1: \def\defstyle#1#2{#1}
2: %!!!Uncomment the following row to make revtex styled paper
3: %\def\defstyle#1#2{#2}
4:
5: %GREEK LETTERS
6: \def\a{\alpha}\def\b{\beta}\def\g{\gamma}\def\d{\delta}\def\e{\epsilon}
7: \def\E{\varepsilon} \def\k{\kappa}\def\l{\lambda}\def\L{\Lambda}
8: \def\s{\sigma}\def\S{\Sigma} \def\Th{\Theta}\def\th{\theta}\def\om{\omega}
9: \def\Om{\Omega}\def\G{\Gamma} \def\und{\underline}\def\r{\rho}\def\n{\nu}
10: \def\m{\mu}\def\t{\tilde} \def\o{\overline} \def\z{\zeta}
11: \def\P{\Psi}\def\Ph{\Phi} \def\ph{\phi}
12:
13: \def\imo{i}
14:
15: \def\be{\begin{equation}}
16: \def\ee{\end{equation}}
17: \def\bea{\begin{eqnarray}}
18: \def\eea{\end{eqnarray}}
19:
20: \def\Re#1{\mathrm{Re}(#1)}
21: \def\Im#1{\mathrm{Im}(#1)}
22:
23: \newcommand{\p}[1]{(\ref{#1})}
24: \let\pd\partial
25: \def\mysection#1{{\bf #1}}
26: %%%%%%%%%%%%%%%%%%%%%%%%%%
27:
28: \defstyle{%
29: \documentclass[10pt]{article}
30: \def\auth#1{{\large #1}}
31: \def\affiliation#1{{\\\vspace{2mm}\it #1\\\vspace{5mm}}}
32: \def\mailbox#1{\footnote{Electronic address: #1}}
33: }{
34: \documentclass[twocolumn]{revtex4}
35: \let\auth\author
36: \let\mailbox\email
37: }%
38:
39: \usepackage[dvips]{graphicx}
40: \topmargin=-1cm\textheight=24cm\textwidth=18cm\oddsidemargin=-1cm\evensidemargin
41: =-1cm
42: \begin{document}
43: \hyphenation{Schwar-zschild}
44:
45: \title{High overtones of Dirac perturbations of a Schwarzschild black hole}
46:
47: \defstyle{%
48: \maketitle
49: \vspace{5mm}
50: \renewcommand{\thefootnote}{\fnsymbol{footnote}}
51: \begin{center}
52: %\vspace{-3cm}
53: %(This title style does not coincide with RevTeX styled one. But contains all the fields.)\\
54: %\vspace{3cm}
55: }{%
56: }%
57: \auth{K. H. C. Castello-Branco}
58: \mailbox{karlucio@fma.if.usp.br}
59: \affiliation{Universidade de S\~ao Paulo, Instituto de F\'\i sica
60: \\ Caixa Postal 66318, 05315-970, S\~ao Paulo-SP, Brazil.}
61:
62:
63: \auth{R. A. Konoplya}
64: \mailbox{konoplya@fma.if.usp.br}
65: \affiliation{Universidade de S\~ao Paulo, Instituto de F\'\i sica
66: \\
67: Caixa Postal 66318, 05315-970, S\~ao Paulo-SP, Brazil.}
68:
69: \auth{A. Zhidenko}
70: \defstyle{\mailbox{Z\_A\_V@ukr.net}}{\mailbox{Z_A_V@ukr.net}}
71: \affiliation{Department of Physics, Dniepropetrovsk National University
72: \\
73: St. Naukova 13, Dniepropetrovsk 49050, Ukraine.}
74:
75: \defstyle{%
76: \maketitle
77: \end{center}
78: \vspace{2mm}
79: \renewcommand{\thefootnote}{\arabic{footnote}}\setcounter{footnote}0
80: }{%
81: }%
82: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
83:
84:
85: \begin{abstract}
86: Using the Frobenius method, we find high overtones of the Dirac
87: quasinormal spectrum for the Schwarzschild black hole. At high
88: overtones, the spacing for imaginary part of $\omega_{n}$ is
89: equidistant and equals to $\Im{\omega_{n+1}}-\Im{\omega_{n}}
90: =i/8M$, ($M$ is the black hole mass), which is twice less than
91: that for fields of integer spin. At high overtones, the real
92: part of $\omega_{n}$ goes to zero. This supports the suggestion
93: that the expected correspondence between quasinormal modes and
94: Barbero-Immirzi parameter in Loop Quantum Gravity is
95: just a numerical coincidence.
96: \end{abstract}
97:
98: \defstyle{%
99: \thispagestyle{empty}
100: %\newpage
101: \twocolumn
102: }{%
103: \maketitle
104: }%
105:
106:
107: The Quasinormal Mode (QNM) spectrum is an important characteristic
108: of a black hole. It dominates the late time response of a black
109: hole to an external perturbation, and, at the same time, does not
110: depend on the way of their excitation. Thus, being dependent on
111: black hole parameters only, the QNMs provide us with the
112: "fingerprints" of a black hole, feasible to be seen in the
113: detection of gravitational waves (for reviews, see
114: \cite{kokkotas-review}).
115:
116:
117: The importance of the QN spectrum is not limited by the above
118: observational aspects of gravitational waves phenomena, since QNMs
119: have interpretation in conformal field theory through AdS/CFT
120: and dS/CFT correspondence \cite{Birmingham2}. There is also
121: suggestion that the asymptotic QNMs of black holes are connected
122: with the so-called Barbero-Immirzi parameter in Loop Quantum
123: Gravity, which must be fixed to reproduce the Bekenstein-Hawking
124: entropy formula within this theory \cite{Dreyer}.
125: All this stimulated considerable interest to study the QNMs of
126: black holes in flat, dS and AdS backgrounds \cite{kucha-mala}.
127: In particular, the QNMs of the Dirac field for different black holes were considered in the
128: papers \cite{first-Cardoso}, \cite{Cho},
129: \cite{Zhidenko}, \cite{charged-dirac}. Nevertheless, the study of
130: the Dirac QNMs of four-dimensional black holes were limited
131: by $low~ overtones$. The low overtones were found for a Schwarzschild black hole in
132: \cite{Cho}, with the help of the third-order WKB method
133: \cite{IyerWill}. There it
134: was observed that for low overtones the real part of the
135: frequency decreases as the damping grows. Soon, the analysis was
136: extended to the case of Schwarzschild-de Sitter black hole
137: \cite{Zhidenko} using the sixth-order WKB method
138: \cite{KonoplyaWKB6}, and also to the case of the charged Dirac
139: field in the background of a charged black hole
140: \cite{charged-dirac}.
141:
142: In the present paper, we are trying to cover this gap in the study
143: of the Dirac QNMs and shall investigate the $high$ overtone
144: behavior of the Dirac perturbations of a Schwarzschild black hole.
145: We observed that at high overtones the imaginary part of the QN
146: spectrum is equidistant with the spacing
147: $\Im{\omega_{n+1}}-\Im{\omega_{n}} =i/8M$. It is also shown that
148: when the overtone number $n$ is increasing, the real part of
149: $\omega$ approaches zero.
150:
151: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
152:
153: The Dirac equation in an arbitrary curved background has the form
154: \cite{Birell}:
155: \be\label{Dirac}%
156: (\g^ae_a^{~\m}(\pd_\m+\G_\m)+m)\P=0,
157: \ee%
158: where $m$ is the mass of the Dirac field, $e_a^{~\m}$ are
159: tetrads, $\G_\m$ are spin connections \cite{Birell}. The
160: Schwarzschild metric has the form:
161: \bea\label{metric}%
162: ds^2 = f(r)dt^2 - \frac{dr^2}{f(r)} - r^2 (d\th^2 + \sin^2\th
163: d\ph^2),
164: \eea%
165: where $f(r) =1 - (2 M/r)$, $M$ is the black hole mass. In this
166: background, the equation for the massless Dirac field can be
167: reduced, after some algebra,
168: to the wave-like equation:
169: \be\label{Wave-like-equation}%
170: \left(\frac{d^2}{dr^{*2}} + \om^2 - V(r^*)\right)\Psi(r^*) = 0, \ee%
171: with the effective potential \cite{Wheeler}
172: \bea\label{potential}%
173: V(r)=f(r)\m\left(\frac{\m}{r^2}\pm\frac{d}{dr}\sqrt{\frac{f(r)}{r^2}}\right),
174: \eea%
175: which vanishes at the both boundaries: $V(r^*=\pm\infty)=0$. The
176: parameter $\m$ corresponds to a multipole number. Under the choice of the positive sign for the
177: real part of $\omega$ ($\omega =\omega_{Re} - i \omega_{Im}, \quad
178: \omega_{Re} > 0$), QNMs satisfy the following boundary conditions
179:
180: \be\label{bounds} \P(r^*) \sim C_\pm \exp(\pm\imo\om r^*), \qquad
181: r^{*}\longrightarrow \pm\infty,
182: \ee%
183: corresponding to purely in-going waves at the black hole event
184: horizon and purely out-going waves at infinity. Then, following
185: Leaver \cite{Leaver}, we can choose
186: \be\label{u-definition}%
187: \P(r^*)=\exp(\imo\om r^*)u(r^*),
188: \ee%
189: where $u(r)$ has a regular singularity at the event horizon and
190: is finite at $r^*\longrightarrow\infty$.
191:
192: The appropriate Frobenius series is
193: \be\label{Frobenius}%
194: u(r) = f(r)^{2s}\sum_{n=0}^\infty a_n f(r)^{n/2},
195: \ee%
196: %where $s$ satisfies the equation
197: \be f(r)^s\propto \exp(-\imo\om
198: r^*), s = -2M\imo\om, \quad r^*\longrightarrow-\infty.
199: \label{equation-s}
200: \ee
201: Substituting (\ref{u-definition}) and (\ref{Frobenius}) in
202: (\ref{Wave-like-equation}) we obtain the five-term recurrence
203: relation, which, after a series of Gaussian
204: eliminations can be reduced to the three-term recurrence relation
205:
206: $$a_{n+1}\a_n(\omega)+a_n\b_n(\omega)+a_{n-1}\g_n(\omega)=0.$$
207:
208:
209: When $\omega$ is the quasinormal frequency, the ratio of the
210: series coefficients is finite and can be found from the standard
211: continued fractions \cite{Leaver}.
212:
213: %\bea\label{ratio}%
214: %\frac{a_{n+1}}{a_n}= \frac{\g_{n}}{\a_n}\frac{\a_{n-1}}{\b_{n-1}
215: %-\frac{\a_{n-2}\g_{n-1}}{\b_{n-2}-\a_{n-3}\g_{n-2}/\ldots}}-\frac{\b_n}{\a_n}=\\\nonumber=-\frac{\g_{n+1}}{\b_{n+1}-\frac{\a_{n+1}\g_{n+
216: %2}}{\b_{n+2}-\a_{n+2}\g_{n+3}/\ldots}}.
217: %\eea%
218:
219: The QNMs are the roots of the inverted continued fraction:
220:
221: $$\b_n-\frac{\a_{n-1}\g_{n}}{\b_{n-1}
222: -\frac{\a_{n-2}\g_{n-1}}{\b_{n-2}-\a_{n-3}\g_{n-2}/\ldots}}=
223: \frac{\a_n\g_{n+1}}{\b_{n+1}-\frac{\a_{n+1}\g_{n+2}}{\b_{n+2}-\a_{n+2}\g_{n+3}/\ldots}}$$
224: that can be solved numerically as soon as
225: $\a_n(\omega)$, $\b_n(\omega)$, $\g_n(\omega)$ are found.
226:
227:
228: Note, that for the above effective potential, we are not able to
229: use the Nollert expansion \cite{Nollert}. That is why in spite of
230: the slow convergence of the continued fractions in the unmodified
231: Leaver procedure we had to be limited by it. This certainly
232: requires much greater computing time to perform the computations.
233:
234:
235:
236: \mysection{Low overtones.} The low overtones can be obtained either with the
237: help of the Frobenius method or, when the overtone number $n$ is
238: less than the multipole number $\mu$, applying the WKB formula
239: \cite{IyerWill}, \cite{KonoplyaWKB6}.
240: \begin{table}
241: \centerline{$\mu=1$\hspace{3.5cm}$\mu=2$}
242: \centerline{\begin{tabular}{|c|c|c|}
243: \hline
244: % after \\: \hline or \cline{col1-col2} \cline{col3-col4} ...
245: $n$ & $\Re\omega$ & $-\Im\omega$ \\ \hline
246: \hline
247: $0$ & $0.182963$ & $0.096982$\\ \hline
248: $**$& $0.182639$ & $0.094938$\\ \hline
249: $*$ & $0.176452$ & $0.100109$\\ \hline
250: $1$ & $0.147822$ & $0.316928$\\ \hline
251: % $**$ & $0.144722$ & $0.314386$\\ \hline
252: % $*$ & $0.146099$ & $0.320954$\\ \hline
253: $2$ & $0.146458$ & $0.424157$\\ \hline
254: $3$ & $0.138885$ & $0.549184$\\ \hline
255: $4$ & $0.134650$ & $0.674318$\\ \hline
256: $5$ & $0.131726$ & $0.799895$\\ \hline
257: $6$ & $0.128822$ & $0.925363$\\ \hline
258: $7$ & $0.126064$ & $1.050710$\\ \hline
259: $8$ & $0.123570$ & $1.176020$\\ \hline
260: \end{tabular}\quad
261: \begin{tabular}{|c|c|c|}
262: \hline
263: % after \\: \hline or \cline{col1-col2} \cline{col3-col4} ...
264: $n$ & $\Re\omega$ & $\-Im\omega$ \\ \hline
265: \hline
266: $0$ & $0.380037$ & $0.096405$\\ \hline
267: $**$& $0.380068$ & $0.096366$\\ \hline
268: $*$ & $0.378627$ & $0.096542$\\ \hline
269: $1$ & $0.355833$ & $0.297497$ \\ \hline
270: $**$& $0.355857$ & $0.297271$\\ \hline
271: $*$ & $0.353604$ & $0.298746$\\ \hline
272: $2$ & $0.326095$ & $0.421677$ \\ \hline
273: $3$ & $0.313180$ & $0.548007$ \\ \hline
274: $4$ & $0.305740$ & $0.673165$ \\ \hline
275: $5$ & $0.300096$ &$0.798686$ \\ \hline
276: $6$ & $0.294725$ & $0.924309$\\ \hline
277: $7$ & $0.289734$ & $1.049817$ \\ \hline
278: $8$ & $0.285201$ & $1.175236$\\ \hline
279: \end{tabular}
280: }%
281: \caption{First ten quasinormal frequencies of Dirac perturbations for
282: $\mu=1, 2$ found by the Frobenius method. For modes with $n <\mu$
283: the 6th order ($**$) and 3th order ($*$) WKB values are given.}
284: \end{table}
285: Let us remind that the potentials with opposite chirality produce
286: the same spectrum (see for instance \cite{Zhidenko} and
287: references therein). That is why we shall consider only positive
288: values of $\mu$. The WKB approach has been used recently when
289: studying QNMs of different black holes (see \cite{WKB} and
290: references therein) and shown good agreement with numerical
291: computations for lower overtones. As can be seen from the tables
292: above we observe very good agreement between the WKB and
293: Frobenius data, and, the higher the WKB order, the closer the
294: obtained WKB values to accurate numerical results.
295:
296:
297: \mysection{High overtones.}
298: The main difference from what we know on the high damping regime
299: for perturbations of fields of integer spin (scalar,
300: gravitational, and electromagnetic) is that now the spacing in
301: imaginary part is not $i/4 M$, as it takes place for integer spin
302: perturbations, but $i/8 M$. The real part of $\omega$ (See
303: Fig.~\ref{fig1} and Fig.~\ref{fig2}) falls down quickly to tiny
304: values which already cannot be found with reasonable accuracy by
305: the Frobenius method. This occurs already at about $n=400-500$ for
306: $\mu=1$ multipole and at about $n =8000$ for $\mu=2$. Note that
307: the greater the ratio $Im\omega_{n}/Re\omega_{n}$, the more slowly
308: the Frobenius series converge, so we had to be careful when
309: finding high overtones and to check the convergence by increasing
310: of the ``length'' of the continued fraction until the result for
311: $\omega_{n}$ will not change. This required, for instance, the
312: length of the continued fraction up to 5 millions for $n = 7000$.
313: That is because we were not able to use the Nollert procedure for
314: the considered effective potential, and therefore, a lot of
315: computer time was required to get the convergence of the Leaver
316: procedure. That is the reason why we could not extend our
317: computation to higher $n$ than that shown on Fig~\ref{fig1}~and
318: Fig~\ref{fig2}. Note also, that for $n$ larger then $500$ it
319: is very time consumingly to compute the modes one by one, so
320: we had to ``skip''through the hundreds of overtones as it is shown on Fig~\ref{fig2}.
321:
322: Thus, as can be seen from Fig.~\ref{fig1}~and~\ref{fig2}, QNMs
323: demonstrate the following asymptotic behavior:
324: \begin{equation}
325: \Re\omega_{n} \approx 0 \quad \textrm{as} \quad n \rightarrow \infty,
326: \quad
327: \end{equation}
328: \begin{equation}
329: \Im{\omega_{n+1}} - \Im{\omega_{n}} \approx -i/8 M \quad
330: \textrm{as}
331: \quad n\rightarrow\infty.
332: \end{equation}
333:
334: The formula for the spacing of the imagianary part can also be
335: reproduced following \cite{Pad}. Thus, for highly damping modes we
336: can use the Born approximation, where the scattering amplitude is
337: given by the formula \cite{Pad}:
338: \begin{equation}
339: S(k)= \int_{- \infty}^{+\infty} V[r(r^{*})] e^{2 i k r^{*}}dr^{*}.
340: \end{equation}
341: The above integral gets significant contribution only near the
342: event horizon. Using the Eq. \ref{potential} for $V[r(r^{*})]$, we find
343: $$ S(k) \sim~combinations~of~\Gamma (4 i k M)~ and~ \Gamma ((1/2) + 4 i k M ).$$
344: The poles of the amplitude occur when $(1/2) + 4 i M k =-n$, or $
345: 4 i M k =-n$ ($n \geq 0$ and is integer), i.e. $k_{n} = i n/8 M$
346: is the required high frequency asymptotic for imaginary part. The
347: same result can be obtained either using the Taylor expansion of
348: the effective potential near the event horizon, or without such
349: expansion, but in the latter case the expression for the
350: scattering amplitude will have a cumbersome form. The
351: singularities of the scattering amplitude are shown on
352: Fig~\ref{fig0}.
353:
354:
355: \begin{figure*}
356: \vspace{-2cm}
357: \begin{center}
358: \centerline{\includegraphics{D_figura3}}
359: \caption{The absolute value of the scattering amplitude $S(k)$ as a
360: function of $k$ up to a constant factor. $M=1$, $\mu=1$.}
361: \label{fig0}
362: \end{center}
363: \end{figure*}
364:
365:
366: %An important question is: which is the criteria that the above
367: %picture represents the true asymptotic regime \mbox{$n \rightarrow
368: %\infty$?} Certainly, there is no mathematical definitiness within
369: %the numerical approach we used, since the falling of the real part
370: %to tiny values does not guaranty that at greater overtone number
371: %the real part will not grow. However such a sharp change in
372: %asymptotic behavior would look very exotic, after the stable
373: %monotonic decreasing we observed. After all, if there were modes
374: %with considerable real part at greater overtones, then it would
375: %possible to catch them with Frobenius method, since the more the
376: %$\Im\omega / \Re\omega$ ratio, the greater the length of the
377: %continued fruction we must compute.
378:
379:
380: \mysection{Conclusion.}
381: In this paper we have studied the high overtones of the Dirac
382: quasinormal spectrum for a Schwarz-schild black hole. The spacing
383: in imaginary part is two times less than that for integer spin
384: fields when $n \rightarrow \infty$.
385: As was shown both numerically and analytically in \cite{referee2}
386: the real part of QNMs for scalar and gravitational perturbations
387: asymptotically approaches a constant equal to $\log 3/8 \pi$.
388: On the contrary, for the electromagnetic perturbations
389: the real part goes to zero \cite{referee3}.
390: The real part for Dirac perturbations goes to zero,
391: as it happens for electromagnetic
392: perturbations. This supports the suggestion that the expected
393: corresondence between quasinormal modes and the Barbero-Immirzi
394: parameter in Loop Quantum Gravity is just a numerical coincidence
395: \cite{referee}.
396:
397:
398: The above analysis can easily be extended to the case of massive
399: Dirac field \cite{Cho}, \cite{Cho2}. At low overtones, massive
400: Dirac pertubations \cite{Cho}, similar to massive scalar
401: pertubations \cite{ya}, lead to greater oscillation frequencies
402: and slower damping. At high overtones, similarly
403: to the treatment in \cite{K-Zh} for massive scalar field, one
404: can easily anticipate that the massive
405: term will not affect the asymptotic quasinormal behavior.
406:
407:
408:
409:
410: \mysection{Acknowledgements.} The work of K. C-B. and R. K. was supported by
411: FAPESP (Brazil).
412:
413:
414: \bigskip
415:
416:
417:
418:
419: \begin{thebibliography}{99}
420: \bibitem{kokkotas-review} K.Kokkotas and B.Schmidt, "Quasi-normal modes of
421: stars and black holes", \texttt{Living.Reviews.Relativ.} {\bf{2}},
422: 2 (1999). H-P. Nollert, Class.Quant.Grav. 16, 159 (1999)
423: \bibitem{Birmingham2} G.T.Horowitz and V.Hubeny, \texttt{Phys. Rev.}
424: {\bf{D62}} 024027 (2000); D. Birmingham, I. Sachs, and S.N.
425: Solodukhin,
426: \texttt{Phys. Rev. Lett.} {\bf 88}, 151301 (2002); A.Starinets,
427: Phys.Rev.D66:124013, 2002; V. Cardoso, R. Konoplya and J.
428: P. S. Lemos, \texttt{Phys.Rev.} {\bf D 68} 044024 (2003); E. Abdalla, B.
429: Wang, A. Lima-Santos, W.G. Qiu, Phys.Lett.B538:435-441,2002; R.A. Konoplya,
430: Phys.Rev.D66:044009,2002; E.Abdalla, K.H.C. Castello-Branco, A. Lima-Santos,
431: Phys.Rev.D66:104018,2002.
432: \bibitem{Dreyer} S.Hod \texttt{Phys.Rev.Lett.} {\bf{81}} 4293 (1998);
433: O. Dreyer, gr-qc/0211076.
434: \bibitem{kucha-mala} E. Berti, K.D. Kokkotas, Phys.Rev.D67:064020,2003
435: [gr-qc/0301052]; C.Molina, D.Guigno, E.Abdalla and A.Saa Phys.
436: Rev. {\bf D69} 104013 (2004); R.A.Konoplya and A.Zhidenko, JHEP
437: (06) 037 (2004); R.A.Konoplya, \texttt{Phys.Rev.} {\bf{D 66}}
438: 084007 (2002); Phys.Rev.D70:047503,2004; G. Siopsis, hep-th/0409262;
439: \bibitem{first-Cardoso} V.Cardoso and J.P.S.Lemos, \texttt{Phys.Rev.} {\bf{D63}}
440: 124015 (2001);
441: \bibitem{Cho} H. T. Cho, \texttt{ Phys.Rev. D} 68, 024003, 2003.
442: \bibitem{Zhidenko} A.Zhidenko, \texttt{Class.Quant.Grav.} {\bf 21} 273 (2004)
443: \bibitem{charged-dirac} Wei Zhou, Jian-Yang Zhu,
444: Int.J.Mod.Phys.D13:1105-1118,2004.
445: \bibitem{IyerWill} S. Iyer and C. M. Will, \texttt{Phys.Rev.} {\bf{D35}} 3621
446: (1987)
447: \bibitem{KonoplyaWKB6} R.A.Konoplya, \texttt{ Phys.Rev.} {\bf D68} 024018 (2003)
448: \bibitem{Birell} N.D. Birrell and P.C.W. Davies, Quantum Fields in
449: Curved Space, Cambridge Univ. Press, Cambridge (1982)
450: \bibitem{Wheeler} D. R. Brill, J. A. Wheeler,
451: Rev.Mod.Phys.29:465-479,1957; S. Chandrasekhar, The Mathematical
452: Theory of Black Holes (Clarendon, Oxford, 1983).
453: \bibitem{Leaver} E.M. Leaver, Proc.Roy.Soc.Lond.A402:285-298,1985
454: \bibitem{Nollert} Nollert H-P 1993 Phys. Rev. D 47 5253
455: \bibitem{WKB} R.A. Konoplya, Phys.Rev.D68:124017,2003;
456: \bibitem{Pad} T.R.Choudhury and T.Padmanabhan, Phys.Rev. D69, 064033
457: (2004); A.J.M. Medved, D. Martin and Matt Visser,
458: \texttt{Class.Quant.Grav.} {\bf 21} 1393 (2004)
459: \bibitem{referee2} L.Motl, Adv.Theor.Math.Phys. 6: 1135, 2003:
460: E.Berti and K.Kokkotas, Phys.Rev.D68:044027, 2003
461: \bibitem{referee3} L.Motl and A.Neitzke, Adv. Theor. Math. Phys.
462: 7, 307 (2003); V.Cardoso, J.Lemos and S.Yoshida, Phys.Rev.D69:044004,
463: 2004.
464: \bibitem{referee} E.Beri, V.Cardoso and S.Yoshida, Phys.Rev. D69:
465: 124018, 2004: J.Natario and R.Schiappa, hep-th/0411267.
466: \bibitem{Cho2} H. T. Cho, and Y-C. Lin gr-qc/0411090
467: \bibitem{ya} R.A. Konoplya, Phys.Lett.B550:117-120,2002.
468: \bibitem{K-Zh} R.A. Konoplya, A.V. Zhidenko, gr-qc/0411059.
469:
470: \end{thebibliography}
471:
472: \begin{figure*}
473: \vspace{-2cm}
474: \begin{center}
475: \centerline{\includegraphics{Dfigura1}}
476: \caption{Real part of $\omega$ as a function of imaginary part for $\mu=1$
477: multi-pole.}
478: \label{fig1}
479: \end{center}
480: \end{figure*}
481:
482:
483: \begin{figure*}
484: \vspace{-2cm}
485: \begin{center}
486: \centerline{\includegraphics{Dfigura2}}
487: \caption{Real part of $\omega$ as a function of imaginary part for $\mu=2$
488: multi-pole.}
489: \label{fig2}
490: \end{center}
491: \end{figure*}
492:
493:
494:
495:
496: \end{document}
497: