astro-ph0009437/ms.tex
1: %                                                               l-aa.dem
2: % L-AA vers. 3.0, LaTeX style file for Astronomy & Astrophysics
3: % Demo file
4: %                                                 (c) Springer-Verlag HD
5: %-----------------------------------------------------------------------
6: %\documentclass[referee]{aa}
7: \documentclass{aa}
8: \usepackage{graphics}
9: \usepackage{psfig}
10: 
11: \begin{document}
12: 
13:    \thesaurus{11.03.1;	 % galaxies: clusters: general
14: 	      12.04.1;   % cosmology: dark matter
15: 	      14.25.2}	 % X-rays: galaxies
16: 
17: \title{Reconstruction of radial temperature profiles of galaxy clusters} 
18: 
19: 
20: \author{Yan-Jie Xue and Xiang-Ping Wu}
21: 
22: \offprints{Y.-J. Xue}
23: \mail{wxp@class1.bao.ac.cn}
24: 
25: \institute{Beijing Astronomical Observatory and 
26: National Astronomical Observatories,
27: Chinese Academy of Sciences, Beijing 100012, China}
28: 
29: 
30:    \date{Received 28 June, 2000; accepted 27 July, 2000}
31: 
32:    \titlerunning{Temperature Profiles of Clusters}
33:    \maketitle
34: 
35: 
36: \begin{abstract}
37: In this {\sl Letter} we present the radial temperature profiles of 
38: three X-ray clusters (A119, A2255 and A2256)
39: reconstructed from a combination of the X-ray
40: surface brightness measurements and the universal density profile
41: as the underlying dark matter distribution. Our algorithm is based
42: on the hydrostatic equilibrium for intracluster gas and the 
43: universality of the total baryon fraction within the virial radius.
44: The scaled temperature profiles of these three clusters appear to be
45: remarkably similar in shape, reflecting the 
46: underlying structural regularity, although they are inconsistent with
47: either isothermality or a significant decline with increasing
48: radius.  Nevertheless, we find a good agreement between our derived 
49: temperature profiles and the recent analysis of 11 clusters observed 
50: with BeppoSAX (Irwin \& Bregman 2000), which provides a useful clue to
51: resolving the temperature profile discrepancy raised recently in
52: literature. A comparison of our derived temperature
53: profiles with future spatially-resolved spectral measurements
54: may constitute a critical test for the standard model of
55: structure formation and the conventional scenario for dynamical
56: properties of clusters.
57: \end{abstract}
58: 
59: \keywords{cosmology: dark matter --- galaxies: clusters: general ---  
60:           X-rays: galaxies}
61: 
62: %
63: %  14.Sep.'90: Demo-Vs.
64: %________________________________________________________________
65: 
66: 
67: \section{Introduction}
68: 
69: The lack of robust constraints on the radial temperature profiles of hot gas 
70: contained within galaxy clusters 
71: is probably the major uncertainty in the present
72: determination of dynamical properties of clusters, which hinders 
73: clusters from acting as an ideal laboratory of testing theories of
74: formation and evolution of structures in the universe including 
75: a direct estimate of the cosmic mass density parameter $\Omega_{\rm{M}}$ 
76: by combining the baryon fraction measurement and the Big Bang
77: Nucleosynthesis. Indeed, previous studies have arrived at
78: conflicting results regarding the radial temperature gradients in
79: clusters. By analyzing 30 clusters observed with ASCA, 
80: Markevitch et al. (1998) claimed a significant temperature 
81: decline with radius quantified by a polytropic index 
82: of $1.2$-$1.3$ on the average. However, subsequent studies have soon 
83: raised doubt about the ubiquity and steepness of these temperature decline:
84: Irwin, Bregman \& Evrard (1999) carried out an analysis of the 
85: color profiles of the same clusters used by Markevitch et al. (1998)
86: but found an essentially flat temperature profile. 
87: Applying the spectral-imaging deconvolution method to 
88: a large sample of 106 ASCA clusters, White (2000) has showed that
89: 90 percent of the temperature profiles are actually consistent 
90: with isothermality.  Further argument against the nonisothermality 
91: of intracluster gas has been put forward recently by Irwin \& Bregman (2000),
92: who reported the detection of a flat and even increasing temperature 
93: profile out to $\sim30\%$ of the viral radius for a sample of 
94: 11 clusters observed with BeppoSAX.  
95: 
96: 
97: Theoretically, it deserves an investigation into the possibility
98: of deriving the radial temperature profiles of intracluster gas
99: from the well-motivated physical mechanisms, incorporated with 
100: the X-ray imaging observations. This may provide a valuable clue to
101: resolving the above temperature profile discrepancy. There are 
102: two well-established facts on which we can rely today:
103: (1)The gravitational potential of a cluster is dominated by the 
104: dark matter distribution which can be described by the so-called universal
105: density profile, as suggested by a number of high-resolution simulations
106: (Navarro, Frenk \& White 1995 and hereafter NFW), 
107: although the innermost slope is still under debate. 
108: (2)The azimuthally-averaged X-ray surface brightness of a cluster 
109: is reliably measurable out to several or even $\sim10$ times as large 
110: as the X-ray core radius,
111: for which a good approximation is provided by the conventional $\beta$ model 
112: (Cavaliere \& Fusco-Femiano 1976). These two facts, along with
113: the hydrostatic equilibrium hypothesis and a reasonable choice of the boundary 
114: conditions, permit a unique determination of   
115: the gas temperature profile (Wu \& Chiueh 2000). On the other hand, 
116: a comparison of the theoretically expected temperature profile 
117: with the result from the X-ray spectroscopic measurement constitutes 
118: a critical test for the validity of the NFW profile and 
119: the hydrostatic equilibrium in clusters.  
120: 
121: 
122: In this {\sl Letter}, we will attempt for the first time 
123: to derive the temperature profiles of 3 well-defined clusters with good X-ray
124: imaging observations extending to relatively large radii, based on 
125: the method developed by Wu \& Chiueh (2000). 
126: Our derived temperature profiles will be compared with 
127: the recent results of 11 clusters  observed with BeppoSAX 
128: (Irwin \& Bregman 2000).  We will examine the possible similarity in the
129: gas temperature profiles as a result of the underlying structural regularity
130: (e.g. Neumann \& Arnaud 1999). The implication of our results  
131: for the reported temperature profile discrepancy will be discussed. 
132: Throughout the {\sl Letter} we assume $H_0=50$ km s$^{-1}$ Mpc$^{-1}$ and a 
133: flat cosmological model with $\Omega_{\rm{M}}=0.3$ and $\Omega_{\Lambda}=0.7$.
134: 
135: 
136: 
137: \section{Theoretical expectations}
138: 
139: 
140: We briefly summarize the mathematical treatment of
141: the intracluster gas tracing the underlying dark matter distribution 
142: of clusters. First, if the X-ray surface brightness profile 
143: of a cluster can be well approximated by the conventional $\beta$ model 
144: (Cavaliere \& Fusco-Femiano 1976)
145: %1
146: \begin{equation}
147: S_{\rm{x}}(r)=S_0\left(1+\frac{r^2}{r_{\rm{c}}^2}\right)^{-3\beta+1/2},
148: \end{equation}
149: this would indicate (Cowie, Henriksen \& Mushotzky 1987),
150: %2
151: \begin{equation}
152: n_{\rm{e}}(r)T^{1/4}(r)=n_{\rm{e}0}T_0^{1/4}
153: \left(1+\frac{r^2}{r_{\rm{c}}^2}\right)^{-3\beta/2}
154: \end{equation}
155: for an optically-thin, thermal bremsstrahlung emission, where 
156: $n_{\rm{e}}$ and $T$ are the electron number density and temperature,
157: respectively. The central electron number
158: density $n_{\rm{e}0}$, temperature $T_0$ and X-ray surface brightness $S_0$ 
159: are connected by 
160: %3
161: \begin{equation}
162: n_{\rm{e}0}^2=\frac{4\pi^{1/2}}{\alpha(T_0)\mu_{\rm{e}} g}
163:         \frac{\Gamma(3\beta)}{\Gamma(3\beta-1/2)}
164:         \frac{S_0(1+z)^4}{r_{\rm{c}}},
165: \end{equation}
166: where $\alpha(T_0)=(2^4e^6/3m_{\rm{e}}\hbar c^2)(2\pi kT_0/3m_{\rm{e}}c^2)^
167: {1/2}$,
168: $\mu_{\rm{e}}=2/(1+X)$ with $X$ being the primordial hydrogen mass fraction,
169: $g\approx1.2$ is the average Gaunt factor, and $z$ is the cluster redshift.
170: The total mass in gas within $r$ is simply
171: %4
172: \begin{equation}
173: M_{\rm{gas}}(r)=4\pi\mu_{\rm{e}} m_p n_{\rm{e}0} \int 
174: \left(\frac{T_0}{T}\right)^{1/4}
175: \left(1+\frac{r^2}{r_{\rm{c}}^2}\right)^{-3\beta/2} r^2 dr.
176: \end{equation}
177: Secondly, if the intracluster gas is in hydrostatic equilibrium with the
178: underlying dark matter distribution, we have
179: %5
180: \begin{equation}
181: \frac{GM_{\rm{DM}}(r)}{r^2}=
182: -\frac{1}{\mu m_{\rm{p}} n_{\rm{e}}}   \frac{d(n_{\rm{e}}kT)}{dr}.
183: \end{equation}
184: where $\mu=0.585$ is the average molecular weight. For NFW profile 
185: %6
186: \begin{equation}
187: M_{\rm{DM}}(r)=4\pi\rho_{\rm{s}} r_{\rm{s}}^3\left
188: [\ln\left(1+\frac{r}{r_{\rm{s}}}\right)-
189:           \frac{r}{r+r_{\rm{s}}}\right].
190: \end{equation}
191: Here we have neglected the self-gravity of the gas. Using the normalized 
192: gas temperature $\tilde{T}(r)\equiv T(r)/T_0$ and 
193: the volume-averaged baryon fraction $f_{\rm{b}}(r)\equiv 
194: M_{\rm{gas}}(r)/M_{\rm{DM}}(r)$
195: as the two variables, we obtain the following two first-order 
196: differential equations
197: %7,8
198: \begin{eqnarray}
199: \frac{d\tilde{T}}{dx}=\frac{4\beta x \tilde{T}}{x^2+a^2}-
200:  \frac{4\alpha_0}{3x^2}\left[\ln(1+x)-\frac{x}{1+x}\right];\\
201: \frac{df_{\rm{b}}}{dx}=\frac{b\tilde{T}^{-1/4}
202:                      (1+x^2/a^2)^{-3\beta/2}x^2
203:                 -f_{\rm{b}} x/(1+x)^2}
204:                 {\ln(1+x)-x/(1+x)},
205: \end{eqnarray}
206: where $x=r/r_{\rm{s}}$, $a=r_{\rm{c}}/r_{\rm{s}}$, $b=\mu_{\rm{e}}
207: n_{\rm{e}0}m_{\rm{p}}/\rho_{\rm{s}}$ and
208: $\alpha_0=4\pi G \mu m_{\rm{p}} \rho_{\rm{s}} r_{\rm{s}}^2/kT_0$.  
209: The first equation
210: can be straightforwardly solved with $\tilde{T}(0)=1$:
211: %9
212: \begin{equation}
213: \tilde{T}(x)=\left(1+\frac{x^2}{a^2}\right)^{2\beta}
214:              \left[1-\frac{4\alpha_0}{3}\int_0^x 
215:                \frac{\ln(1+x)-x/(1+x)}{x^2(1+x^2/a^2)^{2\beta}}dx\right].
216: \end{equation}
217: In order to solve the second equation and determine the free parameters
218: $a$, $b$ and $\alpha_0$, we use the following boundary conditions
219: %10,11
220: \begin{eqnarray}
221: f_{\rm{b}}(r_{\rm vir})=f_{\rm{b,BBN}};\\
222: \frac{df_{\rm{b}}}{dx} \left|_{x=r_{\rm{vir}}/r_{\rm{s}}}=0. \right.
223: \end{eqnarray}
224: Namely, we demand that  the baryon fraction should 
225: asymptotically match the universal value of $f_{\rm{b,BBN}}$ 
226: at the virial radius $r_{\rm{vir}}$ defined by
227: %12
228: \begin{equation}
229: M_{\rm{DM}}(r_{\rm{vir}})=\frac{4\pi}{3}r_{\rm{vir}}^3 \Delta_{\rm{c}} 
230: \rho_{\rm{crit}},
231: \end{equation}
232: where $\Delta_{\rm{c}}$ represents the overdensity of 
233: dark matter with respect to
234: the average background value $\rho_{\rm{crit}}$, for which we take 
235: $\Delta_{\rm{c}}=178\Omega^{0.45}_{\rm{M}}(z)$  and 
236: $\Omega_{\rm{M}}(z)=\Omega_{\rm{M}}(1+z)/\{1+z\Omega_{\rm{M}}+
237: [(1+z)^{-2}-1]\Omega_{\Lambda}\}$.
238: We now come to the free parameters involved in eqs.(8) and (9). 
239: With the X-ray imaging observation, we can obtain the best-fit 
240: values of $\beta$, $r_{\rm{c}}$ and $S_0$. 
241: If, on the other hand, the X-ray spectroscopic measurement can set 
242: a useful constraint on the central temperature $T_0$, we will be able to 
243: derive the central electron density from  eq.(3).  As a result, there
244: are only two free parameters in the above equations: $\rho_{\rm{s}}$ 
245: (or equivalently $\delta_{\rm{c}}=\rho_{\rm{s}}/\rho_{\rm{crit}}$) and 
246: $r_{\rm{s}}$. These two parameters can be
247: fixed during the numerical searches for the solution of eqs.(8) and (9)
248: using the boundary conditions eqs.(10) and (11).  
249: This will allow us to work out simultaneously the radial profiles of gas 
250: density and temperature, and fix
251: the dark matter (NFW) profile of the cluster characterized by 
252: $\rho_{\rm{s}}$ and $r_{\rm{s}}$. 
253: 
254: 
255: \section{Application to X-ray clusters}
256: 
257: Since the reconstruction of gas temperature profile is sensitive 
258: to the initial input of $S_{\rm{x}}$ especially the $\beta$ parameter, 
259: whether or not we can reliably derive the temperature profile 
260: depends critically on the goodness of the single $\beta$ model fit 
261: to the X-ray surface brightness profile. 
262: We thus restrict ourselves to  the X-ray 
263: flux-limited sample of 45 clusters published recently by 
264: Mohr, Mathiesen \& Evrard (1999), in which there are sufficiently large 
265: data points to set robust constraints on the $\beta$ model fit.
266: The inclusion of a cluster is based on the following two criteria:
267: (1)The X-ray surface brightness profile can be well fitted by 
268: a single $\beta$ model with $0.8\leq \chi_{\rm{{\nu}}}^2\leq1.25$;
269: (2)The maximum extension ($r_{\rm{m}}$) of the X-ray observed surface 
270: brightness profile should be large enough to guarantee
271: the validity of the $\beta$ model at the outermost regions of clusters. Here
272: we set $r_{\rm{m}}\geq1.5$ Mpc. Unfortunately, it turns out that 
273: there are only three clusters which meet our criteria 
274: (Table 1): A119, A2255 and A2256.  In fact, our first criterion 
275: implies that the effect of cooling flows in the central regions of clusters
276: should be negligibly small. This explains the fact that the three selected
277: clusters all have large core radii. Note that the presence of cooling flows
278: may lead to the failure of a single $\beta$ model fit to the X-ray  
279: surface brightness profiles. In other words, our method cannot be applied to 
280: the clusters with strong cooling flows.  While the X-ray imaging data  
281: of the clusters can be somewhat accurately acquired, the present X-ray spectral
282: measurements have yielded the emission-weighted temperatures rather than 
283: the central values $T_0$ appearing in the $\alpha_0$ parameter.
284: Therefore, we have to use the emission-weighted temperature 
285: as a first approximation of $T_0$. Alternatively, we adopt 
286: the universal baryon fraction $f_{\rm{b,BBN}}=1/6$ to 
287: reconcile our cosmological
288: model of $\Omega_{\rm{M}}=0.3$ (for $\Omega_{\rm{b}}=0.05$). 
289: 
290: 
291: 
292: 
293:  \begin{table}
294:  \vskip 0.2truein
295:  \scriptsize
296:  \begin{center}
297:  \caption{Cluster Sample}
298:  \begin{tabular}{llll}
299:  \hline
300:   & & &  \\ 
301:  cluster & A119 & A2255 & A2256 \\
302:    & & &  \\ 
303:   $z$       & 0.0438  & 0.0808  & 0.0581 \\
304:   $T_0$ (keV) & 5.80    & 7.30    & 7.51 \\
305:   $S_0^*$ & 1.18 & 1.68 & 4.41 \\
306:  $\beta$   & 0.662 &  0.792 & 0.828 \\
307:  $r_{\rm{c}}$ (Mpc) & 0.494  & 0.608     &  0.500 \\
308:  $n_{\rm{e}0}$ ($10^{-3}$cm$^{-3}$) & $1.37 $ & 1.67 & 2.94 \\
309:  $b$  &  4.15 & 3.48 & 3.31 \\
310: $\alpha_0$ & 17.80(12.30)$^+$ & 14.66(24.80)$^+$ & 13.15(13.21)$^+$\\
311: $\delta_{\rm{c}}$ & 130(490)$^+$     & 184(230)$^+$     & 345(1220)$^+$ \\
312: $r_{\rm{s}}$ (Mpc) & 5.77(2.59)$^+$ & 4.82(5.99)$^+$    & 3.43(2.03)$^+$\\
313:   &  &  & \\
314:  \hline
315:  \end{tabular}
316:  \end{center}
317: 
318: \parbox{8.5in}{$^*$In units of $10^{-13}$ erg s$^{-1}$ cm$^{-2}$ 
319:                arcmin$^{-2}$ for energy band 0.5-2.0 keV;}
320: \parbox{8.5in}{$^+$The result for an isothermal gas distribution.}
321:   \end{table}
322: 
323: 
324: Using the available X-ray data of the three clusters from Mohr et al. (1999),
325: we have performed numerical searches for the solutions of eqs.(8) and (9)
326: by iterations until the boundary conditions eqs.(10) and (11)
327: are satisfied. The resulting parameters $\alpha_0$,  
328: $\delta_{\rm{c}}$ and $r_{\rm{s}}$ are summarized in Table 1, together with a 
329: comparison with the corresponding values for an isothermal
330: gas distribution estimated in previous work (Wu \& Xue 2000).
331: Most importantly, such a procedure enables us to completely fix
332: the radial profiles of gas density, 
333: temperature and baryon fraction for the three clusters. 
334: Here we have no intention to illustrate the radial variations of 
335: $n_{\rm{e}}(r)$ and $f_{\rm{b}}(r)$, 
336: which essentially follow the theoretical expectations
337: (Wu \& Chiueh 2000). Rather, we display in Fig.1 the radial  
338: profiles of the emission-weighted temperatures for 
339: the three clusters constructed from our algorithm. 
340: Surprisingly, none of the temperature profiles of these three clusters
341: are consistent with the conventional speculations, and a visual examination
342: of Fig.1 reveals that they are neither characterized by isothermality nor 
343: represented simply by the polytropic equation of state. 
344: Nevertheless, these temperature profiles indeed demonstrate a similar radial 
345: variation, reflecting probably the underlying structural regularity.
346: Basically, the radial variation of the gas temperature resembles a
347: distorted `S' in shape: 
348: There exist two turnover points roughly at $0.1r_{\rm{vir}}$ 
349: and $0.4r_{\rm{vir}}$, respectively, where $dT/dr=0$, which separate 
350: the temperature curve $T(r)$ into three parts -- a decreasing $T(r)$ 
351: with radius inside the cluster core of $\sim0.1r_{\rm{vir}}$,  following
352: a slightly increasing $T(r)$ until $\sim0.4r_{\rm{vir}}$, 
353: and finally a moderately
354: decreasing $T(r)$ out to the virial radius. Overall, the absolute values of 
355: the gas temperature do not demonstrate a dramatic change within 
356: clusters.
357: 
358: %\placefigure{fig1}
359:     \begin{figure} 
360: 	\psfig{figure=fig1.ps,width=88mm,bbllx=80pt,bblly=120pt,bburx=540pt,bbury=650pt,clip=,angle=270}
361: 	\caption{A comparison of the derived radial temperature profiles
362: of three clusters (A119, A2255 and A2256) with the results of 11
363: clusters observed with BeppoSAX (Irwin \& Bregman 2000).
364: The observed data are normalized  by the mean temperature for each
365: cluster, while the derived temperature curves are scaled by $1.32T_0$
366: for comparison. The horizontal axis is in units of the virial radius. }
367: \label{fig1}
368:    \end{figure}
369: 
370: The azimuthally-averaged radial temperature profiles of 11 clusters 
371: derived by Irwin \& Bregman  (2000) from an analysis of the BeppoSAX data
372: are superimposed on Fig.1. It appears that our derived temperature profiles 
373: are in good agreement with their observed ones over entire radius range. 
374: In fact, the significant temperature discrepancy raised in different 
375: X-ray spectral measurements occurs in the inner parts of clusters.
376: In the outer regions, it seems that many observations have provided   
377: a moderately decreasing temperature profile, 
378: which is essentially consistent with our theoretical predictions.  
379: Alternatively, our result is also compatible with the gas temperature 
380: distribution at large radii revealed by numerical simulations that 
381: demonstrate a temperature
382: decline of $\sim30\%$ of the central value at the virial radius
383: (Frenk et al. 1999). 
384: 
385: 
386: \section{Discussion and conclusions}
387: 
388: 
389: In the absence of the detailed information on the radial temperature
390: profiles of clusters from X-ray spectroscopic measurements, we have 
391: made an attempt to derive the gas temperature profiles by combining
392: the X-ray surface brightness measurements and
393: the NFW profile as the underlying dark matter distribution of 
394: clusters. This has become possible when the intracluster gas is required  
395: to satisfy the hydrostatic equilibrium and the volume-averaged baryon 
396: fraction within the viral radius is required to asymptotically match 
397: the universal value. Consequently, we have obtained semi-analytically  
398: the temperature profiles of three clusters selected carefully from 
399: the ROSAT observed cluster sample. 
400: These derived temperature profiles are consistent with 
401: the new observations of 11 BeppoSAX clusters (Irwin \& Bregman 2000) 
402: and other measurements made at large cluster radii 
403: (e.g. Markevitch et al. 1998) as well as the result given by
404: numerical simulations (e.g. Frenk et al. 1999). 
405: 
406: 
407: 
408: Regardless of the small sample, the three clusters exhibit a 
409: temperature profile similar in shape when the length scales are normalized to
410: their virial radii,  perhaps indicative of the underlying structural 
411: regularity.  
412: The present study provides a helpful clue to resolving the temperature
413: profile discrepancy: It is very likely that the lack of the high-quality 
414: data of the spatially resolved spectral observations 
415: would yield an emission-weighted temperature roughly close to isothermality 
416: within $\sim80\%$ of the virial radius, which does not exclude
417: the possibility that a slightly increasing 
418: temperature profile may be marginally detectable in the range of  
419: $0.1r_{\rm{vir}}<r<0.4r_{\rm{vir}}$. This explains the recent observations of 
420: Irwin \& Bregman (2000) and other studies (e.g. Kikuchi et al. 1999;
421: White 2000; etc.). However, our model does not predict
422: the flat temperature profile toward the inner regions of clusters 
423: as reported particularly by Markevitch et al. (1998), although 
424: a moderately decreasing temperature
425: profile will ultimately take place in the outer clusters ($r>0.4r_{\rm{vir}}$).
426: 
427: 
428: A conclusive test for the universality of our derived temperature profiles 
429: can be provided by future X-ray spectroscopic measurements.  Indeed, 
430: it will be useful to apply the present method to other 
431: X-ray clusters with good X-ray surface brightness profiles 
432: measured to large radii and  high-quality data of the spatially-resolved 
433: spectral observations at least within the central regions.
434: This may allow us to further justify our model and include the measurement 
435: uncertainties which have been neglected in the present study.
436: The inconsistency of the predicted temperature profiles
437: with the X-ray spectroscopic results will challenge the prevailing
438: models of structure formations as well as the conventional scenario of
439: cluster dynamics such as the hydrostatic equilibrium.
440: Finally, we should point out that our proposed method to obtain
441: the temperature profiles of clusters can be 
442: significantly contaminated by nongravitational heating
443: processes especially from the supernova-driven protogalactic winds.
444: Recall that the asymptotic tendency of the derived temperature
445: profiles at large radii depends sensitively on the $\beta$
446: parameter, while the energy injection of supernovae and active galaxies 
447: into the intracluster gas will result in a shallower 
448: X-ray surface brightness distribution (David et al. 1990; Ponman, Cannon \&
449: Navarro 1999; Llyod-Davies, Ponman \& Cannon 2000). Without
450: correction to this effect the theoretically predicted temperature profiles
451: may rise too rapidly at large radii.  
452: Note that at large radii the NFW mass profile 
453: diverges logarithmically with $r$, which differs significantly from
454: the variation of the gas mass profile (roughly $M_{\rm{gas}}\propto r$) 
455: expected from the assumption of isothermality.  
456: For a cluster with smaller $\beta$, $r_{\rm{c}}$ and $r_{\rm{s}}$, 
457: an increasing temperature
458: profile near virial radius is thus required to maintain the 
459: universality of the cluster baryon fraction. 
460: Therefore, a robust, theoretical determination of
461: the temperature profiles of clusters  should also allow 
462: nongravitational heating processes to be included.
463: 
464: 
465: 
466: \begin{acknowledgements}
467: We gratefully acknowledge Tzihong Chiueh for useful 
468: discussion and comments, and an anonymous referee for valuable suggestions.
469: This work was supported by 
470: the National Science Foundation of China, under Grant 19725311.
471: \end{acknowledgements}
472: 
473: 
474: 
475: \begin{thebibliography}{}
476: 
477: \bibitem{}Cavaliere A.,  Fusco-Femiano R., 1976, A\&A, 49, 137
478: \bibitem{}Cowie L. L., Henriksen M., Mushotzky R. F., 1987, 
479:             ApJ, 317, 593
480: \bibitem{}David L. P., Arnaud K. A., Forman W., Jones C., 
481:                 1990, ApJ, 356, 32 
482: \bibitem{}Frenk C. S., et al., 1999, ApJ, 525, 554
483: \bibitem{}Irwin J. A., Bregman J. N., 2000, ApJ, in press
484: \bibitem{}Irwin J. A., Bregman J. N.,  Evrard A. E., 1999, 
485:                 ApJ, 519, 518
486: \bibitem{}Kikuchi K., et al., 1999, PASJ, 51, 301
487: \bibitem{}Lloyd-Davies E. J., Ponman T. J., Cannon D. B., 
488:                 2000, MNRAS, in press
489: \bibitem{}Markevitch M., Vikhlinin A., Forman W. R.,  
490: 	     Sarazin C. L., 1998, ApJ, 527, 545  
491: \bibitem{}Mohr J. J., Mathiesen B., Evrard A. E., 1999, ApJ, 
492: 	    517, 627 
493: \bibitem{}Navarro J. F., Frenk C. S., White S. D. M., 1995, MNRAS, 
494:             275, 720 (NFW)
495: \bibitem{}Neumann D. M., Arnaud M., 1999, A\&A, 348, 711
496: \bibitem{}Ponman T. J., Cannon D. B., Navarro J. F., 1999, 
497:                 Nature, 397, 135
498: \bibitem{}White D. A., 2000, MNRAS, 312, 663
499: \bibitem{}Wu X.-P.,  Chiueh T., 2000, ApJ, in press
500: \bibitem{}Wu X.-P,  Xue, Y.-J.,  2000, ApJ, 529, L5 
501: 
502: \end{thebibliography}
503: 
504: \end{document}
505: 
506: 
507: