relevant for the lhc given that most of the interesting hard processes involve partons with low fractional momenta .
forward jet production in neutral current ( nc ) dis has been studied extensively at hera . as an example ,
measurements of forward jet production with @xmath9 gev , polar angle @xmath10 between @xmath11 and @xmath12 , @xmath13 and @xmath14 , where @xmath15 is the proton - beam energy , in the kinematic region defined by @xmath16 and @xmath17 gev@xmath18 are shown in figure [ figura1]a and exhibit a strong rise towards low @xmath1 @xcite .
perturbative qcd apparently does not account for such a rise : the leading - order ( lo ) qcd calculation does not predict any rise while the one predicted by nlo @xcite is still much too low ( see figure [ figura1]a ) . some of the feynman diagrams that are accounted for in the lo ( @xmath19 ) and nlo ( @xmath20 ) calculations are shown in figure [ figura1]b .
the former has no additional gluon radiation and helps to understand why in the lo calculation there is hardly any phase space left for forward jet production .
in contrast , the nlo calculation accounts for the radiation of one additional gluon . this explains why there is such a huge increase from lo to nlo : due to the opening of a new channel , namely gluon - exchange in the @xmath21 channel .
however , that means that the nlo calculation is effectively a `` lo '' calculation , since no corrections are included .
the nlo calculation should thus have large theoretical uncertainties from higher orders .
a variation of the renormalisation scale around @xmath22 do not give rise to such large theoretical uncertainties ( see figure [ figura1]a ) .
however , if @xmath3 is instead chosen as the renormalisation scale , the resulting theoretical uncertainties are large , as pointed out in @xcite and shown in figure [ figura1]c : measurements of forward jet production @xcite with @xmath23 gev , pseudorapidity in the range @xmath24 , @xmath25 and @xmath26 in the kinematic region given by @xmath27 and @xmath28 gev@xmath18 are compared to nlo qcd calculations @xcite with @xmath29 .
large theoretical uncertainties which arise from higher orders in the pqcd calculations prevent a firm conclusion .
further progress can be made by making measurements for which genuine nlo calculations are available , i.e. one gluon radiation at lo and two additional radiated gluons at nlo .
that is the case for three - jet production , which was already studied in @xcite and has been investigated more thoroughly in @xcite .
the latter is discussed next .
( 10.0,8.0 ) ( -5.0,0.0 ) ( 2.0,2.0 ) ( 7.1,-0.3 ) ( -1.1,-0.3)(a ) ( 4.1,-0.3)(b ) ( 10.1,-0.3)(c ) ( 2.3,1.3)lo @xmath19 ( 5.0,1.3)nlo @xmath30
some of the feynman diagrams accounted for in the pqcd calculations for three - jet production in nc dis are shown in figure [ figura2]a .
a measurement of the differential cross section @xmath31 as a function of @xmath1 for three - jet production @xcite is presented in figure [ figura2]b .
the jets are reconstructed in the @xmath32 frame and are required to fulfill the following conditions : the transverse momentum of each jet @xmath33 gev , the sum of the highest and next - to - highest @xmath34 jets above 9 gev , the jet pseudorapidity in the laboratory frame to lie between -1 and 2.5 and at least one of the jets in the central region ( @xmath35 ) .
the kinematic region is defined by @xmath36 , @xmath37 gev@xmath18 and @xmath38 , where @xmath39 is the inelasticity variable .
perturbative qcd calculations are compared to the data in figure [ figura2]b : the lo ( @xmath20 ) calculation still falls short of the data , but the nlo ( @xmath40 ) calculation @xcite improves dramatically the description of the data at low @xmath1 . the inclusion of the @xmath40 qcd corrections does not only improve the description of the measured rate but also that of the topology of the events .
measurements have been made of the distributions in the variables used to describe the topology of three - jet events in the three - jet centre - of - mass frame : the scaled energy of the jets @xmath42 ( @xmath43 ; @xmath44 ) and the two angles @xmath45 and @xmath46 ( see figure [ figura2]c ) .
the measurements are shown in figure [ figura3 ] and compared to nlo calculations @xcite .
the inclusion of additional gluon radiation provides an improved description of the data , for example , in the @xmath47 distribution : the nlo follows the data and exhibits peaks at @xmath48 and @xmath49 , whereas the lo calculation flattens out at @xmath48 .
further investigations of low-@xmath1 parton dynamics have been made by studying transverse - energy and angular correlations in dijet and trijet production in nc dis @xcite .
jets are reconstructed in the hadronic centre - of - mass ( hcm ) frame and required to fulfill the following conditions : @xmath50 gev , @xmath51 gev and @xmath52 .
the kinematic region is given by @xmath36 , @xmath53 gev@xmath18 and @xmath54 .
one of the most interesting angular correlations is provided by the variable @xmath55 , which is defined as the azimuthal separation of the two jets with largest @xmath56 . for dijet events , @xmath19 kinematics constrain @xmath55 to @xmath57 and @xmath30 calculations provide the lo contribution ; @xmath41 calculations give the nlo correction .
measurements of the doubly differential cross section @xmath58 for dijet production in different regions of @xmath1 are presented in figure [ figura4 ] .
the @xmath30 predictions increasingly deviate from the data as @xmath1 decreases , whereas @xmath41 calculations @xcite provide a good description of the data even at low @xmath1 . in summary ,
parton dynamics at low @xmath1 is vigorously pursued at hera .
precise measurements of multijet production in nc dis have been made down to @xmath59 in terms of jet rates , topologies and correlations .
comparison of perturbative qcd calculations with these measurements demonstrate the big impact of initial - state gluon radiation .
perturbative qcd at @xmath40 reproduces succesfully the measurements .
however , the theoretical uncertainties are still significant and the precision of the data demands next - to - next - to - leading - order corrections to be included .
gribov and l.n .
lipatov , sov .
( 1972 ) 438 ; + l.n .
lipatov , sov .
( 1975 ) 94 ; + yu.l .
dokshitzer , sov .
jetp 46 ( 1977 ) 641 ; + g. altarelli and g. parisi , nucl .
b126 ( 1977 ) 298 .
kuraev , l.n .
lipatov and v.s .
fadin , sov .
jetp 45 ( 1977 ) 199 ; + ya.ya .
balitski and l.n .
lipatov , sov .
( 1978 ) 822 .
mueller , nucl .
c18 ( 1991 ) 125 ; + a.h .
mueller , j. phys .
g17 ( 1991 ) 1443 .
j. bartels , a. de roeck and m. loewe , z. phys .
c54 ( 1992 ) 645 ; + j. kwiecinski , a.d .
martin and p.j .
sutton , phys .
b287 ( 1992 ) 254 ; + w.k .
tang , phys .
b278 ( 1992 ) 363 .
h1 collaboration , a. aktas et al .
j. c46 ( 2006 ) 27 . s. catani and m.h .
seymour , nucl .
b485 ( 1997 ) 291 .
zeus collaboration , s. chekanov et al .
j. c52 ( 2007 ) 515 .
h1 collaboration , f.d .
aaron et al . , eur .
j. c54 ( 2008 ) 389 .
z. nagy and z. trocsanyi , phys .
87 ( 2001 ) 082001 .
zeus collaboration , s. chekanov et al .
b786 ( 2007 ) 152 . | recent measurements of multijet production in neutral current deep inelastic @xmath0 scattering at hera are presented .
emphasis is put on parton dynamics at low @xmath1 . |
the formation of black hole ( bh ) seeds and their evolution have been the subject of several theoretical investigations .
the `` flow - chart '' presented rees ( 1978 ) still stands as a guideline for the possible paths leading to formation of bh seeds in the center of galactic structures .
one possibility is the direct formation of a massive bh ( mbh ) from a collapsing gas cloud ( haehnelt & rees 1993 ; loeb & rasio 1994 ; eisenstein & loeb 1995 ; bromm & loeb 2003 ; koushiappas , bullock & dekel 2004 ; begelman , volonteri & rees 2006 ) .
the gas can condenses to form a central massive object .
the mass of the seeds predicted by different models vary , but typically are in the range @xmath8 .
alternatively , the seeds of bhs can be associated with the remnants of the first generation of stars , formed out of zero metallicity gas .
the first stars are believed to form at @xmath9 in halos which represent high@xmath10 peaks of the primordial density fie ld .
the absence of metals might lead to a very top heavy initial stellar mass function , and in particular to the production of very massive stars with masses @xmath11 ( carr , bond , & arnett 1984 ) .
if very massive stars form above 260 @xmath12 , they would rapidly collapse to bhs with little mass loss ( fryer , woosley , & heger 2001 ) , i.e. , leaving behind seed bhs with masses @xmath13 ( madau & rees 2001 ) .
the subsequent growth of bhs from these initial seeds has been investigated in the hierarchical framework , typically associating episodes of accretion to galaxy mergers ( haiman 2004 , yoo & miralda - escud 2004 , shapiro 2005 ; volonteri & rees 2005 , 2006 ; lapi et al .
2006 ) .
only a few observational constraints are currently available .
many are limited to the brightest sources , probing only the very bright end of the luminosity function of quasars at @xmath7 and therefore only the upper end of the massive black hole mass function ( fan et al .
the observation of very luminous quasars , powered by billion solar masses bhs , at @xmath14 in the sdss survey ( fan et al .
2001 ) , implies that a substantial population of smaller accreting bhs must exist at earlier times .
deeper but limited constraints have been provided for more typical agn , with @xmath15 2 orders of magnitude smaller mbhs , using deep x - ray observations ( e.g. , alexander et al .
2001 ; barger et al . 2003b ; koekemoer et al . 2004
; mainieri et al .
however , these constraints lack the detail and depth required to understand and determine the global evolution of bh seeds and of the average mbh population at very high redshifts .
accreting bhs are observed as x ray emitters up to @xmath16 ( barger et al . 2003 , steffen et al .
2006 ) , and there are no reasons to believe that higher redshift agn would be any different .
the detection of very high redshift agn is one of the goals of future space missions .
in particular , x ray telescopes , such as the planned _ xeus _ and _ constellation x _ , are expected to detect these sources even at @xmath0 .
moreover , x ray observatories can identify even heavily obscured agn activity , due to the penetrating nature of hard x - rays .
additional , albeit less direct , limits on the early mbh population can already be placed by the requirement that the cumulative emission from the predicted high redshift sources does not saturate the observed unresolved x - ray background ( dijkstra , haiman & loeb 2004 ; salvaterra , haardt , ferrara 2005 ) .
we show that future space - borne missions will be able to constrain different proposed models for the accretion history of mbhs at early times .
we focus here on a specific model for mbh formation , which traces mbh seeds to the first generation of metal free stars .
we then compare how different accretion histories reflect onto the detectability of x ray sources .
the mass growth of the most massive bhs at high redshift must proceed very efficiently to explain the luminosity function of luminous quasars at @xmath17 in the sloan digital sky survey ( sdss , e.g. fan et al .
volonteri & rees ( 2005 ) explore the conditions which allow a sufficient growth of mbhs by @xmath18 under the assumption that accretion is triggered by major mergers . at such high redshift
an investigation of evolution of the mbh population must take into account the dynamical evolution of mbhs .
black hole mergers , in fact , can give a net negative contribution to the early black hole growth , due to gravitational effects which can kick bhs out of their host halos ( gravitational recoil , due to to the non - zero net linear momentum carried away by gravitational waves ) .
adopting recent estimates for the recoil velocity ( baker et al .
2006 ) , volonteri & rees ( 2005 ) find that if accretion is always limited by the eddington rate via a thin disc , the maximum radiative efficiency allowed to reproduce the lf at @xmath18 is @xmath19 ( corresponding to an upper limit to the mbh spin parameter of 0.8 ) .
if , instead , high - redshift mbhs can accrete at super - critical rate during an early phase ( volonteri & rees 2005 , begelman , volonteri & rees 2006 ) , then reproducing the observed mbh mass values , is not an issue .
the constraints from the lf at @xmath18 are still very weak , so either a model with a low @xmath20 or a model with supercritical accretion can not be ruled out based on these results only . in a previous paper
we have investigated how current observations can constrain the late accretion history of mbhs , at @xmath21 ( volonteri , salvaterra & haardt 2006 ) . in this paper
we study by means of a detailed model of mbh assembly , the detectability of very high redshift agn both at x ray and near - ir wavelengths by future missions .
we discuss the contribution of these objects to the unresolved x ray background ( in particular in the 0.52 kev band ) .
finally , we discuss the role of an early population of agn in the reionization of the universe and their contribution to the uv background at high redshift .
the paper is organized as follows . in sect .
2 we describe briefly the merger tree model for the formation and evolution of mbh in the early universe .
3 presents the basic equations we need in the number counts and background calculation and , in sect . 4 , we discuss the adopted spectrum of agn . in sect .
5 we present our results and discuss the role of future space mission in the observations of high redshift agn . in sect .
6 we compare our results to those of a model in which the early evolution of mbh seeds is characterized by a phase of super critical accretion . finally , in sect . 7
, we briefly summarize our results . throughout the paper we use the ab magnitude system , where @xmath22 is the spectral energy density within a given passband in units of erg s@xmath3 @xmath23 hz@xmath3 ] and the standard @xmath25cdm cosmology ( spergel et al .
the main features of a plausible scenario for the hierarchical assembly , growth , and dynamics of mbhs in a @xmath25cdm cosmology have been discussed in volonteri , haardt & madau ( 2003 ) .
dark matter halos and their associated galaxies undergo many mergers as mass is assembled from high redshift to the present .
the halo merger history is tracked backwards in time with a monte carlo algorithm based on the extended press - schechter formalism .
seed " holes are assumed to form with intermediate masses in the rare high @xmath26 peaks collapsing at @xmath27 ( madau & rees 2001 ) as end - product of the very first generation of stars . as our fiducial model
we take @xmath28 at @xmath29 , which ensures that galaxies today hosted in halos with mass larger than @xmath30 are seeded with a mbh .
the assumed threshold allows efficient formation of mbhs in the range of halo masses effectively probed by dynamical studies of mbh hosts in the local universe . as a reference ,
we adopt here a conservative model assuming eddington - limited accretion which is able to reproduce the bright end of the optical lf , as traced by observations in the sdss ( e.g. fan et al 2001 ) .
we then discuss in section 6 a simple model which considers super - eddington accretion rates for high - redshift mbhs and one in which more massive seeds form late , as in koushiappas et al .
( 2004 ) , and evolve through eddington limited accretion .
the fraction of active bhs ( i.e. , agn ) at early times is large in our models , reaching almost unity for the most massive black holes ( @xmath31 ) at @xmath32 .
the fraction of agn with respect to the whole bh population decreases by at least one order of magnitude at @xmath33 and further at lower masses . under our model assumptions , detection of active sources
can therefore provide robust constraints on the whole bh population .
the number of sources observed per unit solid angle at redshift @xmath34 in the flux range @xmath35 and @xmath36 at frequency @xmath37 is @xmath38 where @xmath39 is the comoving volume element per unit redshift per unit solid angle , and @xmath40 is the comoving density of sources at redshift @xmath41 with observed flux in the range @xmath42 $ ] .
the specific flux of a source observed at @xmath34 is given by @xmath43 where @xmath44 is the specific luminosity of the source ( in units of erg s@xmath3 hz@xmath3 ) _ averaged over the source lifetime _
, which is assumed to be only a function of the mass of the central bh . in the above eq .
( 2 ) , @xmath45 , @xmath46 is the luminosity distance , and @xmath47 is the effective optical depth of the intergalactic medium ( igm ) at @xmath37 between redshifts @xmath34 and @xmath41 .
shortward of the observed lyman@xmath48 , the emergent spectrum is strongly modified by igm absorption ( see sect . 2.2 of salvaterra & ferrara 2003 for a full description of the igm modeling ) . for sources at @xmath49 , as those we are considering in this paper , igm absorption in the observed 0.510 kev band ( corresponding , for sources at @xmath49 as those considered here , to rest frame energy @xmath50 kev ) can be safely neglected .
the radiation background @xmath51 observed at redshift @xmath34 at frequency @xmath37 , is @xmath52 where @xmath53 is the comoving specific emissivity , and @xmath54 is the proper line element .
the source term @xmath55 is given by @xmath56 the second approximated equality holds once we consider the source light curve averaged over the typical source lifetime @xmath57 , and assuming the source formation rate per unit mass as constant over such timescale .
the physical characterization of the source is encoded in its spectral energy distribution ( sed ) , @xmath58 .
we model the uv part of the sed of unabsorbed agn ( @xmath59 , referred as type i ) as a multicolor disk black body ( shakura & sunyaev 1973 ) .
assuming eddington limited accretion , the maximum disk temperature is @xmath60 .
the characteristic multicolor disk spectrum is broadly peaked at @xmath61 , and follows a power law with @xmath62 for energies @xmath63 , and exponentially rolls off for @xmath64 . in the x
ray , the spectrum can be described by a power law with photon index @xmath65 ( @xmath66 ) , and an exponential cutoff at @xmath67 kev ( marconi et al .
the averaged x
ray sed of absorbed agn ( type ii ) is described by the same type i spectrum for @xmath68 kev , and , in the range @xmath69 kev , by a power law ( continuously matched ) with photon index @xmath70 , obtained convolving the type i spectrum with a lognormal distribution of absorption column density centered at @xmath71 ( sazonov et al .
uv emission from type ii agn is assumed to be negligible .
we further assume a type i / type ii ratio of 1/4 , independently of redshift and luminosity , since detailed modelling of intrinsic absorption at @xmath49 are not currently available .
note that the observed x rays correspond , for sources at @xmath72 , to rest
frame energies between @xmath73 and @xmath74 kev , where the emission properties of type i and type ii agn are thought to be similar ( the rare compton thick sources are not considered here ) .
the x ray emission of type i agn is normalized to the optical , adopting an optical to x rays energy index @xmath75 ( see eq . 5 of steffen et al .
2006 ) , where @xmath76 is the monochromatic luminosity at @xmath77 (rest frame ) . by definition , the x
ray luminosity at the rest frame energy of 2 kev , @xmath78 , is @xmath79 .
the scaling of @xmath80 with redshift and luminosity has been obtained by steffen et al .
( 2006 ) combining data from the sdss , combo-17 and _ chandra _ surveys .
the final sample consists of 333 agn extending out to @xmath81 spanning five decades in uv luminosity and four decades in x ray luminosity .
since only a mild dependence on redshift is found , we extrapolate this result also to @xmath49 .
we further assume eddington limited accretion ( see section 2 ) .
[ fig : xcount ] we compute the soft ( 0.52 kev ) and hard ( 210 kev ) x ray number counts from mbhs shining at @xmath72 predicted by our model of mbh assembly and evolution .
results are shown in fig . 1 , and are compared to available observational data ( moretti et al .
tie indicates results from the fluctuation analysis of the _ chandra _ deep field ( miyaji & griffiths 2002 ) .
different lines refer to different ranges in bh masses . at the flux limit of current surveys , @xmath82 in the soft ( hard ) band ( alexander et al . 2003 ) , the contribution to the logn /
logs from sources at @xmath72 is @xmath83 in the soft x
rays , @xmath84 in the hard band . in the 0.5 - 8 kev band ,
we expect @xmath85 source at @xmath72 in the _ chandra _ deep field north with fluxes exceeding @xmath86 , to be compared to an upper limit of 7 sources with extreme x ray / optical flux ratios ( exo , koekemoer et al .
2004 ) , that are candidate high redshift agn .
the optical identification of these objects is problematic , owing to absorption ( both internal and by the igm ) and to their very low optical flux .
barger et al . ( 2003a , 2003b ) searched for optical counterparts ( at @xmath87 confidence ) with @xmath41band magnitude @xmath88 of agn of the _ chandra _ deep field north exposure .
apart from a source at @xmath89 with @xmath90 , no other @xmath91 candidate was identified .
this is consistent with our results , where the majority of the very high redshift sources detected in the deepest _ chandra _ observations should have a z band magnitude fainter than 27 .
barger et al . (
2003b ) used the lack of optical identifications to derive limits on the number of objects at @xmath91 .
they found that in a field of view ( f.o.v . ) corresponding to 6 arcmin radius circle , only @xmath92 sources with fluxes exceeding @xmath93 in the 0.46 kev band should lie at @xmath94 .
our model is consistent with this limit , predicting that only @xmath95 sources at @xmath72 should be found in the same field of view . at fainter fluxes ,
a significant fraction of the sources identified in the fluctuation analysis of the deepest _ chandra _ data might be agn at @xmath7 .
our model predicts that high redshift agn contribute to the logn / logs at a level @xmath96% in the soft x ray band , @xmath97% in the hard x ray band at the flux limit of the fluctuation analysis ( @xmath98 sources deg@xmath99 at the flux limit @xmath100 for the soft band , and @xmath101 sources deg@xmath99 at the flux limit @xmath102 for the hard band ) .
direct observations of such sources is among the main goals of the next generation of x ray telescopes ( e.g. _ xeus _ and _ constellation x _ ) .
the _ xeus _
mission is expected to have sufficient sensitivity to measure the x ray spectra of sources as faint as @xmath103 in the 0.52 kev energy range , while the photometric limiting sensitivity is expected to be @xmath104 . in the hard x band , the limiting sensitivity , both spectroscopic and photometric , will be larger by almost an order of magnitude . at the spectroscopic flux limit of _ xeus _ , we predict almost @xmath105 ( @xmath106 ) agn in the soft ( hard ) x ray band , within a 1 deg@xmath107 f.o.v .. at such flux limits , @xmath108 ( @xmath109 ) sources deg@xmath99 are type i objects , indicating that , because of obscuration , deep surveys in the soft ( hard ) x ray band will miss nearly 90% ( 20% ) of type ii agn . at the photometric flux limits
, we expect @xmath110 ( @xmath111 ) sources deg@xmath99 in the soft ( hard ) x ray band . in this case
the type ii missing fraction is 70% ( 7% ) .
_ xeus _ will be directly probing the lower end of the mass function of accreting mbhs at @xmath49 , @xmath112 ( i.e. for luminosity @xmath113 erg s@xmath3 in the rest - frame 210 kev energy band ) .
the main contribution to the number counts is still expected from sources at @xmath114 , but almost @xmath115 deg@xmath99 sources , i.e. , 1% of the sources , are expected to be detected at @xmath116 .
we note here that , assuming an half energy width of the psf of 2 " , _ xeus _ will be confusion limited at a sensitivity of @xmath117 , saturating the logn / logs at a level of @xmath118 deg@xmath99 ( arnaud et al .
2000 ) . for bright sources ,
redshift determination should be possible from the detection of x ray emission lines ( e.g. , 6.4 kev k@xmath119 ) . for fainter sources , however , redshift determination in the x rays might be very challenging , and a combination of deep x
ray observations and ultra - deep optical / near - ir spectroscopy is probably required ( see section 5.3 ) . in conclusion , we find that the next generation of x ray missions will be able to investigate the early stages of mbh build up and to provide fundamental information on the faint - end of the luminosity function of agn even at very high redshift . [ fig : xback ] according to moretti et al .
( 2003 ) , the intensity of the total x ray background ( xrb ) is @xmath120 and @xmath121 in the 0.52 kev , and 210 kev energy bands , respectively . a large fraction , @xmath122% , of the xrb in the 0.52 kev band has been attributed to sources with fluxes exceeding @xmath123 , while @xmath124% of xrb in the 210 kev band is resolved into sources whose flux is @xmath125 ( moretti et al .
more recently , hickox & markevitch ( 2006 ) estimate the unaccounted fraction of the xrb due to extragalactic unresolved sources as @xmath126 in the soft x ray energy band ( 0.52 kev ) and @xmath127 in the hard x ray energy band ( 28 kev ) .
we compute the contribution to the residual unresolved xrb in the soft and hard ( 0.52 kev and 28 kev , respectively ) energy band from the population of agn predicted by our model to exist at @xmath128 .
note that almost all these sources are below the source detection limit used by moretti et al .
results are shown in fig . 2 , where the cumulative contribution to the xrb of sources at redshift @xmath129 is considered .
different line
styles refer to different mbh mass ranges , as in fig . 1 .
the integrate contribution to the soft xrb from agn at @xmath72 is found to be @xmath130 , corresponding to @xmath131% of the total , and to @xmath132% of the unaccounted fraction .
sources with masses @xmath133 give the largest contribution to the xrb , whereas mbhs with @xmath134 produce a negligible background .
moreover , the major contribution to the unresolved x ray background is from mbhs shining at @xmath135 . in the 28 kev band , the xrb from unresolved ,
@xmath72 agn is @xmath136 , corresponding to @xmath131% of the total observed hard xrb , and to @xmath137% of the unresolved fraction .
mbhs with masses @xmath138 give the main contribution also in the hard band , indicating that the unresolved fraction of the xrb can be used to constrain the faint end of the x - ray luminosity function of agn at very high redshift .
we find that agn at @xmath49 contribute significantly to the unaccounted xrb , although their contribution is still well below the available constraints .
other faint unresolved x - ray sources at @xmath139 may contribute to the x - ray background , including galaxies , starbursts ( e.g. , bauer et al . 2004 ) , and a population of faint agn ( e.g. , volonteri , salvaterra & haardt 2006 ; see table 1 ) .
it is interesting to note that _ at least _
@xmath6% of the observed xrb ( 0.5 - 10 kev ) will remain unaccounted even after _
observations will become available , as due to sources at @xmath49 below the flux detection limit . only even more sensitive
ray observatories such as the proposed _ generation
x _ will assess these extremely faint sources .
[ fig : jwst ] [ fig : uvback ] deep field observations in the near - ir might be able , in principle , to detect agn at very high redshifts .
the @xmath140band acs data in the hubble ultra deep field ( hudf ) reach sensitivity as the 28th mag ( beckwith et al . 2006 ) .
our model predicts that , at this limiting magnitude , the sky density of agn with @xmath72 is @xmath141 arcmin@xmath99 . for comparison ,
the observed sky density of @xmath49 galaxies in the hudf is @xmath142 arcmin@xmath99 ; bouwens et al . 2005 ) , corresponding to an agn fraction of @xmath143 , so that the probability to detect these sources in a field as small as the hudf ( @xmath144 arcmin@xmath145 ) is extremely low .
given the actual size of the hudf ( bouwens et al . 2005 ) , an area three times larger has to be observed at the same magnitude limit in order to detect one agn at @xmath49 . in order to have a statistically significant number of very high redshift agn ,
lower detection limits are needed .
the james webb space telescope ( _ jwst _ ) is expected to detect sources down to a flux limit of @xmath95 njy ( i.e. , @xmath146 ) , in a f.o.v .
of @xmath147 arcmin@xmath145 .
we have computed the number counts predicted by our model for near - ir deep - field observations .
results are shown in fig .
3 . at the expected sensitivity of _ jwst
_ almost @xmath148 deg@xmath99 sources at @xmath72 should be detected , down to a mass limit for the accreting mbhs of @xmath149 , almost the same mass scale accessed by the deepest future x ray surveys . in spite of this good result , we have to note that _ jwst _ observations will be able to reveal only the un
absorbed population of agn , whereas x - ray survey will probe the entire agn population .
moreover , the detection of high redshift agn in the deepest _ jwst _ data could be hampered by the difficulty to select high-@xmath41 agn from the more numerous population of high redshift galaxies . in conclusion ,
x ray surveys are a better tool to investigate the early stage of mbh assembly , allowing to detect all ( type i and type ii ) agn to the highest redshifts and to the lowest mass range .
_ jwst _ deep surveys of agn might be useful to investigate the optical part of the sed , and the evolution of the type i / type ii ratio at very high redshifts .
however , optical / near - ir follow up ( with _ jwst _ and the forthcoming 30 meter class telescopes ) of x ray selected faint sources can be used to redshift determination .
unobscured agn should be recognizable both from the presence of broad emission lines and distinctive optical - near - ir colours while obscured agn should in general be identifiable from the detection of high - excitation optical / near - ir emission lines . for @xmath49
agn the strongest line will be lyman @xmath119 at 1216 .
further in the infrared civ ( 1549 ) , heii ( 1640 ) , ciii ( 1909 ) and mgii ( 2800 ) , will be all detectable by jwst , provided the gas has substantial metallicity .
of these lines the civ line is probably the strongest . a combination of _ jwst _
, 30 meter class telescopes and deep x - ray observations should provide the most effective identification strategy of agn .
we have to note here that mbhs shining at high redshifts , similarly to miniquasars ( salvaterra , haardt & ferrara 2005 ) , are not expected to contribute significantly to the cosmic background in the near - ir , and do not represent a viable solution of the near - ir background excess problem ( salvaterra & ferrara 2003 , 2006 ) .
we have also computed ( see eqs . 3 and 4 ) the background intensity at the lyman limit ( 912 ) as a function of @xmath150 using the opacity of fardal , giroux & shull ( 1998 ; model a1 ) .
results are shown in fig .
4 . as a reference
, we show the uv background ( blue solid line ) as computed with an updated version of the code cuba ( haardt & madau 1996 ) , extrapolating the observed agn luminosity function and igm opacity .
we find a good agreement between the two different approaches . in the bottom panel of fig .
4 we show the corresponding hi ionization rates .
the uv background is dominated by the most massive , i.e. @xmath151 , mbhs up to @xmath152 . at higher redshifts ,
these objects are rarer , and the background is mostly due to mbhs in the mass range @xmath153 . at @xmath154
the background intensity falls rapidly .
[ fig : reion ] finally , we may ask whether high redshift agn can contribute significantly to the reionization of the universe . in order to answer this question , we compute the redshift evolution of the filling factor of hii regions as ( barkana & loeb 2001 ) , @xmath155 where @xmath156 and @xmath157 are the present - day number densities of hydrogen and baryons ( @xmath158 is the hydrogen mass fraction ) , and @xmath159 is the production rate of ionizing photons .
the function @xmath160 takes into account the effect of recombinations . assuming a time - independent volume - averaged clumping factor @xmath161 , common to all hii regions , we can write , @xmath162,\ ] ] and @xmath163 where @xmath164 is the hydrogen recombination rate . the evolution of @xmath165 as function of redshift is shown in fig . 5 for @xmath166 ( dotted line ) , @xmath167 ( solid line ) and @xmath168 ( dashed line ) . for @xmath166 , agn alone
are able to reionize completely the universe at @xmath169 . for higher values of @xmath161 , although the model fails to reach @xmath170 , we find that the agn contribution is not negligible and might help in substaining the ionization of the igm at @xmath81 .
[ fig : lfx ] in this section , we discuss possible differences between the eddington limited model and a model allowing mbhs to accrete at super critical rate during the early phases of their evolution . in fig . 6 we plot the lf in the rest
frame hard x ray band [ 2 - 10 kev ] at @xmath18 for the two models ( top panel : eddington limited model ; bottom panel : rapid growth model ) . since both models reproduce the observed optical lf at @xmath18 ( volonteri & rees 2005 , begelman et al .
2006 ) , the x ray lfs are also quite similar . for the eddington
limited model the differential x ray lf can be described by a simple power
law with power index @xmath171 , whereas for the super critical accretion model @xmath172 . the open triangle in fig . 6 shows the estimated number density of @xmath173 agn as obtained by barger et al .
( 2003b ) in the _ chandra _ deep field north .
both models are in good agreement with the observed number density .
although the two models share similar results at @xmath18 , the lf at higher redshift shows significant differences . for example , the number density of bright quasars at @xmath174 ( i.e. with luminosity in the rest - frame hard x ray band in the range @xmath175 erg s@xmath3 , corresponding to mbhs with @xmath176 ) is @xmath177 mpc@xmath178 for the super critical accretion model , whereas in the eddington
limited model this density decreases by almost an order of magnitude .
this result is confirmed by fig . 7 , where the solid line shows the redshift distribution of sources with fluxes above @xmath179 ( the planned spectroscopic flux limit of future x ray missions ) in the observed soft x ray band for the two models . a model that allows an early phase of super critical accretion will have a redshift distribution of sources pushed towards higher redshifts with respect to the eddington
limited model .
dotted lines in fig .
7 show the redshift distribution of sources with x ray fluxes above @xmath179 that may also be observable with future near - ir facilities such as _ jwst _
( i.e. , at the source detection limit of @xmath180 njy ) , showing that future x ray observations along with optical ( near - ir ) identifications by the next generation of space and 30meter class telescopes might be able to discriminate the two different accretion histories . also shown in fig .
7 is the redshift distribution of sources with fluxes above @xmath181 , easily achieved by deep _
chandra _ observations ( e.g. , alexander et al .
furthermore , we note here that at the planned soft x
ray photometric limit of _ xeus _ , only @xmath182% of observable sources is expected to lie at @xmath183 , if bhs growth is eddington limited , whereas this fraction increases up to @xmath184% in case of super critical accretion .
it is unlikely that sufficient @xmath49 agn will be identified by the chandra deep field observations to discriminate between these two models .
however , we can provide indirect constraints by considering the contribution that each make to the unresolved x - ray background .
we find that the contribution to the unresolved xrb increases by a factor @xmath185 : the predicted unresolved background from sources at @xmath72 is @xmath186 in the 0.52 kev band , and @xmath187 in the 28 kev band .
results are summarized in table 1 . as a net result ,
once that the contribute to the unresolved xrb of sources at @xmath21 is considered , rapid growth model saturates the observed unaccounted background . note that , according to worsley et al .
( 2005 ) , most of the unaccounted 28 kev xrb actually lies at energies @xmath188 kev , strengthening our conclusions . we also find that the x ray background is saturated in a model in which massive seeds form late , as in koushiappas et al .
( 2004 ) . in this model seed mbh form from the low angular momentum tail of material in halos with efficient gas cooling . in first approximation , the koushiappas et al .
( 2004 ) model indicates that seed mbh form in halos with mass above the threshold @xmath189 , with a mass @xmath190 .
the largest seed mass allows for a less efficient accretion , in particular , we have assumed that the accretion rate is eddington limited , and the maximum radiative efficiency is 20% ( gammie et al 2004 ) .
also in the late formation scenario , the unresolved xrb is saturated , being @xmath191 in the 0.52 kev band , and @xmath192 in the 28 kev band .
xrb saturation could be a problem for such models , as a substantial contribution to the xrb from low
redshift ( i.e. @xmath21 ) , unresolved faint sources ( volonteri , salvaterra & haardt 2006 ) and/or galaxies ( bauer et al .
2004 ) can not be excluded .
lccl sources & & reference + & 0.52 kev & 28 kev & + faint agn ( @xmath21 ) & 0.7 & 3.5 & volonteri et al .
2006 + galaxies & 0.4 in c.g.s units ] & 0.2 in c.g.s units ] & bauer et al .
2004 + + eddington limited & 0.4 & 0.9 & this paper + rapid growth & 1.2 & 2.3 & this paper + massive seeds & 1.5 & 3.2 & this paper + + & [email protected] & [email protected] & hickox & markevitch 2006 + [ tab : xrb ] [ fig : zdist ]
in this paper we have assessed , using monte carlo simulations of dm halo merger history coupled with semi analytical recipes for the assembly of mbhs within galaxy spheroids ( volonteri et al .
2003 ) , the possibility of constraining the agn population at @xmath72 with currently available and planned space borne missions . in particular , we have considered ultra deep x ray and near - ir surveys .
we claim that , among the unresolved sources in _ chandra _ deep fields , few of them are agn at @xmath72 , whose bh masses are @xmath194 , while in the near - ir , an are three times larger than the hudf area is required in order to detect just one of these high z sources .
future x ray missions , such as _ xeus _ , and near - ir facilities such as _ jwst _ , will have the technical capabilities to detect accreting mbhs at @xmath72 down to a mass limit as low as @xmath153 . since _ jwst _
observations will reveal only the un absorbed agn population , x ray deep surveys are definitely the best tool suited for investigation of the early stages of mbh assembly . in particular , the next generation of x ray telescopes might detect as many as @xmath110 sources / deg@xmath107 at their photometric sensitivity limits in the 0.52 kev band .
further constraints on the population of high redshift agn will be provided by future 21 cm experiments like lofar ( rhook & haehnelt 2006 ) .
we have shown that our predicted population of high redshift agn would account for a significant fraction of the unresolved xrb ( 0.58 kev ) .
almost 5% of the measured xrb ( or @xmath5% of the unresolved one ) should come from sources at @xmath72 .
moreover , in our model of @xmath49 agn , the major contribution actually comes from sources at @xmath135 , with fluxes @xmath195 , i.e. that can not be detected even by the next generation of x ray telescopes ; we find that at least 3 - 4% of the measured xrb should remain unresolved even after _
observations become available .
these constraints become much more severe for a model in which super critical accretion is allowed in the early stages of the mbh growth ( volonteri & rees 2005 ) .
the model reproduces well the observational constraints at @xmath18 ( optical and x ray lfs ) , and predicts a more extended tail of sources observable at @xmath196 .
however , we find that in the rapid growth model , the predicted xrb is three times higher than in the eddington limited accretion model , saturating the unresolved fraction of the xrb in both the 0.52 kev and 28 kev energy bands . since faint sources at @xmath21 are expected to contribute substantially to the unaccounted xrb ( bauer et al .
2004 ; volonteri et al .
2006 ) , this result suggests that the occurrence and effectivity of super - critical accretion should be investigated in much more detail .
in particular , super
eddington accretion could be much less efficient in few@xmath10 peaks halos , due to the gas evacuation from the ionizing radiation emitted by the mbh seed pop iii star progenitor ( johnson & bromm 2006 ) .
eddington accretion consequently is biased towards the highest density peaks , which experience the largest number of mergers with halos containing pristine gas to replenish the gas reservoir .
models in which seeds are much more massive than pop iii star remnants , as in koushiappas et al .
( 2004 ) , saturate the unresolved fraction of the xrb in both the 0.52 kev and 28 kev energy bands as well .
we note here that different models for mbh seed formation , although predicting rather large bh masses , can have a lower formation efficiency , which can ease the constraints given by the xrb ( see eisenstein & loeb 1995 , and begelman , volonteri & rees 2006 ) .
our constraints on theoretical models are conservative , as we adopted the largest value of the unresolved xrb available in literature .
finally , we have computed the evolution of the uv background produced by the modeled population of high redshift agn . later than @xmath152 ,
the ionizing intensity is dominated by relatively massive bhs , @xmath197 , while lighter bhs contribute mostly at earlier epochs .
the uv background from agn rapidly declines at @xmath116 .
we compute the contribution of these sources to the reionization of the universe , showing that agn alone can provide enough ionizing photons only in the ( improbable ) case of an almost completely homogeneous inter galactic medium .
nevertheless , for a more clumpy medium , the agn contribution to the ionizing background is never negligible .
we note that high redshift agn can not contribute significantly to the near - ir background .
we wish to thank a. ferrara for fruitful discussions on reionization issues , and the referee , david alexander , for thoughful comments that improved the quality of the paper .
this research was supported in part by the national science foundation under grant no .
phy99 - 07949
. nsf - kitp-06 - 89 pre - print .
alexander d. m. , brandt w. n. , hornschemeier a. e. , garmire g. p. , schneider d. p. , bauer f. e. , griffiths r. e. , 2001 , aj , 122 , 2156 alexander d. m. et al .
, 2003 , aj , 126 , 539 arnaud m. et al . , 2000 , in _ x - ray evolving - universe spectroscopy the xeus science case _ , esa publication division .
baker j.g .
, centrella j.,choi d .- i . , koppitz m. , van meter j.r . ,
miller m.c . , 2006 , submitted , astro - ph/0603204 barkana r. , loeb a. , 2001 , phr , 349 , 125 barger a.j .
et al . , 2003a , aj , 126 , 632 barger a.j .
et al . , 2003b , apj , 584 , l61 bauer f.e . ,
alexander d.m .
, brandt w.n . , schneider d.p . ,
treister e. , hornschemeier a.e .
, garmire g.p . , 2004 ,
aj , 128 , 2048 beckwith s. v. w. et al . , 2006 ,
aj , in press , astro - ph/0607632 begelman m. c. , volonteri m. , rees m. j. 2006 , mnras in press , astro - ph/0602363 bouwens r. j. , illingworth g. d. , blakeslee j. p. & franx m. , 2005 , apj in press , astro - ph/0509641 brandt w. n. , hasinger g. , 2005 , ara&a , 43 , 827 bromm v. , loeb a. , 2003 , apj , 596 , 34 carr , b. j. , bond j. r. , arnett w. d. , 1984 , apj , 277 , 445 dijkstra m. , haiman z. , loeb a. , 2004 , apj , 613 , 646 eisenstein d. j. , loeb a. , 1995 , apj , 443 , 11 fan x. et al . , 2001 , apj , 122 , 2833 fan x. et al . , 2004 , apj , 128 , 515 fardal m. a. , giroux m. l. , shull j. m. , 1998 , aj , 115 , 2206 fryer c. l. , woosley s. e. , heger a. , 2001 , apj , 550 , 372 gammie , c. f. , shapiro , s. l. & mckinney , j. c. 2004 , apj , 602 , 312 johnson , j. l. & bromm , v. 2006 , mnras submitted , astro - ph/0605691 koekemoer a. m. et al . , 2003 , apj , 600 , 123 haardt f. , madau p. , 1996 ,
apj , 461 , 20 haiman z. , 2004 , apj , 613 , 36 haehnelt m. g. , rees m. j. , 1993 , mnras , 263 , 168 koushiappas s. m. , bullock j. s. , dekel a. , 2004 , mnras , 354 , 292 lapi a. , shankar f. , mao j. , granato g. l. , silva l. , de zotti g. , danese l. , 2006 , apj accepted , astro - ph/0603819 loeb a. , rasio f. a. , 1994 , apj , 432 , 52 madau p. , rees m. j. , volonteri m. , haardt f. , oh s. p. , 2004
, apj , 604 , 484 mainieri v. , et al . , 2005 , a&a , 437 , 805 marconi a. , risaliti g. , gilli r. , hunt l. k. , maiolino r. , salvati m. , 2004 , mnras , 351 , 169 miyaji t. , griffiths r. e. , 2002 , apj , 564 , 5 moretti a. , campana s. , lazzati d. , tagliaferri g. , 2003 , apj , 588 , 696 rhook k. j. , haehnelt m. g. , 2006 , mnras in press , astro - ph/0609081 salvaterra r. , ferrara a. , 2003 , mnras , 339 , 973 .
salvaterra r. , ferrara a. , 2006 , mnras , 367 , l11 .
salvaterra r. , haardt f. , ferrara a. , 2005 , mnras , 362 , l50 scannapieco e. , schneider r. , ferrara a. , 2003 , apj , 589 , 35 sazonov s. y. , ostriker j. p. , sunyaev r. a. , 2004 , mnras , 347 , 144 shakura n.i . ,
sunyaev r.a . , 1973 , a&a , 24 , 337 shapiro s. l. , 2005 , apj , 620 , 59 spergel d. n. et al . , 2003 ,
apjs , 148 , 175 .
steffen a. t. , strateva i. , brandt w. n. , alexander d. m. , koekemoer a. m. , lehmer b. d. , schneider d. p. , vignali c. , 2006 , aj , 131 , 2826 volonteri m. , haardt f. , madau p. , 2003
, apj , 582 , 559 volonteri m , madau p , quataert e. , rees m. j. , 2005 , apj , 620 , 69 volonteri m. , rees m.j . , 2005 , apj , 633 , 634 volonteri m. , salvaterra r. , haardt f. , 2006 , mnras in press , astro - ph/0606675 .
ueda y. , akiyama m. , ohta k. , miyaji t. , 2003 , apj , 598 , 886 .
yoo j. , miralda - escud j. , 2004 , apj , 614 , l25 willott c. j. , percival w. j. , mclure r. j. , crampton d. , hutchings j. b. , jarvis m. j. , sawicki m. , simard l. , 2005 , apj , 626 , 657 worsley m. a. , et al .
, 2005 , mnras , 357 , 1281 | we study , by means of dedicated simulations of massive black hole build - up , the possibility to constraint the existence and nature of the agn population at @xmath0 with available and planned x ray and near infrared space telescopes .
we find that x ray deep
field observations can set important constraints to the faint
end of the agn luminosity function at very high redshift . planned x
ray telescopes should be able to detect agn hosting black holes with masses down to @xmath1 ( i.e. , x ray luminosities in excess of @xmath2 erg s@xmath3 ) , and can constrain the evolution of the population of massive black hole at early times ( @xmath4 ) .
we find that this population of agn should contribute substantially ( @xmath5% ) to the unresolved fraction of the cosmic x
ray background in the 0.510 kev range , and that a significant fraction ( @xmath6% ) of the total background intensity would remain unaccounted even after future x ray observations . as byproduct , we compute the expected uv background from agn at @xmath0 and we discuss the possible role of agn in the reionization of the universe at these early epochs , showing that agn alone can provide enough ionizing photons only in the ( improbable ) case of an almost completely homogeneous inter galactic medium .
finally , we show that super
eddington accretion , suggested by the observed qsos at @xmath7 , must be a very rare event , confined to black holes living in the highest density peaks .
cosmology : theory black holes galaxies : evolution quasars : general |
given samples from two distributions , one fundamental and classical question to ask is : how close are the two distributions ?
first , one must specify what it means for two distributions to be close , for which many different measures quantifying the degree of these distributions have been studied in the past .
they are frequently called distance measures , although some of them are not strictly metrics .
the divergence measures play an important role in statistical theory , especially in large theories of estimation and testing .
they have been applied to different areas , such as medical image registration ( @xcite ) , classification and retrieval . in machine learning ,
it is often convenient to view training data as a set of distributions and use divergence measuires to estimate dissimilarity between examples .
this idea has been used in neuroscience , where the neural response pattern of an individual is modeled as a distribution , and divergence meaures is used to compare responses across subjects ( see , e.g @xcite ) .
later many papers have appeared in the literature , where divergence or entropy type measures of information have been used in testing statistical hypotheses . for more examples and other possible applications of divergence measures ,
see the extended technical report ( @xcite ) . for these applications and others , it is crucial to accurately estimate divergences .
+ the class of divergence measures is large ; it includes the rnyi-@xmath0 ( @xcite ) , tsallis-@xmath0 ( @xcite ) , kullback - leibler ( kl ) , hellinger , bhattacharyya , euclidean divergences , etc .
these divergence measures can be related to the csiszr-@xmath1 divergence ( @xcite ) .
the kullback - leibler , hellinger and bhattacharyya are special cases of rnyi-@xmath0 and tsallis-@xmath0 divergences .
but the kullback leibler one is the most popiular of these divergence measures .
1ex in the nonparametric setting , a number of authors have proposed various estimators which are provably consistent .
krishnamurthy and kandasamy @xcite used an initial plug - in estimator by estimates of the higher order terms in the von mises expansion of the divergence functional . in their frameworks , they proposed tree estimators for rnyi-@xmath0 , tsallis-@xmath0 , and euclidean divergences between two continuous distributions and establised the rates of convergence of these estimators .
1ex the main purpose of this paper is to analyze estimators for divergence measures between two continuous distributions .
our approach is similar on those of krishnamurthy and kandasamy @xcite and is based on plug - in estimation scheme : first , apply a consistent density estimator for the underlying densities , and then plug them into the desired formulas . unlike of their frameworks , we study the uniform bandwidth consistent estimators of these divergences .
we introduce a method to establish consistency of kernel - type estimators divergences between two continuous distributions when the bandwidthh is allowed to range in a small interval which may decrease in length with the sample size .
our results will be immediately applicable to proving uniform bandwidth consistency for nomparametric estimation of divergenge measures .
1ex the rest of this paper is organized as follows : in section 2 , we introduce divergence measures and we construct their nonparametric estimators . in section 3 , we study the unfiform bandwidth consistency of the proposal estimators .
section 4 is devoted on the proofs .
let us begin by standardizing notation and presenting some basic definitions .
we will be concerned with two densities , @xmath2 : @xmath3 $ ] where @xmath4 denotes the dimension .
the divergence measures of interest are rnyi-@xmath0 , tsallis-@xmath0 , kullback - leibler , hellinger , bhattacharyya are defined respectivelly as follows @xmath5 whenever the integrals in the underlying expressions are meaningful .
these quantities are nonnegative , and they are zero iff @xmath6 almost surely ( @xmath7 ) . these expressions can be used to measure the distance between two distributions .
remark that , the divergences @xmath8 and @xmath9 are respectively special cases of @xmath10 and @xmath11 .
we easily check that @xmath12 for the following , we focus only on the estimation of @xmath10 and @xmath11 .
the kullback - leibler , hellinger and bhattacharyya can be deducing immediately .
+ we will next provide consistent estimator for the following quantity @xmath13 whenever this integral is meaningful . plugging it estimates into the appropriate formula immediately leads to consistent estimator for the divergence measures @xmath11 , @xmath10 .
+ 1ex now , assuming that for all the rest of the paper , the density @xmath14 satisfies : @xmath15 is finite , this implies that @xmath16 is finite .
next , consider @xmath17 a sequence of independent and identically distributed @xmath18-valued random vectors , with cumulative distribution function @xmath19 a density function @xmath20 with respect to lebesgue measure on @xmath18 .
we start by giving some notation and conditions that are needed for the forthcoming sections . to construct our divergence estimators
we define , in a first step , a kernel density estimator for @xmath20 , and then substituting @xmath20 by its estimator in the divergence like functional of @xmath20 . towards this aim ,
we introduce a measurable function @xmath21 fulfilling the following conditions .
1ex ( * k.1 * ) @xmath21 is of bounded variation on @xmath22 1ex ( * k.2 * ) @xmath21 is right continuous on @xmath22 1ex ( * k.3 * ) @xmath23 1ex ( * k.4 * ) @xmath24 1ex the well known akaike - parzen - rosenblatt ( refer to @xcite and @xcite ) kernel estimator of @xmath20 is defined , for any @xmath25 , by @xmath26 where @xmath27 is the smoothing parameter . for notational convenience
, we have chosen the same bandwidth sequence for each margin .
assuming that the density @xmath1 is contiuous , one obtains a strongly consistent estimator @xmath28 of @xmath1 , that is , one has with probability @xmath29 , @xmath30 , @xmath31 .
there are also results concerning uniform convergence and convergence rates . for proving such results one usually writes the difference @xmath32 as the sum of a probabilistic term @xmath33 and a deterministic term @xmath34 , the so - called bias .
on can refer to @xcite , among other authors .
1ex in a second step , given @xmath35 , we estimate @xmath16 by setting @xmath36 where @xmath37 and @xmath38 is a sequence of positive constant .
thus , using [ divestim ] , the associated divergences @xmath11 and @xmath10 can be estimated by : @xmath39 the appraoch use to define the plug - in estimators is also develloped in @xcite in order to introduce a kernel - type estimators of shannon s entropy .
the uniform bandwidth of these divergences is related on those of the kernel estimator @xmath35 .
1ex the limiting behavior of @xmath35 , for appropriate choices of the bandwidth @xmath40 , has been studied by a large number statisticians over many decades . for good sources of references
to research literature in this area along with statistical applications consult @xcite and @xcite .
in particular , under our assumptions , the condition that @xmath41 together with @xmath42 is necessary and sufficient for the convergence in probability of @xmath43 towards the limit @xmath44 , independently of @xmath31 and the density @xmath20 .
various uniform consistency results involving the estimator @xmath43 have been established .
we refer to @xcite and the references therein . in the next section
, we will use their methods to establish convergence results for the estimates @xmath45 and deduce the convergence results of @xmath46 and @xmath47 .
we first study the strong consistency of the estimator @xmath45 defined in ( [ divestim ] ) .
we shall consider another , but more appropriate and more computationally convenient , centering factor than the expectation @xmath48 which is delicate to handle .
this is given by @xmath49 1ex [ lem1 ] let @xmath21 satisfy ( * k.1 - 2 - 3 - 4 * ) and let @xmath20 be a continuous bounded density .
then , for each pair of sequence @xmath50 , @xmath51 such that @xmath52 , together with @xmath53 , @xmath54 as @xmath55 , for any @xmath56 one has with probability 1 @xmath57 1ex the proof of lemma [ lem1 ] is postponed until section [ proof ] . [ lem2 ]
let @xmath21 satisfy ( * 3 - 4 * ) and let @xmath20 be a uniformly lipschitz and continuous density .
then , for each pair of sequence @xmath50 , @xmath51 such that @xmath52 , together with @xmath53 , as @xmath55 , for any @xmath56 we have @xmath58 1ex the proof of lemma [ lem2 ] is postponed until section [ proof ] . [ theo1 ] let @xmath21 satisfy ( * k.1 - 2 - 3 - 4 * ) and let @xmath20 be a uniformly lipschitz , bounded and continuous density .
then , for each pair of sequence @xmath50 , @xmath51 such that @xmath52 , together with @xmath53 , @xmath54 as @xmath55 , for any @xmath56 one has with probability 1 @xmath59 this , in turn , implies that @xmath60 the proof of theorem [ theo1 ] is postponed until section [ proof ] .
+ the following corollaries handle respectively the uniform deviation of the estimate @xmath61 and @xmath62 with respect to @xmath10 and @xmath11 .
[ coro1 ] assuming that the assumptions of theorem [ theo1 ] hold .
then , we have @xmath63 this , in turn , implies that @xmath64 1ex the proof of corollary [ coro1 ] is postponed until section [ proof ] .
1ex [ coro2 ] assuming that the assumptions of theorem [ theo1 ] hold .
then , we have @xmath65 this , in turn , implies that @xmath66 1ex the proof of corollary [ coro2 ] is postponed until section [ proof ] .
+ note that , the main problem in using the divergence estimates such as ( [ divestim ] ) is to choose properly the smoothing parameter @xmath40 .
the result given in ( [ amza1 ] ) , ( [ amza2 ] ) and ( [ amza3 ] ) show that any choice of @xmath67 between @xmath68 and @xmath69 ensures the consistency of the underlying divergenge estimates .
in other word , the fluctuation of the bandwidth in a small interval do not affect the consistency of the nonparametric estimator of these divergences .
1ex now , we shall establish another result in a similar direction for a class of compactly supported densities .
we need the following additional conditions .
1ex @xmath70 @xmath20 has a compact support say @xmath71 and is is @xmath72-time continuously differentiable , and there exists a constant @xmath73 such that @xmath74 1ex ( * k.5 * ) @xmath21 is of order @xmath72 , i.e. , for some constant @xmath75 @xmath76 and @xmath77 under ( * f.1 * ) the expression @xmath16 may be written as follows @xmath78 [ theo2 ] assuming conditions ( * k.1 - 2 - 3 - 4 - 5 * ) hold .
let @xmath20 fulfill ( * f.1 * ) . then for each pair of sequences @xmath52 with @xmath53 , @xmath79 as @xmath80 , for any @xmath56 , we have @xmath81 where @xmath82 1ex the proof of theorem [ theo2 ] is postponed until section [ proof ] . [ coro3 ] assuming that the assumptions of the theorem [ theo2 ] hold .
then , @xmath83 1ex [ coro4 ] assuming that the assumptions of the theorem [ theo2 ] hold .
then , for any @xmath84 we have @xmath85 1ex the proof of corollaries [ coro3 ] and [ coro4 ] are given in section [ proof ] .
1ex using the techniques developed in @xcite , the corollaries ( [ coro3 ] ) and ( [ coro4 ] ) lead to the construction of asymptotic @xmath86 certainty intervals for the true divergences @xmath10 , @xmath11 .
now , assume that there exists a sequence @xmath87 of strictly nondecreasing compact subsets of @xmath71 , such that @xmath88 for the estimation of the support @xmath71 we may refer to ( @xcite ) and the references therein . throughout
, we let @xmath89 $ ] , where @xmath90 and @xmath91 are as in corollaries ( [ coro3 ] ) and ( [ coro4 ] ) .
chose an estimator of @xmath92 in the corollaries ( [ coro3 ] ) and ( [ coro4 ] ) as the form @xmath93 thus , we have @xmath94 consequently , by defining the quantities @xmath95 we get from corollaries ( [ coro3 ] ) and ( [ coro4 ] ) , @xmath96 and @xmath97 thus , we obtain asymptotic certainty interval for @xmath10 and @xmath11 in the following sense . + for each @xmath98 , we have , as @xmath99 , @xmath100\right)\longrightarrow1.\ ] ] and @xmath101\right)\longrightarrow1.\ ] ] finally , we will say that the intervals @xmath102,\ ] ] and @xmath103,\ ] ] provide asymptotic @xmath86 certainty intervals for the divergences @xmath10 and @xmath11 .
we have addressed the problem of nonparametric estimation of a class of divergence measures .
we are focusing on the rnyi-@xmath0 and the tsallis-@xmath0 divergence measures . under our study , one can easily deduced kullback - leibler , hellinger and bhattacharyya nonparmetric estimators .
the results presented in this work are general , since the required conditions are fulfilled by a large class of densities .
we mention that the estimator @xmath45 in ( [ divestim ] ) can be calculated by using a monte - carlo method under the density @xmath14 .
and a pratical choice of @xmath104 is @xmath105 where @xmath106 and @xmath107 .
+ it will be interesting to enrich our results presented here by an additional uniformity in term of @xmath104 in the supremum appearing in all our theorems , which requires non trivial mathematics , this would go well beyond the scope of the present paper .
another direction of research is to obtain results , in the case where the continuous distributions @xmath1 and @xmath14 are both unknown .
* proof of lemma [ lem1 ] . * to prove the strong consistency of @xmath45 , we use the following expression @xmath108 where @xmath109 and @xmath38 is a sequence of positive constant . define @xmath110 we have @xmath111 since @xmath112 is a 1-lipschitz function , for @xmath113 then @xmath114 . + therefore for @xmath115 , we have @xmath116 where @xmath117 denotes , as usual , the supremum norm , i.e. , @xmath118 .
hence , @xmath119 finaly , @xmath120 1ex we now impose some slightly more general assumptions on the kernel @xmath21 than that of theorem [ theo1 ] . consider the class of functions @xmath121 for @xmath122 , set @xmath123 , where the supremum is taken over all probability measures @xmath124 on @xmath125 , where @xmath126 represents the @xmath127-field of borel sets of @xmath18 . here , @xmath128 denotes the @xmath129-metric and @xmath130 is the minimal number of balls @xmath131 of @xmath128-raduis @xmath132 needed to cover @xmath133 .
+ we assume that @xmath133 satisfies the following uniform entropy condition .
1ex ( * k.6 * ) for some @xmath134 and @xmath135 , @xmath136 1ex ( * k.7 * ) @xmath137 is a pointwise measurable class , that is there exists a countable sub - class @xmath138 of @xmath137 such that we can find for any function @xmath139 a sequence of functions @xmath140 in @xmath141 for which @xmath142 this condition is discussed in @xcite .
it is satisfied whenever @xmath143 is right continuous .
1ex _ remark that condition ( * k.6 * ) is satisfied whenever ( * k.1 * ) holds , i.e. , @xmath21 is of bounded variation on @xmath18 ( in the sense of hardy and kauser , see , e.g. @xcite and @xcite . condition ( * k.7 * ) is satisfied whenever ( * k.2 * ) holds , _
i.e. _ , @xmath21 is right continuous ( refer to @xciteand the references therein ) .
_ 1ex from theorem 1 in @xcite , whenever @xmath21 is measurable and satisfies ( * k.3 - 4 - 6 - 7 * ) , and when @xmath20 is bounded , we have for each pair of sequence @xmath144 , @xmath145 such that @xmath146 , together with @xmath147 and @xmath148 as @xmath55 , with probability 1 @xmath149 since @xmath150 , in view of ( [ terme1 ] ) and ( [ mason2005 ] ) , we obtain with probability 1 @xmath151 it concludes the proof of the lemma .
* proof of lemma [ lem2 ] .
* + let @xmath152 be the complement of @xmath153 in @xmath18 ( _ i.e _ , @xmath154 ) .
we have @xmath155 with @xmath156 and @xmath157 .
repeat the arguments above in the terms @xmath158 with the formal change of @xmath28 by @xmath1 .
we show that , for any @xmath159 , latexmath:[\[\label{result2 } which implies latexmath:[\[\label{term2 } on the other hand , we know ( see , e.g,@xcite ) , that since the density @xmath20 is uniformly lipschitz and continuous , we have for each sequences @xmath161 , with @xmath162 , as @xmath99 , @xmath163 thus , @xmath164 .
it is obsious to see that @xmath165 thus , @xmath166 hence , @xmath167 thus , in view of ( [ mason20051 ] ) , we get @xmath168 finaly , in view of ( [ form1 ] ) and ( [ form2 ] ) , we get @xmath169 it concludes the proof of the lemma . * proof of theorem [ theo1 ] . *
we have @xmath170 combinating the lemmas ( [ lem1 ] ) and ( [ lem2 ] ) , we obtain @xmath171 it concludes the proof of the theorem .
* proof of corollary [ coro1 ] . *
remark that @xmath172 using the theorem ( [ theo1 ] ) , we have @xmath173 and the corollary [ coro1 ] holds * proof of corollary [ coro2 ] . *
a first order taylor expansion of @xmath174 arround @xmath175 and @xmath176 gives @xmath177 remark that from theorem [ theo1 ] , @xmath178 which turn , implies that @xmath179 thus , for all @xmath180 consequently @xmath181 and the corollary [ coro2 ] holds .
* proof of theorem [ theo2 ] .
* under conditions @xmath182 , @xmath183 and using taylor expansion of order @xmath72 we get , for @xmath184 , @xmath185 where @xmath186 and @xmath187 thus a straightforward application of lebesgue dominated convergence theorem gives , for @xmath188 large enough , @xmath189 let @xmath190 be a nonempty compact subset of the interior of @xmath71 ( say @xmath191 ) . + first , note that we have from corollary 3.1.2 .
p. 62 of viallon ( 2006 )
( see also , @xcite , statement ( 4.16 ) ) .
@xmath192 set , for all @xmath193 , @xmath1940,1 [ , \\ \label{valiron2d } & \leq & \displaystyle \sup_{x\in \mathrm{j } } \vert \widehat{f}_{n}(x)- f(x ) \vert^{\alpha } \int_{\mathrm{j } } g^{1-\alpha}(x)dx,\\ \label{valiron2 } \displaystyle & \leq & \sup_{x\in \mathrm{j } } \vert \widehat{f}_{n}(x)- f(x ) \vert^{\alpha } \int_{\mathrm{r}^d } g^{1-\alpha}(x)dx.\end{aligned}\ ] ] one fined , by combining ( [ valliron ] ) and ( [ valiron2 ] ) @xmath195 @xmath196 let @xmath197 be a sequence of nondecreasing nonempty compact subsets of @xmath198 such that @xmath199 now , from ( [ dem ] ) , it is straightforward to observe that @xmath200 the proof of theorem [ theo2 ] is completed .
* proof of corollary [ coro3 ] . *
a direct application of the theorem [ theo2 ] leeds to the corollary [ coro3 ] .
* proof of corollary [ coro4 ] . * here
again , set , for all @xmath193 , @xmath201 a first order taylor expansion of @xmath202 leeds to @xmath203 using condition @xmath204 , @xmath20 is compactly supported ) , @xmath20 is bounded away from zero on its support , thus , we have for @xmath188 enough large , there exists @xmath205 , such that @xmath206 , for all @xmath207 in the support of @xmath20 . from ( [ valiron2d ] )
, we have @xmath208 hence , @xmath209 one fined , by combining the last equation with ( [ valliron ] ) @xmath210 @xmath211 @xmath212 the proof of corollary is completed .
9 clarkson , j. a. and adams , c. r. ( 1933 ) . on definitions of bounded variation for functions of two variables . _ trans .
_ , 35 ( 4 ) , 824 - 854 .
csiszr , i. ( 1967 ) .
information - type measures of differences of probability distributions and indirect observations .
_ studia sci .
hungarica , _ 2 : 299 - 318 .
deheuvels , p. ( 2000 ) .
uniform limit laws for kernel density estimators on possibly unbounded intervals .
_ in recent advances in reliability theory _
( bordeaux , 2000 ) , stat .
technol . , pages 477 - 492 .
birkha boston , deheuvels , p. and mason , d. m. ( 2004 ) .
general asymptotic confidence bands based on kernel - type function estimators .
inference stoch . process .
_ , 7(3 ) , 225 - 277 .
deroye , l. and gyorfi , l. ( 1985 ) .
nonparametric density estimation .
_ wiley series in probability and mathematical statistics _ : tracts on probability and statistics .
john wiley @xmath213 sons inc .
, new york .
the l1 view .
rnyi , a. ( 1961 ) . on measures of entropy and information .
_ in fourth berkeley symposium on mathematical statistics and probability .
_ rnyi , a. ( 1970 ) .
probability theory .
_ publishing company , amsterdam . _ villmann , t. and haase , s. ( 2010 ) .
mathematical aspects of divergence based vector quantization using frechet - derivatives .
_ university of applied sciencesmittweida . _
vituskin , a. g. ( 1955 ) .
o mnogomernyh variaciyah .
tehn.teor . lit . , | we propose nonparametric estimation of divergence measures between continuous distributions .
our approach is based on a plug - in kernel - type estimators of density functions .
we give the uniform in bandwidth consistency for the proposal estimators . as a consequence ,
their asymptotic 100% confidence intervals are also provided . , , + |
historically , string theory is originated as an attempt to provide a theory of strong interactions .
the famous veneziano amplitude is shown to be derivable from the vibrating string which sweeps out a two dimensional world sheet in space time .
the scattering amplitude is calculated from the correlation function on the riemann surface which can be regarded geometrically as punctured riemann surface . two dimensional conformal field theory ( cft )
plays a central role in the development of the string theory .
the most important dynamical principle of this approach @xcite is the associativity of the operator product expansion ( ope ) which in turn leads to the crossing symmetry of the four particle scattering amplitude .
this crossing symmetry is the underlying reason to eliminate the high energy divergences of tree level diagrams . on the other hand ,
modular invariance of the one - loop partition function of conformal cft leads a miraculous way to eliminate high energy divergences of one loop diagrams in string theory .
furthermore , people were trying to understand the strong coupled dynamics of four dimensional quantum field theory from the srong - weak duality ( s dualtiy ) of the quantum field theory itself @xcite . during the past two decades
, we gained a lot of understandings about four dimensional gauge theories from the s dualities @xcite .
s dualities are also related to geometric langlands duality @xcite .
these two seemingly unrelated subjects are put together by gaiotto s observation @xcite on @xmath0 superconformal field theory ( scft)(see recent developments in ( @xcite ) .
it is shown that four dimensional @xmath0 scft can be realized as six dimensional @xmath3 theory compactified on a punctured riemann surface .
the coupling constants are determined by the complex structure of this punctured riemann surface while different s dual frames are determined by various degeneration limits of this surface .
the flavor groups are determined by the puncture types which are labeled by young tableaux .
everything about the gauge theory is encoded into this punctured riemann surface .
recalling our discussion about the conformal field theory , it is natural to wonder if there is any connection between those two subjects .
agt @xcite proposed a remarkable conjecture that for @xmath1 theory the partition function of the gauge theory is equivalent to the correlation function of the liouville theory .
liouville theory falls into the general framework of cft @xcite .
one may wonder why a four dimensional theory can be equivalent to a two dimensional theory .
this might be related to a remarkable property of the liouville theory : the central charge is adjustable .
then the degrees of freedom of liouville theory is rather mysterious and makes a four dimensional correspondence possible .
it is interesting to further explore the correspondence between gauge theory and liouville theory . in this note
, we make the observation that we can determine the transformation law under s - duality of the partition function of @xmath0 gauge theory with four fundamentals from the crossing symmetry of the liouville four point function .
the transformation law for @xmath2 @xmath1 theory under s - duality can be determined similarly from the one - loop partition function of liouville theory .
let s first review how the dual string model was proposed @xcite .
consider an elastic scattering amplitude with incoming spinless particles of momentum @xmath4 , @xmath5 and outgoing spinless particles of momentum of @xmath6 , @xmath7 ( see figure 1 ) .
the conventional mandelstam variables are @xmath8 , @xmath5 and outgoing particles of momentum @xmath6 , @xmath7 .
we indicate the contribution from @xmath9 channel and @xmath10 channel .
the field theory amplitude is constructed from the sum of those contributions .
, width=336 ] consider first the @xmath10 channel contribution .
there are various particles with mass @xmath11 and spin @xmath12 which might be exchanged : @xmath13 we have the following similar amplitude if we consider @xmath9 channel : @xmath14 two remarkable properties of the scattering amplitude are that the above sums are infinite and these two amplitudes are equal to each other @xmath15 . the last property which is called @xmath16 duality motivates the proposed veneziano amplitude .
it is well known that the veneziano amplitude can be derived from two dimensional ( world sheet ) string theory .
the infinite sum is due to the infinite number of states in mass spectrum .
the @xmath16 duality is simply the crossing symmetry of the four point function of conformal field theory .
this crossing symmetry is also equivalent to the associativity of the ope on the world sheet : @xmath17 we briefly summarize some properties of cft .
the virasoro algebra is @xmath18=l_{m+n}+{c\over12}n(n^2 - 1)\delta_{n+m,0},\ ] ] here @xmath19 is the central charge and @xmath20 are generators of conformal symmetry .
the representations of this algebra are labeled by primary states which satisfy : @xmath21 @xmath22 is the conformal dimension of this primary state .
the other states of this representation are represented as : @xmath23 here @xmath24 .
these secondary states have conformal weights @xmath25 ; here @xmath26 is the total boxes of the young tableaux with rows @xmath27 .
the correlation functions involving the energy momentum tensor and secondary states are expressed by the correlation functions of the primary states .
the ope of two primary states are given as @xcite : @xmath28 the most important dynamical information are @xmath29 and the conformal dimensions .
the four point function has the form @xmath30 here we fixed the positions of three vertex operators as @xmath31 and @xmath32 is the projective invariant variable . the crossing symmetry is @xmath33 using ope of @xmath34 , the four point function is @xmath35 here @xmath36 is the conformal block which is entirely determined by the conformal symmetry(we give the s channel contribution ) .
the crossing symmetry relates the contributions of different channels ( see for instance figure 2 ) : next let s consider the one - loop partition function defined on a torus : @xmath37 here @xmath38 runs over all the states in cft and @xmath39 is the fermion number , @xmath40 .
it is well known that this partition function is needed to be invariant under the @xmath41 modular group transformation of the torus .
the high energy density states of this theory is determined by the central charge for compact unitary conformal field theory .
for liouville theory , there is a modification to this result , see @xcite .
it is proposed in @xcite that a strongly coupled gauge theory can be described by a weakly coupled theory in which the elementary particles are the monoples of the original theory .
this proposal is naturally realized in @xmath2 supersymmetric gauge theory and is extended to a @xmath41 duality group .
later , @xmath42 @xmath0 @xmath1 gauge theory with four fundamental hypermultiplets has also been shown to have the @xmath41 duality @xcite .
gaiotto found an extremely useful way to describe the gauge structure of different s - dual frames of this theory .
we use the @xmath43 subgroup of full @xmath44 flavor group .
this theory can be realized as the six dimensional @xmath3 @xmath45 theory compactified on a sphere with four punctures .
the different s dual frames are realized as the different degeneration limits of this punctured sphere , see figure 3 .
@xmath1 gauge theory with four fundamentals.,width=336 ] the seiberg - witten ( sw ) curve of this theory is written as @xmath46 we have fixed the positions of three punctures and left an unfixed puncture which is identified with the gauge coupling constant .
the sw curve for the mass deformed theory is @xmath47 , where @xmath48 for each puncture , we associate a mass parameter @xmath49 to it ; the physical masses of the fundamentals are given by @xmath50 the partition function of this theory on @xmath51 is given by @xcite : @xmath52e^{-{4\pi^2r^2\over g_{ym}^2}(a , a)}z_{1loop}|z_{inst}^n(r^{-1},r^{-1},a)|^2 \label{2}\ ] ] here @xmath53 is the parameter for the coulomb branch and @xmath54 is the radius of the sphere .
the 1-loop part is given by : @xmath55 here @xmath56 is given by barnes s
g function@xcite @xmath57 ; @xmath58 is the mass parameters for the four fundamentals and @xmath53 is the coulomb branch parameter .
the instanton part of the partition function is identified with the nekrasov instanton partition function @xcite @xmath59 .
notice that for the @xmath51 case @xmath60 .
next , let s study the @xmath2 su(2 ) theory .
it is given by the six dimensional @xmath45 theory compactified on a smooth torus .
the @xmath41 duality of the gauge theory is interpreted as the @xmath41 modular invariance of this torus .
the full partition function of this theory is of the same form as formula ( [ 2 ] ) .
it is interesting to note that for @xmath2 @xmath61 gauge theory the one - loop part is trivial @xmath62 and the full partition function for @xmath61 gauge theory is : @xmath63 where @xmath64 , @xmath65 .
agt @xcite made a conjecture that the full partition function of the above @xmath0 @xmath1 scft is equivalent to the correlation function of liouville theory .
it is shown in @xcite that the instanton part of the gauge theory partition function is equal to the conformal block of the correlation function and the one - loop part and classical part correspond to the structure constant part of the correlation function .
it is also argued that the energy momentum tensor of liouville theory is related to the operator ( [ 5 ] ) ( this can be seen from the classical uniformization problem with the punctured sphere ) .
the relation between the deformation parameters and the parameters in liouville field theory is @xmath66 here @xmath67 and @xmath68 are the deformation parameters in nekrasov s instanton partition function .
notice that in order to use the partition function on @xmath51 , we need to set @xmath69 .
we associate a exponential vertex operator @xmath70 to each puncture .
we also associate a intermediate state @xmath71 to weakly coupled @xmath1 group with coulomb parameter @xmath53 .
the exact relations between the parameters in gauge theory and liouville theory are @xmath72 here @xmath73 is the conventional parameter for the liouville theory @xmath74 . see figure 4 for the correspondence .
gauge theory in one particular s dual frame , the coulomb branch parameter is @xmath53 .
one the right hand side , we draw the @xmath9 channel contribution to the four point function of liouville theory.,width=336 ] now the crossing symmetry ( [ 1 ] ) of cft states that the correlation function of different channels are related . when we consider the gauge theory , the different channels mean different @xmath75 dual frames ( see figure 3 ) . with the identification between the partition function of gauge theory and correlation function of cft , we conclude that : the partition function of this four dimensional scft in different s dual frames are related as in formula @xmath76 .
notice that the gauge coupling is identified with the position of the unfixed coordinate of the vertex operator , and the second identity in @xmath76 relates theories with gauge couplings @xmath77 .
so this identity implies that the partition function of one strongly coupled theory is determined by another weakly coupled theory as predicted by s - duality !
we have determined the exact relations between them from liouville theory .
next let s consider @xmath2 @xmath1 theory , then we are tempting to identify the gauge theory partition function with the partition function of liouville theory , the liouville partition function can be calculated from ( [ 3 ] ) : @xmath78 here @xmath79 is the zero mode contribution and is independent of @xmath80 , this is identified with the gauge theory partition function ; we also need to identify @xmath81 for intermediate state and the gauge coupling is @xmath82 which is consistent with our previous identification . comparing to ( [ 4 ] ) with @xmath83 , we can see that the @xmath84 part contribution to gauge theory partition function is @xmath85 this result can be generalized to @xmath86 case in which the @xmath87 conformal toda field theory is involved .
the one - loop partition function of @xmath87 theory is equivalent to @xmath88 free scalars so the partition function has a factor @xmath89 , comparing with ( [ 4 ] ) , the @xmath84 part contributes a factor @xmath90 .
in this note , we derive the transformation law for the partition function of certain su(2 ) scft under the s duality transformation from liouville theory point of view .
it is easy to generalize to higher rank gauge theory ( for higher rank gauge theory , the cft side is the conformal toda field theory @xcite ) . consider a six dimensional @xmath3 theory compactified on a punctured riemann surface @xmath91 , we get a four dimensional gauge theory on a four manifold @xmath92 ; on the other hand , we can first compactify six dimensional theory on a four manifold @xmath92 ( with possible singularities ) , we can get a cft on @xmath91 .
it is interesting to identify @xmath91 and @xmath92 and the corresponding gauge theory and conformal field theory .
we may also have the beautiful correspondence between gauge theory and cft .
finally , liouville theory plays a fundamental role in non - critical string theory , and it has interesting application to cosmology@xcite .
any understanding of the dynamics of the theory is interesting .
the dynamical information of the theory is encoded in the structure constant of the theory .
this can be calculated from the one - loop part of the partition function .
it is interesting to see what we can learn from gauge theory about the property of liouville theory , for instance : the spectrum , the seiberg bound ... 99 a. a. belavin , a. m. polyakov and a. b. zamolodchikov , nucl.phys.b241:333-380,1984 .
p.goddard , j.nuyts , and d.i.olive , `` nucl.phys.b125(1977)1-28 . c. montonen and d. olive , phys.lett.b72:117,1977 . n.seiberg and e.witten , nucl.phys .
b 431 ( 1994 ) 484 [ arxiv : hep - th/9408099 ] .
edward witten [ arxiv:0905.2720][hep - th ] .
davide gaiotto,[arxiv:0904.2715 ] [ hep - th ] .
p.c.argyres and n.seiberg , jhep 0712 ( 2007 ) 088 [ arxiv:0711.0054][hep - th ] .
d.gaiotto and j.maldacena , [ arxiv : 0904.4466][hep - th ] .
yuji tachikawa , [ arxiv : 0905.4074 ] ; francesco benini , sergio benvenuti , yuji tachikawa,[arxiv:0906.0359 ] .
a.marshakov , a.mironov , a. morozov,[arxiv:0907.3946][hep - th ] ; andrei mironov , sergei mironov , alexei morozov , andrey morozov , [ arxiv:0908.2064][hep - th ] ; a.mironov , a.morozov , [ arxiv:0908.0307][hep - th ] ; davide gaiotto , [ arxiv:0908.0307][hep - th ] .
kazunobu maruyoshi , masato taki , seiji terashima , futoshi yagi , [ arxiv:0907.2625][hep - th ] .
nadav drukker , david r. morrison , takuya okuda , [ arxiv:0907.2593][hep - th ] .
dimitri nanopoulos , dan xie,[arxiv:0907.1651][hep - th ] .
davide gaiotto , gregory w. moore , andrew neitzke , [ arxiv:0907.3987][hep - th ] .
luis f. alday , davide gaiotto , yuji tachikawa , [ arxiv:0906.3219 ] .
a.b.zamolodchikov , al.b.zamolodchikov , nucl.phys .
b477 ( 1996 ) 577 - 605 , [ arxiv:9506136][hep - th ] . m.b.green , j.h.schwarz , e.witten , ' ' superstring theory .
vol . 1 : introduction " , cambridge , uk : univ .
dimitri v. nanopoulos , dan xie , phys.rev.d78:044038,2008 .
vasily pestun,[arxiv:0712.2824][hep - th ] .
v.s.adamchik,[arxiv:0308086][math ] .
nikita a. nekrasov , adv.theor.math.phys.7:831-864,2004 , [ arxiv:0206161][hep - th ] .
niclas wyllard , [ arxiv:0907.2189][hep - th ] .
i. antoniadis , c. bachas , john ellis , d. v. nanopoulos , phys.lett.b211:393,1988 , nucl.phys.b328:117-139,1989 ; j. ellis , n.e .
mavromatos , m. westmuckett , d.v .
nanopoulos , int.j.mod.phys .
a21 ( 2006 ) 1379 - 1444 . | in this note , we explore the relation between crossing symmetry and modular invariance in conformal field theory and s - duality in gauge theory . it is shown that partition functions of different s dual theories of @xmath0 @xmath1 gauge theory with four fundamentals can be derived from the crossing symmetry of the liouville four point function .
we also show that the partition function of @xmath2 @xmath1 gauge theory can be derived from the liouville partition function on torus . |
the epoch of reionization ( eor ) , during which high energy photons produced by the first luminous objects reionized the neutral hydrogen in the intergalactic medium ( igm ) , was an important milestone in cosmic history @xcite .
the latest measurements from the planck satellite suggest that the universe was reionized at redshift @xmath1 @xcite , in agreement with the seven - year results of wilkinson microwave anisotropy probe ( wmap ) @xcite , while studies of gunn - peterson absorption @xcite of high - redshift quasars ( qsos ) suggest that reionization began as early as @xmath2 and ended at @xmath3 @xcite .
the reionization history strongly constrains not only the formation of the first generation of galaxies and qsos , but also their feedback and impacts on structure formation at later times @xcite . over the past few years
, impressive progress has been made in detecting distant objects .
recent observations using both broad - band colors ( e.g. , @xcite ) and narrow - band ly@xmath4 emission ( e.g. , @xcite ) have detected hundreds of galaxies at @xmath5 .
meanwhile , over two dozen luminous qsos have been detected at @xmath3 ( e.g. , @xcite ) .
while it is generally believed that early star - forming galaxies played an important role in reionizing the universe @xcite , the ionization process and the actual contributions from different ionizing sources remain poorly understood .
the 21 cm hyperfine line of neutral hydrogen has been proposed as a powerful tool to probe the eor , as it traces the thermal history of the igm and the ionization structures ( e.g. , @xcite ) .
recent advances in radio instrumentation and techniques will soon make it possible to measure the highly redshifted 21 cm line from gas during the first billion years after the big bang , as a number of radio interferometers are currently being built or planned , such as murchison widefield array ( mwa ; * ? ? ? * ) , the low frequency array ( lofar ; * ? ? ? * ) , the precision array to probe the epoch of reionization ( paper ; * ? ? ? * ) , the giant meter - wave radio telescope ( gmrt ; * ? ? ? * ) , and the square kilometre array ( ska ; * ? ? ? * ) . in order to understand and interpret observations from these instruments ,
it is critical to understand the 21 cm structures from different ionizing sources . it has been suggested by advanced numerical simulations that the first stars ( so - called `` population iii '' stars ) started to form in mini - halos of @xmath6 as early as @xmath7 @xcite .
the first galaxies are believed to have formed in low - metallicity halos via gas accretion or mergers around @xmath2 thanks to feedback and metal enrichment from population ( pop ) iii stars @xcite .
meanwhile , the massive black holes ( bhs ) may have formed from remnants of massive pop iii stars or direct collapse of gas clumps or supermassive stars ( e.g. , @xcite , leading to the emergence of the first qsos @xcite .
these objects provided strong uv radiation that reionized the neutral hydrogen . to date , a number of theoretical works have studied the 21 cm signals of pop iii stars @xcite , galaxies @xcite , and qsos @xcite .
the 21 cm signal can be emission or absorption against the cosmic microwave background ( cmb ) , if spin temperature ( @xmath8 ) is above or below the cmb temperature ( @xmath9 ) .
the scattering process of @xmath0 photons is the main mechanism which causes @xmath8 decouple from @xmath9 via the wouthuysen - field effect @xcite .
therefore , it is crucial to take into account @xmath0 scattering in the calculation of 21 cm signals . the @xmath0 radiation field in a moving medium like the hubble flow is difficult to estimate analytically , because the diffuse approximation is no longer valid in a partially ionized or moving medium with higher relative velocity @xcite , and so must be determined through detailed radiative transfer ( rt ) calculations .
however , despite the importance of the wouthuysen - field effect , the @xmath0 radiation field was simply modeled or completely ignored in most of the previous work ; only a few actually included @xmath0 rt in the estimate of @xmath8 ( e.g. , * ? ? ?
* ; * ? ? ?
. in particular , @xcite and @xcite combined cosmological simulations and post - processing @xmath0 rt to calculate the 21 cm signals from uv and x - ray sources .
they found that the absorption phase of the 21 cm survives throughout the eor even in the presence of strong x - ray sources , and that the brightness temperature fluctuation of the 21 cm signal evolves strongly with redshift , with a higher amplitude in the early reionization phase . in order to investigate the wouthuysen - field effect from different ionizing sources and the resulting 21 cm structures , we carry out here a comparative study of the ionization history of three types of sources : pop iii stars , primordial galaxies and qsos .
we perform rt calculations , which include both ionization and @xmath0 scattering , on individual sources embedded in the igm .
different intrinsic spectral energy distributions ( seds ) appropriate for each source type are used .
we follow the evolution of the ionization structures and temperature to derive the 21 cm signals .
the paper is organized as follows . in 2 , we describe the model and methodology underlying our simulations . in 3 , we present the results , which include the structure of ionization , temperatures and the 21 cm signal around different sources , and their time evolution .
we discuss in 4 the detectability of the 21 cm structures of these sources by upcoming facilities such as mwa , lofar and ska , and we summarize in 5 .
in this work , we carry out rt calculations that include ionization and @xmath0 scattering on three types of ionizing sources embedded in the igm : pop iii stars , galaxies , and qsos . the igm is modeled as spherical shells around the source .
the boundary of the sphere is set to @xmath10 , where @xmath11 is the radius of strmgren sphere , and it is linearly divided into 300 bins . for each source type , we consider a range of masses in the redshift range @xmath12 . for pop iii stars , @xmath13 ; for galaxies , @xmath14 , and for qsos , @xmath15 , where @xmath16 is total stellar mass . for pop iii stars with @xmath17 , we assume multiple star formation of @xmath18 in a halo ( e.g. , * ? ? ? * ) , or a cluster of halos with pop iii stars .
the fluctuation in 21 cm intensity ( or fluctuation of brightness temperature ) from different regions of the igm at a given redshift @xmath19 depends sensitively on the gas properties , including density , velocity gradients , gas temperature , gas spin temperature and ionization state @xcite .
we follow the procedure of previous works ( e.g. , * ? ? ?
* ; * ? ? ?
* ) which calculate the fluctuation of the brightness temperature as @xmath20 where @xmath21 is the neutral hydrogen fraction , @xmath22 is over density , @xmath8 and @xmath9 are the gas spin and cmb temperature respectively .
the igm is assumed to be uniform , i.e. , @xmath23 , as recent simulations suggest small clumpiness in the igm at high redshift @xcite .
the contribution of the gradient of the proper velocity is not considered in this work .
the gas spin temperature is controlled by thomson scattering of cmb photons , @xmath0 photon pumping , and collisions by the gas particles , as formulated by @xcite : @xmath24 where @xmath25 is the color temperature of the @xmath0 line , @xmath26 is the kinetic temperature of the gas , @xmath27 and @xmath28 are the coupling coefficients by @xmath0 photon scattering and gas collision respectively , which are calculated for each spherical shell from rt simulations of ionizing and @xmath0 photons as follows @xmath29 where @xmath30 , @xmath31 is the spontaneous emission factor of the 21 cm transition , @xmath32 is the number of scatterings of @xmath0 photons per atom per second , and @xmath33 , @xmath34 and @xmath35 are the de - excitation rates due to collision with neutral atoms , protons , and electrons , respectively .
we use the fitting formula given by @xcite and @xcite for the de - excitation rates .
since @xmath25 quickly settles into @xmath26 owing to the recoil effect of @xmath0 photon scattering , @xmath36 is assumed in our calculations .
the spin temperature depends sensitively on the coupling due to @xmath0 scattering and collision . when the coupling is strong , it decouples from the cmb temperature and becomes @xmath37 .
we consider different seds for the three types of sources , popiii stars , galaxies , and qsos in our calculations . for popiii stars , we assume a black body spectrum , and follow the formulae from @xcite to calculate the effective temperature and bolometric luminosity of a star with mass @xmath38 : @xmath39 haloes hosting pop iii stars can be metal - enriched quickly due to type - ii or pair - instability supernovae @xcite . as a result , metallicity can exceed the threshold of pop iii star formation , @xmath40 @xcite , and then pop ii stars form in halos ( e.g. , * ? ? ?
on the other hand , recent simulations showed that multiple pop iii stars form in each halo @xcite , and haloes are clustered close together @xcite . in addition , if the mass of pop iii stars is @xmath41 , they collapse directly into black holes @xcite , as a result , the halo can keep making pop iii stars even after several myr .
hence , in this work , we assume pop iii halos which consist of pop iii stars alone .
effects of mixed systems of pop iii and pop ii stars on 21 cm signal will be investigated by detailed hydrodynamics simulations in future work .
the primordial galaxies in our model are assumed to consist of only young , metal - poor stars which follow a salpeter initial mass function ( imf ) @xcite .
we assume a stellar age of 10 myr and a metallicity of @xmath42 , which are consistent with observations of high - redshift @xmath0 emitting galaxies ( e.g. , @xcite ) .
the seds of galaxies are then generated using the stellar population synthesis code pgase v2.0 @xcite for a given stellar mass with the above parameters of stellar age , metallicity and imf . in practice , the sed changes with evolution time .
however , the star formation history of high - redshift galaxies is as yet poorly understood . in this work , for simplicity
, we use the above sed with constant age , metallicity and stellar mass . note that
, however , even if we consider an evolving sed with a constant star formation model like @xmath43 ( e.g. , * ? ? ?
* ) , the differences of sizes of hii bubbles are within a factor of @xmath44 .
the sed of qsos is estimated by adding a broken power - law spectrum of massive black holes to the galaxy s sed .
the power - law spectrum is assumed , following @xcite , to be @xmath45 recent observations of the most distant qsos show that the bh accretion rates are close to the eddington limit @xcite , and simulations of early qsos by @xcite showed that the mass ratio of bhs to stars in the hosts is @xmath46 , similar to that of local galaxies ( e.g. , @xcite ) .
therefore , in our qso model , the bh mass is set to be @xmath47 , and its luminosity is derived assuming an eddington limit . for ionization and heating , we integrate the above sed up to @xmath48 .
contribution of higher - energy photons @xmath49 is negligible because of the small energy fraction and low ionization cross section .
in addition , free electrons made by x - ray ionization are highly energetic , and they make @xmath0 photons via collisions with neutral hydrogen .
roughly 40 per cent of x - ray energy can be used for the production of @xmath0 photons , and they can enhance the 21 cm signal @xcite .
we also include the additional @xmath0 photons by converting 40 per cent of x - ray energy at @xmath50 .
the seds of these three types of sources are shown in figure [ fig : sed ] .
pop iiii stars have a hard sed with a peak at @xmath51 .
the sed of galaxies drops sharply at wavelengths shortward of the lyman limit @xmath52 , while that of qso has a power - law tail from x - rays to the lyman limit owing to radiation from an accreting bh . as we will show later
, the difference in the seds at @xmath53 would result in different ionization structures of hydrogen by the three ionizing sources . on the other hand ,
both galaxies and qsos have much higher continuum flux between @xmath0 and ly@xmath54 frequencies than pop iii stars , which would lead to stronger effects of @xmath0 scattering . at @xmath55 .
the three vertical dashed lines indicate the wavelengths of heiii , heii and hii , respectively . ]
we calculate the ionization of hydrogen and helium in the igm by central sources using one - dimensional rt of ionizing photons .
recent simulations showed that the escape fraction of ionizing photons ( @xmath56 ) of high - redshift galaxies or pop iii halos can be @xmath57 @xcite .
hence , we simply assume that half of ionizing photons are absorbed by the interstellar medium ( ism ) and the rest escape from halos and ionize igm .
about 0.68 of ionizing photons absorbed by ism or igm are converted to @xmath0 photons via recombination processes . here
we ignore detailed ionization structure in halos of which the spatial scale is much smaller than the typical size of hii bubbles in the igm .
in addition , we focus on the early phase when the radiation field is localized , hence the radiation background from external sources is not included .
the time evolution of the ionization of hydrogen and helium is estimated by the following equations : @xmath58 here , @xmath21 , @xmath59 , @xmath60 , and @xmath61 are the ionization fractions of neutral hydrogen , neutral helium , ionized hydrogen and ionized helium , respectively , @xmath62 is the electron density , @xmath63 is the case b recombination rate , @xmath64 is the collisional ionization rate .
we use the fitting formula of @xcite for @xmath63 , and that of @xcite for @xmath64 .
the photo - ionization rate @xmath65 in each shell is estimated by @xmath66 where @xmath67 is the volume of the gas shell , @xmath68 is the lyman limit frequency , @xmath69 is the optical depth from the central source to the shell , and @xmath70 is optical depth of the shell .
the optical depth is calculated by @xmath71 where @xmath72 is the ionization cross section .
we use the fitting formula from @xcite , @xmath73 , @xmath74 $ ] , @xmath75 .
the evolution of ionization would lead to change of gas temperature in each shell with time as follows , @xmath76\ ] ] where @xmath77 is the boltzmann constant , @xmath78 is the total number density of hydrogen and helium , @xmath79 is the heating rate , and @xmath80 is the cooling rate .
we follow the procedures of @xcite to calculate the heating and cooling rates : we consider photo - heating for @xmath79 , while for @xmath80 , we include recombination , collisional ionization and excitation cooling processes . the upper limit of the gas temperature is set to @xmath81 , and heating by @xmath0 photon scattering is not included , as it was shown to be insignificant by @xcite .
we follow the evolution of ionization and temperature up to @xmath82 , and we calculate the @xmath0 radiation field and the 21 cm signal of the snapshots at @xmath83 and @xmath84 yr . our fiducial runs use the snapshot at @xmath85 yr at redshift @xmath86 .
the rt of @xmath0 photons in igm is a highly complex process .
it depends strongly on the @xmath0 resonant scattering , and on the density distribution and ionization state of the medium .
the frequency change resulting from the scattering is difficult to estimate analytically .
@xcite has developed a three - dimensional , monte carlo @xmath0 rt code which couples the continuum and ionization of hydrogen .
here we use the 1-d version of the code to numerically simulate the @xmath0 rt in spherical shells of the igm . in the scattering process , the outgoing frequency in the laboratory frame @xmath87 is calculated as @xmath88 where @xmath89 is the incoming frequency in the rest frame of scattering medium , @xmath90 is the line center frequency @xmath91 hz , @xmath92 is the atom velocity , @xmath93 corresponds to the doppler frequency width , @xmath94 is the velocity dispersion of the maxwellian distribution describing the thermal motions , i.e. , @xmath95 , and @xmath96 and @xmath97 are the incoming and outgoing propagation directions , respectively .
we consider @xmath0 photons from both continuum radiation of sources and recombination processes in ionized ism and igm .
the intrinsic @xmath0 luminosity is estimated by @xmath98 where the continuum spectrum is considered in the frequency range from the @xmath0 line ( @xmath99 ) to lyman limit ( @xmath100 ) , @xmath101 is the absorption fraction in the calculation boxes , @xmath102 is the conversion fraction to @xmath0 photons via the cascade decay from higher quantum levels to the ground state @xcite , and @xmath103 is the @xmath0 frequency .
the @xmath0 from ism is simply proportional to number of absorbed ionizing photons because of short recombination time - scale in ism , while that from igm depends on the volume of the ionized bubble @xmath104 and the density of the ionized hydrogen and electrons .
photons in this continuum range can be absorbed by neutral hydrogen which is then excited to higher levels of the lyman series , and some of hydrogen atoms can transit from the 2p to the 1s state by emitting @xmath0 photons .
since we focus on the early phase of ionization up to @xmath105 yr in this work , which is shorter than the recombination time scale in igm ( @xmath106 ) , the ionized volume @xmath104 in equation [ eq : la ] is generally smaller than the strmgren sphere at ionization equilibrium .
hence , the @xmath0 emission from igm is much smaller than the other components .
the @xmath0 emission from ism is dominant for pop iii stars due to the high effective temperature , while the contribution of the continuum radiation is dominant for galaxies and qsos .
we simulate the @xmath0 rt using the structure of ionization and temperature at times @xmath83 and @xmath84 yr , by assuming that the travel time of the @xmath0 photons @xmath107 .
we divide the @xmath0 radiative transfer calculations into two parts .
first , we calculate the rt of @xmath0 photons from the recombination process which are emitted from central sources .
these @xmath0 photons diffusely propagate outward .
we performed a convergence test and found an optimal number of photon packets for the rt calculations , @xmath108 , of which the number of @xmath0 photons @xmath109 . in these calculations , since we need to estimate the rate of scattering of @xmath0 photons in each shell precisely to derive the spin temperature , we can not use the `` core skipping '' technique in the @xmath0 rt which can accelerate the calculation , as used in most @xmath0 rt simulations ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
hence , the calculations are very expensive even with @xmath110 photon packets .
each simulation took a few days running on 64 processors .
second , we consider continuum photons between @xmath0 and lyman limit frequencies .
total energy of these photons is larger than that of the recombination photons in the cases of galaxies and qsos .
unlike the @xmath0 photons from the recombination process , the continuum photons can travel for long distances until they are absorbed at the lyman series frequencies after cosmological redshifting @xcite .
for example , at @xmath111 , the photons for which the frequency is just below ly@xmath54 can travel @xmath112 before the frequency is shifted to @xmath0 .
once they are trapped , some fraction are converted to @xmath0 photons after the cascade decay @xcite and experience numerous scatterings .
@xcite showed the number of scattering at the trapped place could be approximated by @xmath113 , ignoring spatial diffusion .
we use this approximation for the continuum photons . for uniform density and temperature ,
the scattering rate per atom by the continuum photons decreases with the @xmath114 profile as the radial distance from sources becomes large @xcite . as shown in figure [ fig : r2 ] , we also test the distribution of the coupling coefficient by continuum photons and show the roughly @xmath114 distribution for a case of qso of @xmath115 in uniform igm on hubble flow at @xmath86 , which is similar to the test calculation of @xcite ( other test calculations were shown in * ? ? ?
the slight difference from the @xmath114 profile is due to the not flat seds shown in figure [ fig : sed ] . at @xmath86 .
uniform igm density on hubble flow is assumed . ]
we performed a set of rt simulations on three types of ionizing sources , pop iii star , primordial galaxy and qso , in the redshift range @xmath12 .
a wide range of masses were considered for these sources : @xmath13 for pop iii stars or cluster of pop iii stars , @xmath116 for galaxies , and @xmath15 for qsos . ,
red ) , a galaxy ( @xmath117 , blue ) , and a qso ( @xmath118 , green ) , respectively . the black dotted line is the analytical solution from @xcite . ] photons , and differential brightness temperature as a function of distance from the central source at redshift @xmath86 for a pop iii star ( @xmath119 , left column ) , a galaxy ( @xmath117 , middle column ) , and a qso ( @xmath118 , right column ) , respectively .
the top row panels show the fraction of hii , heii , and heiii at @xmath120 , respectively .
the second row panels show the temperatures of spin , gas and cmb , respectively .
the third row panels are the fluctuations of brightness temperature .
the bottom row panels represent the coupling coefficients @xmath121 ( by @xmath0 photon scattering , solid line ) and @xmath28 ( by gas collision , dotted line ) .
the dashed line represents unity .
when the coupling coefficients are higher than unity , the spin temperature can decouple from the cmb temperature .
the vertical dash - dot line indicates the location of the ionization front at @xmath122 , while the long - dash - dot line indicates the location where @xmath123 . in region beyond the long - dash - dot line ,
if the coupling coefficient is higher than unity , then it would cause absorption , as is the case of the galaxy model . ]
the propagation of the ionization front of a pop iii star , a primordial galaxy , and a qso are shown in figure [ fig : ifront ] , in comparison with the analytical solution from @xcite .
the position of the ionization front is measured where the ionized hydrogen fraction @xmath122 .
the size of the ionized region from the galaxy and qso follows the analytical one closely , but that of the pop iii star appears to be smaller , owing to helium ionization by pop iii stars . as shown in figure [ fig : sed ] , the sed of pop
iii stars peaks around the lyman limit frequency of heii , hence a fraction of the ionizing photons is consumed by helium ionization , resulting in a smaller hii region compared to the analytical solution which assumes that all the ionizing photons are absorbed by neutral hydrogen . the resulting structures of the ionization , temperatures and differential brightness temperature of the above ionizing sources are shown in figure [ fig : tb ] .
the ionization structures of both the pop iii star and qso show smooth transitions from ionized to neutral state for hydrogen , while the galaxy shows a sharp transition .
this is because the high - energy photons of popiii stars and qsos have long mean free paths , and they can partially ionize the gas , making the transition smooth .
moreover , the hard seds of pop iii stars and qsos can also produce ionized heii regions , which is absent in the modeled galaxy without accreting black holes .
the size of the ionized heiii region is a fraction of @xmath124 of the hii bubble in pop iii stars and @xmath125 in qsos . at @xmath120 , defined as the position at @xmath122 . _
middle panel _ : surface area of emission @xmath126 ( filled symbols ) and absorption ring @xmath127 ( open symbols ) .
_ lower panel _
: ratio of @xmath127/@xmath126 as a function of stellar mass of the source .
the dotted lines are artificial lower limits . ]
similarly , the temperatures show different structures from these different sources .
the kinetic temperature of the gas , @xmath26 , shows a smooth transition from a hot ( @xmath128 ) to cold ( @xmath129 a few k ) state around the pop iii star and the qso , in contrast to the sharp transition in the galaxy , due to partial ionization and heating from photoionization .
the spin temperature @xmath8 shows different patterns in these sources . as indicated in equation [ eq : ts ]
, @xmath8 depends strongly on the coupling efficiency due to @xmath0 scattering and gas collision .
if the coupling is strong , @xmath8 is coupled to @xmath26 which is in general different from the cmb temperature @xmath9 . as shown in the second row panels of figure [ fig : tb ] , for the pop iii star , @xmath8
is weakly coupled to @xmath26 within the hii region due to the small number of the @xmath0 scatterings resulted from low number of @xmath0 photons in its sed . as the @xmath0 scattering becomes sparse beyond the hii region , @xmath8 is decoupled from @xmath26 and it takes the value of @xmath9 . for the galaxy , @xmath8
is completely coupled to @xmath26 at @xmath130 due to strong @xmath0 scattering , it is thus completely decoupled from @xmath9 . for the qso , @xmath8 is tightly coupled to @xmath26 within @xmath131 ( @xmath129 three times of the size of hii bubble ) , but outside of which it decouples from @xmath26 and becomes @xmath9 due to the sharp decline of @xmath0 scattering . as a result of the different ionization and temperature structures around the three sources ,
the 21 cm signal shows different features , as shown in @xmath132 in the lower panels in figure [ fig : tb ] .
the modeled galaxy shows a narrow emission ring surrounded by an extended absorption trough owing to efficient coupling between @xmath8 and @xmath26 . in the case of pop iii star and qso , since the travel of most @xmath0 photons from the recombination process is confined to the transition region with numerous resonant scatterings within the limited @xmath120 , the @xmath26 at transition region is higher than @xmath9 , so they both show extended emission in @xmath132 in the transition region , and the qso shows the extended absorption region due to continuum flux .
however , due to the dilution of flux , the absorption signal is weak ( @xmath133 k ) .
the 21 cm structure of qsos from our model differs from that of @xcite , which showed extended deep absorption shell in the outer cold region under the assumption of @xmath134 .
in addition , @xcite used the assumption of @xmath135 , which leads to emission even in the region far from qsos . in the case of pop iii stars ,
the absorption feature from our calculations is shallower than that of @xcite .
this is caused by a smaller @xmath32 due to the limited traveling time of @xmath0 photons . to further illustrate the physical process behind the emission and absorption features of the 21 cm signal , we show the radial profiles of the coupling coefficients caused by @xmath0 scatterings , @xmath121 , and by gas collisions , @xmath28 , respectively .
when the coupling coefficient is higher than unity , the spin temperature can decouple from the cmb temperature . as can be seen , @xmath28 is around unity in high - temperature region , and becomes very small in region with neutral hydrogen gas . as a result , gas collision has little effect on the change of @xmath8 , the @xmath132 , in neutral region .
on the other hand , @xmath121 shows a much higher value than unity even in the outer neutral region , and it decreases steeply with distance , because the scattering cross section decreases due to high relative velocity , and @xmath0 photons emitted near sources can not propagate to outer region within the limited travel time . on the other hand
, the continuum photons can propagate long distance until they are trapped as lyman series photons due to the doppler shift .
consequently , due to the contribution of @xmath0 photons traveling near the sources , the distribution of @xmath121 shows complicated structure , not the simple @xmath114 profile . in the case of pop iii stars ,
@xmath121 is @xmath136 in the partially ionized and moderately high temperature region @xmath137 .
therefore , in such a region , @xmath8 is much higher than @xmath9 , resulting in emission in @xmath132 .
however , at the location of @xmath138 , @xmath121 becomes @xmath139 , and less than unity in the cold region , leading to the shallower absorption structure .
the @xmath121 of the qso case shows a similar trend with that of the pop iii star .
however , since the continuum flux is much higher than for pop iii stars , these photons propagate a long distance and make the extended absorption region .
in addition , we can see the @xmath114 profile in the outer region with these continuum photons . in the case of the galaxy
, @xmath121 appears to be very large ( @xmath140 ) at the location of @xmath123 , and it drops to @xmath141 at @xmath142 , resulting in the extended absorption region over a few hundreds kpc . the resulting 21 cm maps of the pop iii star , galaxy , and qso at @xmath86 are shown in figure [ fig : map ] . clearly , the 21 cm structures of these different sources are distinctively different .
the @xmath132 of the popiii star shows a ring structure with emission in the inner region and absorption in the outer region .
the galaxy shows a very thin emission ring but a deep , extended absorption region , while the qso shows an extended emission ring and an outer weak absorption region . to further investigate the dependence of 21 cm structures on source properties , we show the emission and absorption structures of the three sources at different stellar masses in figure [ fig : ring ] , as represented by the radius of the ionized bubble @xmath143 , the surface area of emission ( @xmath126 ) and absorption ( @xmath144 ) , and their ratio .
first , the size of the ionized region increases with the mass of the source , as shown in figure [ fig : ring ] ( top panel ) .
this can be understood since the total number of ionizing photons is simply proportional to the cube - root of the total stellar mass in our models , @xmath145 .
@xmath143 is in the range of @xmath146 for popiii stars , @xmath147 for galaxies , and @xmath148 for qsos .
the size of the region ionized by qsos is consistent with that from @xcite .
pop iii stars have a higher ionizing power because their hard seds produce more ionizing photons and higher effective temperature @xmath149 .
of course , the ionized region would appear as a zero - signal hole " ( @xmath150 ) in the 21 cm structure as shown in figure [ fig : map ] .
the middle panel of figure [ fig : ring ] shows the surface areas of the emission ( @xmath126 ) and absorption ( @xmath127 ) regions in a two dimensional slice containing the central sources .
for all models , @xmath126 monotonically increases with the stellar mass of the source ; the @xmath127 , however , is clearly different depending on the source type and the size of ionized region .
the @xmath127 of galaxies are much higher than the @xmath126 , which produces the strong absorption trough as seen in figure [ fig : map ] .
in the case of qsos , @xmath127 decreases with increasing stellar mass at @xmath151 , since even continuum photons can not propagate to the outside cold , neutral gas region in the limited time .
the ratio between @xmath127 and @xmath126 is shown in the lower panel of figure [ fig : ring ] .
the ratio is in the range of @xmath152 for pop iii stars , @xmath153 for the galaxies , and @xmath154 for the qsos .
these results demonstrate that pop iii stars , the first galaxies , and the first qsos have clearly different 21 cm structures , owing to different temperature and ionization structures resulting from different photon seds and propagation of @xmath0 photons in these regions . and @xmath82 , respectively .
the mass of these three sources is @xmath155 , respectively . ] , surface area of emission @xmath126 and absorption @xmath127 , and their ratio @xmath127/@xmath126 , at different times for pop iii stars ( red triangles ) , galaxies ( blue squares ) , and qsos ( green circles ) .
the masses of these three different sources are @xmath155 , respectively .
the symbols are the same as in figure [ fig : ring ] . ]
strmgren sphere .
we derived the normalized emission volume of each stellar mass , and stacked the pdfs . ]
the evolution of the @xmath132 signal at different times is shown in figure [ fig : tbtime ] . as the ionizing bubble grows with time ,
the @xmath0 photons propagate into the more extended cold region through scattering , resulting in the stronger absorption signal . at @xmath156 , the @xmath132 of pop iii stars
has no absorption region . at @xmath157 ,
all simulations show 21 cm absorptions .
note that for pop iii stars , the evolution of ionization may not last @xmath82 due to their short lifetimes and fast metal enrichment ( e.g. , * ? ? ?
however , since the recombination time scale of igm is longer than @xmath82 , the @xmath0 photons can travel in the relic hii region and the cold region outside , which may produce similar absorption features as in figure [ fig : tbtime ] .
figure [ fig : ring_time ] shows the evolution of the size of the ionizing bubble @xmath143 , surface area of emission @xmath126 and absorption @xmath127 , and their ratio @xmath158 , at different time @xmath159 and @xmath82 for pop iii stars , galaxies , and qsos .
@xmath143 increases with @xmath160 , @xmath161 , where @xmath162 is recombination time scale .
the @xmath126 and @xmath127 of galaxies and qsos show the slow increase with time , and the ratio of the two becomes nearly constant after @xmath163 . in the case of pop iii stars , however , @xmath127 increases dramatically after @xmath163 , leading to a strong absorption feature .
this is because @xmath0 photons from the recombination process are the dominant source of @xmath0 photon scatterings , and they are trapped near the source at @xmath164 yr and can propagate to the outer cold region at @xmath165 yr .
the probability distribution function ( pdf ) of the 21 cm emission volume is shown in figure [ fig : pdf ] .
as time evolves , more @xmath0 photons propagate to cold gas region , producing the extended tail of the pdf with the absorption signal . at @xmath166
, the pdf of pop iii stars is confined to @xmath167 mk , and it extends to @xmath168 mk at @xmath82 . in the case of the qso , the pdf has the more extended tail at @xmath169 mk at @xmath166 due to the continuum photons .
at @xmath86 with different stellar masses as shown in figure [ fig : ring ] , which is estimated in the spherical top hat beam with size of the distance from the source to the outer edge of @xmath170 .
for each stellar mass , we calculate 2d @xmath132 maps with widths of @xmath171 and @xmath172 mhz , and take the mean values of the four slices . with increasing stellar mass , the symbols shift to higher @xmath173 due to the larger sizes of the hii region and the 21 cm ring .
the open symbols are the mean @xmath132 of the lowest stellar masses for each source at @xmath174 .
the yellow shaded regions indicate the sensitivity for @xmath86 by upcoming facilities with @xmath175 and @xmath176 estimated by equation ( [ eq : sensitivity ] ) , while its width corresponds to the integration time from @xmath177 to @xmath178 hours with @xmath179 . ]
the detectability of the 21 cm emission from the first luminous objects with future instruments is of great interest .
here we calculate the detectability of the 21 cm emission from pop iii stars , first galaxies and quasars with upcoming observatories such as mwa , lofar , and ska .
the noise per resolution element is estimated by @xcite as follows , @xmath180 where @xmath181 is the effective collect area , @xmath173 is the angular diameter , @xmath182 is the band width , and @xmath183 is the integration time . for the redshifted 21-cm line at @xmath184 , the contribution from the polarized galactic synchrotron foreground is dominant over the system temperature , and it has @xmath185 .
the sensitivity curves are shown in figure [ fig : detectability ] .
the yellow shade with @xmath186 roughly corresponds to the sensitivity of mwa and lofar , while that with @xmath187 corresponds to the sensitivity of ska . here
, we measure the size of the emission or absorption spheres accompanying the modeled sources with boundaries of @xmath188 .
we calculate the @xmath189 map with thickness of @xmath171 and @xmath190 around central sources , and we estimate @xmath191 by taking the absolute value of the spatial mean of @xmath189 at @xmath192 .
the @xmath191 is not sensitive to the different thicknesses , and we take the mean values of the @xmath191 for the four different thicknesses . the results of this are shown in figure [ fig : detectability ] .
different symbols represent the results of the different stellar masses as shown in figure [ fig : ring ] .
the size @xmath173 increases with the stellar mass and the size of hii bubbles . at @xmath86 ,
pop iii stars show @xmath193 , while galaxies have @xmath194 and qsos have @xmath195 . in pop iii case , there are absorption spheres comparable to emission ones .
hence the net signal becomes small due to the offset .
for the qso case , the emission spheres are relatively small to the central zero - signal holes , hence the net signal decreases . in addition , the net signal decreases for massive qsos from the offset between emission and absorption , because the absorption region becomes small due to the limited propagation distance of photons . on the other hand , in the case of galaxies ,
there are deep extended absorption spheres relative to the central holes and inner emission spheres .
thus , galaxies appear to have much stronger 21 cm signals than the pop iii stars and qsos .
the hii regions around luminous qsos and massive galaxies can be detected by mwa and lofar with @xmath196 hours integration . for low - mass galaxies , ska or an @xmath197 class telescope
will be needed to detect the hii regions around them . in the case of pop iii stars with typical mass @xmath198
, however , it appears to be difficult to detect the hii region even with ska . at @xmath174 , the @xmath191 of a pop iii star with @xmath199 and a galaxy with @xmath200
is enhanced by factor @xmath201 due to the decrease of the sizes of their hii regions and the higher gas density . on the hand , the sensitivity gets lower from @xmath86 to @xmath202 by factor @xmath203 .
hence , the detectability of pop iii stars and low - mass galaxies does not change significantly . in the case of the qso with @xmath204 , the continuum radiation is significantly diluted in cold neutral region due to the larger hii region .
the increase of @xmath205 is not large compared to pop iii stars , and hence @xmath191 does not change as much .
the redshift evolution of the 21 cm structure is also discussed in section [ sec : discussion ] our results suggest that lofar , mwa or ska may be able to detect the 21 cm signal around galaxies or qsos , and may be able to distinguish the natures of sources through the difference of emission or absorption and the size . here
, we optimistically assume that the beam size of observations is comparable to the size of the outer emission / absorption shells .
in practice , the beam size can be larger than the one used in our models . as a result ,
the net signal may decrease with the beam size .
however , the decrease will follow the same gradient of the sensitivity curve @xmath206 , hence it may not significantly affect the detectability in the above discussion .
in addition , note that we have ignored uv background radiation in this work .
if there is uv background radiation and @xmath121 becomes @xmath207 at everywhere , even igm far from sources will show emission or absorption ( e.g. , * ? ? ?
this leads to a uniform 21 cm signal over the whole sky , and makes the identification of individual sources difficult .
for such a case , only high - redshift qsos can be identified via detection of giant holes in 21 cm sky map @xcite .
in addition , the x - ray background made by black holes and supernovae affects the 21 cm signal by heating the igm @xcite .
the strength of uv and x - ray background depends on the cosmic star formation history and the formation of massive black holes , which are still under the debate .
recent observation of high - redshift galaxies have suggested that the cosmic star formation rate steeply decreases with increasing redshift at @xmath208 ( e.g. , * ? ? ?
if so , @xmath121 is unlikely to be much higher than unity at @xmath209 .
therefore , the 21 cm structure around sources at @xmath210 should be studied along with formation of galaxies and black holes in a large volume .
we will investigate the 21 cm structure s relationship with galaxy formation throuth cosmological simulation in future work .
) , galaxy ( @xmath211 ) and qso ( @xmath212 ) at @xmath213 and @xmath202 .
the snap shots at @xmath120 . ] in this work , we focus on the 21 cm signal at @xmath86 . however , the size of the ionizing bubble and the propagation distance of @xmath0 change at different redshifts .
we test the change of @xmath132 structure at @xmath214 and @xmath202 with same source properties .
figure [ fig : tbz ] shows the @xmath132 structure with @xmath215 at different redshifts .
the size of ionized bubbles decreases with increasing redshift , while the ionizing front becomes closer to the strmgren radius in the same propagation time . as a result
, the expansion rate of the ionizing front is estimated by @xmath216^{1/3}\ ] ] where @xmath217 .
hence , the size ratio of @xmath218 to @xmath86 is @xmath219 . for pop iii stars at @xmath174
, there is a deep extended absorption area . with decreasing redshift ,
the absorption signal becomes shallower , because the distance to the cold - neutral region becomes large , leading to a larger travel time to there . in the case of galaxies ,
the 21 cm structure is almost the same , however , the absorption area is relatively small at lower redshift , because the hii bubble is bigger and the transition from ionized to neutral becomes more smooth due to the lower density .
note that the qso at @xmath174 shows a clear absorption signal , despite the fact that the hii region is much larger than that of pop iii star at @xmath214 which does not show a clear absorption signal .
this is due to the higher @xmath0 luminosity of the continuum spectrum than the pop iii star .
in this paper , we investigate the redshifted 21 cm signal from different ionizing sources by combining idealized models of the first stars , the first galaxies , and quasars with radiative transfer simulations that include both ionization of neutral hydrogen and resonant scattering of @xmath0 photons .
we find that the 21 cm signal depends significantly on the source sed and stellar mass .
the pop iii stars and quasars produce a smooth transition from an ionized and hot state to a neutral and cold one , owing to their hard sed with abundant ionizing photons , in contrast to the sharp transition in galaxies .
the hii bubble size is typically @xmath220 for popiii stars , @xmath221 for galaxies , and @xmath222 for qsos .
furthermore , @xmath0 scattering plays a dominant role in producing the 21 cm signal as it determines the relation between hydrogen spin temperature and gas kinetic temperature .
the @xmath0 photons can produce both emission and absorption region around the small hii bubbles of popiii stars , extended absorption region around the first galaxies , and extended emission around qsos .
we predict that future surveys from large radio arrays such as mwa , lofar and ska may be able to detect the 21 cm signals of primordial galaxies and quasars , but not likely pop iii stars due to their small angular diameter .
these results are based on idealized models of the first stars , the first galaxies , and quasars .
we will investigate the 21 cm structures in more realistic scenarios including inhomogeneous density , metal enrichment and black hole growth by using cosmological simulations in future works .
support from nsf grants ast-0965694 , ast-1009867 ( to yl ) , and ast-0807075 ( to ta ) is gratefully acknowledged .
we acknowledge the research computing and cyberinfrastructure unit of information technology services at the pennsylvania state university for providing computational resources and services that have contributed to the research results reported in this paper ( url : http://rcc.its.psu.edu ) .
the institute for gravitation and the cosmos is supported by the eberly college of science and the office of the senior vice president for research at the pennsylvania state university . | observations of the redshifted 21 cm line with upcoming radio telescopes promise to transform our understanding of the cosmic reionization .
to unravel the underlying physical process , we investigate the 21 cm structures of three different ionizing sources , population ( pop ) iii stars , the first galaxies , and the first quasars , by using radiative transfer simulations that include both ionization of neutral hydrogen and resonant scattering of @xmath0 photons .
we find that pop iii stars and quasars produce a smooth transition from an ionized and hot state to a neutral and cold one , owing to their hard spectral energy distribution with abundant ionizing photons , in contrast to the sharp transition in galaxies .
furthermore , @xmath0 scattering plays a dominant role in producing the 21 cm signal as it determines the relation between hydrogen spin temperature and gas kinetic temperature .
this effect , also called wouthuysen - field coupling , depends strongly on the ionizing source .
it is the strongest around galaxies , where the spin temperature is highly coupled to that of the gas , resulting in extended absorption troughs in the 21 cm brightness temperature . on the other hand , in the case of pop iii stars , the 21
cm signal shows both emission and absorption regions around a small hii bubble . for quasars , a large emission region in the 21 cm signal
is produced , and the absorption region decreases as the size of the hii bubble becomes large due to the limited traveling time of photons .
we predict that future surveys from large radio arrays such as mwa , lofar and ska may be able to detect the 21 cm signals of primordial galaxies and quasars , but not likely pop iii stars due to its small angular diameter .
[ firstpage ] radiative transfer diffuse radiation galaxies : evolution galaxies : formation galaxies : high - redshift quasars : supermassive black holes |
molecular maser emission associated with newly formed massive stars commonly comprises many small spots spread over a site of typical extent 30 milliparsec , enveloping the star .
hydroxyl masers are especially useful probes of the immediate surroundings of the star since the zeeman effect leads to a readily measurable separation of left and right circularly polarized emission in magnetic fields of a few mg . at most maser sites
, it requires the high spatial resolution of very long baseline interferometry to unequivocally recognise individual zeeman patterns , and to explore the homogeneity of the magnetic field within the masing region .
pioneer observations began more than two decades ago , and there has now been increased recent activity from the vlba , with new results for 18 targets ( fish , reid , argon & zheng 2005 ) .
the target properties are found to be diverse , but hint at clearly recognisable classes of object , demonstrating a need for similar studies of the numerous potential targets in the southern hemisphere .
the southern hemisphere long baseline array ( lba ) of the australia telescope national facility ( atnf ) allows high frequency resolution over a total bandwidth spanning both the 1665- and the 1667-mhz ground - state oh transitions .
using three stations , the spatial resolution of approximately 100 milliarcsec ( mas ) is quite well suited to observations of masers at distances of a few kpc ; it allows clear position discrimination between the numerous tight clusters of spots ( of typical size @xmath1 mas ) that are commonly spread over a total extent of an arcsec . our program with the lba to study the rich variety of masers in the southern sky began with oh 323.459 - 0.079 ( caswell & reynolds 2001 ) ;
the present study is similar , but improved by a lower noise level , and analysed at higher spectral resolution .
observations were made over a 15h period 2000 august 17 .
the array comprises the parkes 64-m antenna , complemented by a 22-m paraboloid at mopra , and the compact array of the australia telescope ( atca ) at narrabri operating as a tied array ( equivalent in sensitivity to a dish of 46-m diameter ) .
these yield baselines of approximately 119 km , 203 km and 321 km , predominantly north - south .
the s2 tape system recorded both right- and left - hand senses of circular polarization ( rhcp , lhcp ) simultaneously , with a bandpass centred at 1667 mhz , and limited to a 4-mhz bandwidth by a digital filter with excellent flat response in amplitude and phase .
our strategy was to observe a target for 25-min periods , bracketed by 5-min periods on a phase and secondary amplitude calibrator ( 1148 - 671 , 1215 - 457 or 1740 - 517 , with total intensity flux densities of 1.78 , 3.37 and 7.47 jy ) .
flux densities are relative to the absolute calibration derived from 1934 - 638 ( unpolarized , with total intensity of 14.157 jy ) .
the target position of 300.969 + 1.147 ( j2000 ra 12@[email protected]@xmath4 , dec .
-61@[email protected]@xmath7 ) was alternated with another target ( 337.505 - 0.053 , to be discussed in a later paper ) for the period when both were accessible to the parkes telescope .
each baseline required a unique correlator pass in order to achieve the high frequency resolution of 8192 channels across the 4-mhz band .
the output from correlation of the signals from each polarization was binned into 5-sec integration periods . in the subsequent processing
using the aips reduction package , no hanning smoothing was applied so that the final channel separation remains 0.48828125 khz (= 0.088 @xmath0 ) , and the final velocity spectral resolution is 0.105 @xmath0 ( larger than the channel separation by a factor of 1.2 for uniform weighting ) .
the calibrators were used to derive first order phase and amplitude corrections to the target source observations , and this calibration also established the correct phase between rhcp and lhcp ( and subsequently ensures precise registration of their relative positions ) . the program ` cvel ' was used to align channels to the same velocity , and
allow for the varying doppler correction arising from the earth s motion during the observations , adopting rest frequency values of 1665.4018 and 1667.359 mhz .
after perusal of spectra from each polarization , strong channels were selected , from which preliminary maps were made .
= 8.4 cm maps of the field using only the preliminary calibration did not allow the measurement of an absolute position .
however the strongest ( 15-jy ) spectral feature , rhcp 1665 mhz at velocity -39.8 @xmath0 , has a known position derived from atca observations ( caswell 1998 ) of ra 12@[email protected]@xmath4 , dec .
-61@[email protected]@xmath7 , with an uncertainty of 0.4 arcsec
. it would be possible to use this feature as a phase reference to determine positions of all other features .
however , it turns out that a rhcp 1667-mhz feature of somewhat lower amplitude , at velocity -35.8 @xmath0 , is more suitable as a phase reference since it is more clearly distinct from other spectral features .
relative to the atca measured feature , we show later from the current lba observations that it is located at larger ra by 0.03s (= 0.22 arcsec ) and north by 0.67 " , ie at ra 12@[email protected]@xmath4 , dec .
this rhc 1667-mhz feature was then used as the reference feature for the maps .
the uncertainty in this absolute registration remains the same as the atca measurement , approximately 0.4 arcsec .
this final calibration of all data relative to a single rhcp 1667-mhz feature reduces phase errors , establishes a correct absolute position , and maintains correct relative positions between polarizations and between transitions . in the velocity range -44
to -33 @xmath0 , ` cleaned ' maps were made of each spectral line channel ( total 128 channels ) for both transitions , and for both circular polarizations .
the cellsize used was 25 mas , and the restoring beamwidth was 72 x 130 mas ( major axis at position angle 94@xmath8 ) .
the typical rms noise on a map was 8 mjy , allowing reliable detections of features down to 50 mjy , or weaker if present over several channels .
at the few velocities where a strong feature exceeding 2 jy is present , the detection of weak features at some positions is limited by dynamic range ( owing to occasional 5 per cent sidelobe levels ) .
oh spectra in fig . 1 show the total maser emission from the site at 1665 and 1667 mhz using the present observations .
earlier spectra , recorded 1982 february , can be seen in caswell & haynes ( 1987a ) , and are similar to those measured when it was discovered ( robinson et al .
the present spectra from 2000 august ( fig .
1 ) resemble the earlier spectra , but have higher sensitivity .
subsequent spectra from the parkes telescope in 2004 november and 2005 october again resemble earlier spectra . measurements of weak linear polarization in these new spectra will be discussed later insofar as they assist interpretation of the lba maps which were not analysed for linear polarization . from the ` cubes ' of the images of rhcp and lhcp at 1665 and 1667
mhz , all emission features were measured using the ` imfit ' task in aips to derive positions and flux densities , and all appear to be of small angular size ( @xmath9 mas ) relative to the beamwidth of 72 x 130 mas .
the positions measured for 59 individual features , corresponding to maser ` spots ' , are listed in table 1 ; they are spread over an elliptical region elongated 1.66 arcsec north - south .
each row of table 1 lists a narrow spectral feature ( typically 0.4 @xmath0 to halfpower ) , in a single sense of circular polarization , at one of the two oh transitions of 1665 or 1667 mhz .
the maser features are grouped by transition and sense of circular polarization , and then ordered by radial velocity within each grouping .
positional rms uncertainties for the spots were estimated both from the position fit errors in each velocity plane and from the scatter in these values across the velocity width of the feature , demonstrating that the uncertainties are generally less than 0.025 arcsec . the overall spatial pattern of the spots is displayed in fig .
2 . note that 1 second = 7.12 arcsec at this declination .
many coincidences occur between features of different polarization , and between the transitions . for the purposes of position comparison between features , we treat any offset of up to about one - third of the halfpower beamwidth as a coincidence .
position coincidences of features with different polarization , and/or different transition , result in the 59 spectral features being located in just 28 different spots , as described below .
[ cols="^,^ , > , > , > , > , > , < " , ] [ table - lrpos ] ten features of table 1 have a matching stronger feature in the opposite polarization but at the same position and velocity .
for example , 1665 mhz lhc feature q matches stronger rhc feature p. in many similar cases studied in earlier work ( e.g. section 4.3 of fish & reid 2006 ) , detailed investigation shows such features to be elliptically polarized , sometimes with additional unpolarized emission .
our data , without linear polarization analysis , do not distinguish these situations , but for simplicity , we add the comment to q : ` elliptical with p ' , and likewise the comment ` elliptical with q ' is added to p. the weaker component is not a truly separate feature , and is distinguished in table 1 by adding parentheses to both its feature label in column 1 and its comment .
this particular example is of additional interest because it is not part of a recognisable zeeman pair , and it is co - located with an unpaired 1667-mhz lhc feature i ( similarly possessing a matching stronger feature ( e ) in rhc polarization ) .
spatially coincident features of different circular polarization which have different velocities are expected to be the @xmath10-components of a zeeman pattern , the @xmath10-components are most generally elliptically polarized , usually with the circular percentage dominating over the linear percentage .
we interpret 34 features as components of 17 zeeman pairs ( 12 at the 1665-mhz transition and 5 at 1667 mhz ) , as noted in the comments column of table 1 ; the derived magnetic fields are given in the previous column .
the magnetic field values assume zeeman splitting factors such that , in a 1 mg magnetic field , a spectral line will be split into two components of opposite circular polarization , separated in frequency by an amount equivalent to 0.590 @xmath0 for the 1665-mhz transition and 0.354 @xmath0 for the 1667-mhz transition .
three of the 1667-mhz zeeman pairs closely match corresponding 1665-mhz zeeman pairs , as noted in the comments .
zeeman pairs z12 at 1665 mhz and z17 at 1667 mhz are co - located , and match in magnetic field , but differ by 0.5 @xmath0 in their mean velocity .
in addition to the 14 different spot positions hosting zeeman pairs , there are 14 spot positions ( one of them with emission at both 1665 and 1667 mhz ) which have ` unpaired ' spectral features i.e. features that are not part of a zeeman pair . further notes on some of the individual spectral features
are given below ; these include consideration of single dish parkes spectropolarimetry 2005 november , which obtained full polarimetry at 1665 and 1667 mhz with high spectral resolution but low spatial resolution .
+ + [ [ mhz - lhc - g . ] ] 1665 mhz lhc g. single dish spectra from the parkes telescope suggest 35 per cent linear polarization of this feature .
we note that the parkes observations were several years later than the lba observations , and this feature in particular had increased in intensity , but it is useful to explore the possible implications of the parkes data for interpreting the lba data . if 35 per cent linear polarization is present in the lba data , then the maximum possible circular polarization would be 94 per cent , and there would then be an expectation of a lhc component of 0.05 jy . since this is close to our detection limit
, our non - detection of a lhcp feature is not significant .
we conclude that feature g is most likely elliptically polarized , with the rh circular component dominant .
[ [ mhz - lhc - m . ] ] 1665 mhz lhc m. coincidence with feature k is noted as evidence for linear polarization in k. this is supported by the single dish data from parkes indicating at least 10 per cent linear polarization of this feature .
[ [ mhz - lhc - p . ] ] 1665 mhz lhc p. commented as ` elliptical with n ' .
this is supported by single dish spectra from the parkes telescope showing at least 10 per cent linear polarization at this velocity .
the circular polarization is 71 per cent and so if the feature is 100 per cent polarized , then its linear polarization would also be 71 per cent . [
[ mhz - rhc - l . ] ] 1665 mhz rhc l. single dish spectra from the parkes telescope suggest 10 per cent linear polarization at this velocity .
the maximum possible circular polarization would then be 99.5 per cent , and there would then be an expectation of a lhc component of 0.077 jy .
although this is not detected , the absence may result from some maser variability in the several years between lba and parkes observations .
we conclude that , although the rh circular component is dominant , there is likely to be some linear polarization also present , to give net elliptical polarization .
[ [ mhz - rhc - c . ] ] 1667 mhz rhc c. single dish spectra from the parkes telescope suggest 10 per linear polarization at this velocity
. the maximum possible circular polarization would then be 99.5 per cent and the expectation of a 20 mjy lhc component .
this is below our sensitivity level , so its absence is not significant .
+ from the median velocity of detected maser features , we can estimate the systemic velocity of oh 300.969 + 1.147 to be -38 @xmath0 .
the hii region in this direction has a radial velocity of -47 @xmath0 ( caswell & haynes 1987b ) .
the ` tangent point ' velocity , the most negative ` allowed ' velocity in this direction for commonly used galactic rotation models , is -40 @xmath0 .
we conclude that the maser and hii region are close to the tangent point , and thus at a uniquely defined distance of 4.3 kpc ( if the galactic centre is assumed to be at 8.5 kpc from the sun ; or 5.1 kpc distant if the galactic centre were at 10 kpc , the value assumed in older studies ) .
considering the zeeman pairs in more detail , we note that , throughout the whole site , the estimates of magnetic field for all 17 zeeman pairs are notably in the same sense ( towards the observer , as denoted by a negative field ) and range from -1.1 to -4.7 mg , with a median value of -3.5 mg .
further information on the magnetic field is provided by excited - state oh 6035- and 6030-mhz masers .
a position measurement for strong 6035-mhz emission at -37.5 @xmath0 ( caswell 1997 ) yields a position ra 12@[email protected]@xmath4 , dec . -61@[email protected]@xmath7 .
the spectra in fig .
1 of caswell ( 2003 ) show that this 6035-mhz reference feature is a zeeman pair arising in a magnetic field of -5.0 mg , and is accompanied by an oh 6030-mhz pair indicative of a similar field .
two other weak 6035-mhz zeeman pairs corroborate this field strength .
the 6035-mhz position is shown on fig .
2 as a dotted square ; the uncertainty in the position is larger than the square , with contributions from the absolute position registration for both species , but the location lies , encouragingly , amongst a high concentration of 1665- and 1667-mhz maser spots of which several reveal similar magnetic fields . indeed
, it may be closely associated with the oh 1665-mhz zeeman pair z5 which has essentially the same magnetic field ( -4.6 mg ) , same mean velocity , and is coincident ( the offset of 0.23 arcsec is smaller than the registration uncertainty ) .
the 6-ghz atca observations ( caswell 1997 ) also reveal a weak methanol 6668-mhz maser close to the oh 6035-mhz maser , and superposed on a uchii region with flux density 164 mjy at 6 ghz . lower resolution maps of the region at 5 ghz reveal a much higher flux density of 3.1 jy extending over an estimated size of 1 arcmin ( caswell & haynes 1987b ) , so the region is clearly quite complex , with additional emission on a much more diffuse scale than 1 arcsec . in the absence of full mapping of the uchii region ,
its boundary is sketched as a large dotted circle on fig .
2 circumscribing the maser spots and consistent with its absolute position measurement .
we now explore the kinematics , and whether the velocity field , as derived from the combination of the spatial distribution of the maser spots and their radial velocities , might reveal a pattern indicative of rotation or expansion .
however , we ought first to correct for the fact that velocity values at 1665 and 1667 mhz are not purely kinematic and have been significantly affected by the magnetic field . in order to obtain a corrected velocity estimate at each position ,
we first use the mean velocity value for each of the 17 zeeman pairs .
the corrections relative to individual spot velocities range from 0.18 @xmath0 for the lowest - field 1667-mhz feature to nearly 1.5 @xmath0 for the highest - field 1665-mhz feature . for the other ( 15 ) unpaired features
, we must first decide whether they are more likely to be @xmath11-components or lone surviving @xmath10-components from zeeman patterns .
we note that none of them has location and velocity compatible with being the @xmath11-component of any of our recognized zeeman pairs , and regard them as most likely lone surviving @xmath10-components .
we ` correct ' their velocities by assuming that they originated in the same field as their nearest zeeman pair companion .
all ` corrected ' velocities were then inspected over the whole map and the only suggestive pattern was an apparent gradient of velocity with declination . a plot of velocity against declination
was then made , and is shown in fig .
large symbols are used for the zeeman patterns since they represent the information from 2 spots rather than 1 , and have more reliable corrections . fig .
3 reveals a tendency for more negative velocities to be at the south of the distribution , with gradient of approximately 4 @xmath0 over 1 arcsec .
naively we could regard this as indicating rotation about an east - west axis .
with regard to other oh transitions , we note the detection of a loosely associated 1720-mhz maser ( caswell 2004 ) at ( j2000 ) ra 12@[email protected]@xmath4 , dec .
-61@[email protected]@xmath7 ) with uncertainty 0.4 `` . it lies outside the region shown in fig .
2 , with an offset to larger ra by 0.36s (= 2.6 '' ) and south by 0.4 " from the reference location used in this study ( for table 1 and fig .
2 ) . the 1720-mhz maser displays a clear zeeman pattern indicative of a magnetic field of -5 mg . with mean velocity -42.5 @xmath0 .
it lies in a broad absorption dip extending from -39 to -47 @xmath0 , presumably indicative of a dense , more extensive , host molecular cloud .
the spatial offset of the 1720-mhz maser from our current object of study seems large enough to rule out a common source of excitation , but it clearly lies in the same region of massive star formation .
this is corroborated by its radial velocity and its magnetic field value , both of which are in the same range as our 1665 and 1667-mhz measurements .
there has been no detection of oh at the 1612-mhz transition ( caswell 1999 ) .
in addition to the presence of the previously mentioned methanol maser at the 6668-mhz transition , there is a methanol maser at the 12-ghz transition with a weak peak at the same velocity ( caswell et al .
1995 ) .
water maser emission peaking above 40 jy and covering the velocity range -43 to -86 @xmath0 ( caswell et al .
1989 ) was reported at approximately this position but is now found ( caswell , atca unpublished data ) to be offset 18 arcsec , and thus not closely related .
the general location of the star formation complex is at a galactic latitude of more than 1@xmath5 where , with 301.109 + 0.969 ( caswell & haynes 1987b ) , it is one of only two prominent star forming regions in this portion of sky .
their separation of 0.246@xmath5 corresponds to 18 pc
. this comparative isolation will allow detailed future studies to be undertaken unaffected by other nearby companion star clusters , an isolation which can be especially valuable when investigating the more common molecular species that are able to trace the properties of a massive - star - forming environment . summarizing the properties ,
we conclude that the region , although young enough to host masers of oh and methanol , is mature enough to have a prominent uchii region .
the relative weakness of methanol , plus the absence of an intimately associated water maser , are also common properties of these more mature regions ( caswell 1997 ) .
the complex as a whole does , however , also host a water maser , a 1720-mhz oh maser , and a more extensive oh cloud ( responsible for a prominent feature of 1720 mhz absorption against the extended hii region ) .
the consistency of magnetic field direction within the source ( and even at the location of 1720-mhz emission ) also reinforces its status as a stable region which is approaching the end of the short evolutionary period that can support maser emission , according to current ideas ( caswell & reynolds 2001 ) .
the most recent major addition to high resolution mapping of oh masers is from a study of 18 targets with the vlba by fish , reid , argon & zheng ( 2005 ) , discussed in detail by fish & reid ( 2006 ) .
very few targets conform precisely to the canonical picture , in which maser spots appear projected onto a simple uchii region , with magnetic field estimates all of the same sign , and some kinematic order discernible .
most of the vlba targets lie in the galactic first quadrant , longitude 0 to 90@xmath5 , and the only one , 351.775 - 0.536 that lies in the fourth quadrant ( longitude 270 to 360@xmath5 ) is not covered over its full velocity range .
the present observations add another example from the largely untapped resource of southern fourth quadrant masers and establishes another simple morphology against which we can compare the more pathological ones . in a previous study ( caswell & reynolds 2001 ) , we compared the properties of oh 323.459 - 0.079 with those of the archetypical w3(oh ) , and concluded that it was similar , but slightly larger in linear extent and weaker , and thus perhaps slightly older .
the present target also appears larger ( largest separation of maser spots is 36 mpc ) , weaker , and it has relatively weaker methanol maser emission and no water maser .
all of these slight differences conform to expectations for a slightly more evolved maser site .
maser sites show a weak tendency to display a magnetic field direction in the sense of galactic rotation , and indeed the present target conforms to this , with a field towards us in the galactic fourth quadrant .
some counter - examples such as oh 34.257 + 0.154 occur in crowded and apparently disturbed regions of the galactic plane which may be relevant .
however , oh 323.459 - 0.079 shows a magnetic field counter to galactic rotation , and yet lies in an apparently undisturbed region .
indeed , it lies at similar distance , and in the same ( crux - scutum ) spiral arm as oh 300.969 + 0.147 , with its main point of difference being its location at the inner edge of the arm rather than the outer edge .
another notable counter - example is 285.263 - 0.050 which appears to lie close to the centre of the carina arm .
thus at present , there is no evidence that the counter - examples are merely exceptions to a general rule because of their exceptional circumstances .
rather , the relation appears somewhat weak in its significance .
in the common picture of oh masers in massive young stellar objects , a cluster of maser spots is projected against an uchii region of diameter about 30 mpc , with the pattern of maser spots revealing a magnetic field of a few mg , and a small velocity range of less than 10 @xmath0 , with velocity field perhaps indicative of rotation .
the detailed study of w3(oh ) has defined this archetype , but finding additional real examples conforming to this pattern has proved difficult . the present study establishes oh 300.969 + 1.147 as a new example conforming to this simple pattern , and we regard it as a mature region nearing the end of the brief evolutionary period that can support maser emission .
examples of these canonical maser sites are important for comparison with other diverse sites that have no discernible uchii region ( perhaps younger ) , or with maser spots scattered over a larger area ( perhaps multiple sources ) , magnetic fields more complex , and velocity patterns suggestive of outflows .
we thank the atnf staff who make the intricate operation of the lba successful , and fellow astronomers who participated in the observing session .
bhk thanks atnf for hospitality and funding as a distinguished visitor during some of the period when this work was performed . | the southern maser site oh 300.969 + 1.147 has been studied using the long baseline array of the australia telescope national facility . the 1665- and 1667-mhz hydroxyl ground - state transitions
were observed simultaneously .
a series of maps with tenth - arcsec spatial resolution , at velocity spacing 0.09 @xmath0 , and in both senses of circular polarization , reveal 59 small diameter maser spots .
the spots are scattered over 2-arcsec , coincident with a strong ultracompact hii region , at a distance of 4.3 kpc .
17 zeeman pairs of oppositely polarized spots were found , all yielding magnetic field estimates towards us ( negative ) , ranging from -1.1 to -4.7 mg , with a median value of -3.5 mg .
excited state masers of oh at 6035 mhz and 6030 mhz at this site also display zeeman pairs revealing a magnetic field of -5.0 mg .
weak methanol maser emission is intermingled with the oh masers , but there is no detectable closely related water maser .
the consistent magnetic field direction found within this site is a striking feature of several other maser sites associated with strong hii regions studied in comparable detail .
we interpret the site as a mature region nearing the end of the brief evolutionary stage that can support maser emission .
[ firstpage ] masers - stars : formation - hii regions - ism : molecules - radio lines : ism . |
in many physical phenomena percolation effects play an important role @xcite . particularly , some dilute magnets are well described , in what concerns magnetic phase transitions , by uncorrelated diluted models @xcite . in these models , magnetic sites on a lattice are randomly replaced by non - magnetic ones , and a bond links each pair of occupied ( magnetic ) first neighbours . at zero temperature ,
the problem is purely geometrical and is described in the following way .
sites on a lattice are randomly occupied with probability @xmath1 , while all bonds are considered to be present .
a cluster is then defined as a collection of present sites which are connected to each other through steps between occupied first - neighbours . as @xmath1 increases
, an infinite cluster appears for the first time at a critical probability @xmath12 , which is lattice dependent . in analogy with thermal critical phenomena , some quantities are singular at the critical point , following a power - law behavior near @xmath12 @xcite .
typical examples are the probability @xmath13 that a site belongs to the infinite cluster , which behaves as @xmath14 near @xmath4 , and the correlation length @xmath15 , which diverges at @xmath4 , according to @xmath16 . nevertheless , some systems were found to be better described by _ correlated _ percolation models , where the presence of sites ( or bonds ) depends also on their neighbourhood .
typical examples of correlated percolation models are bootstrap percolation @xcite and site - bond correlated percolation @xcite .
the physical motivation for the introduction of the former model comes from diluted magnetic systems where competition between exchange ( which favour a magnetic ground state ) and crystal - field ( which leads to a non - magnetic ground state ) interactions takes place . to mimic this competition at zero temperature , the bootstrap percolation model was introduced @xcite : in this model , sites on a lattice are randomly occupied with probability @xmath1 but only those with
at least @xmath3 occupied first - neighbours remain occupied . in the stable final configuration ,
all occupied sites have at least @xmath3 occupied first - neighbours or the whole lattice is empty . as our purpose in this paper is to study one implementation of the bootstrap percolation model , we briefly review some of its properties in what follows ( for a thourough discussion of the results up to 1990 , see reference @xcite ) .
the @xmath7 case regains the usual ( uncorrelated ) site percolation model , where the transition is continuous and @xmath17 in two and three dimensions .
more specifically , the most precise evaluations of the critical exponents for three - dimensional lattices are @xmath18 and @xmath19 from simulation @xcite and @xmath20 and @xmath21 from series @xcite . in the @xmath22 case ,
only isolated sites are removed by the culling process : these sites do not contribute to the critical behaviour and @xmath4 and all critical exponents remain the same as for usual percolation .
for @xmath23 , on the other hand , isolated clusters with two sites are eliminated , as well as some dangling structures of more compact clusters .
the elimination of these dangling structures , however , does not break the infinite cluster ( whenever present ) and , therefore , the critical probability is the same as for usual percolation .
moreover , as the exponent @xmath24 is also connected to the formation of this infinite cluster , its value is the same for @xmath23 and @xmath7 . in what concerns `` field ''
exponents ( @xmath25 , for instance ) , previous results for two - dimensional systems indicate a higher value of this exponent for @xmath23 than for usual percolation @xcite . nevertheless , it has been shown later , through simulations on bigger lattices in two dimensions @xcite and general arguments applied to both two and three dimensions @xcite , that the exponent @xmath25 is the same for uncorrelated and @xmath23 bootstrap percolation models .
let us now turn our atention to higher values of @xmath3 .
it is generally believed that , for any value of @xmath3 where only infinite clusters can survive the culling process , @xmath26 .
this is indeed the case for @xmath27 on hypercubic lattices ( @xmath28 is the dimension of the lattice ) @xcite , for @xmath29 on cubic @xcite and triangular @xcite lattices and for @xmath30 on the triangular lattice @xcite .
moreover , it has been shown that , for these cases , the usual finite - size scaling relation does not hold .
this finite - size scaling predicts that , if a suitable definition of a finite - lattice `` critical '' point , @xmath31 , is made , this point will approach the critical point in the thermodynamic limit , @xmath4 , as : @xmath32 where @xmath33 is the linear size of the finite system , such that @xmath34 , and @xmath24 is the usual critical exponent .
this finite - size behaviour is indeed observed for @xmath35 , but fails for high values of @xmath3 . for @xmath27 on hypercubic lattices , it has been proven that @xmath36 @xcite , e.g. , proportional to @xmath37 for @xmath38 dimensions . also for @xmath29 on the cubic lattice
, the correct finite - size behaviour is @xmath39 , with @xmath26 @xcite .
these results for high @xmath3 have been conjectured or tested in numerical simulations @xcite .
the results stated in the previous paragraphs do not apply to the @xmath0 case on the cubic lattice . for this model ,
one expects @xmath4 to be above the value for uncorrelated percolation ( where @xmath40 @xcite ) , since the infinite cluster for usual percolation at @xmath4 is not stable with respect to the culling process for @xmath0 . on the other hand , since finite clusters are still stable for this value of @xmath3 on the cubic lattice , it is expected that @xmath41 .
numerical simulations confirmed this scenario @xcite , although the relative small sizes used in those works indicate that the values might not be precise ( we will return to this point later ) . in what concerns the critical exponents ,
previous results indicate that the exponent @xmath24 is the same as for usual percolation @xcite but @xmath25 is higher than its uncorrelated counterpart @xcite .
however , in neither of these works a extrapolation to the thermodynamic limit is attempted , leaving the possibility that finite - size effects might be the reason for the discrepancy in the values of @xmath25 .
this possibility was first proven to be right , in the context of two - dimensional bootstrap percolation models , in references @xcite and later confirmed for @xmath0 on the triangular lattice @xcite .
the possibility of a new universality class for bootsrap percolation with @xmath0 on the cubic lattice is the problem we address in this work .
we resort to numerical simulation methods , which , together with finite - size scaling analysis , allowed us to obtain more precise values for the critical parameters .
we also study the so - called critical spanning probability , i.e. , the probability that a given lattice has a cluster connecting its boundaries at criticality @xcite .
this quantity shows some degree of universality : it depends on the dimension and shape of the system and on the specific boundary condition but not on the lattice type ( simple cubic of f.c.c .
, for instance ) and on the particular kind of percolation ( site or bond ) @xcite .
it is then interesting to see whether it remains invariant for percolation models such that long - range correlation is involved , like bootstrap percolation .
the remainder of the paper is organized as follows . in the next section we present the method and discuss some technical details , as well as the results for the critical parameters . in section [ csp ]
we discuss some previous results concerning the critical spanning probability for percolation and our results for bootstrap percolation .
finally , we summarize our results in the last section .
the method we use is connected to real - space renormalization - group and finite - size scaling procedures @xcite .
the approach needs that precise values of the physical quantities are available ; these are only obtained for high values of the linear system size @xmath33 .
we then studied finite systems of size @xmath42 , with @xmath43 ; from the results for @xmath34 , it is possible to use finite - size scaling techniques to extrapolate to the thermodynamic limit ( @xmath44 ) .
the critical probability @xmath4 and the critical exponent @xmath24 are calculated as follows . for a lattice of size @xmath33
, we occupy each site with probability @xmath1 , apply the bootstrap condition and test the lattice for percolation ( here we define percolation as the presence of a cluster which connects the bottom and top planes of the finite cubic lattice ; we discuss this and other technical points below ) .
our finite - size estimate of the critical probability , @xmath45 , is taken as the value of @xmath1 at which the cell percolates for the first time , when @xmath1 is increased from zero ; this procedure is made for @xmath46 different runs ( which correspond to @xmath46 different seeds to the random number generator ) .
each run leads to a different value of @xmath45 , since the lattice is finite .
we take the average of the @xmath46 values of @xmath45 as our estimate of @xmath31 ( see section [ intro ] ) .
it is then assumed that @xmath31 will approach @xmath4 as given by equation [ fss1 ] . moreover , it is possible to calculate the width @xmath47 , which behaves as ( @xmath48 stands for the average of @xmath49 over the @xmath46 realizations ) : @xmath50 from the previous equation and a @xmath51 plot of @xmath52 , it is then possible to obtain the value of the critical exponent @xmath24 .
tha data is depicted in figure [ sig ] ; from the slope of the straight line we obtain @xmath53 .
we compare this value with other evaluations in table [ crit ] ; within the numerical precision , this exponent is the same for bootstrap percolation with @xmath0 and ordinary percolation on the cubic lattice and agrees with previous evaluations of @xmath24 for bootstrap percolation with @xmath0 on the cubic lattice .
note from the graph that the straight line regime is achieved for @xmath54 , while for uncorrelated percolation this regime sets in for smaller values of @xmath33 .
this is expected , since finite - size effects are stronger for correlated percolation problems than for their uncorrelated counterparts @xcite .
therefore , we neglect the data for @xmath55 and used only lattices with @xmath56 in our linear regression analysis .
= 9.5 cm in order to calculate the critical threshold , we resort to equations [ fss1 ] and [ fss2 ] , which imply that , for @xmath34 : @xmath57 this is a convenient way to calculate @xmath4 , since it does not depend on the value of @xmath24 ; such dependence would appear if equation [ fss1 ] was used . it is shown in figure [ sigma ] a plot of @xmath58 : @xmath4 is given by the linear coeficient and the value obtained is @xmath59 .
this value is slightly above the one calculated in reference @xcite , in which the value of @xmath33 varied from 10 to 110 .
if we use the same range in our calculation the extrapolated value of @xmath4 is consistent with the result of @xcite ( see table [ crit ] ) .
= 9.5 cm let us know discuss some technical points .
an important quantity is the value of @xmath1 at which the lattice percolates for the first time , when @xmath1 is increased from zero .
one needs to define what `` percolate '' means for a finite lattice . in this work
, we used the rule called @xmath60 in reference @xcite , i.e. , a lattice percolates if , after the bootstrap culling process , there is a path of present sites which links the boundaries of the lattice in a fixed direction ( vertical , say ) .
there are other possible definitions ( see @xcite ) and it is expected that all of them lead to the same value of @xmath4 in the thermodynamic limit .
note , however , that the value of the critical spanning probability does depend on this definition , as we will discuss i m section [ csp ] .
the numerical procedure we used to test for percolation is the hoshen - kopelman algorithm @xcite : for usual percolation , it requires the storage of only one plane . for bootstrap percolation ,
on the other hand , the bootstrap iteration needs the storage of the whole lattice , due to correlation effects . to cope with this drawback , we store the lattice in bits , instead of words : this saves memory and time , since the updates connected to the bootstrap rule can be made in parallel for a set of 32 sites @xcite . to define all six first - neighbours of sites at the boundaries of the lattice ,
periodic boundary conditions are used .
it is expected that the boundary condition does not affect the values of the critical parameters in the thermodynamic limit , since it is a `` surface '' effect .
finally , let us mention that the number of realizations @xmath46 ( see section [ intro ] ) varied from 12000 for @xmath9 to 640 for @xmath10 ; the errors were calculated as three times the standard deviation for subsets of the total number of realizations .
.critical parameters for @xmath0 bootstrap percolation and comparison with previous results for this model and for usual percolation . [ cols="<,<,<",options="header " , ] + @xmath61 present work .
@xmath62 reference @xcite .
@xmath63 reference @xcite .
+ @xmath64 reference @xcite .
@xmath65 reference @xcite . @xmath66 reference @xcite [ crit ] to have access to the `` magnetic '' scaling power @xmath67
, one possible way is to define a `` ghost''- site , which is linked to all sites of the lattice with probability @xmath68 @xcite . within a real - space renormalization group framework ,
it is possible to calculate the eigenvalue @xmath69 through , where @xmath70 is the average number of occupied sites linked to one of the boundary planes of the finite cubic lattice at the critical threshold , @xmath71 .
the value of @xmath4 is obtained as explained above and is then used to calculate @xmath70 and hence @xmath72 , averaging over @xmath73 configurations . the value of @xmath73 varied from @xmath74 , for the smaller lattices , to @xmath75 , for the bigger ones .
we use two procedures to obtain @xmath6 in the thermodynamic limit .
the first one is based on the fact that , for @xmath76 : @xmath77 with @xmath6 independent of @xmath33 ; this equation leads to a straight line in a @xmath51 plot of @xmath78 , for @xmath34 . as depicted in figure [ yh ] , this is indeed the case for @xmath56 . from a linear fitting of the data for this range of @xmath33 , and resorting to the scaling relation @xmath79 , we obtained the value @xmath80 ( see table [ crit ] ) .
alternatively , we could take into account correction - to - scaling terms , through : @xmath81 from this equation , we see that local slopes of @xmath82 provide estimates of @xmath83 , which , when extrapolated to @xmath84 , leads to an evaluation of @xmath6 in the thermodynamic limit .
we have applied this procedure for @xmath43 , using three consecutive points to calulate the local slopes and extrapolating to @xmath76 through a @xmath85 graph .
the value thus obtained for @xmath25 is the same as for the first procedure . in table
[ crit ] we see that our estimate of @xmath25 disagrees with the values calculated in references @xcite and @xcite . in evaluating this exponent , the former uses lattices of linear size @xmath86 , while the latter uses @xmath87 ; neither attempted to make an extrapolation to the thermodynamic limit . from table
[ crit ] , we can infer that our estimate of @xmath25 for the bootstrap model we study is , within the error bars , equal to the corresponding value of this exponent for usual ( uncorrelated ) percolation .
since we have already seen that @xmath24 is also the same for both models , we can draw the conclusion that usual percolation and bootstrap percolation with @xmath0 on the cubic lattice belong to the same universality class .
this result contradicts references @xcite and @xcite ; we believe this is caused by the small lattices used in those works .
it has been established some time ago that the critical spanning probability , @xmath8 , defined as the probability of spanning a lattice at the critical point , shows some degree of universality @xcite .
more precisely , this quantity does not depend on the lattice type and on the kind of percolation .
this result contradicts the assumption made on early applications of real - space renormalization group procedures to percolation . in those
, it was assumed that the critical spanning probability is equal to the critical percolation threshold , @xmath4 and , therefore , lattice dependent @xcite .
however , later numerical tests confirmed and expanded the universality proposal @xcite .
nevertheles , no study has been made on correlated models , to the best of our knowledge . while it is expected that _ short - range _ correlations do not change the universality scenario @xcite , it is not clear whether @xmath8 changes when _ long - range _ correlations are introduced .
a convenient model to test these possibilities is the bootstrap percolation one . while the behaviour of @xmath8 is trivial for the cases where @xmath26 , for @xmath0 on the simple cubic lattice the presence of correlation may lead to a non trivial behavior .
we study this possibility using numerical simulation on simple cubic lattices of size @xmath42 , with @xmath43 .
the programs and algorithms used are essentially the same as the ones described in the previous section .
the basic procedure is to generate @xmath73 independent runs for each lattice size and compute the fraction of those which percolate after the bootstrap condition is used and a stable configuration is reached .
the values of @xmath73 are the same as those used in the calculation of @xmath6 ( see previous section ) .
the results are depicted in figure [ rpc ] : we can infer that @xmath11 , where the error bar is a rough estimate .
there are two previous calculation of @xmath8 for the uncorrelated site percolation on the cubic lattice with free boundary conditions : they lead to @xmath88 @xcite and @xmath89 @xcite .
our value agrees , within the numerical error , with the first one but the result of reference @xcite can not be ruled out .
it is then reasonable to infer that usual percolation and @xmath0 bootstrap percolation on the cubic lattice belong to the same universality class , also in what regards the critical spanning probability .
we calculate , using numerical simulations and finite - size scaling techniques , the critical parameters of the bootstrap percolation model with @xmath0 on the cubic lattice , using finite lattices of size @xmath42 , with @xmath43 .
our evaluations of @xmath24 and @xmath25 strongly support the conclusion that usual percolation and the model we study belong to the same universality class .
this result disagrees with previous calculations @xcite ; we believe this is due to the small sizes used in those works . to support this assumption ,
we note that the finite - size scaling assumptions hold for lattices of linear size @xmath54 , which are higher then the sizes studied in previous works . the critical spanning probability , @xmath8 , is also calculated .
it has been shown that this quantity shows some degree of universality , but , to the best of our knowledge , no study concerning correlated percolation models have been done so far .
our result for bootstrap percolation with @xmath0 on the cubic lattice , @xmath11 , is , within the numerical accuracy , the same value as for usual percolation with free boundary conditions .
therefore , we can infer that @xmath8 is not sensitive to short range correlations and even to some long range correlations , like the one studied in this paper .
stauffer d and aharony a 1992 _ introduction to percolation theory _ ,
2nd edition ( london : taylor and francis ) stinchcombe r 1983 _ phase transitions and critical phenomena _ , vol 7 ed c domb and m s green ( london : academic ) p 151 adler j 1991 _ physica a _ * 171 * 453 and references therein .
de aguiarj a o , engelsberg m and guggenheim h j 1986 _ j. magn .
mater . _ * 54 - 57 * 107 + de aguiar j a o , brady - moreira f g and engelsberg m 1986 _ phys .
b _ * 33 * 652 + branco n s and machado k d 1993 _ phys . rev . b _ * 47 * 493 chalupa j , leath p l and reich g r 1981 _ j. phys .
c : solid state phys .
_ * 12 * l31 branco n s , de queiroz s l a and dos santos r r 1986 _ j. phys .
c : solid state phys .
* 19 * 1909 ziff r m and stell g 1988 _ university of michigan report _ number lasc 88 - 4 adler j , meir y , aharony a and harris a b 1990 _ phys . rev .
b _ * 41 * 9183 branco n s , dos santos r r and de queiroz s l a 1984 _ j. phys .
c : condens . matter _
* 17 * l37 chaves c m and koiller b 1995 _ physica a _ * 218 * 271 + medeiros m c and chaves c m 1996 _ physica a _ * 234 * 604 branco n s 1993 _ journal of statistical physics _ * 70 * 1035 schonmann r h 1990 _ journal of statistical physics _ * 58 *
1239 kogut p m and leath p l 1981 _ j. phys .
c : condens
. matter _ * 14 * 3187 aizenman m and
lebowitz j l 1988 _ j. phys . a : math .
* 21 * 3801 cerf r and cirillo e n m 1998 _ preprint cond - mat/9812077 _ amd referebces therein van enter a c d , adler j and duarte j a m s 1991 _ journal of statistical physics _ * 60 * 323 + van enter a c d , adler j and duarte j a m s 1990 _ journal of statistical physics _ * 62 * 505 acharyya m and stauffer d 1998 _ int .
c _ * 9 * 643 adler j and stauffer d 1990 _ j. phys . a : math .
* 23 * l1119 stauffer d and de arcangelis l 1996 _ int .
j. of mod .
c _ * 7 * 739 langlands r p , pichet c , pouliot ph and saint - aubin y 1992 _ journal of statistical physics _ * 67 * 553 lorenz c d and ziff r m 1998 _ j. phys . a : math .
gen . _ * 31 * 1 lin c - y , hu c - k and chen j - a 1998 _ j. phys . a : math .
gen . _ * 31 * l111 reynolds p j , stanley h e and klein w 1980 _ phys .
b _ * 21 * 1223 branco n s , de queiroz s l a and dos santos r r 1988 _ j. phys .
c : condens .
matter _ * 21 * 2463 j. hoshen and r. kopelman 1976 _ phys .
b _ * 14 * 3428 stauffer d 1991 _ j. phys . a : math .
* 24 * 909 cardy j l 1992 _ j. phys .
a : math gen . _ * 25 * l210 + ziff r m 1992 _ phys
lett . _ * 69 * 2670 | we study the @xmath0 bootstrap percolation model on a cubic lattice , using monte carlo simulation and finite - size scaling techniques . in bootstrap percolation ,
sites on a lattice are considered occupied ( present ) or vacant ( absent ) with probability @xmath1 or @xmath2 , respectively .
occupied sites with less than @xmath3 occupied first - neighbours are then rendered unoccupied ; this culling process is repeated until a stable configuration is reached .
we evaluate the percolation critical probability , @xmath4 , and both scaling powers , @xmath5 and @xmath6 , and , contrarily to previous calculations , our results indicate that the model belongs to the same universality class as usual percolation ( i.e. , @xmath7 )
. the critical spanning probability , @xmath8 , is also numerically studied , for systems with linear sizes ranging from @xmath9 up to @xmath10 : the value we found , @xmath11 , is the same as for usual percolation with free boundary conditions .
0.7 cm key words : correlated percolation ; phase transition ; universality classes .
0.7 cm pacs numbers : 64.60.ak , 64.60.fr + 0.7 cm |
a real lie algebra can be easily complexified by tensoring it with the complex numbers over the reals , and extending the lie bracket bilinearly . complexifying a lie group
is a much more subtle problem , for which a solution does not always exist .
all compact lie groups admit a complexification , but the proof of that is not trivial ( see for example @xcite , 106 ) .
let @xmath2 be a khler manifold , and let @xmath1 denote the group of hamiltonian symplectomorphisms of @xmath3 , with lie algebra @xmath4 .
@xmath1 is known to be morally " an infinite - dimensional analogue of a compact group , and one can wonder whether it has a complexification .
( one can rigorously prove that @xmath1 is a diffeological lie group , @xcite . ) from a physical point of view this would correspond to finding a sensible way to associate a dynamical system to a complex - valued hamiltonian , @xmath5 , in a manner that extends the notion of hamilton flow in case @xmath6 is real - valued .
this issue was raised and taken on by donaldson in a series of papers beginning with @xcite , @xcite , in connection with a set of lovely problems in khler geometry , and has generated a lot of research . from this point of view ,
the interest in finding a complexification of _ ham _ is based on the fact , discovered independently by atiyah and guillemin and sternberg , that if a compact lie group acts in a hamiltonian fashion on a khler manifold then the the action extends to the complexified group in an interesting way that can be understood .
donaldson pointed out that _ ham _ acts on the infinite - dimensional space of khler potentials of @xmath7 , and the moment map is a hermitian scalar curvature .
the appeal of extending the finite - dimensional picture to this infinite - dimensional setting of khler metrics is that it was hoped to lead to a way of constructing extremal metrics . in this note
we will not enter into the connections with khler geometry , other than to point out here that exponentiating purely - imaginary hamiltonians leads to geodesics in the space of khler potentials .
we will explore this in an expanded version of this note .
our focus is on the following issue : donaldson has put forward a notion of formal lie group " to conceptualize his notion of a formal complexification of _
ham_. briefly , a formal lie group with lie algebra @xmath8 is a manifold " ( the notion is of interest only in infinite dimensions ) , @xmath9 , together with a trivialization of its tangent bundle of the form @xmath10 , such that the map @xmath11corresponding vector field @xmath12 on @xmath9 is a lie algebra homomorphism ( with respect to the commutator of vector fields ) .
the vector fields @xmath12 should be thought of as left- invariant " , though no group structure on @xmath9 exists .
( again this definition can be made completely rigorous in the language of diffeologies , @xcite . ) in the present case @xmath13 and the exponential map in our title refers to the problem of constructing the flow of the fields @xmath12 .
corollary [ cormain ] states that the family of diffeomorhpisms @xmath14 there exponentiate the complex - valued hamiltonian @xmath6 , in one of donaldson s models for the complexification of @xmath9 ( see @xcite 4 ) .
our construction is based on the following existence result : let @xmath15 be a real analytic khler manifold of real dimension @xmath16 .
there exists a holomorphic complex symplectic manifold @xmath17 of complex dimension @xmath16 and an inclusion @xmath18 such that @xmath19 , and with the following additional structure : 1 .
an anti - holomorphic involution @xmath20 whose fixed point set is the image of @xmath21 and such that @xmath22 .
2 . a holomorphic projection @xmath23 , @xmath24 , whose fibers are holomorphic lagrangian submanifolds .
the local existence is simple : we take @xmath25 to be a neighborhood of the diagonal in @xmath26 , with the complex structure @xmath27 .
@xmath28 is the holomorphic extension of @xmath29 and @xmath30 .
finally , the projection is simply @xmath31 .
however there exist natural complexifications that make our results below much more global in some cases .
for example , if @xmath32 is a generic coadjoint orbit of a compact simply connected lie group @xmath33 ( one through the interior of a weyl chamber ) , then one can take for @xmath25 the orbit of the complexification , @xmath34 , of @xmath33 through the same element .
if we let @xmath35 be the iwasawa decomposition of @xmath34 , then the orbits of the action of @xmath36 on @xmath25 are the fibers of a holomorphic projection @xmath37 .
( this picture extends to non - generic orbits . ) to describe our results we need some notation . given a function @xmath5 whose real and imaginary parts are real analytic
, there is a holomorphic extension @xmath38 perhaps only defined near @xmath39 , but we will not make a notational distinction between @xmath25 and such a neighborhood , as our results are local in time .
let @xmath6 and @xmath40 be as above .
the fibers of @xmath41 are the leaves of a holomorphic foliation , @xmath42 , of @xmath25 .
denote by @xmath43 the hamilton flow of @xmath44 ( where @xmath44 is the real part of @xmath40 ) with respect to the real part of @xmath28 .
we denote by @xmath45 the image of the foliation @xmath42 under @xmath46 ( so that @xmath47 ) we assume that there exists @xmath48 an open interval containing the origin such that @xmath49 the leaves of @xmath45 are the fibers of a projection @xmath50 .
we will denote @xmath51 the fiber of @xmath45 over @xmath52 .
[ main ] let @xmath48 be an open set as above .
let @xmath53 be defined by @xmath54 then , @xmath49 : 1 .
there is a complex structure @xmath55 such that @xmath56 ( and @xmath57 ) .
@xmath58 is holomorphic ( @xmath59 ) .
the infinitesimal generator of @xmath60 is @xmath61 where @xmath62 , denotes the hamilton vector field of @xmath63 with respect to @xmath29 , etc .
[ ck ] conditions ( 2 ) and ( 3 ) ( together with the initial condition @xmath64 ) characterize the family @xmath65 .
it is in fact not hard to see that in local real - analytic coordinates @xmath66 on @xmath32 , these conditions show that the family solves a system of first - order non - linear pdes of the form @xmath67 with @xmath68 and where the @xmath69 are real - analytic .
therefore local existence and uniqueness follow from the cauchy - kowalewsky theorem .
suppose @xmath34 is a compact lie group acting on @xmath32 in a hamiltonian fashion and preserving @xmath70 .
then the action extends to a holomorphic action to the complexification @xmath71 ( c.f .
@xcite , 4 ) .
the extended action is as follows : if @xmath72 are two components of the moment map of the @xmath34 action , then the infinitesimal action corresponding to @xmath73 is the vector field @xmath74 .
the corresponding one - parameter group of diffeomorphisms , @xmath75 , satisfies ( 2 ) and ( 3 ) of theorem [ main ] , with @xmath76 for all @xmath77 . by the uniqueness part of the previous remark we must have @xmath78 .
in other words , our construction is an extension of the process of complexifying the action of a compact group of symmetries of @xmath0 .
we now explain the geometry behind the construction of @xmath60 summarized by ( [ lapapa ] ) . to find the image of @xmath79 under @xmath60 one flows the leaf @xmath80 of the foliation @xmath81 by @xmath46 , and intersects the image leaf with @xmath32 . in other words , ( [ lapapa ] )
can be stated equivalently as : @xmath82 the definition of @xmath83 is summarized in the following figure , where @xmath84 : in this figure @xmath32 is represented by the horizontal segment . under @xmath46
the fibers of the foliation @xmath85 are transformed into the fibers of @xmath45 , and @xmath86 .
as we will see the following is an easy consequence of the previous result : [ cormain ] let @xmath87 and @xmath88 the symplectic form defined by @xmath89 .
then @xmath90 where @xmath91 , denotes the hamilton vector field of @xmath63 with respect to @xmath88 , etc .
it is not hard to check that @xmath92 , therefore in the ( relatively rare ) cases when @xmath60 is a one - parameter subgroup of diffeomorphisms one has @xmath93 .
the present construction is motivated by semiclassical analysis . ignoring ( possibly catastrophic ) domain issues , the notion of the exponential of a non - hermitian _ quantum _ hamiltonian is clear .
( for example , if @xmath32 is compact and planck s constant is fixed , this amounts to exponentiating a matrix . )
therefore a very natural approach to exponentiating a non - hermitian classical hamiltonian is to first quantize it , exponentiate it on the quantum side , and then take the semiclassical limit .
this approach has been developed by rubinstein and zelditch , @xcite , and raises a number of interesting but difficult analytical questions . in our construction
we bypass these analytic difficulties by considering the following geometric remnants of quantization : each leaf of the foliation @xmath85 corresponds to a quantum state ( element of the projectivization of the quantum hilbert space ) represented by a _ coherent state _ centered at a point on the leaf , that is , an element in the hilbert space that semiclassically concentrates at the intersection of the leaf with @xmath32 .
the fact that @xmath41 is holomorphic says the the coherent states are associated to the metric of @xmath94 . on the quantum side
the evolution of a coherent state remains a coherent state , whose lagrangian is simply the image of the one at time @xmath95 by the complexified classical flow . as explained above our maps @xmath65
simply follow the evolution of the real center .
finally , we mention that the paper @xcite by graefe and schubert contains a very clear and detailed account of the case when @xmath32 is equal to @xmath96 , @xmath6 is a quadratic complex hamiltonian , and the lagrangian foliations are by complex - linear positive subspaces , corresponding to standard gaussian coherent states with possibly complex centers . by explicit calculations on both quantum and classical sides , they show that the the evolution of a coherent state centered at @xmath52 is another coherent state whose center may be complex , but that represents the same quantum state as a suitable gaussian coherent state centered at @xmath97 .
let @xmath98 denote the infinitesimal generator of @xmath46 .
the following is easy to check in a local trivialization of the foliation of @xmath25 by fibers of @xmath99 : introduce coordinates @xmath105 in a neighborhood @xmath106 centered at @xmath107 so that @xmath108 is defined by @xmath109 and the projection @xmath99 is just @xmath110
. note that , since @xmath102 is a one - parameter local subgroup of diffeomorphisms , for @xmath111 small @xmath112 for @xmath111 near zero denote the map @xmath113 in coordinates by @xmath114 ( in a smaller neighborhood of @xmath107 ) . for each @xmath111
the image of the @xmath115-axis under @xmath113 locally parametrize the fiber @xmath116 , namely @xmath117 therefore we can write @xmath118 , where @xmath119 is defined implicitly by @xmath120 and @xmath121 .
it follows that @xmath122 however @xmath123 is the identity , so that @xmath124 and therefore @xmath125 . to proceed
further we will need some notation .
we regard @xmath25 as a real manifold of real dimension @xmath126 with an integrable complex structure @xmath127 .
let us write @xmath128 for the real and imaginary parts of @xmath28 .
thus the @xmath129 are real symplectic forms on @xmath25 and @xmath32 is @xmath130-symplectic and @xmath131-lagrangian .
let us write @xmath132 for the real and imaginary parts of @xmath40 .
recall that , by definition , @xmath98 is the hamilton field of @xmath133 with respect to @xmath130 .
if @xmath6 is real then @xmath142 ( by uniqueness of analytic continuation of @xmath6 ) , so @xmath44 is @xmath143-invariant . since @xmath22 , @xmath130 is also @xmath143-invariant , and therefore @xmath143 maps @xmath98 to itself and so @xmath98 has to be tangent to the fixed - point set of @xmath143 .
for the second part just note that @xmath144 .
\(1 ) since the fibers of @xmath99 are the leaves of a holomorphic foliation , there is a well - defined complex structure in the abstract normal bundle to the fibers .
the inclusion @xmath18 realizes @xmath32 as a cross - section to the foliation and identifies @xmath145 with the normal bundle to the foliation along @xmath32 .
therefore it inherits a complex structure that makes @xmath99 holomorphic .
\(2 ) this follows from the interpretation of the structures @xmath146 as arising from the normal bundle structures together with the fact that @xmath46 is holomorphic , or can be checked directly as follows .
let @xmath147 , then @xmath148 , where @xmath149 . since @xmath150 is holomorphic
, one has @xmath151 but @xmath152 since @xmath46 maps fibers of @xmath85 to fibers of @xmath45 .
therefore , the previous relation implies that @xmath153 which precisely says that @xmath60 is holomorphic . by the previous lemma ,
if @xmath6 is real , @xmath158 and there is nothing more to prove .
suppose now that @xmath6 is purely imaginary .
by the second relation in ( [ hamfi ] ) , @xmath159 is the hamilton field of @xmath160 ( see [ re+iim ] ) , and by the lemma [ realcase ] @xmath159 is tangent to @xmath25 and its restriction to @xmath25 is the hamilton field of @xmath161 with respect to @xmath29 .
therefore in this case we can write @xmath162 and it suffices to apply @xmath70 to both sides to get the result .
the general case follows by @xmath163-linearity of the composition @xmath164 , where the first arrow is analytic continuation .
this concludes the proof of theorem [ main ] .
_ proof of corollary [ cormain]_. recalling that @xmath92 , so we wish to compute @xmath165 .
differentiating with respect to time the identity @xmath166 , we get @xmath167_{\phi_{-t}(x)}\ ] ] now it is not hard to check that @xmath168 = \xi^{\omega_t}_{\re h\circ f_t^{-1}}$ ] and that @xmath169 ( using that @xmath170 is holomorphic ) .
therefore @xmath171_{f^{-1}_{t}(x)}.\ ] ] s. donaldson , remarks on gauge theory , complex geometry and 4-manifold topology , _ fields medallists lectures _ , 384403 , world sci .
20th century math .
, 5 , world sci .
publishing , river edge , nj , 1997 . | let @xmath0 be a khler manifold and @xmath1 its group of hamiltonian symplectomorphisms .
the complexification of @xmath1 introduced by donadson is not a group , only a
formal lie group " .
however it still makes sense to talk about the exponential map in the complexification . in this note
we show how to construct geometrically the exponential map ( for small time ) , in case the initial data are real - analytic .
the construction is motivated by , but does not use , semiclassical analysis . |
gl 618 is a young , bipolar planetary nebula ( pn ) .
ground - based optical and near - ir imaging of this object reveal two lobes of emission ( each about 3@xmath0 in extent ) separated by a dark lane .
the central regions of the nebula are hidden from direct view at optical wavelengths by the lane of obscuring material .
the spectrum of gl 618 is composed of a faint continuum and a variety of low - excitation emission lines .
trammell , dinerstein , & goodrich ( 1993 ) used spectropolarimetry to study gl 618 and found that the continuum and part of the permitted line emission are reflected from deep in the nebula .
the low - excitation , forbidden line flux and remainder of the permitted line emission are produced in the bipolar lobes . the emission produced in the bipolar lobes is indicative of shock heating ( @xmath1 = 50@xmath2100 kms@xmath3 ) .
long - slit optical spectroscopy of gl 618 confirms that the shock emission is associated with out - flowing gas ( carsenty & solf 1982 ) and near - ir spectroscopy of gl 618 has revealed the presence of thermally excited h@xmath4 emission ( thronson 1981 ; latter et al .
gl 618 exhibits [ fe ii ] emission also thought to be associated with the shock - heated gas ( kelly , latter , & rieke 1992 ) and more recent observations establish that this emission is associated with an outflow ( kelly , hora , & latter 1999 ) .
shock - excited emission dominates the spectra of the lobes of gl 618 .
we present wfpc2 images of this object that demonstrate that the source of this shock - excited emission is a set of highly collimated outflows originating in the central regions of the object .
we have obtained wfpc2 images of gl 618 as part of a hst cycle 6 program .
gl 618 was centered in the planetary camera which has a 36@xmath0 @xmath5 36@xmath0 field of view and a plate scale of 0.0455per pixel@xmath3 .
images were obtained through four filters : f631n ( isolating [ o i]@xmath66300 line emission ) , f656n ( isolating h@xmath7 line emission ) , f673n ( isolating [ s ii ] @xmath6@xmath66717,31 line emission ) , and f547 m ( a continuum band ) .
these filters were chosen so that we could study the morphology of the shock - excited emission in the lobes of gl 618 .
the images were processed through the hst data reduction pipeline procedures and cosmic rays were removed by combining several exposures of each object .
the images of gl 618 were obtained on 23 october 1998 and exposure times ranged from 15 to 45 minutes .
the overall morphologies seen in the [ s ii ] and [ o i ] images are similar ( figure 1 , panels ( a ) and ( b ) ) .
these images trace the morphology of the shock - excited forbidden line emission in gl 618 .
three highly collimated outflows , or jets , are seen in both images .
the brightest emission occurs near the tip of each of the outflows and there is no forbidden line emission seen in the central regions of gl 618 .
ripple - like morphology is seen in the outflows in both the [ s ii ] and [ o i ] images .
these ripples might be the result of instabilities in the flow and/or an interaction with the surrounding nebular material .
the morphology observed in the h@xmath7 image ( figure 1 , panel ( c ) ) differs slightly from the forbidden line morphology .
h@xmath7 emission is seen associated with the outflows , but in addition , a significant amount of h@xmath7 emission is seen towards the central regions of gl 618 .
spectropolarimetric observations indicate that part of the h@xmath7 emission is reflected and part of this emission is produced by shocks in the lobes ( trammell et al .
we have spatially separated these components in the hst images . the h@xmath7 emission associated with the central regions of the object
is probably reflected emission from an h ii region buried deep in the nebula .
a high density h ii region has been observed at the center of gl 618 at radio wavelengths ( e.g. kwok & feldman 1981 ) and in the reflected optical spectrum ( trammell et al . 1993 ) .
the h@xmath7 emission coincident with the outflows is the shock - excited component of the permitted line emission .
the [ o i ] to [ s ii ] line ratios in the bullet - like structures at the tips of the outflows are approximately [email protected] . by comparing these
observed line ratios with the predictions of planar shock models ( hartigan , raymond , & hartmann 1987 ) , we estimate the shock velocity in these regions to be approximately 80 kms@xmath3 .
this is consistent with the range in shock velocities estimated from previous spectropolarimetric observations ( trammell et al .
1993 ) .
careful examination of the bullet - like structures at the tip of the outflow in the upper lobe in figure 1 reveals an excitation gradient across this region .
h@xmath7 is brightest on the side of the spot facing away from the central regions of gl 618 .
[ s ii ] and [ o i ] are brighter on the side closest to the central source .
this type of gradient is expected for a jet flowing away from the central source and impinging on the surrounding nebular material .
the bright spots near the tops of the outflows are not clumps of material being overrun by a wind or outflow .
hst observations ( e.g. trammell & goodrich 1996 ; sahai & trauger 1998 ) and ground - based imaging surveys ( e.g. balick 1987 ; schwarz , corradi , & melnick 1992 ) have revealed the presence of collimated outflows , fliers , and a myriad of other small - scale structures in pn .
the origins of these structures and their role in the overall development of pn remain puzzling .
the debate concerning the origin of these small - scale structures , and also the formation of aspherical pn in general , centers on whether binary or single stars are responsible for producing aspherical mass loss .
both models of binary star interaction ( e.g. soker & livio 1994 ) and magnetic confinement ( e.g. garcia - sergura 1997 ) , while providing a scheme for producing the overall aspherical structure in pn , may also provide mechanisms to produce the highly collimated outflows . the complex , mulitpolar outflow geometry seen in gl 618 may be difficult for either of these types of models to explain .
our observations demonstrate that jets can be present during the early phases of pn development and may play an important role in the early shaping of these objects .
futher , these collimated outflows may set the stage for the development of other small scale structures seen in more evolved objects . | we present narrow - band hubble space telescope images of the young planetary nebula gl 618 .
these images have allowed us to study the detailed morphology of shock - excited emission present in the bipolar lobes of this object .
these images reveal the presence of three highly collimated outflows emanating from the central regions of gl 618 .
we discuss the significance of the detection of these outflows and the possible origins of these features .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
plasmonic nano - antennas ( na ) have emerged as a powerful tool for sensing and monitoring variations of their local environment such as the binding of molecules at the nano - antenna surface@xcite , hydrogen gas absorption@xcite , chemical reactions @xcite , phase - changing processes@xcite . such sensing properties are due to the frequency shift capability of localized surface plasmon resonances under tiny changes in the effective dielectric constant of the surrounding medium . in all the above - cited examples , nano - antennas have monitored chemical and physical processes in contact to the nano - antennas , by taking advantage of the highly localized plasmon fields that are resonantly generated within of the nanostructures . the sensing ability of nano - antennas to changes happening at a distance from the nanostructures has been hardly discussed .
a hybrid structure combining a plasmonic nano - antenna and a dielectric fabry perot cavity have been demonstrated to remotely sense with nanometer spatial accuracy important film parameters in thin - film nanometrology @xcite but , to our knowledge , no investigation of remote sensing with single plasmonic nano - antennas have been reported so far .
moreover , applications of na - based refractive index sensing are restricted to the extraction of weak nanoscale plasmonic information from a strong background signal .
direct background - free optical detection of plasmonic nanosensors is highly desirable because it would avoid the use of complex detection mechanisms such as absorption / extinction or dark - field scattering spectroscopies @xcite ..
this letter addresses the concept of fiber - integrated bowtie aperture nano - antenna ( bna)@xcite as a metallic nanoresonator for remote sensing applications .
this aperture - type nano - antenna @xcite opened at the end of an opaque metal coated snom fiber tip offers the unique ability of a direct in - fiber and background - free detection of the far - field optical signal emitted by the bna under resonance .
this moveable nano - antenna is demonstrated in the remote ( i.e. , non - contact ) sensing of a silicon surface with direct in - fiber detection , bringing innovative solution to nanopositioning .
given the extremely high sensitivity of bnas to changes of their environment , the tip - to - surface distance can be monitored and controlled over a range at least one order of magnitude larger than that of mechanical force detection devices used in shear - force and atomic force microscopy ( afm ) , with sub - nanometer positionning accuracy and subwavelength lateral sensing area .
in order to produce the bna - on - tip , polymer tips are first grown by photopolymerization at the cleaved end facet of a monomode ( @xmath1 wavelength ) glass fiber @xcite .
next , the probes are metal - coated with @xmath2 thick aluminum layer .
finally , the bna is fabricated at the apex of the tip by fib milling ( dual beam fei helios 600i with a raith elphy multibeam attachment ) .
figure [ fig : bna](a ) displays scanning electron micrographs of a resulting fiber device .
= 1.55 @xmath3 m ) , for two perpendicular polarization directions of the incoming wave ( see arrows ) .
this enhancement is the ratio between the field intensity diffracted by the bna placed at 2 nm beyond a si substrate and the intensity calculated without the presence of the tip.[fig : bna ] ] two bnas of different sizes are considered in this study .
the first one ( bna1 ) is @xmath4 wide with a square gap about @xmath5 large and @xmath6 flare angles ( see fig.[fig : bna](a ) ) . with this geometry ,
the bna s optical resonance is centered to a wavelength of @xmath7 when it is located at a distance of @xmath8 away from a silicon substrate @xmath9 .
the second one ( bna2 ) has a smaller gap of about @xmath10 and is @xmath11 large so that its resonance is redshifted with respected to the first bna , centered to @xmath12 at @xmath8 far from the si surface .
the design process is performed using three - dimensional finite difference time domain method ( 3d fdtd ) @xcite .
details on the polarization - dependent resonance process of the bna can be found in refs .
we just note that the resonant mode of the bna is strongly bound to its nanometer - scale feed gap which strongly enhances the optical electric field by optical capacitive effect ( see fig . [
fig : bna](c ) for the first bna ) .
the extent of the resulting light confinement beyond the tip apex does not exceed @xmath5 ( @xmath13 calculation of the longitudinal intensity profile along the tip axis ) .
this tiny optical `` hot spot '' vanishes when the input polarization is turned by @xmath14 , due to resonance cancelation ( cf .
[ fig : bna](d ) for the first bna ) .
figure [ fig : bna](b ) displays the polarization diagram of the first bna on tip used in collection mode ( red curve ) . a weakly focused linearly polarized beam
is directly projected onto the nano - antenna and the signal collected by the tip ( which is due exclusively to the bna resonance ) is measured at the fiber end as a function of the incident polarization direction .
we see that this diagram shows a two - lobes structure which fits the polarization diagram of an ideal oriented dipole detector described by a cosine function ( blue curve ) .
this proves that the bna can be described optically by an electric dipole moment .
the first bna shows a polarization ratio of about 1:20 .
the polarization ratio of the second bna is measured to a value of 1:8 .
these two nano - antennas are approached to a si surface and studied optically with the experimental set - up shown in fig . [
fig : scheme](a ) .
the bnas on fiber tip are mounted onto a tuning fork and placed into a snom head from nt - mdt company ( ntegra system ) .
a shear - force detection based feedback loop is used to position the tip in close proximity of the surface of a si wafer .
the si sample is mounted onto a calibrated piezo stage ( pi company ) to be accurately moved away from the tip along the longitudinal direction ( perpendicular to the sample surface ) .
the bna on tip is used in collection mode .
laser light at @xmath1 ( @xmath15 ) is slightly focused with a lens onto the tip apex through the si wafer . the fiber output
is coupled to an ingaas photodiode ( thorlabs ) and a synchronous detection scheme is used to enhance signal - to - noise ratio .
the output voltage is directly monitored with a numerical oscilloscope connected to a computer .
the bna is resonantly excited by aligning the polarization direction of the incident optical waves along the bna s polarization axis , ie .
the symmetry axis of the bna that passes through each metal triangle s tip .
figure [ fig : scheme ] ( b ) and ( c ) show the optical signal collected by the two bna - on - tips considered in this study , as a function of the spacing ( @xmath16 ) between the nanostructures and the surface . to this end , the tip is set in close proximity of the sample with shear force distance control facility . then , the sample is moved away from the surface by steps of a few nanometers with the piezo stage and at each step the signal is recorded .
the minimum spacing , for which the feedback loop is triggered , is known to be about @xmath17@xmath18 @xmath8 . to preserve the tip
, we did not investigate this zone , this is why the last @xmath17 before contact appear as a shaded area in the figures .
the experimental approach curves ( circles ) are compared with a numerical analysis of the problem ( solid line ) . in the simulation , the bna
is excited with an incoming plane wave in the si substrate at normal incidence that is polarized along the bna s polarization axis .
data are obtained with poynting vector flow calculation through the output cross - section of the 2-micron - long tip portion considered in the simulation . with ( b ) the first and ( c ) the second bna - on - tip : experimental results ( circles ) and numerical predictions ( solid line ) .
the bna - on - tips are used in collection mode .
inset of ( b ) : far - field images of the first bna used in emission mode ( light is injected into the fiber and the bna is imaged with an objective coupled to an infrared camera ) .
inset of ( c ) : experimental approach curve with a non - resonant excitation of the second bna ( the incident polarization is perpendicular to the bna s polarization axis).,scaledwidth=99.0% ] we see that the approach curves obtained with the two bna configurations show important discrepancies . with the first bna ( fig . [
fig : scheme](b ) ) , the collected signal increases continuously when the spacing @xmath16 decreases , whereas for the second bna ( fig .
[ fig : scheme](c ) ) , the collected signal is increased for decreasing spacings down to 45 nm and is then strongly decreased for smaller distances .
the origin of this discrepancy is evidenced in the spectral analysis of the two bnas shown in fig .
[ fig : spectre ] .
figures [ fig : spectre](a ) and ( b ) show the theoretical resonance spectra of the bnas at proximity of a si substrate , for various values of bna - to - surface spacings @xmath16 smaller than @xmath0 . the bnas are excited with a pulsed incoming plane wave in the si substrate at normal incidence .
the spectra are obtained by fourier transforming the electric field recorded during the simulation at a single grid cell , located at the center of the bnas .
they are normalized by the spectra calculated at the same point without the tip to provide information about intensity enhancement .
the experimental wavelength @xmath19=1550 nm is emphasized with dashed lines into the figures .
we see in figs .
[ fig : spectre](a ) and ( b ) that the presence of a silicon surface close to the bnas leads to a dramatic redshift of the nano - antenna resonance that exceeds @xmath20 . for a spacing varying from @xmath0 down to @xmath8 , the resonance wavelength shifts from @xmath21 to @xmath22 with the first bna ( fig .
[ fig : spectre](a ) ) and from @xmath23 to @xmath24 with the second bna ( fig .
[ fig : spectre](b ) )
. therefore , with the first bna , the resonance peak stays centered to wavelengths smaller than @xmath1 over the experimental values of @xmath16 comprised between @xmath0 and @xmath17 .
this results , at a wavelength of @xmath1 , in an intensity enhancement at the bna s gap that continuously increases while the tip approaches the si surface .
such a phenomenon is clearly shown in fig .
[ fig : spectre](c ) which displays the value of the spectra of the two bnas at @xmath25 , for various value of @xmath16
. given the electric dipolar properties of the bna , optical emission into the tip should be continuously increased during the tip approach , which is consistent with fig .
[ fig : scheme](b ) .
the inset of fig .
[ fig : scheme](b ) show far - field images of the first bna - on - tip used in emission mode , for two tip - to - surface spacings of @xmath17 ( right inset ) and @xmath11 ( left inset ) . to this end
, laser light @xmath26w@xmath27 is injected into the fiber so that the bna is resonantly excited and the magnified far - field images of the bna through the si wafer are realized with a long working distance objective ( @xmath2860 , 0.7 ) coupled to an infrared camera .
we see that the light spot intensity right at the bna is increased close to the surface , which means that the overall bna emission is modified by the sample , not only the bna emission into the tip .
this property avoids changes in the collected signal due to modifications of the dipolar emission pattern of the bna close to a high index dielectric substrate @xcite .
note that this study stays qualitative since ir cameras undergo noticeable non - linearity in their optical response . from fig .
[ fig : spectre](b ) , we see that the resonance peak of the second bna is centered to @xmath25 when it is positioned @xmath29 away from the si surface . as a result ,
the electric field enhanced by the nano - antenna at @xmath25 during tip approach towards the sample surface is maximum at @xmath30 , as shown in fig .
[ fig : spectre](c ) , which is fully consistent with experimental results given the electric dipole emission properties of the bna .
when the bna is excited off - resonance , by rotating the incident polarization by @xmath14 , the approach curve does not show maximum anymore and the collected signal is 6 times weaker ( see inset of fig .
[ fig : scheme](c ) ) .
the approached curve is described by an exponential like function typical to that near field microscopy .
therefore , variations of the collected signal during tip approach are mainly governed by the resonance redshift of the bna , rather than other optical processes that could be in play such as tip - to - surface interferences . these variations ,
strongly dependent on bna geometry , are thus the signature that the bna is capable of sensing the presence of a flat sample even when it is located at distances from the sample surface well beyond its resonant optical confinement ( @xmath31 field amplitude ) .
nano - positioning techniques used in scanning probe microscopies are generally based on processing low voltage variations that results from the detection of mechanical or optical signals . in our case ,
the detection voltage range has been set to a few volts together with low noise with the synchronous detection scheme ( @xmath32 integration time ) . in that case , the tangential slope of the experimental curves of fig .
[ fig : scheme](b ) is measured to values about @xmath33 , @xmath34 , @xmath35 and @xmath36 at tip - to - sample spacings of @xmath17 , @xmath37 , @xmath2 and @xmath38 , respectively .
the second experimental curves of fig .
[ fig : scheme](c ) show local tangential slopes about @xmath39 , @xmath40 , @xmath41 and @xmath42 at tip - to - surface spacings of @xmath17 , @xmath43 , @xmath2 and @xmath38 , respectively . since the millivolt range is beyond the noise level of the current highly sensitive optical detection devices , one can expect with bna - on - tips distance control schemes of sub - nanometer accuracy along the longitudinal direction perpendicular to the sample surface .
the on - tip aperture type nature of the bna leads to surface sensing in subwavelength areas , in compact , fiber and plug - and - play optical architectures that require modest power consumption .
surface sensing is usually achieved on the nanometer scale by using mechanical probes aimed at detecting locally surface forces ( shear force and atomic force microscopy ) .
if transverse spatial resolution can be of the order of the angstrom , the spatial sensing depth is limited to a few nanometers . in our case
, one can expect distance regulation over distances at least two orders of magnitude higher and with subwavelength sensing lateral area .
in our paper we have demonstrated the remote sensing capability of using a bna engineered at the end of a snom fiber probe .
the bna has been shown to sense the presence of a si surface up to distance of @xmath0 .
we thus prove that na s have the ability of detecting variation in their large scale environment .
such sensing is due to the high sensitivity of the bna to a small variation in the effective dielectric constant of the medium surrounding it .
in addition , our experimental result have presented a high agreement with the theoretical one , and showed good voltage variation relative to the variation of the distance .
consequently , a single bna engraved at the apex of a snom tip may represent an alternative solution for nano - positioning and distance control instead of nano - mechanical and optical interference - based techniques used nowadays
this work is funded by the agence nationale de la recherche ( anr ) under contract nanoec ( anr-07-nano-036 ) and baltrap ( anr-10-nano-002 ) .
it is supported by the `` ple de comptitivit microtechnique '' , the labex action , and the french network of technology platforms `` renatech '' .
tip metal coating and fib fabrication are realized at mimento technology platform .
d. y. lei , k. appavoo , y. sonnefraud , r. f. haglund jr , and s. a. maier , `` single - particle plasmon resonance spectroscopy of phase transition in vanadium dioxide , '' opt .
lett . * 35*(23 ) , 39883990 ( 2010 ) .
m. a. schmidt , d. y. lei , l. wondraczek , v. nazabal , and s. a. maier , `` hybrid nanoparticle
microcavity - based plasmonic nanosensors with improved detection resolution and extended remote - sensing ability , '' nat . commun . * 3 * , 1108 ( 2012 ) .
m. mivelle , i. a. ibrahim , f. baida , g. w. burr , d. nedeljkovic , d. charraut , j .- y .
rauch , r. salut , and t. grosjean , `` bowtie nano - aperture as interfacebetween near - fields and a single - modefiber , '' opt . express * 18*(15 ) , 15,96415,974 ( 2010 ) .
vo , m. mivelle , s. callard , a. rahmani , f. baida , d. charraut , a. belarouci , d. nedeljkovic , c. seassal , g. burr , and t. grosjean , `` near - field probing of slow bloch modes on photonic crystals with a nanoantenna , '' opt .
express * 20*(4 ) , 41244135 ( 2012 ) .
m. mivelle , t. s. van zanten , l. neumann , n. f. van hulst , and m. f. garcia - parajo , `` ultrabright bowtie nanoaperture antenna probes studied by single molecule fluorescence , '' nano lett . * 12*(11 ) , 59725978 ( 2012 ) .
m. mivelle , p. viktorovitch , f. i. baida , a. el eter , z. xie , t .-
vo , e. atie , g. w. burr , d. nedeljkovic , j .- y .
rauch , s. callard , and t. grosjean , `` light funneling from a photonic crystal laser cavity to a nano - antenna : overcoming the diffraction limit in optical energy transfer down to the nanoscale , '' opt .
express * 22*(12 ) , 15,075 ( 2014 ) .
a. el eter , n. m. hameed , f. i. baida , r. salut , c. filiatre , d. nedeljkovic , e. atie , s. bole , and t. grosjean , `` fiber - integrated optical nano - tweezer based on a bowtie - aperture nano - antenna at the apex of a snom tip , '' opt .
express * 22*(8 ) , 10,07210,080 ( 2014 ) .
r. bachelot , c. ecoffet , d. deloeil , p. royer , and d .- j .
lougnot , `` integration of micrometer - sized polymer elements at the end of optical fibers by free - radical photopolymerization , '' appl . opt . * 40 * , 58605871 ( 2001 ) .
w. lukosz , `` light emission by magnetic and electric dipoles close to a plane dielectric interface .
radiation patterns of dipoles with arbitrary orientation , '' j. opt .
* 69*(11 ) , 14951503 ( 1979 ) . | plasmonic nano - antennas have proven the outstanding ability of sensing chemical and physical processes down to the nanometer scale .
sensing is usually achieved within the highly confined optical fields generated resonantly by the nano - antennas , i.e. in contact to the nanostructures . in this paper
, we demonstrate the sensing capability of nano - antennas to their larger scale environment , well beyond their plasmonic confinement volume , leading to the concept of `` remote '' ( non contact ) sensing on the nanometer scale . on the basis of a bowtie - aperture nano - antenna ( bna ) integrated at the apex of a snom fiber tip
, we introduce an ultra - compact , moveable and background - free optical nanosensor for the remote sensing of a silicon surface ( up to distance of @xmath0 ) .
sensitivity of the bna to its large scale environment is high enough to expect the monitoring and control of the spacing between the nano - antenna and a silicon surface with sub - nanometer accuracy .
this work paves the way towards a new class of nanopositionning technique , based on nano - antenna resonance monitoring , that are alternative to nanomechanical and optical interference - based devices . |
the adiabatic bond charge model ( bcm ) has been quite successful in explaining the phonon dispersion curves of group iv elemental semiconductors@xcite and partially ionic iii - v semiconducting materials@xcite with a zincblende structure . in recent years
it has been successfully applied to study phonons in semiconducting superlattices@xcite , optical properties of al@xmath0ga@xmath1as@xcite , open semiconductor surfaces@xcite and even sp@xmath2-bonded materials like graphite@xcite and fullerenes.@xcite quite recently , a modified version of the bond charge model has been applied@xcite to study the second order raman spectra of alas and alsb . in view of the similar dispersion curves of tetrahedrally
connected iii - v and ii - vi materials , it is surprising that no attempt to extend bcm to the latter has appeared in the literature
. this may be partly due to the comments indicating that early attempts in this direction were not successful because it was found@xcite that in the case of ii - vi materials the asymmetry of the bond charge position became too large to find a stable equilibrium position for the bond charges .
these conclusions were based on the studies of the valence electron charge density using local pseudopotentials@xcite , which suggested a nearly complete charge transfer from the cation to the anion .
these calculations actually overestimated the ionicity of these compounds and produced valence band spectra in strong disagreement with experimental photoemission results.@xcite later , more accurate calculations using nonlocal pseudopotentials@xcite showed better agreement with the experiments and yielded charge densities indicating a strong shift of the bond maximum rather than complete charge transfer .
in fact , the charge density plots for iii - v and ii - vi compounds@xcite are nearly identical except that the charge maxima in the latter appear to be slightly shifted toward the anion , indicating that , despite their greater ionicity , ii - vi compounds are dominantly covalent in nature . in view of these results
, it is expected that the bcm should give a good account of the phonons in ii - vi materials provided the bond charges are placed at physically reasonable places suggested by the pseudopotential calculations .
traditionally , the lattice dynamics of these materials has been done using rigid ion or shell models.@xcite these models give good fits to the observed phonon dispersion curves at the cost of a large number of adjustable parameters ( 10 or more ) , some of which have no physical interpretation .
recently , _
ab initio _ calculations of phonon spectra have appeared,@xcite but they are are not feasible for studying large systems such as alloys like cd@xmath0hg@xmath1te or thick superlattices .
therefore it is desirable to have a realistic model with fewer , physically meaningful , parameters that is easy to extend to more complex systems . in this paper
we show that the 6-parameter bcm provides a good description of the phonons and other lattice dynamical quantities such as elastic constants and specific heat in ii - vi materials .
the rest of the paper is organized as follows . in sect .
ii we provide a brief overview of the bond charge model.@xcite in sect .
iii we discuss the results for six ii - vi compounds and in sect .
iv trends in the parameters are discussed .
the adiabatic bond charge model ( bcm ) for homopolar semiconductors@xcite and partially ionic iii - v compounds@xcite is the simplest empirical lattice dynamical model that correctly describes the phonon dispersion curves of covalent crystals . in the bcm the valence electron charge density
is represented by massless point particles , the bond charges ( bc s ) , that follow the ionic motion adiabatically .
the bcm unit cell consists of two ions and four bond charges that are placed along the bonds between the ions . in homopolar covalent crystals
the bond charges are placed midway between the neighboring atoms while in iii - v compounds the bc divides the bond length in the ratio of 5:3 .
this is consistent with nonlocal pseudopotential calculations for the valence electron charge density@xcite that indicate that the charge density maximum in iii - v compounds shifts toward the group v element .
this shift is even stronger in the case of ii - vi compounds , reflecting their more ionic character .
the bcm parameter @xmath3 which measures the polarity of the bond is defined in terms of the ratio in which the bc position divides the bond length .
if @xmath4 is the bond length and @xmath5 and @xmath6 are the two ion - bc distances then @xmath7 ( @xmath8 ) for homopolar materials and @xmath9 ( @xmath10 ) for iii - v compounds . in our extension of the bcm to ii - vi materials
we have chosen to use @xmath11 corresponding to the ratio @xmath12 which is based on the results@xcite of microscopic calculations .
this choice will be discussed in more detail in sect .
[ sec : trends ] .
the cation and the anion interact with one another and with the bond charges via central potentials @xmath13 , @xmath14 , and @xmath15 , respectively . the bond charges centered on a common ion interact via a three body keating potential,@xcite @xmath16 , where @xmath17 is the distance vector between ion @xmath18 ( @xmath18 = 1 , 2 ) and bc @xmath19 , @xmath20 is the force constant and @xmath21 is the equilibrium value of @xmath22 .
the bond charges centered on a particular ion also interact directly with one another through a central potential , @xmath23 , where @xmath24 is the distance between the bond charges centered on the cation ( @xmath25 ) or the anion ( @xmath26 ) .
finally , the ions and the bc s interact via the coulomb interaction characterized by a single parameter @xmath27 where @xmath28 is the charge of a bc , and @xmath29 is the dielectric constant .
each of the ions is presumed to have a charge @xmath30 so that the net charge in the unit cell is zero . to reduce the number of parameters it is assumed@xcite that @xmath31 , @xmath32 and @xmath33 . if we use these constraints on the total lattice energy per unit cell @xmath34 - \alpha_{\scriptscriptstyle m } { ( 2z)^2 \over \epsilon}{e^2 \over t } + 6[v_{bb}^1 + v_{bb}^2 + \psi_1(r_{bb}^{(1 ) } ) + \psi_2(r_{bb}^{(2)})]\ ] ] along with the equilibrium conditions , @xmath35 and @xmath36 , we find@xcite @xmath37 the conditions for stable equilibrium , @xmath38 and @xmath39 , further yield @xmath40 and @xmath41 the madelung constant @xmath42 of the model is defined by writing the coulomb energy per unit cell as @xmath43 . for @xmath44
the values of @xmath42 , @xmath45 and @xmath46 are found numerically@xcite to be 5.0598 , 4.0539 and 17.46 , respectively . in eq .
( [ eq : stable1 ] ) and eq .
( [ eq : stable2 ] ) the force constants are in units of @xmath47 , where @xmath48 is the unit cell volume .
with @xmath49 , @xmath50 , @xmath51 , @xmath52 , @xmath53 , @xmath54 and @xmath55 given as above , the six free parameters of the model are @xmath56 , @xmath57 , @xmath58 , @xmath59 , @xmath60 and @xmath61 , which we adjust to fit the neutron scattering data and the measured elastic constants .
the phonon eigenfrequencies and eigenvectors are found by diagonalizing the dynamical matrix@xcite constructed from the bcm equations of motion @xmath62{\bf u } + [ { \bf t } - 2{(ze)^2 \over \epsilon } { \bf c}_{\scriptscriptstyle t } ] { \bf v}}\nonumber\\ % { \bf 0 } & = & { [ { \bf t}^+ - 2{(ze)^2 \over \epsilon } { \bf c}_{\scriptscriptstyle t}^+]{\bf u } + [ { \bf s } + { ( ze)^2 \over \epsilon } { \bf c}_{\scriptscriptstyle s}]{\bf v}}\end{aligned}\ ] ] here * m * is the mass matrix for the ions and * u * and * v * are the vectors formed by the displacements of the ions and the bc s , respectively . the matrices * r * , *
t * , and * s * are the dynamical matrices for the short range ion - ion , ion - bc and bc - bc interactions and * c*@xmath63 , * c*@xmath64 and * c*@xmath65 are the corresponding coulomb matrices which are evaluated by ewald s method.@xcite explicit forms of * r * , * t * , and * s * can be found in the appendices of refs . and .
figure 1 shows the dispersion curves for six ii - vi materials along with the existing neutron scattering data .
figure 2 shows the corresponding densities of states ( dos ) .
we present in table i the bcm parameters and in table ii the calculated and measured elastic constants . in all cases
the overall agreement with the experimental data is fairly good and is of the same quality as for bcm fits for iii - v compounds.@xcite for comparison we have also included the dispersion curves for gaas and insb calculated using the parameters from ref . . for cdte
our six parameter fit is as good as the 14 parameter shell model fit of rowe _
et al._@xcite and the 11 parameter rigid ion model fit of talwar _
et al._@xcite all three models show a slight upward bend in the to branch in the ( 100 ) direction , which is in contrast with the results of a recent _ ab initio _
calculation.@xcite the three models give very similar predictions for the phonon density of states .
the principal difference is that the shell model does not predict a gap in the dos between the acoustic and optical contributions , while the bcm has a smaller gap than that seen in the rigid ion model .
this underscores the fact that the bcm is , in many ways , an intermediate model between the shell model and the rigid ion model .
the shell model takes care of the electronic polarizability explicitly by attaching deformable shells to the ions .
the bcm partially accounts for the electronic polarizability through the adiabatic motion of the bond charges , while the rigid ion model ignores it completely . for zns and
znte our fits are comparable with the 10-parameter valence shell model results of vagelatos _
et al._@xcite for znte the bcm predicts a large dispersion in the lo branch near the zone edge .
however , the maximum deviation from the experimentally measured frequency at the x point is only about 7 percent . for both zns and znte the shape of the optical branches in the ( 110 ) direction is different from the results of ref .
but is similar to that predicted@xcite by _ ab initio _ calculations . for znse also
the agreement with the neutron data@xcite and the measured elastic constants@xcite is fairly good .
mercury compounds , because of their semi - metallic nature , deserve a separate discussion . because of the zero band - gap , the energy for electronic transitions from the valence band to the conduction band is comparable to the optical phonon energy . for hgte raman measurements@xcite at 90 k and infrared reflectivity measurements@xcite at 77
k yield @xmath66 @xmath67 whereas infrared spectra at 8 k gave@xcite @xmath68 @xmath67 .
this difference was attributed to the large number of carriers at higher temperatures .
the same reason is invoked to explain the degenerate values of @xmath69 and @xmath70 found in neutron scattering experiments.@xcite because of this controversy we did not use optical phonon frequencies near the zone center in our fit for hgte .
it is seen that the agreement with the acoustic and transverse optical phonons is good . however , the fit for the lo branch is not of the same quality , although the deviation from the experimental points is only a few percent . in the ( 111 ) direction
the bcm predicts that the lo branch will dip downward instead of the upward trend observed experimentally.@xcite for hgse the agreement with the available neutron data on acoustic phonons@xcite and optical measurements@xcite and measured elastic constants@xcite is very good . because of the lack of neutron data on optical phonons we can not comment on the accuracy of the optical branches .
however , it should be mentioned that the bcm and the 11 parameter rigid ion model@xcite give similar behavior for the optical branches . to further check the parameters we have calculated the specific heats@xcite for all the six materials using the parameters given in table i. the results are shown in fig . 3 as plots of log @xmath71 _ vs. _ log @xmath72 along with the experimental data .
in every case good agreement is obtained with the experiments , giving further support for the parameters used and the calculated density of states .
we have thus demonstrated that the 6-parameter bcm provides a good description of the phonons and other lattice dynamical quantities such as elastic constants and specific heat for ii - vi compounds with zincblende coordination .
the overall agreement with the neutron data is very good .
however , some discrepancies remain , particularly in the lo branch near the zone edge , where the bcm predicts a large dispersion in almost every material including the iii - v compounds .
this is most obvious in hgte and is indicative of the failure of the bcm to account for the polarizability of the ion core .
because of the associated macroscopic field , the lo phonons are more affected than the other branches .
the calculated@xcite static dielectric function @xmath73 for iii - v and ii - vi compounds is known to have considerable structure , so including a charge form factor should remedy this discrepancy.@xcite
some trends in the parameters presented in table i are immediately obvious . one notices that as one goes from group iv elements to iii - v compounds to ii - vi compounds the parameters involving bc s change uniformly . for group
iv elements the bond charge is situated midway along the bond and the ion - bc and bc - ion - bc force constants are equal for the two ions . however , in iii - v compounds the bc shifts toward the anion which results in higher values for @xmath74 and @xmath60 than @xmath75 and @xmath59 respectively .
this trend continues as we move to ii - vi compounds in which the bc is even closer to the anion .
this pattern in the values of these parameters can be traced to the equations linking @xmath50 and @xmath51 to @xmath42 and @xmath45 .
it should be noted that the values for @xmath42 and @xmath76 for ii - vi compounds are higher than those for iii - v compounds .
apart from these obvious features , there are no other discernible trends in the parameters with respect to the ionicity or the bond length .
however , it is seen that the ion - bond parameters @xmath75 and @xmath74 are considerably lower for mercury compounds than for other materials .
this is reasonable in view of the semi - metallic nature of these materials and the fact that these parameters represent off - diagonal contributions to the dielectric function .
the effect of ionicity on the phonon dispersion curves can be investigated by studying the isoelectronic sequence of materials in which the bond lengths and the average mass in the unit cell are almost same .
two such sequences are ge - gaas - znse and @xmath77sn - insb - cdte .
increased ionicity results in a general lowering of all frequencies and elastic constants and a lifting of the lo - to degeneracy at the @xmath78 point and the lo - la degeneracy at the x point .
a glance at the bcm parameters for these materials shows that the only pattern is a decrease in the magnitude of the bond charge @xmath79 and , in the case of ge - gaas - znse , a decrease in @xmath80 with increased ionicity . for the other sequence
, @xmath80 is almost same for @xmath77-sn and insb but it is smaller for cdte .
a cross comparison of the corresponding materials in the two sequences shows that moving down the periodic table , with its concomitant increase in metallicity , yields an increase in the ion - ion interaction @xmath80 while the parameters involving the bond charges decrease .
we should also comment on the choice of the equilibrium positions of the bond charges . in principle , @xmath3 should be treated as the seventh adjustable parameter of the model .
however , we decided to use the physically reasonable value of 1/3 for @xmath3 .
this corresponds to dividing the bond length in a ratio 2:1 and is consistent with the pseudo - potential calculations of the valence electron charge density.@xcite however , we were also able to find values for the six parameters which still satisfied the stability conditions ( [ eq : stable1 ] ) and ( [ eq : stable2 ] ) and which gave satisfactory fits for @xmath3 as high as 0.55 .
the parameters that varied most with @xmath3 were the ion - bc force constants ; increasing @xmath3 led to a larger @xmath74 and a smaller @xmath75 .
the remaining parameters also changed , but by much less . in every case
, the acoustic phonon curves showed very good agreement with the neutron scattering data .
however , the agreement with the optical phonons slightly worsened as @xmath3 increased .
these results highlight the arbitrariness involved in defining the equilibrium position for the bond charges . our choice of @xmath11 , which coincides roughly with the position predicted by the pseudopotential calculations@xcite , still gave the best overall agreement with the experimental results .
we have applied the adiabatic bond charge model to study phonons in six ii - vi compounds with a zincblende structure .
the theoretical predictions of the 6-parameter bcm are in good agreement with the available neutron data and the experimentally measured elastic constants and specific heats .
some minor discrepancies in the lo branch near the zone edge are believed to be due to the incomplete description of the electronic polarizability of the ions .
these deviations are larger , though still only a few percent , in the case of hgte as expected from its semi - metallic nature and consequently stronger screening effects . in conclusion
, we have found that the 6-parameter adiabatic bond charge model provides a satisfactory description of the lattice dynamics of tetrahedrally connected ii - vi compounds .
the agreement with the experimental data is of the same quality as for iii - v compounds .
we have also discussed some broad trends seen in the parameters .
we find that the parameters for the bond charge - cation short range interactions decrease with a corresponding increase in the parameters for the bond charge - anion interactions as one goes from group iv elements to iii - v to ii - vi compounds .
cccccccc & @xmath80 & @xmath75 & @xmath74 & @xmath59 & @xmath60 & @xmath61 & @xmath79 + si & 6.21 & 6.47 & 6.47 & 8.60 & 8.60 & 0.1800 & 1.47 + ge & 6.61 & 5.71 & 5.71 & 8.40 & 8.40 & 0.1620 & 1.61 + @xmath77-sn & 7.43 & 5.59 & 5.59 & 7.80 & 7.800 & 0.163 & 1.98 + + alas & 5.80 & 2.27 & 15.48 & 5.79 & 8.54 & 0.1800 & 1.21 + gap & 6.04 & 2.4 & 17.91 & 5.20 & 10.0 & 0.2030 & 1.36 + gaas & 6.16 & 2.36 & 16.05 & 5.36 & 8.24 & 0.1870 & 1.43 + gasb & 6.77 & 2.37 & 13.10 & 6.28 & 7.08 & 0.1600 & 1.52 + inp & 7.16 & 2.95 & 21.62 & 3.43 & 8.37 & 0.2490 & 1.55 + inas & 7.31 & 2.64 & 17.86 & 3.99 & 7.30 & 0.2100 & 1.60 + insb & 7.47 & 2.33 & 14.09 & 4.56 & 6.24 & 0.1720 & 1.64 + + zns & 5.74 & 0.79 & 29.90 & 0.83 & 15.40 & 0.2130 & 1.05 + znse & 5.01 & 1.19 & 22.82 & 1.21 & 15.65 & 0.1790 & 1.03 + znte & 5.51 & 1.06 & 22.93 & 1.07 & 17.00 & 0.1800 & 1.05 + cdte & 6.85 & 0.77 & 23.34 & 0.39 & 15.44 & 0.1830 & 1.15 + hgse & 5.32 & 0.15 & 14.01 & 0.35 & 17.50 & 0.1095 & 0.91 + hgte & 6.46 & 0.081 & 13.46 & 1.08 & 15.60 & 0.1062 & 1.03 + | we extend the adiabatic bond charge model , originally developed for group iv semiconductors and iii - v compounds , to study phonons in more ionic ii - vi compounds with a zincblende structure .
phonon spectra , density of states and specific heats are calculated for six ii - vi compounds and compared with both experimental data and the results of other models .
we show that the 6-parameter bond charge model gives a good description of the lattice dynamics of these materials .
we also discuss trends in the parameters with respect to the ionicity and metallicity of these compounds . |
the origin of the generation structure realized in nature is an engrossing subject which has been discussed for a long time , but still remains veiled . within the standard electroweak theory ,
masses and mixing angles of fermions originate from the yukawa interactions with the higgs boson , responsible for the electroweak symmetry breaking .
while , the relative strength of the gauge interactions for various fermion species are controlled by gauge invariance , yukawa couplings are not governed by any principle .
they bring a multitude of free parameters into the theory and even have ambiguities in reconstructing their values from experiments .
a viable approach is , therefore , to search for mass matrices which are taking suggestive forms in light of model building ingredients , such as symmetry amongst generations .
a direct scheme in this spirit is `` texture zero '' in the mass matrices of fermions . in this framework
, it is assumed that the mass matrices have several elements which are anomalously small compared to the others .
initially this approach was studied in the quark sector @xcite and it was found that the existing relations among masses and mixings of quarks can be explained by the vanishing matrix elements .
the available textures presented in the literature provide foundations of model building and insights into the generation puzzle . as for the lepton sector
, recent progress in neutrino physics makes it possible to discuss feasible forms for mass matrices . in @xcite and many subsequent papers , the texture zeros in lepton sector have been discussed in the context of various types of mass matrices such as the majorana mass matrix of the left - handed neutrinos @xcite , both the charged lepton and neutrino mass matrices @xcite , the dirac and the majorana mass matrices in the seesaw mechanism @xcite and amalgamated yukawa couplings in grand unification @xcite .
all these results also provide useful information to infer the structure lying behind the yukawa interactions .
it is to be noted , that , for the mass matrices in the lepton sector , the situation is different from that of the quark sector because of the weak hierarchy of neutrino mass spectrum ( for normal hierarchy ) and two large mixing angles .
in addition , the neutrino spectrum can also have inverted hierarchy and can even be quasi - degenerate which has no analogue in the quark sector .
one of the most distinct features amongst observations in the neutrino sector is the small 1 - 3 angle against possible maximal 2 - 3 mixing .
this is the origin of the @xmath4-@xmath5 symmetric nature of the neutrino mass matrix @xcite . in the generation basis where the charged - lepton mass matrix is diagonal
, the exact @xmath4-@xmath5 symmetric limit means that there are simple equalities among the matrix elements of the neutrino mass matrix .
moreover , the @xmath4-@xmath5 symmetric neutrino mass matrix contains the tri - bimaximal mixing @xcite as a special case , which is obtained by requiring further special relations among the matrix elements .
thus , the equality between the matrix elements might play an important role in understanding the properties of neutrinos and become a simple and direct approach to search for `` suggestive forms '' of the mass textures .
the equalities among left - handed majorana mass matrix elements were discussed in @xcite in a bottom - up way . in @xcite ,
hybrid structures , with the coexistence of equalities among matrix elements and texture zeros for the left - handed majorana mass matrix were studied . in this paper
we follow the approach of @xcite , but consider the equalities and zeros in the dirac and right - handed majorana mass matrices in the seesaw mechanism .
we perform a thorough classification of such hybrid textures and identify the left - handed majorana mass matrices that can be reconciled with the inputs obtained from neutrino oscillation data . in particular , we consider the case where only two right - handed neutrinos take part in the seesaw mechanism @xcite .
texture analysis in the context of such a minimal seesaw scheme has been accomplished in @xcite .
an advantage of this choice is that there are less number of free parameters .
thus the mass matrices get simple forms and rich predictions compared to the standard three heavy neutrino models . by exhausting all possibilities
, we find a novel class of textures which have only two adjustable parameters to fit the low - energy data .
we discuss the possibility of having leptogenesis in these textures and explore the connection between high and low energy cp violation if any .
the layout of the paper goes as follows . in section 2 , we present the formulations which are needed for the subsequent discussions . in section 3
, we study equalities amongst mass matrix elements and enumerate general possible forms of the mass matrices with equality relations . in section 4
, we perform a combined analysis of equalities and texture zeros and discuss the viable minimal textures which are compatible with the current neutrino data . in section 5 , we discuss cp violation , paying particular attention to possible connections between leptogenesis and cp violation in neutrino oscillation . in section 6 , we briefly comment on the renormalization group effects for the mass matrices .
section 7 is devoted to conclusion and summary .
we assume that the tiny neutrino masses are generated through type - i seesaw mechanism @xcite by the suppression effect of the large mass scale of the right - handed neutrinos . by integrating out the heavy right - handed neutrinos ,
the majorana mass matrix of the left - handed neutrinos is obtained as , @xmath6 where @xmath7 denotes the dirac mass matrix after the electroweak symmetry breaking , and @xmath8 is the majorana mass matrix for the right - handed neutrinos . in this paper , we consider the case where we have only two right - handed neutrinos .
accordingly , the dirac mass matrix @xmath7 is a @xmath9 rectangular matrix , while @xmath8 is given by a @xmath10 symmetric matrix .
it is noteworthy that the two generations of fermions are matched well with the idea of the doublet ( irreducible representation ) of discrete groups , which have recently been utilized to address the observed lepton mixings and masses .
the mixing matrix in the lepton sector , the so - called pontecorvo - maki - nakagawa - sakata ( pmns ) matrix @xmath11 can be defined as the unitary matrix which diagonalizes the left - handed majorana mass matrix @xmath12 in the generation basis where the charged - lepton mass matrix is diagonal : @xmath13 where @xmath14 .
the neutrino oscillation experiments are capable of determining the squared mass differences @xmath15 and the three mixing angles and one cp violating phase in @xmath11 .
we parameterize the pmns matrix @xmath11 as @xmath16 where @xmath17 and @xmath18 stand for @xmath19 and @xmath20 .
the phase @xmath21 represents one phase degree of freedom which is responsible for cp violation phenomena at low - energy , while @xmath22 and @xmath23 are the majorana phases .
@xmath24 is a diagonal phase matrix which is to be removed by the redefinition of the left - handed fields .
so far , oscillation experiments have determined the two mass squared differences and the two angles , and have provided an upper bound on the third mixing angle @xmath25 .
the three generation analysis of the current data suggests .the present best - fit values and the 3@xmath23 ranges of oscillation parameters from @xcite . [ cols="^,^,^",options="header " , ] let us first discuss the solution i ; @xmath26 after the seesaw mechanism , the majorana mass matrix for the left - handed neutrinos becomes @xmath27 where we introduce the parameter @xmath28 and @xmath29 to simplify the notation .
note that this matrix obeys the so called scaling property between the second and third rows and second and third columns @xcite .
this matrix has @xmath4-@xmath5 symmetry and a zero eigenvalue such that the mass ordering is predicted to be inverted .
moreover , the reactor and the atmospheric angles are given as @xmath30 and @xmath31 .
a nontrivial consequence of this solution is thus the relation between masses and the solar angle .
the two nonzero eigenvalues are @xmath32 we find that these eigenvalues should be identified as @xmath33 and @xmath34 in order to fit the observations .
the parameters @xmath35 and @xmath36 are therefore fixed in terms of the two mass differences as @xmath37 where @xmath38 .
here we are taking a combination of the solutions which is viable with the solar neutrino observations .
the solar angle is also fixed by the two mass differences .
it is given by @xmath39 where @xmath40 by expanding ( [ sin12 ] ) in powers of @xmath41 , we find @xmath42 it is interesting to observe that , in the 0-th order , the solar mixing angle is predicted to be @xmath43 , which is the same as in the tri - bimaximal mixing scenario .
in fact , the whole mixing matrix @xmath11 can be written as @xmath44 up to @xmath45 .
thus , this is given by the tri - bimaximal mixing matrix multiplied by the small correction matrix .
this type of lepton mixing is naturally realized by a class of flavor models which utilize the scherk - schwarz twisted boundary conditions @xcite .
it might be interesting to study this type of deviation from the tri - bimaximal mixing and its implication for future neutrino experiments systematically .
the effective neutrino mass @xmath46 which is responsible for the neutrino - less double beta decay is given by the element @xmath35 itself .
thus in this texture we find @xmath46 is predicted as @xmath47 at the best fit of the mass difference . with this value , neutrino - less double beta decay may be detectable in forthcoming experiments .
since there are only two mixing angles in @xmath11 , cp must be conserved at low energy . however , at high energy there is one unremovable phase which can violate cp .
such cp violation is conveniently measured by the weak basis invariant @xmath48 $ ] where @xmath49 and @xmath50 @xcite . for the texture
i , we find @xmath51 in the basis where only @xmath52 is complex ; @xmath53 .
the leptogenesis @xcite is thus possible , with the lepton asymmetry proportional to this quantity .
next we discuss the solution ii ; @xmath54 after the seesaw mechanism , the majorana mass matrix for the left - handed neutrinos becomes @xmath55 where we define @xmath56 and @xmath57 to simplify the notation .
as in the previous case , this matrix also has @xmath4-@xmath5 symmetry and a zero eigenvalues .
the mass hierarchy is thus predicted as inverted ordering , and the reactor and the atmospheric angles are given as @xmath30 and @xmath31 .
the two nonzero eigenvalues are @xmath58 we find that these eigenvalues should be identified as @xmath33 and @xmath34 in order to fit the observations . a solution of this equation system is found to be @xmath59 here we show approximate expressions for @xmath60 and @xmath61 because the exact expressions are too long and complicated to present here .
the solar angle is also fixed by the two mass differences .
it is written in terms of @xmath60 and @xmath61 , and the mass differences as @xmath62 we can see that the prediction for the solar angle agrees with the current @xmath63 range although it is close to its lower bound . the averaged neutrino mass @xmath46 which is responsible for the neutrino - less double beta decay is given by @xmath64 in this case .
we find @xmath46 is predicted as @xmath47 at the best fit of the mass difference and a signal can be expected in future we have a slightly better chance to detect neutrino - less double beta decay measurements .
cp violation is possible at high - energy as the weak - basis invariant @xmath65 is non - vanishing .
in fact , it becomes @xmath66 in the basis where only @xmath67 is complex ; @xmath68 . on the other hand ,
no cp violation occurs in the neutrino oscillation because of the vanishing @xmath25 .
finally let us discuss the solution iii .
the viable texture is given by @xmath69 after the seesaw mechanism , the majorana mass matrix for the left - handed neutrinos becomes @xmath70 where we introduce the parameter @xmath71 and @xmath72 to simplify the notation .
this has one zero element in the final matrix and has been discussed in @xcite . unlike the two solutions found in the previous subsection , there is no @xmath4-@xmath5 symmetry in ( [ miii ] ) . therefore the reactor and the atmospheric angles no longer satisfy @xmath30 and @xmath31 , and all nonzero values of the mixing angles can be described as functions of the mass differences .
we note that , because of the lack of the @xmath4-@xmath5 symmetry , there appears an additional solution in this case .
namely , the @xmath7 which is obtained by 2 - 3 column exchange of ( [ soliii ] ) can also be compatible with the data . although these two solutions are independent in the sense that they are not related to each other by basis rotations , their physical consequences are almost the same .
we thus discuss only ( [ soliii ] ) to illustrate important features of these solutions .
the two nonzero eigenvalues of ( [ miii ] ) are @xmath73 we find that these eigenvalues should be identified as @xmath74 and @xmath75 in order to fit the observations . the parameters @xmath60 and @xmath61 are therefore fixed in terms of the two mass differences . however , the exact expressions are too long and complicated and it is not appropriate to present them here . instead ,
we shall approximate them in powers of @xmath41 ; @xmath76 here we are picking up a combination of the solutions for @xmath74 and @xmath75 , which gives the correct mixing angles .
note that we represent the leading and the first order coefficients by decimal numbers .
these terms can also be represented as functions of integers as in ( [ abii ] ) , but the expressions are not so simple as in the case of the other two solutions .
the three angles are also fixed by the two mass differences .
they are given by @xmath77 here we note again that the decimals in the above expressions represent definite irrational numbers which are too long and complicated to be presented in closed form .
it is interesting that the predictions for @xmath78 and @xmath79 are very close to the current best fit values indicated in table 1 .
we see that this texture predicts a relatively large @xmath25 which can be measured in forthcoming experiments like double - chooz .
the effective neutrino mass @xmath46 which is responsible for the neutrino - less double beta decay is given by @xmath80 and is predicted as @xmath47 at the best fit of the mass difference .
thus , neutrino - less double beta decay will be detectable in forthcoming experiments .
finally we comment on cp violation . since @xmath25 is nonzero with the solution iii
, one might expect that there exists corelation between cp violation phenomena at high and low - energy scale . however , this is not the case .
the invariant measure @xmath65 is given by @xmath81/|b|^4 $ ] , but @xmath82 is real valued for the @xmath7 in ( [ soliii ] ) .
thus the leptogenesis does not occur with the solution iii . on the other hand ,
the phase @xmath21 in the pmns matrix is not vanishing . the cp violation caused by @xmath21
can be conveniently described with , for example , @xmath83 ^ 3 $ ] , where @xmath84 is the charged - lepton mass matrix @xcite .
it is easily seen that the basis - independent quantity @xmath85 takes the form @xmath86 + c_2 \cdot { \rm im}\big [ a^2 b^*{}^2 \big]$ ] , where @xmath87 and @xmath88 are nonzero real values , and @xmath85 is therefore nonzero in general .
to summarize , with the solution iii , cp violation in the oscillation can be measured , while leptogenesis is not possible .
interestingly , the three solutions i , ii and iii are consistent only with the inverted mass ordering for neutrino mass spectrum . from this fact ,
we conclude that the hybrid property prefers the inverted hierarchy .
however , there exist three quasi viable textures iv , v and vi with normal hierarchy .
we call them quasi - viable as certain predictions of these textures are marginally consistent with the @xmath63 range presented in table 1 , this can be seen , for example , with v and vi in which the element @xmath89 is vanishing . as a general consequence of @xmath90 and
the normal hierarchy with @xmath91 , the solar and reactor angles correlate with @xmath41 as @xmath92 .
this relation needs small @xmath78 and @xmath41 , and large ( just below the current @xmath63 bound ) @xmath25 . with the solutions
iv and v , there are is no cp violating phase for the right - handed mass matrix since the heavy neutrino are degenerate in mass .
such degeneracy can be relaxed by some corrections , for example , renormalization group evolution from the texture scale ( at which the textures are assumed ) down to the right - handed neutrino scale . however , even with relaxations of the degeneracy , the lepton asymmetries for iv and v are small because the off - diagonal components of @xmath93 are real valued .
it may be possible to generate a non - zero value for this by introducing two - loop renormalization group effects @xcite . in this paper
we do not consider these effects . on the other hand , we have nonzero @xmath65 and @xmath85 within the solution vi . therefore the texture vi can give rise to a connection between leptogenesis and cp violation in neutrino oscillation .
we discuss this connection in detail in the next section .
with the solution vi , both @xmath65 and @xmath85 are non - vanishing , and there is a correlation between cp violation phenomena at high and low - energy scales .
it is interesting if we can predict the amount of cp violation at low energy in terms of the phase of the yukawa coupling responsible for successful baryogenesis via leptogenesis . in this section
, we address this issue and study the connection between leptogenesis and cp violation at low energy with the interesting example of the solution vi . to perform this program , it is convenient to parameterize the dirac mass matrix @xmath7 as @xcite @xmath94 where @xmath95 are positive - real parameters , and @xmath96 is an unitary matrix which contains only one dirac type phase ( conveniently parameterized as ( [ pmns ] ) without majorana phases ) , and @xmath97 is a diagonal phase matrix . taking account of the texture zeros and equalities in @xmath7 of the solution vi , the general form ( [ sahu ] )
is reduced to @xmath98 where @xmath99 and @xmath100 where @xmath101 .
therefore ( [ sahu2 ] ) contains two real parameter @xmath102 and @xmath103 , and one phase @xmath104 , as it should do . with the parameterization ( [ sahu2 ] ) ,
the seesaw formula is written as @xmath105 where @xmath106 is a complex - symmetric @xmath107 matrix of @xmath108 , but it has nonzero elements only in the lower - right @xmath10 block .
this is an advantage of the form ( [ sahu ] ) because diagonalization of a @xmath109 matrix is significantly easier than @xmath110 case .
the pmns matrix is given by @xmath111 where @xmath112 is a unitary matrix which diagonalize @xmath106 such that @xmath113 .
we should identify the two mass eigenvalues @xmath114 as @xmath115 and @xmath116 , where @xmath117 is the eigenvalue of @xmath118 which is a function of @xmath119 and @xmath104 .
thus the mass ratio @xmath120 gives a constraint on the @xmath119-@xmath104 space . on the other hand ,
the overall scale of the neutrino mass gives a relation between @xmath102 and @xmath121 as @xmath122 .
the three angle and the mass ratio @xmath41 are controlled by only two parameters @xmath119 and @xmath104 .
since the analytic expressions for the low - energy observables are complicated , we check the constraints on @xmath104-@xmath119 space numerically . fig .
[ range ] shows the allowed region on @xmath104-@xmath119 plane .
the upper(lower ) green(solid ) curves are the @xmath63 lower(upper ) bound of the ratio @xmath41 , which are obtained from @xmath120 .
more important constraints come from the reactor bound @xmath123 and the lower bound of the solar angle @xmath124 , which are shown by red(dotted ) and blue(dashed ) curves respectively .
both @xmath25 and @xmath78 get increased as @xmath119 increases , so that the shaded region remains allowed by the current oscillation data .
the upper bound on @xmath78 draws a curve above the shaded region and it does not reduce the allowed space . in the following , we fix @xmath125 to assess maximal impact on the low - energy cp violation . for @xmath125 ,
the possible range of the phase @xmath104 is @xmath126 . with these parameters ,
the three mixing angles are predicted as @xmath127 , @xmath128 , @xmath129 and @xmath130 .
the dirac type phase @xmath21 in @xmath11 can be measured in long - baseline experiments @xcite .
the cp violation arises in the difference of transition probability @xmath131 .
the difference is proportional to the leptonic version of the jarlskog invariant @xcite @xmath132.\end{aligned}\ ] ] as discussed above , the mixing matrix @xmath11 depends only on the two parameters @xmath119 and @xmath104 . with @xmath125 ,
@xmath133 is bounded as @xmath134 for @xmath126 .
the cp asymmetry @xmath135 with the heavy neutrino decay ( lighter one ) is given by @xmath136 } { ( \bar{m}_d^ * \bar{m}_d^{\rm t})_{11}v^2 } \cdot f \ ! \left(\frac{m_2 ^
2}{m_1 ^ 2}\right),\end{aligned}\ ] ] where @xmath137)$ ] and @xmath138 is the vacuum expectation value of the higgs field .
the right - handed neutrino masses are denoted by @xmath139 and @xmath140 with @xmath141 and @xmath142 .
we neglect the contribution from the self - energy diagram which is small compared to the vertex one for the heavy neutrino scale of @xmath143 in hierarchical case . in the above , the dirac mass matrix @xmath144 is in the basis where @xmath8 is diagonal .
the baryon - to - photon ratio is given by @xmath145 where the factor @xmath146 represents the sphaleron conversion and the dilution due to the photon productions from the onset of leptogenesis until recombination .
the factor @xmath147 is the final efficiency factor which we are taking @xmath148 in this case .
[ jcpetab ] shows @xmath149 and @xmath133 as a parametric plot with respect to @xmath104 .
we put details about the plot in the caption .
we can see a sharp correlation between @xmath149 and @xmath133 .
in particular , the sign of @xmath150 is predicted to be negative ; the disappearance probability of the anti - neutrino @xmath151 will be observed greater than that of the ordinary @xmath152 .
it is also clear that there is a lower bound of the mass scale @xmath153 as @xmath154 .
if the cp violation @xmath155 is measured , then it is an indirect measurement of the right - handed mass scale under the assumption that the leptogenesis is solely responsible for the baryon asymmetry of the universe .
it is to be noted that the majorana mass matrices obtained through seesaw diagonalization are implicitly at some high scale which depends on the mass of the heavy neutrinos . consequently the mixing angles and the mass eigenvalues are the corresponding quantities at the high scale . to obtain the values at the low scale ,
renormalization group ( rg ) induced running effects need to be incorporated @xcite .
impact of rg running with tri - bimaximal mixing at high scale has been considered in @xcite .
it was found that these effects are typically small for hierarchical spectrum .
considerable running can be possible for quasi - degenerate neutrinos depending on the values of majorana phases @xcite .
however since the rg induced corrections to the mass matrix elements are multiplicative in nature it is expected that a zero in the mass matrix @xmath156 will remain a zero @xcite . it is also shown in @xcite that a mass matrix obeying scaling properties are stable against rg corrections .
therefore it is plausible that the textures which we find as allowed will be stable against rg corrections .
however it is possible that certain textures which are disallowed marginally may get allowed if one included rg effects . in this paper
we do not attempt to classify such textures .
renormalization effects for texture zero mass matrices have been discussed in @xcite .
in particular , @xcite discussed radiative generation and stability of texture zeros in the context of type - i seesaw models for running from low to high scale and reached the same conclusion that the rg effects can not make an allowed texture forbidden but the converse may be possible .
thus we do not expect the allowed patterns to get excluded by rg effects .
in this work we consider simultaneous presence of equalities and texture zeros in the elements of dirac and majorana mass matrices in the context of the minimal seesaw model containing two heavy right - handed neutrinos .
it is well known that because of the symmetric nature of the majorana mass matrix ( @xmath8 ) the off - diagonal elements are equal . in the present study
, we impose additional equalities among the elements of the majorana mass matrix as well as on the elements of the dirac mass matrix ( @xmath7 ) at some high scale .
equalities among matrix elements of neutrino mass matrices can arise for example due to @xmath4-@xmath5 exchange symmetry which predicts @xmath25=0 and @xmath157 in the basis where the charged lepton mass matrix is diagonal .
such equalities reduce the number of free parameters in the theory and hence increase its predictive power . another way to reduce the number of free parameters
is the postulation of texture zeros which can also be motivated by certain class of flavor symmetries in the mass matrix .
we classify and enumerate the general possibilities of mass matrices with equalities among elements .
then we perform a hybrid texture analysis combining both equalities and zeros .
our aim is to identify the left - handed majorana mass matrices obtained by seesaw diagonalization , that are compatible with the neutrino oscillation data .
we study a large number of independent options ( more than 400 ) and find that at the level of minimal number of free parameters ( _ i.e. _ with maximum number of conditions imposed on the elements of the dirac and majorana mass matrices ) , only 6 textures stand out to be consistent with global neutrino oscillation data .
these 6 patterns , presented in table 2 are thus , quite special and rare from the point of view of the parameter sets realized in nature .
these textures are characterized by two free parameters ( ignoring the phases ) so that there exist 3 relations among 5 oscillation parameters in each solution .
we formulate these relations by taking the two mass squared differences as input and 3 mixing angles as output parameters . in two out of the three cases
the elements of the pmns matrix are found to be given by irrational but simple algebraic numbers , to the leading order in the small parameter @xmath158 .
all the six solutions in table 2 have one physical phase .
we study the the possibility of obtaining leptogenesis in these models and explore if there is any connection between the phase responsible for generation of lepton asymmetry and the low energy cp phase .
we find that there is only one solution in which a connection between leptogenesis and low energy cp violation is possible ignoring radiative effects .
it is interesting to observe that the first 3 solutions in table 2 are consistent with the data only with the inverted mass hierarchy .
a priori , there is no reason that some texture must belong to a particular hierarchy .
the basic principles which we take are the equalities among matrix elements , texture zeros and minimality of the parameters .
thus we conclude that the minimal seesaw mechanism prefers the inverted hierarchy under the constraints which are likely to stem from physics beyond the standard model .
in fact , many authors have tried to explain the generation structure invoking discrete or other symmetries , where the equalities and vanishing elements in yukawa couplings are often realized as direct consequences of imposed flavor symmetries or secondary products of model constructions @xcite .
while the inverted hierarchy seems somewhat special in the sense that it shows sharp contrast to all the other fermions , the nature seems to be open to the inverted hierarchy in the context of hybrid texture .
the authors thank a. mohanty for a careful reading of the draft .
s.g . and a.w .
acknowledges support from the neutrino project under the xith plan of harish - chandra research institute .
in this appendix , we show the detailed classification of the dirac mass matrix @xmath7 . since @xmath7 has six entries , we can impose equalities on @xmath7 up to five .
we shall start with 1 equality in @xmath7 . by imposing 1 equality among 6 matrix elements ,
the 6 elements are divided into 5 groups , that is , for instance @xmath159 , and other 4 matrix elements .
this situation can be symbolized by @xmath160 , where each entry means the `` slot '' of the independent parameter .
since we impose 1 equality among 6 elements , the number of the independent parameters is reduced to 5 .
therefore we have 5 entries in @xmath160 .
the number of each entry in the first bracket denotes the number of the matrix elements included in each group .
the sum of the entries must be equal to 6 .
the @xmath160 case includes @xmath161 patterns of different textures .
the `` representatives '' are @xmath162 here `` representative '' means that the other patterns can be generated by the permutation of the rows and the columns from the above three matrices . in other words , the above three matrices are not related to each other by permutations of the rows and the column , so that they compose a set of `` primary '' matrices in this category . in this case , there are @xmath163 mass matrices . if we regard the first two groups of ( 2,2,1,1 ) as identical , then the total number is reduced to @xmath164 patterns .
the representatives are @xmath165 all 45 patterns can be generated from these 9 patterns . it should be noted again that we regard the textures which is related by the label exchange of the first two entries of ( 2,2,1,1 ) as identical .
the classification of the above 9 patterns is similar to the general possibilities for the 2 zero textures for @xmath7 .
we have @xmath166 general possibilities and three representatives in this category . @xmath167
all 20 patterns can be generated from these 3 patterns . an easy way to understand these 3 patterns comes from the analogy with the 3 zero textures in @xmath7 .
here we consider 3 equalities in @xmath7 .
since we have 3 equalities , the matrix elements are divided into 3 groups .
there are three types of distributions ; ( 3,2,1 ) , ( 4,1,1 ) and ( 2,2,2 ) .
let us see in turn .
there are @xmath163 patterns in this category .
it is helpful to remember the case of ( 2,2,1,1 ) in 2 equalities . this case
is obtained by imposing equalities between the last two entries of ( 2,2,1,1 ) . as in the case of ( 2,2,1,1 ) , we should identify the three entries of ( 2,2,2 ) .
then the total number is reduced to @xmath172 patterns .
the representatives are given by @xmath173 all 15 textures are generated from the above 5 representatives by the exchange of the rows and the columns .
there are @xmath177 patterns of textures in this case .
however 20 patterns contain redundancy .
we can reproduce all 20 patterns from fundamental 10 patterns by exchanging the two entries of ( 3,3 ) .
the 10 patterns can be obtained from the three representatives .
they can be taken as @xmath178 all 10 textures are generated from the above 3 representatives by the exchange of the rows and the columns . in this case , all the matrix elements in @xmath7 are equal and the resultant left - handed majorana mass matrix is of democratic form .
this provides two massless neutrinos together with a nonzero @xmath8 .
thus we can exclude @xmath7 with 5 equalities .
h. fritzsch , phys .
b * 73 * , 317 ( 1978 ) ; g. f. giudice , mod .
a * 7 * , 2429 ( 1992 ) [ arxiv : hep - ph/9204215 ] ; p. ramond , r. g. roberts and g. g. ross , nucl .
b * 406 * , 19 ( 1993 ) [ arxiv : hep - ph/9303320 ] ; g. c. branco and j. i. silva - marcos , phys .
b * 331 * , 390 ( 1994 ) ; t. k. kuo , s. w. mansour and g. h. wu , phys . rev .
d * 60 * , 093004 ( 1999 ) [ arxiv : hep - ph/9907314 ] ; h. fritzsch and z. z. xing , prog .
* 45 * , 1 ( 2000 ) [ arxiv : hep - ph/9912358 ] ; r. g. roberts , a. romanino , g. g. ross and l. velasco - sevilla , nucl .
b * 615 * , 358 ( 2001 ) [ arxiv : hep - ph/0104088 ] ; h. d. kim , s. raby and l. schradin , phys .
d * 69 * , 092002 ( 2004 ) [ arxiv : hep - ph/0401169 ] ; n. uekusa , a. watanabe and k. yoshioka , phys .
d * 71 * , 094024 ( 2005 ) [ arxiv : hep - ph/0501211 ] ; s. tatur and j. bartelski , phys .
d * 74 * , 013007 ( 2006 ) [ arxiv : hep - ph/0605261 ] .
p. h. frampton , s. l. glashow and d. marfatia , phys .
b * 536 * , 79 ( 2002 ) [ arxiv : hep - ph/0201008 ] .
z. z. xing , phys .
b * 530 * , 159 ( 2002 ) [ arxiv : hep - ph/0201151 ] ; z. z. xing , phys .
b * 539 * , 85 ( 2002 ) [ arxiv : hep - ph/0205032 ] ; a. merle and w. rodejohann , phys .
d * 73 * , 073012 ( 2006 ) [ arxiv : hep - ph/0603111 ] ; s. dev , s. kumar , s. verma and s. gupta , phys .
d * 76 * , 013002 ( 2007 ) [ arxiv : hep - ph/0612102 ] .
z. z. xing and h. zhang , phys .
b * 569 * , 30 ( 2003 ) [ arxiv : hep - ph/0304234 ] ; z. z. xing , int .
j. mod .
a * 19 * , 1 ( 2004 ) [ arxiv : hep - ph/0307359 ] ; s. zhou and z. z. xing , eur .
j. c * 38 * , 495 ( 2005 ) [ arxiv : hep - ph/0404188 ] ; z. z. xing and s. zhou , phys .
b * 593 * , 156 ( 2004 ) [ arxiv : hep - ph/0403261 ] ; m. randhawa , g. ahuja and m. gupta , phys . lett .
b * 643 * , 175 ( 2006 ) [ arxiv : hep - ph/0607074 ] .
g. k. leontaris , s. lola , c. scheich and j. d. vergados , phys .
d * 53 * , 6381 ( 1996 ) [ arxiv : hep - ph/9509351 ] ; s. m. barr and i. dorsner , nucl .
b * 585 * , 79 ( 2000 ) [ arxiv : hep - ph/0003058 ] ; a. kageyama , s. kaneko , n. shimoyama and m. tanimoto , phys .
b * 538 * , 96 ( 2002 ) [ arxiv : hep - ph/0204291 ] ; p. h. frampton , s. l. glashow and t. yanagida , phys .
b * 548 * , 119 ( 2002 ) [ arxiv : hep - ph/0208157 ] ; r. barbieri , t. hambye and a. romanino , jhep * 0303 * , 017 ( 2003 ) [ arxiv : hep - ph/0302118 ] ; a. ibarra and g. g. ross , phys .
b * 591 * , 285 ( 2004 ) [ arxiv : hep - ph/0312138 ] ; s. chang , s. k. kang and k. siyeon , phys .
b * 597 * , 78 ( 2004 ) [ arxiv : hep - ph/0404187 ] ; c. hagedorn and w. rodejohann , jhep * 0507 * , 034 ( 2005 ) [ arxiv : hep - ph/0503143 ] ; :1 a. watanabe and k. yoshioka , jhep * 0605 * , 044 ( 2006 ) [ arxiv : hep - ph/0601152 ] ; w. l. guo ,
z. z. xing and s. zhou , int .
j. mod .
e * 16 * , 1 ( 2007 ) [ arxiv : hep - ph/0612033 ] ; g. c. branco , d. emmanuel - costa , m. n. rebelo and p. roy , phys . rev .
d * 77 * , 053011 ( 2008 ) [ arxiv:0712.0774 [ hep - ph ] ] .
h. georgi and c. jarlskog , phys .
b * 86 * , 297 ( 1979 ) ; z. berezhiani and a. rossi , jhep * 9903 * , 002 ( 1999 ) [ arxiv : hep - ph/9811447 ] ; k. s. babu , j. c. pati and f. wilczek , nucl .
b * 566 * , 33 ( 2000 ) [ arxiv : hep - ph/9812538 ] ; k. matsuda , t. fukuyama and h. nishiura , phys .
d * 61 * , 053001 ( 2000 ) [ arxiv : hep - ph/9906433 ] ; m.
bando , t. kugo and k. yoshioka , prog .
phys . * 104 * , 211 ( 2000 ) [ arxiv : hep - ph/0003220 ] ; w. buchmuller and d. wyler , phys .
b * 521 * , 291 ( 2001 ) [ arxiv : hep - ph/0108216 ] ; m. c. chen and k. t. mahanthappa , phys .
d * 68 * , 017301 ( 2003 ) [ arxiv : hep - ph/0212375 ] ; s. raby , phys .
b * 561 * , 119 ( 2003 ) [ arxiv : hep - ph/0302027 ] ; m.
bando and m. obara , prog .
phys . *
109 * , 995 ( 2003 ) [ arxiv : hep - ph/0302034 ] .
t. fukuyama and h. nishiura , [ arxiv : hep - ph/9702253 ] ; r. n. mohapatra and s. nussinov , phys .
d * 60 * , 013002 ( 1999 ) [ arxiv : hep - ph/9809415 ] ; c. s. lam , phys . lett .
b * 507 * , 214 ( 2001 ) [ arxiv : hep - ph/0104116 ] ; w. grimus and l. lavoura , jhep * 0107 * , 045 ( 2001 ) [ arxiv : hep - ph/0105212 ] ; w. grimus and l. lavoura , acta phys .
b * 32 * , 3719 ( 2001 ) [ arxiv : hep - ph/0110041 ] ; p. f. harrison and w. g. scott , phys .
b * 547 * , 219 ( 2002 ) [ arxiv : hep - ph/0210197 ] ; e. ma , phys .
d * 66 * , 117301 ( 2002 ) [ arxiv : hep - ph/0207352 ] ; s. choubey and w. rodejohann , eur .
j. c * 40 * , 259 ( 2005 ) [ arxiv : hep - ph/0411190 ] ; c. s. lam , phys .
d * 71 * , 093001 ( 2005 ) [ arxiv : hep - ph/0503159 ] ; w. grimus , s. kaneko , l. lavoura , h. sawanaka and m. tanimoto , jhep * 0601 * , 110 ( 2006 ) [ arxiv : hep - ph/0510326 ] ; i. aizawa and m. yasue , phys . rev .
d * 73 * , 015002 ( 2006 ) [ arxiv : hep - ph/0510132 ] ; k. fuki and m. yasue , phys .
d * 73 * , 055014 ( 2006 ) [ arxiv : hep - ph/0601118 ] ; z.
z. xing , h. zhang and s. zhou , phys .
b * 641 * , 189 ( 2006 ) [ arxiv : hep - ph/0607091 ] ; n. haba and w. rodejohann , phys .
d * 74 * , 017701 ( 2006 ) [ arxiv : hep - ph/0603206 ] ; p. f. harrison , d. h. perkins and w. g. scott , phys .
b * 530 * , 167 ( 2002 ) [ arxiv : hep - ph/0202074 ] ; p. f. harrison and w. g. scott , phys .
b * 535 * , 163 ( 2002 ) [ arxiv : hep - ph/0203209 ] .
m. frigerio and a. y. smirnov , nucl .
b * 640 * , 233 ( 2002 ) [ arxiv : hep - ph/0202247 ] .
s. f. king , phys .
b * 439 * , 350 ( 1998 ) [ arxiv : hep - ph/9806440 ] ; s. f. king , nucl .
b * 562 * , 57 ( 1999 ) [ arxiv : hep - ph/9904210 ] ; r. kuchimanchi and r. n. mohapatra , phys . rev .
d * 66 * , 051301 ( 2002 ) [ arxiv : hep - ph/0207110 ] ; t. endoh , s. kaneko , s. k. kang , t. morozumi and m. tanimoto , phys .
lett . * 89 * , 231601 ( 2002 ) [ arxiv : hep - ph/0209020 ] ; m. raidal and a. strumia , phys .
b * 553 * , 72 ( 2003 ) [ arxiv : hep - ph/0210021 ] ; s. f. king , phys .
d * 67 * , 113010 ( 2003 ) [ arxiv : hep - ph/0211228 ] ; s. raby , phys .
b * 561 * , 119 ( 2003 ) [ arxiv : hep - ph/0302027 ] ; b. dutta and r. n. mohapatra , phys .
d * 68 * , 056006 ( 2003 ) [ arxiv : hep - ph/0305059 ] ; v. barger , d. a. dicus , h. j. he and t. j. li , phys .
b * 583 * , 173 ( 2004 ) [ arxiv : hep - ph/0310278 ] ; w. l. guo and z. z. xing , phys .
b * 583 * , 163 ( 2004 ) [ arxiv : hep - ph/0310326 ] ; w. rodejohann , eur .
j. c * 32 * , 235 ( 2004 ) [ arxiv : hep - ph/0311142 ] ; k. bhattacharya , n. sahu , u. sarkar and s. k. singh , phys .
d * 74 * , 093001 ( 2006 ) [ arxiv : hep - ph/0607272 ] ; b. brahmachari and n. okada , phys .
b * 660 * , 508 ( 2008 ) [ arxiv : hep - ph/0612079 ] ; r. barbieri , t. hambye and a. romanino ; a. ibarra and g. g. ross ; w. l. guo , z. z. xing and s. zhou ; s. chang , s. k. kang and k. siyeon in ref . @xcite .
p. minkowski , phys .
b * 67 * , 421 ( 1977 ) ; t. yanagida , in _ proceedings of the workshop on the unified theory and the baryon number in the universe _ ( o. sawada and a. sugamoto , eds . ) , kek , tsukuba , japan , 1979 , p. 95 ; m. gell - mann , p. ramond , and r. slansky , _ complex spinors and unified theories _ , in _
( p. van nieuwenhuizen and d. z. freedman , eds . ) , north holland , amsterdam , 1979 , p. 315 ; s. l. glashow , _ the future of elementary particle physics _
, in _ proceedings of the 1979 cargse summer institute on quarks and leptons _ ( m. lvy , j .- l .
basdevant , d. speiser , j. weyers , r. gastmans , and m. jacob , eds . ) , plenum press , new york , 1980 , pp .
687713 .
t. schwetz , m. tortola and j. w. f. valle , arxiv:0808.2016 [ hep - ph ] .
a. s. joshipura , eur .
j. c * 53 * , 77 ( 2008 ) [ arxiv : hep - ph/0512252 ] ; j. c. gomez - izquierdo and a. perez - lorenzana , phys .
d * 77 * , 113015 ( 2008 ) [ arxiv:0711.0045 [ hep - ph ] ]
. k. inoue and n. yamatsu , prog .
phys . * 119 * , 775 ( 2008 ) [ arxiv:0712.2938 [ hep - ph ] ] ; k. inoue and n. yamatsu , [ arxiv:0806.0213 [ hep - ph ] ] .
r. n. mohapatra and w. rodejohann , phys .
b * 644 * , 59 ( 2007 ) [ arxiv : hep - ph/0608111 ] ; a. blum , r. n. mohapatra and w. rodejohann , phys .
d * 76 * , 053003 ( 2007 ) [ arxiv:0706.3801 [ hep - ph ] ] .
n. haba , a. watanabe and k. yoshioka , phys .
lett . * 97 * , 041601 ( 2006 ) [ arxiv : hep - ph/0603116 ] ; t. kobayashi , y. omura and k. yoshioka , arxiv:0809.3064 [ hep - ph ] .
g. c. branco , t. morozumi , b. m. nobre and m. n. rebelo , nucl .
b * 617 * , 475 ( 2001 ) [ arxiv : hep - ph/0107164 ] .
j. bernabeu , g. c. branco and m. gronau , phys .
b * 169 * , 243 ( 1986 ) ; g. c. branco , m. n. rebelo and j. i. silva - marcos , phys .
b * 633 * , 345 ( 2006 ) [ arxiv : hep - ph/0510412 ] .
k. s. babu , y. meng and z. tavartkiladze , arxiv:0812.4419 [ hep - ph ] .
k. bhattacharya , n. sahu , u. sarkar and s. k. singh in @xcite .
m. tanimoto , phys .
d * 55 * , 322 ( 1997 ) [ arxiv : hep - ph/9605413 ] ; h. minakata and h. nunokawa , phys .
d * 57 * , 4403 ( 1998 ) [ arxiv : hep - ph/9705208 ] ; s. m. bilenky , c. giunti and w. grimus , phys .
d * 58 * , 033001 ( 1998 ) [ arxiv : hep - ph/9712537 ] ; j. arafune , m. koike and j. sato , phys .
d * 56 * , 3093 ( 1997 ) [ erratum - ibid .
d * 60 * , 119905 ( 1999 ) ] [ arxiv : hep - ph/9703351 ] . c. jarlskog , phys .
lett . * 55 * , 1039 ( 1985 ) .
p. h. chankowski and z. pluciennik , phys .
b * 316 * , 312 ( 1993 ) [ arxiv : hep - ph/9306333 ] ; k. s. babu , c. n. leung and j. t. pantaleone , phys .
b * 319 * , 191 ( 1993 ) [ arxiv : hep - ph/9309223 ] ; s. antusch , j. kersten , m. lindner and m. ratz , nucl .
b * 674 * , 401 ( 2003 ) [ arxiv : hep - ph/0305273 ] ; s. antusch , j. kersten , m. lindner , m. ratz and m. a. schmidt , jhep * 0503 * , 024 ( 2005 ) [ arxiv : hep - ph/0501272 ] . f. plentinger and w. rodejohann , phys .
b * 625 * , 264 ( 2005 ) [ arxiv : hep - ph/0507143 ] .
a. dighe , s. goswami and w. rodejohann , phys .
d * 75 * , 073023 ( 2007 ) [ arxiv : hep - ph/0612328 ] ; a. dighe , s. goswami and p. roy , phys .
d * 76 * , 096005 ( 2007 ) [ arxiv:0704.3735 [ hep - ph ] ] .
a. dighe , s. goswami and p. roy , phys .
d * 73 * , 071301 ( 2006 ) [ arxiv : hep - ph/0602062 ] .
g. bhattacharyya , a. raychaudhuri and a. sil , phys .
d * 67 * , 073004 ( 2003 ) [ arxiv : hep - ph/0211074 ] ; c. hagedorn , j. kersten and m. lindner , phys .
b * 597 * , 63 ( 2004 ) [ arxiv : hep - ph/0406103 ] ; m. honda , s. kaneko and m. tanimoto , jhep * 0309 * , 028 ( 2003 ) [ arxiv : hep - ph/0303227 ] .
the paper by c. hagedorn , j. kersten and m. lindner in @xcite . for typical example , n. haba and k. yoshioka , _ nucl .
phys . _ * b739 * , 254 ( 2006 ) [ arxiv : hep - ph/0511108 ] ; s. kaneko , h. sawanaka , t. shingai , m. tanimoto and k. yoshioka , _ prog . theor .
* 117 * , 161 ( 2007 ) [ arxiv : hep - ph/0609220 ] ; s. kaneko , h. sawanaka , t. shingai , m. tanimoto and k. yoshioka , [ arxiv : hep - ph/0703250 ] . | in the context of minimal seesaw framework , we study the implications of dirac and majorana mass matrices in which two rigid properties coexist , namely , equalities among mass matrix elements and texture zeros . in the first part of the study
, we discuss general possibilities of the dirac and majorana mass matrices for neutrinos with such hybrid structures .
we then classify the mass matrices into realistic textures which are compatible with global neutrino oscillation data and unrealistic ones which do not comply with the data . among the large number of general possibilities , we find that only 6 patterns are consistent with the observations at the level of the most minimal number of free parameters .
these solutions have only 2 adjustable parameters , so that all the mixing angles can be described in terms of the two mass differences or pure numbers .
we analyze these textures in detail and discuss their impacts for future neutrino experiments and for leptogenesis .
hri - p-08 - 11 - 004 + srubabati goswami@xmath0 , subrata khan@xmath1 , atsushi watanabe@xmath2 0.5 cm @xmath3 , + @xmath2 + |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.