1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: % INSTITUTE OF PHYSICS PUBLISHING %
3: % %
4: % `Preparing an article for publication in an Institute of Physics %
5: % Publishing journal using LaTeX' %
6: % %
7: % LaTeX source code `ioplau2e.tex' used to generate `author %
8: % guidelines', the documentation explaining and demonstrating use %
9: % of the Institute of Physics Publishing LaTeX preprint files %
10: % `iopart.cls, iopart12.clo and iopart10.clo'. %
11: % %
12: % `ioplau2e.tex' itself uses LaTeX with `iopart.cls' %
13: % %
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
15: %
16: %
17: % First we have a character check
18: %
19: % ! exclamation mark " double quote
20: % # hash ` opening quote (grave)
21: % & ampersand ' closing quote (acute)
22: % $ dollar % percent
23: % ( open parenthesis ) close paren.
24: % - hyphen = equals sign
25: % | vertical bar ~ tilde
26: % @ at sign _ underscore
27: % { open curly brace } close curly
28: % [ open square ] close square bracket
29: % + plus sign ; semi-colon
30: % * asterisk : colon
31: % < open angle bracket > close angle
32: % , comma . full stop
33: % ? question mark / forward slash
34: % \ backslash ^ circumflex
35: %
36: % ABCDEFGHIJKLMNOPQRSTUVWXYZ
37: % abcdefghijklmnopqrstuvwxyz
38: % 1234567890
39: %
40: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
41: %
42: \documentclass{iopart}
43: % Uncomment next line if AMS fonts required
44: \usepackage{iopams}
45: \usepackage{graphicx}
46: \def\be{\begin{equation}}
47: \def\ee{\end{equation}}
48:
49: \begin{document}
50:
51: \title{
52: Lattice calculations of meson correlators
53: and spectral functions at finite temperature
54: }
55:
56: \author{P\'eter Petreczky
57: \footnote[1]{
58: Goldhaber Fellow\\
59: This work has been authored under contract number DE-AC02-98CH10886 with
60: the U.S. Department of Energy
61: }
62: }
63: \address{
64: Nuclear Theory Group, Department of Physics, Brookhaven
65: National Laboratory, Upton, New York, 11973\\
66: email: petreczk@bnl.gov
67: }
68:
69:
70:
71: \begin{abstract}
72: I review recent progress in relating meson spectral
73: function to imaginary time correlation function at finite temperature
74: calculated on isotropic as well as on anisotropic
75: lattices. Special attention is payed for the lattice artifacts
76: present in calculation of meson spectral functions.
77: Results in the case of light quarks as well as heavy
78: quarks are reviewed which indicate in particular that
79: even in the chiral limit meson spectral functions have non-trivial
80: structure and the ground state quarkonia survive up to
81: temperature $1.5T_c$.
82: \end{abstract}
83:
84: %Uncomment for PACS numbers title message
85: \pacs{11.15.Ha, 11.10.Wx, 12.38.Mh, 25.75.Nq}
86:
87: % Uncomment for Submitted to journal title message
88: \submitto{\JPG}
89:
90: % Comment out if separate title page not required
91: %\maketitle
92:
93: \section{Introduction}
94: Spectral functions of mesonic operators play an important role
95: in finite temperature QCD. Many experimental results
96: in high energy heavy ion collisions ( e.g. low invariant mass
97: dilepton enhancement, anomalous $J/\psi$ suppression etc.)
98: can be understood in terms of medium modifications of meson
99: spectral functions \cite{rapp}.
100:
101: It is commonly believed that lattice QCD is capable only for calculation
102: of static quantities at finite temperatures, such as the transition temperature,
103: equation of state, screening lengths etc. However, it was shown by
104: Asakawa, Hatsuda and Nakahara that using the {\em Maximum Entropy Method}
105: one can in principle reconstruct also meson spectral functions.
106: The method was successfully applied at zero temperature \cite{asakawa01,cppacs}
107: and later also at finite temperature \cite{karsch02,wetzorke02,karsch03,
108: datta03,asakawa02,umeda2,wetzorke03}. Though systematic uncertainties in
109: the spectral function calculated on lattice are not yet completely
110: understood, it was shown in Ref. \cite{karsch02} that precise determination
111: of the imaginary time correlator can alone provide stringent constraints
112: on the spectral function at finite temperature.
113:
114: In this contribution I am going to review recent results on
115: meson correlators and spectral functions in finite temperature QCD.
116: The rest of the paper is organized as follows. In section 2 I will
117: discuss the relation between the imaginary time correlators and
118: meson spectral functions. In section 3 numerical results for the light
119: quarks will be discussed. Section 4 contains results on charmonia
120: at finite temperature. Finally conclusions are given in section 5.
121:
122:
123:
124:
125:
126: \section{Meson correlators and spectral function in continuum
127: and lattice QCD}
128:
129: In finite temperature field theory one usually considers two
130: types of real time correlation function
131: of some operator $\hat O$
132: \be
133: D^{>}(t,t')=\langle \hat O(t) \hat O(t') \rangle_T,~~~
134: D^{<}(t,t')=\langle \hat O(t') \hat O(t) \rangle_T,
135: \ee
136: where $\langle ... \rangle_T = \langle ... e^{-\hat{H}/T}
137: \rangle_T$ denotes the thermal average \cite{lebellac}.
138: The imaginary time correlation function which can be calculated also
139: by using lattice simulations is defined as simple analytic
140: continuation
141: \be
142: G(\tau)=\langle {\cal T} \hat O(-i \tau) \hat O(0) \rangle.
143: \label{rel1}
144: \ee
145: ( ${\cal T}$ stands for time ordered product).
146: The spectral function can be defined through the Fourier transform of
147: $D^{>(<)}(t)$ as
148: \be
149: \displaystyle
150: \sigma(\omega)=\frac{D^{>}(\omega)-D^{<}(\omega)}{2 \pi}=\frac{1}{\pi} Im D_R(\omega),
151: \ee
152: where $D_R(\omega)$ is the retarded correlator.
153: Using Eq. (\ref{rel1}) and KMS condition on $D^{>(<)}$ \cite{lebellac}
154: one can easily derive the following integral relation between the
155: imaginary time correlator and the spectral function
156: \be
157: \displaystyle
158: G(\tau)=\int_0^{\infty} d \omega \sigma(\omega) \frac{\cosh(\omega (\tau-1/(2T))}
159: {\sinh\frac{\omega}{2T}} \equiv \int_0^{\infty} d \omega \sigma(\omega) K(\omega,\tau)
160: \label{rel2}
161: \ee
162: Using this relation one can in principle reconstruct the spectral function
163: by calculating $G(\tau)$ on lattice.
164: At zero temperature the kernel, $K(\omega,\tau)$ reduces to simple
165: exponential and at large Euclidean times the correlation function
166: picks up the contribution from the lowest lying meson state in $\sigma(\omega)$,
167: i.e $G(\tau)=\exp(-m \tau)$. At finite temperature the analysis of the
168: asymptotic behavior of the correlators is no longer possible as $\tau$ is limited
169: to the interval $[0,~1/T]$ where excited states are equally important as the
170: ground state. Additional complications arise in lattice calculations where
171: correlators are calculated only at finite set of Euclidean times
172: $\tau T=k/N_{\tau}$, $k=0,...,N_{\tau}-1$
173: with $N_{\tau}$ being the temporal extent of the lattice.
174: In order to reconstruct the spectral functions from this
175: limited information it is necessary to include in the statistical
176: analysis of the numerical results also prior information on the
177: structure of $\sigma(\omega)$ (e.g. such as $\sigma(\omega)>0$ for
178: $\omega>0$). This can be done through the application of the
179: {\em Maximum Entropy Method} (MEM) \footnote{Other methods of
180: introducing prior information into the statistical analysis have
181: been also discussed in Ref. \cite{lepage} for the zero temperature
182: case and in Ref. \cite{gupta-v} for finite temperature QCD.}.
183:
184: In order to study meson properties at finite temperature appropriate choice
185: of the operator $\hat O$ should be made. One possible choice is local
186: meson operator bilinear in quark-antiquark fields (current)
187: \cite{asakawa01,cppacs,karsch02,wetzorke02}
188: \be
189: O_H(\tau,\vec{x})=Z_H \bar q(\tau,\vec{x}) \Gamma_H q(\tau,\vec{x}),
190: ~~~\Gamma_H=1, \gamma_5, \gamma_{\mu},
191: \gamma_5 \gamma_{\mu}
192: \label{pointop}
193: \ee
194: for scalar, pseudoscalar, vector and axial vector channels correspondingly.
195: The normalization constant $Z_{H}$ relates the current calculated
196: in lattice regularization scheme to current in $\overline{MS}$ scheme with
197: $\mu_{\overline{MS}}=a^{-1}$ (with $a$ being the lattice spacing).
198: We can define then the temporal correlators at finite spatial momentum
199: $\vec{p}$
200: \be
201: G_{H}(\tau,\vec{p})=\langle O_H(\tau,\vec{p}) O_H^{\dagger}(\tau,-\vec{p}) \rangle,
202: O(\tau,\vec{p})=\sum_{\vec{x}} e^{i \vec{p} \vec{x}} O_H(\tau,\vec{x})
203: \ee
204: One can also consider correlators of extended operators defined as \cite{umeda2,umeda1}
205: \be
206: \tilde O_H(\tau,\vec{x})=\sum_{\vec{y}} \phi(\vec{y}) \bar q(\tau,\vec{x}) \Gamma
207: q(\tau,\vec{x}+\vec{y}).
208: \ee
209: Here
210: \be
211: \phi(\vec{y})=\exp(-b |\vec{y}|^p)
212: \label{triwf}
213: \ee
214: is the trial wave function
215: which controls the size of the meson source and can be regarded as a
216: physical input to the problem. The use of extended operators makes
217: the reconstruction of the spectral function easier as the correlator is
218: dominated by a single peak in the spectral function \cite{umeda2}.
219: However, the corresponding spectral function is not related to a physically
220: observable quantity, it can provide information only about the mass
221: and the width of the resonance but not about the corresponding decay constant.
222: This also only works for sharp resonances.
223: At high temperature "mesons" will appear as broad structures in the
224: spectral function and the above method is no longer applicable.
225:
226: Though the in-medium properties of mesons are encoded primarily in the
227: temporal correlator and the corresponding spectral function,
228: in finite temperature QCD it is customary to study spatial correlators
229: defined as
230: \be
231: G_{H}=\langle O(z) O^{\dagger}(0) \rangle,~~~
232: O(z)=\sum_{\tau,x,y} O(\tau,x,y,z).
233: \ee
234: Since the lattice extent is not limited in the spatial directions
235: one can study their large distance behavior. At large distances
236: the spatial correlators decay exponentially and the exponential
237: decay governed by the so-called screening mass. The spatial
238: correlators can be related to the spectral function via
239: following relation
240: \be
241: \hspace*{-2.1cm}
242: G_{H}(z)=\int_{-\infty}^{\infty} d p_z e^{i p_z z}
243: \int_0^{1/T} G(\tau,0,0,p_z)=\int_{-\infty}^{\infty} d p_z e^{i p_z z}
244: \int_0^{\infty} d \omega \frac{\sigma(\omega,0,0,p_z)}{\omega}
245: \ee
246: In general the screening masses are different from the pole
247: masses, however, from the above equation one can easily see
248: that in the special case when the spectral function is dominated by a single
249: $\delta$-function for small $\omega$ the screening and the pole
250: masses are equal.
251:
252: \begin{figure}
253: \hspace*{-1cm}
254: \includegraphics[width=7cm]{fSPF.eps}
255: \includegraphics[width=7cm]{fSPFNt.eps}
256: \caption{The ratio of the lattice and continuum spectral functions
257: in different channels (left) versus $\omega a$ and the lattice pseudoscalar
258: spectral function versus $\tilde \omega=\omega/T$ for different
259: values of $N_{\tau}$ calculated using Wilson fermions.
260: See Ref. \cite{karsch03a} for further details.}
261: \label{latspf}
262: \end{figure}
263: The relation between the imaginary time correlator and
264: the spectral function (\ref{rel2}) is valid only in the continuum theory.
265: As the correlators are calculated on lattice the question arises
266: whether the spectral representation of the form (\ref{rel2}) can
267: be derived also for the lattice correlator.
268: In ref. \cite{karsch03a} it was shown that this is indeed the case
269: in the free theory and that the cutoff effects present in the
270: correlator at small separations are contained in the spectral
271: function. Because of the asymptotic freedom calculation in the free
272: theory are relevant at high temperature and/or large energies $\omega$.
273: In Fig. \ref{latspf} I show the ratio of of the lattice spectral function
274: to the corresponding continuum spectral function versus $a \omega$
275: as well as the lattice pseudoscalar spectral function versus
276: $\omega/T$ in the free theory. As one can see from the figures the
277: lattice spectral functions start to deviate from the continuum one for
278: $\omega a>0.5$ and in the pseudoscalar and vector channels show
279: a peak like structure around $\omega a \simeq 2$. The lattice
280: spectral functions vanishes for $\omega a > \omega_{max} a \sim 4 N_{\tau}$.
281: As the temporal extent of the lattice $N_{\tau}$ is increased for
282: fixed temperature, i.e. the lattice spacing decreases ($a=1/(N_{\tau} T$))
283: the peak structure moves to larger value of $\omega/T$ (see Fig. \ref{latspf}).
284: Peak structure around $\omega a \simeq 2$ was also observed in the
285: interacting theory \cite{cppacs}.
286: The main reason for the large cutoff effects in the lattice spectral function is
287: that on lattice the quark dispersion relation gets modified
288: at large momenta, $pa >1$, relative to its continuum counterpart.
289: Cutoff effects in the spectral
290: function can be substantially reduced by using an improved fermion action,
291: the so-called truncated perfect action on hypercube \cite{biet96}
292: for which the lattice
293: quark dispersion relation is much closer to the continuum even for
294: $pa >1$ \cite{karsch03a}.
295:
296: \section{Numerical results in the light quark sector}
297:
298: Meson correlators and spectral functions for small
299: quark masses were intensively studied during recent
300: years \cite{karsch02,wetzorke02,karsch03,qcdtaro}.
301: In Refs. \cite{karsch02,wetzorke02,karsch03}
302: spectral functions were studied using non-perturbatively
303: ${\cal O}(a)$ improved Wilson action \cite{luescher}
304: and isotropic $32^3 \times 16$,
305: $48^3 \times 12$, $64^3 \times 16$ and $64^3 \times 24$ lattices.
306: In these studies quark masses corresponding to the pion masses
307: in the range 400MeV to 1Gev were considered in the confined phase,
308: while in the deconfined phase calculation were performed in the
309: chiral limit and the temperature interval was $0.4T_c-3T_c$ with
310: $T_c$ being the deconfinement temperature. The renormalization
311: factors $Z_{H}$ appearing in Eq. \ref{pointop} were
312: determined non-perturbatively in Ref. \cite{luescher} for
313: the vector and axial vector channels, while for the scalar
314: and pseudoscalar channels they were calculated using tadpole
315: improved perturbation theory.
316: \begin{figure}
317: \includegraphics[width=8cm]{ratall_above.eps}
318: \caption{The ratio of meson correlators at $1.5T_c$ in
319: different channels to the corresponding correlators in the
320: free theory calculated on $64^3 \times 16$ lattice.}
321: \label{restor}
322: \end{figure}
323: Asakawa, Hatsuda and Nakahara have studied meson spectral functions
324: using $32^3 \times N_{\tau}$ anisotropic lattices with
325: $a_{\tau}/a_{\sigma}=4$ and $N_{\tau}=96-32$ corresponding to temperatures
326: $0.8T_c-3T_c$ \cite{asakawa02}. The quark masses used in this
327: study correspond to $m_{\pi}/m_{\rho}=0.7$.
328: Below the deconfinement temperature neither the correlators nor
329: the spectral functions indicate in-medium change of meson properties
330: contrary to existing theoretical predictions \cite{rapp}. This
331: is likely due to the quenched approximation and large quark masses
332: used in these studies.
333: \begin{figure}
334: \hspace*{-1cm}
335: \includegraphics[width=7cm]{rat_ps_above.eps}
336: \includegraphics[width=7cm]{rat_v1234_above.eps}
337: \caption{The ratio of meson correlators at $1.5T_c$ and $3T_c$ in
338: the pseudoscalar (left) and vector (right) channels
339: to the corresponding correlators in the
340: free theory calculated on lattice.}
341: \label{corpsvc}
342: \end{figure}
343:
344: Before discussing the results on the spectral function
345: above deconfinement let us first
346: look at the correlators. The correlation functions
347: alone can provide some constraints on the spectral functions,
348: for example the analysis of the vector correlation function
349: provides stringent constraint on the thermal dilepton rate
350: \cite{karsch02}. At high temperature one would expect that the
351: quark propagators should be close to the free ones. Therefore
352: in what follows we will always normalize the meson
353: correlators by the corresponding lattice correlators in the free theory.
354: We will restrict the discussion to the case of zero spatial momentum $\vec{p}=0$.
355: In Fig. \ref{restor} I show the meson correlators at $1.5T_c$ in different
356: channels calculated on $64^3 \times 16$ lattice. As one can see from
357: the figure the scalar-pseudoscalar and vector-axial-vector correlators
358: become degenerate except at very small $\tau T$ where lattice artifacts
359: present in the Wilson formulation explicitly break chiral symmetry.
360: This can be viewed as an indication of the $U_A(1)$ symmetry restoration at
361: high temperature. In Fig. \ref{corpsvc} I show the vector and pseudoscalar
362: correlators for $1.5T_c$ and $3T_c$ for different values of $N_{\tau}$.
363: From the figure it is evident that the vector correlator stays close
364: to its free value even at $1.5T_c$ while the pseudoscalar correlator is
365: considerably enhanced compared to the corresponding free correlators.
366: Note also that in the vector channel the correlators is $N_{\tau}$
367: (i.e. lattice spacing) independent, while
368: in the pseudoscalar case there is a small $N_{\tau}$ dependence
369: which is probably due to the perturbative error in calculation of
370: $Z_{PS}$. In Fig. \ref{spfpsvc} I show the spectral functions
371: for the pseudoscalar and vector channels. The spectral functions
372: in both channels show a peak like structure at $\omega \simeq (5-6)T$
373: which is more pronounced in the
374: pseudoscalar case and probably leads to the large enhancement
375: of the pseudoscalar correlator over the free case discussed above.
376: Note that the position of the peak appears to be proportional to the temperature.
377: Spectral function calculated on anisotropic lattice \cite{asakawa01} are shown in Fig. \ref{asak}
378: and exhibit a similar peak like structure,
379: roughly for the same value of $\omega/T$. However, at lower temperature,
380: $T=1.4T_c$ the corresponding structure in the spectral functions appears
381: to be more sharp (see Fig. \ref{asak}).
382: \begin{figure}
383: \hspace*{-1cm}
384: \includegraphics[width=7cm]{ps_above.eps}
385: \includegraphics[width=7cm]{v1234_above.eps}
386: \caption{The spectral functions in the deconfined phase for
387: the pseudoscalar channel (left) and the vector channel (right)
388: reconstructed using MEM on $64^3 \times 16$ lattice.}
389: \label{spfpsvc}
390: \end{figure}
391: \begin{figure}
392: \hspace*{-1cm}
393: \includegraphics[width=7cm]{asak1.eps}
394: \includegraphics[width=7cm]{asak2.eps}
395: \caption{The spectral functions $\rho(\omega)=\sigma(\omega)/\omega^2$
396: in different channels calculated
397: at $1.4T_c$ (left) and at $1.9T_c$ (right) using anisotropic lattices \cite{asakawa01}
398: Note that normalization of the spectral functions is different from the case on
399: isotropic lattices.}
400: \label{asak}
401: \end{figure}
402:
403: Meson correlators in the spatial directions and the corresponding
404: screening masses have been studied since long time
405: both in full and quenched QCD
406: \cite{de86,de87,born91,bern92,gupta00,laermann,gupta-over,gavai03,prog}.
407: However, reliable determination of
408: the screening masses is available only in quenched approximation
409: \cite{laermann,gavai03,prog} as lattices with temporal extent $N_{\tau}\ge 8$
410: are needed in such analysis \cite{gavai03}. The most recent results
411: for the screening masses are summarized in Fig. \ref{spat1}.
412: The screening masses are close to their asymptotic
413: ($T \rightarrow \infty $) value $2\pi T$ already at temperature for $T=1.5T_c$.
414: This could lead to the conclusion that there is an almost free quark propagation
415: in spatial directions in the deconfined phase at temperatures as low
416: as $1.5T_c$. However, a closer look on the spatial correlators in
417: Fig. \ref{spat1} reveals that this is not the case even at $3T_c$.
418: The correlators calculated by using lattice simulations are several times larger
419: than the corresponding free ones. This is not unexpected since it
420: is well known that the physics at distances $z>1/(g^2 T)$
421: is non-perturbative even at very high
422: temperature \cite{linde}.
423: In Fig. \ref{spat2} I show the vector and pseudoscalar spatial
424: correlators at short distances together with the temporal one.
425: As one can see from the figures at small separations temporal
426: and spatial correlators show very similar behavior.
427: \begin{figure}
428: \hspace*{-1cm}
429: \includegraphics[width=7cm]{meson_scmass.eps}
430: \includegraphics[width=7cm]{rat_spat.eps}
431: \caption{The screening masses (left) and spatial correlators (right) for
432: vector and pseudoscalar channels at $T>T_c$. The screening masses have been
433: calculated in quenched QCD using improved Wilson fermions \cite{laermann,prog}
434: and Kogut-Susskid fermions \cite{gavai03}.}
435: \label{spat1}
436: \end{figure}
437: \begin{figure}
438: \hspace*{-1cm}
439: \includegraphics[width=7cm]{rat_tz_ps.eps}
440: \includegraphics[width=7cm]{rat_tz_v1234.eps}
441: \caption{
442: The ratio of spatial and temporal correlators in the
443: deconfined phase to the corresponding free correlators for
444: the pseudoscalar (left) and vector (right) channels.
445: }
446: \label{spat2}
447: \end{figure}
448:
449: \section{Heavy quarkonia correlators and spectral functions}
450: Heavy quarkonia correlators and spectral functions were studied
451: in Refs. \cite{umeda2,umeda1} using anisotropic lattices
452: and extended operators as well as on isotropic lattices and point
453: correlators \cite{datta03}. In Ref. \cite{datta03} pseudoscalar, vector,
454: scalar and axial vector channels were considered which correspond
455: to $^1 S_0$ ($\eta_c$), $^3 S_1$ ($J/\psi$), $^3P_0$ ($\chi_{c0}$)
456: and $^3P_1$ ($\chi_{c1}$) charmonia states respectively.
457: As in the light quark sector some statements about in-medium
458: properties of charmonia can be made just by analyzing the
459: behavior of the correlators.
460: \begin{figure}
461: \hspace*{-1cm}
462: \includegraphics[width=7cm]{psvc.eps}
463: \includegraphics[width=7cm]{scax.eps}
464: \caption{The ratio of charmonia correlators to the
465: corresponding free meson correlators in pseudoscalar and vector
466: channels (left) as well as for scalar and axial vector (right) channels
467: .}
468: \label{heavycor}
469: \end{figure}
470: In Fig. \ref{heavycor} I show
471: the ratio of different quarkonia correlators to the corresponding
472: free correlators. As one can see from the figures the
473: correlators in the pseudoscalar and vector channels show very little
474: change across $T_c$ while in the case of the scalar and axial vector
475: channels larger changes in the correlators are visible. This
476: implies that the ground state charmonia ($^1 S_0$ and $^3 S_1$)
477: are very likely to survive in the deconfined phase.
478: The spectral function reconstructed using MEM are shown in Fig.
479: \ref{heavyspf} which shows indeed the the ground state peak
480: survives in the deconfined phase while the excited $^3P_0$ state
481: is dissociated already at $T=1.25T_c$.
482: Spectral functions were also reconstructed for $1.5T_c$ and
483: and show again that the ground state charmonia survive while
484: the excited P-states are dissociated.
485: \begin{figure}
486: \hspace*{-1cm}
487: \includegraphics[width=7cm]{ps_heavy.eps}
488: \includegraphics[width=7cm]{ax_heavy.eps}
489: \caption{Spectral function in the pseudoscalar
490: (left) and axial vector channels (right) reconstructed using MEM.
491: }
492: \label{heavyspf}
493: \end{figure}
494: Ground state charmonia ( $^1 S_0$, $^3 S_1$ states) spectral functions
495: were also studied in Ref. \cite{umeda2} using extended operators.
496: The results for the vector channel are shown in Fig. \ref{umedafig1}.
497: The $J/\psi$ mass (the peak position) changes very little across $T_c$
498: while the corresponding peak in the spectral function gets broader above
499: $T_c$. This broadening of the ground state peak is also visible
500: in calculations with local operators. However, such broadening
501: is not necessarily a physical effect and may come from the fact that
502: less data points $N_{\tau}$ are available at higher temperature
503: \cite{asakawa01} and also the temporal extent of the lattice
504: in physical units become smaller \cite{umeda2}. The authors
505: of Ref. \cite{umeda2} have found that the width of the peak
506: in the spectral function also depends on the trial wave function
507: $\phi(y)$ used in constructing the extended operator (see Fig. \ref{umedafig1}),
508: the peak is broader for larger value of $b$ in the trial wave
509: function $\phi(\vec{y})$. Therefore
510: MEM cannot yet provide a reliable estimate of the width of the
511: $J/\psi$ in the plasma. To overcome this difficulty $\chi^2$ -fits
512: to the Breit-Wigner form of the spectral function was used \cite{umeda2}.
513: This led to a first hints for a non-zero
514: widths of charmonium states above $T_c$.
515:
516:
517:
518: \begin{figure}
519: \includegraphics[width=10cm]{umeda_proc.eps}
520: \caption{
521: The $J/\psi$ spectral function reconstructed from the
522: correlation function of extended operators using MEM \cite{umeda2}
523: for $0.9T_c$ and $1.1T_c$ and different input wave functions
524: (different values of $b$).
525: }
526: \label{umedafig1}
527: \end{figure}
528:
529: The existence of ground state charmonia at $1.5T_c$ is in sharp
530: disagreement with predictions of potential models based on color
531: screening \cite{mehr,digal01}. A possible reason for this
532: discrepancy is an oversimplified picture of color screening used
533: in these studies. In fact recent analysis of the free energy of
534: quark-antiquark pair suggest quite complicated medium
535: dependence of inter-quark forces \cite{okacz02,okacz03,digal03}.
536:
537:
538: \section{Conclusions}
539: In this paper recent results on meson spectral functions and
540: correlators calculated on lattice were discussed.
541: Though systematic uncertainties are present in the
542: spectral functions the results appear to be very interesting and
543: even intriguing to some extent.
544: The presence of the peak like structures in the meson spectral
545: functions well above deconfinement temperature is, too large
546: extent, unexpected. These peak structures seems to be present
547: in the spectral functions calculated on isotropic lattices
548: \cite{karsch02,wetzorke02,datta03} as well as on anisotropic
549: lattices \cite{asakawa02,umeda2} and is some case their
550: presence is supported by direct inspection of the corresponding
551: correlation functions (c.f. Fig. \ref{corpsvc} (left) and Fig. \ref{heavycor}).
552: In the case of heavy quarkonia the interpretation of the peaks in the
553: spectral function is clear, they correspond to the quarkonia states
554: which survive in the deconfined phase up to temperatures $1.5T_c$.
555: We have seen that the position of these peaks does not change considerably
556: compared to the zero temperature case. The interpretation of peak-like
557: structures in the light quark sector is less evident as the position
558: of these peaks seems to be proportional to the temperature.
559: Certainly much more work is needed in order to be able
560: to interpret the present findings in physical terms as well as
561: for the detailed understanding of the systematic uncertainties involved in extraction of the
562: spectral functions from lattice correlators.
563:
564: \vspace*{0.5cm}
565:
566: \ack
567: Many results presented in the manuscript were obtained by
568: S. Datta, F. Karsch, E. Laermann, S. Stickan and I. Wetzorke and myself.
569: I would like to thank my colleagues for the collaboration. I would like
570: to thank S. Stickan and F. Karsch for careful reading of the manuscript. Finally
571: e-mail correspondence from T. Yamazaki, M. Asakawa and T. Umeda regarding
572: their results is gratefully acknowledged.
573:
574: \vspace*{0.5cm}
575:
576: \begin{thebibliography}{99}
577: \bibitem{rapp}
578: Rapp R, Wambach J 2000, {\em Adv. Nucl. Phys. } {\bf 25} 1; Rapp R, Grandchamp L, hep-ph/0305143
579: \bibitem{nakahara99}
580: Nakahara Y, Asakawa M and Hatsuda T 1999, \PR {\bf 60} 091503
581: \bibitem{asakawa01}
582: Asakawa M, Hatsuda T, Nakahara Y 2001, {\em Prog. Part. Nucl. Phys.} {\bf 46} 459
583: \bibitem{cppacs}
584: CP-PACS Collaboration, Yamazaki T 2001, \PR {\bf D65} 014501
585: \bibitem{karsch02}
586: Karsch F et al 2002, \PL {\bf B530} 147
587: \bibitem{wetzorke02}
588: Wetzorke I et al 2002, {\em Nucl. Phys. B (Proc. Suppl.)} {\bf 106} 513
589: \bibitem{karsch03}
590: Karsch F et al 2003, {\NP} {\bf A715} 701c
591: \bibitem{datta03}
592: Datta S et al 2003, {\em Nucl. Phys. B (Proc. Suppl.) } {\bf 119} 487 (hep-lat/0208012)
593: \bibitem{asakawa02}
594: Asakawa Y, Hatsuda T, Nakahara Y, hep-lat/0208059
595: \bibitem{umeda2}
596: Umeda T, Nomura K, Matsufuru H, hep-lat/0211003
597: \bibitem{wetzorke03}
598: Wetzorke I, hep-lat/0305012
599: \bibitem{lebellac}
600: Le Bellac M 1996, {\em Thermal Field Theory} ( Cambridge University Press )
601: \bibitem{lepage}
602: Lepage G P et al 2002, {\em Nucl. Phys. B ( Proc. Suppl.)} {\bf 106} 12
603: \bibitem{gupta-v}
604: Gupta S, hep-lat/0301006
605: \bibitem{umeda1}
606: Umeda T et al 2001, {\em Int. J. Mod. Phys. } {\bf A16} 2215
607: \bibitem{karsch03a}
608: Karsch F et al, hep-lat/0303017
609: \bibitem{biet96}
610: Bietenholz W et al 1997,
611: {\em Nucl. Phys. B (Proc. Suppl.)} {\bf 53} 921
612: \bibitem{qcdtaro}
613: QCD-TARO Collaboration, de Forcrand Ph et al 2001, \PR {\bf D63} 054501
614: \bibitem{luescher}
615: L\"uscher M et al 1997, \NP {\bf B491} 344
616: \bibitem{de86}
617: DeGrand T A, DeTar C E 1986, \PR {\bf D34} 2469
618: \bibitem{de87}
619: DeTar C E, Kogut J B 1987, \PR {\bf D36} 2828
620: \bibitem{born91}
621: Born K D et al 1991, \PRL {\bf 67} 302
622: \bibitem{bern92}
623: Bernard C et al 1992, \PRL {\bf 68} 2125
624: \bibitem{laermann}
625: Laermann E, Schmidt P 2001, {\em Eur. Phys. J.} {\bf C20} 541
626: \bibitem{gupta00}
627: Gavai R V, Gupta S 2000, \PRL {\bf 85} 2068
628: \bibitem{gupta-over}
629: Gavai R V, Gupta S 2002, \PR {\bf D65} 094504
630: \bibitem{gavai03}
631: Gavai R, Gupta S 2003, \PR {\bf D67} 034501
632: \bibitem{prog}
633: F Karsch et al, work in progress
634: \bibitem{linde}
635: Linde A 1980, \PL {\bf B96} 289
636: \bibitem{mehr}
637: Karsch F, Mehr M T, Satz H 1988, \ZP {\bf C37} 617
638: \bibitem{digal01}
639: Digal S, Petreczky P, Satz H (2001), \PR {\bf D64} 094015
640: \bibitem{okacz02}
641: Kaczmarek 0 et al 2002, \PL {\bf B543} 41
642: \bibitem{okacz03}
643: Zantow F et al, hep-lat/0301015
644: \bibitem{digal03}
645: Digal S, Fortunato S, Petreczky P, hep-lat/0304017
646: \end{thebibliography}
647: \end{document}
648:
649:
650:
651:
652:
653:
654:
655:
656:
657:
658:
659:
660: