1: %\documentclass[preprint]{aastex}
2: %\documentclass[preprint2]{aastex}
3: %\documentclass[12pt,preprint]{aastex}
4: \documentclass[preprint2]{aastex}
5: % MACROS
6: \def\ap{$\,\approx\,$}
7: \def\uka { \raisebox{-0.5ex} {\mbox{$\stackrel{<}{\scriptstyle \sim}$}}}
8: \def\uga { \raisebox{-0.5ex} {\mbox{$\stackrel{>}{\scriptstyle \sim}$}}}
9: \def\upz { \raisebox{-0.5ex} {\mbox{$\stackrel{\propto}{\scriptstyle \sim}$}}}
10: \def\deg {$^\circ$}
11: \def\nerr#1#2{{+{#1}}-{#2}}
12: \renewcommand{\topfraction}{0.98}
13: \renewcommand{\bottomfraction}{0.98}
14: \renewcommand{\floatpagefraction}{0.98}
15: \setlength{\textheight}{23.5cm}
16: \setlength{\tabcolsep}{0.1cm}
17: \sloppy
18: \def\fluss#1{$10^{#1}$ cm$^{-2}$s$^{-1}$}
19: \def\plusminus#1#2{{\raisebox{-0.5ex} {$\stackrel{+{#1}}{\scriptstyle-{#2}}$}}}
20: \slugcomment{Submitted to Astrophysical Journal Letters}
21: \shortauthors{H.\ Krawczynski}
22: \shorttitle{Spectroscopic Deprojection-Analysis of 3C~129}
23: \begin{document}
24: \title{
25: Spectroscopic Deprojection-Analysis of Chandra Data of the
26: Galaxy~Cluster~3C~129}
27: %
28: \author{H. Krawczynski\altaffilmark{1}}
29: \altaffiltext{1}{Yale University, P.O. Box 208101, New Haven, CT 06520-8101, USA, email: krawcz\@astro.yale.edu}
30: %
31: \begin{abstract}
32: %
33: We report on spectroscopic imaging observations of the nearby ($z$=0.022)
34: galaxy cluster 3C~129 performed with the ACIS detector on board of the Chandra
35: X-ray observatory. Applying a deprojection analysis which fully takes into
36: account the spatially resolved X-ray energy spectra we investigate the
37: radial variation of temperature and particle density of the Intracluster
38: Medium (ICM).
39: %
40: While earlier data indicated a weak cooling flow and a cool cluster core,
41: the Chandra data suggests that the core is hotter than the rest of the
42: cluster: at the center we infer 7.9$^{+4.4}_{-2.4}$~keV while the mean
43: temperature of all the other ICM shells is 5.1$\pm$0.3~keV
44: (all errors on 90\% confidence level).
45: Based on the radial entropy profile we discuss the evidence for shock
46: heating of the ICM at the cluster center.
47: We discuss the possibility to use imaging spectroscopy data of more regular
48: clusters to extract maps of the gravitational dark matter potential.
49: %
50: %
51: \end{abstract}
52: \keywords{galaxies: clusters: individual (3C~129) ---
53: cosmology: dark matter --- X-rays: galaxies: clusters}
54: \section{Introduction}
55: \label{intro}
56: %
57: While observations with the ROSAT and ASCA X-ray telescopes revealed
58: non-homogeneous temperature of the ICM for a number of clusters
59: (see e.g.\ Honda et al.\ 1996), the large scale radial temperature
60: structure remained uncertain due to ROSAT's limited spectral resolution
61: and coverage, and ASCA's non-negligible energy dependent point spread function.
62: %
63: Based on ASCA observations of 30 galaxy clusters, Markevitch et al.\ (1998)
64: found evidence for a general temperature decline with distance from
65: cluster center. However, independent ASCA studies of the same and other
66: galaxy clusters did not confirm this result (e.g., White 2000,
67: Kikuchi et al. 1999).
68: %
69: The new X-ray observatories Chandra and XMM-Newton with broad energy coverage
70: of 0.3-10~keV (Chandra) and 0.1-15~keV (XMM-Newton) and angular
71: resolutions of 0.5~arcsec (Chandra) and 15~arcsec (XMM-Newton) make it now
72: possible to assess spatially resolved ICM X-ray energy spectra with
73: sufficiently broad energy coverage and excellent angular resolutions.
74: %
75: First observations of temperature or metallicity variations have already been
76: reported for a number of clusters, e.g.\ Coma (Arnaud et al.\ 2000), A~2142
77: (Markevitch et al.\ 2000), Hydra~A (McNamara et al.\ 2000), A~1795
78: (Tamura et al.\ 2001), S\'{e}rsic 159-03 (Kaastra et al.\ 2001), and
79: Abell 1835 (Schmidt et al.\ 2001).
80: %
81: On theoretical grounds one expects that the formation of clusters
82: gives rise to substantial temperature, entropy, and metallicity gradients.
83: The interplay of gravitational collapse and heating by accretion
84: shocks produces characteristic radial temperature and entropy profiles
85: (e.g.\ Evrard 1990). Depending on when the ICM is enriched
86: with metals, also radial metallicity gradients could be observed
87: \cite{Metz:94}. Furthermore, merger of major cluster components
88: produce pronounced ICM non-uniformities, and hydro/N-body simulations predict
89: that the gradients could be observed several $10^9$ years after the onset
90: of the merger event \cite{Roet:96}.
91: %
92: The decay of these non-uniformities might be reduced by magnetic fields
93: which organize the ICM into a filamentary structure and thereby
94: substantially reduce the ICM's heat conductivity and particle
95: diffusion (Eilek 1999, but see also Narayan \& Medvedev 2001).
96:
97: In this letter we describe Chandra observations of the
98: nearby rich cluster of galaxies 3C~129.
99: Due to its low galactic latitude ($l$=160$^\circ$.5, $b$=0.3$^\circ$)
100: the cluster has not been studied intensively at optical wavelengths.
101: Based on ROSAT, Einstein, and EXOSAT data, Leahy \& Yin (2000)
102: estimated a total 0.2-10~keV luminosity of
103: 2.7$\times10^{44}$~ergs s$^{-1}$, a total ICM gas mass of
104: 3.6$\times10^{13}$~M$_{\odot}$, and a total cluster mass of
105: $\sim$5$\,\times\,10^{14}$~M$_{\odot}$. The cluster harbors two radio
106: galaxies: the prototypical head-tail galaxy 3C~129, and the
107: weaker FR~I source 3C~129.1.
108: The main emphasis of the Chandra observation had been a study of the
109: pressure balance between the ICM and the radio plasma and this study will
110: be presented elsewhere.
111:
112: In this letter we present a ``spectroscopic deprojection-analysis''
113: of the Chandra data.
114: %
115: The deprojection-approach has been the standard-technique to convert
116: observed X-ray surface brightness profiles into temperature
117: and particle density profiles (see Fabian 1994, and references therein).
118: Lacking spatially resolved X-ray energy spectra, these studies
119: used several assumptions, namely that the ICM had a certain geometry,
120: that the ICM was in hydrostatic equilibrium, and that
121: the dark matter gravitational potential could be described by
122: a certain shape and depth, together with an ICM temperature and
123: metal abundance estimate derived from a non-imaging X-ray observation.
124: %
125: The spatially resolved energy spectra from the Chandra and XMM-Newton
126: observatories now make it possible to directly derive the radial temperature,
127: particle density and metallicity profiles based only on assumptions
128: about the ICM geometry. A similar analysis has independently been
129: developed by Allen et al.\ (2001).
130:
131: %
132: The rest of this letter is organized as follows.
133: After describing the data set in Sect.~\ref{data} and the
134: spectroscopic deprojection-method in Sect.~\ref{method},
135: we will present the results of the observations in
136: Sect.~\ref{results}, and discuss their implications in Sect.~\ref{discussion}.
137: In the following we use $H_0=65$~km~s$^{-1}$~Mpc$^{-1}$ and $q_0=0.5$;
138: the cluster is thus a distance of 100~Mpc and 1~arcmin corresponds to 28.7~kpc.
139: %
140: %
141: \section{Data Set and Data Preparation}
142: \label{data}
143: \begin{figure}[t]
144: \begin{center}
145: \hspace*{-0.8cm}\resizebox{3.5in}{!}{\plotone{figure01.eps}}
146: \end{center}
147: \caption{\label{f1} \small The ACIS~I image (0.5-7.5~keV)
148: with a 32$\times$32-pixel binning.
149: The 4 ellipses are shown which have been used to characterize
150: the symmetry of the ICM's surface brightness. The contour lines indicate the
151: location of the radio galaxy 3C~129.1 at the cluster center and
152: 3C~129 in the west of the cluster, and have been derived from 330~MHz VLA
153: observations (Lane et al. 2001).
154: The small rectangles cover the ACIS~S and ACIS~I field of views
155: and show the sky-regions used for the deprojection analysis.
156: The deprojection analysis uses only the regions with their centers within
157: the outer ellipse.
158: }
159: \end{figure}
160: We use two 3C~129 pointings, one 30~ksec pointing with the ACIS~S CCDs
161: taken on December 9th, 2000, and one 10~ksec pointing with the
162: ACIS~I CCDs taken on January 9th, 2001.
163: We detect X-ray emission near the radio cores of the radio galaxies
164: 3C~129 and 3C~129.1 and exclude the corresponding sky regions,
165: as well as an additional point source in the south-east of the cluster center
166: from the following analysis.
167: %
168: %
169: The radio galaxy 3C~129 has a 15~arcmin long tail which covers a substantial
170: fraction of the field of view of the first observation at radial distances
171: from 10~arcmin to 25~arcmin from the cluster center.
172: We exclude the corresponding sky region from the analysis.
173: %
174: %
175:
176: The analysis is performed with the \verb+CIAO 2.1+ software.
177: We divide the 8.4'$\times$8.4' solid angle regions of the sky covered by each
178: ACIS chip in 4$\times$4 ``analysis regions'' of each 126''$\times$126''
179: (256$\times$256 pixels) solid angle coverage.
180: This choice is a compromise between keeping the analysis feasible and using
181: detector response functions which are as local as possible.
182: %
183: We construct background data sets with the tool \verb+make_acisbg+
184: using the blank field background data sets of M.~Markevitch.
185: Only events in the energy range from 1.2~keV to 7.5~keV are used.
186: We use a rather high lower threshold to minimize the influence
187: of uncertainties in the detector response and in the contribution of
188: the diffuse galactic X-ray emission. Above 7.5~keV the signal to noise
189: ratio is poor. Application of the same requirements on the background
190: rate as used
191: for the production of the background data sets, reduces
192: the exposure time of the first and second pointing to 19.5~ksec
193: and 9.3~ksec, respectively.
194: We use the tools \verb+calcarf+ and \verb+calcrmf+ of A.~Vikhlinin to
195: compute for each analysis region an Auxiliary Response File (ARF) and a
196: Response Matrix File (RMF) averaged over the corresponding detector area.
197: %
198: The analysis uses the ACIS~S chips S2-S4, and the ACIS~I chips I0-I4.
199: Figure \ref{f1} shows the binned ACIS~I picture together with the
200: ACIS~S and ACIS~I fields of view, the analysis regions,
201: and the location of the two radio sources.
202:
203: %
204: \section{Spectroscopic Deprojection Analysis}
205: \label{method}
206: %
207: The iso-brightness contours of 3C~129 show a slightly elliptical
208: shape with the major axis along the east-west direction.
209: Using the Chandra data together with the Einstein IPC contour map of
210: \cite{Leah:00}, we define 4 ellipses with different centers and
211: elasticities to parameterize the ICM surface brightness distribution
212: at different distances from the cluster center (see Fig.~\ref{f1}).
213: Assuming rotational symmetry around the major axis of each ellipse
214: and using an interpolation scheme we define 11 approximately
215: ellipsoidal ICM shells. For each shell $s$ ($s=$1, 2, ..., 11)
216: we compute the volumes $V(s,r)$ along the line of sight of
217: the 128 $\times$ 128 pixel analysis regions $r$
218: ($r=$1, 2, ..., 95) by simple numerical integration.
219: For each shell we define one Raymond-Smith plasma model described
220: by one set of parameters temperature, $T(s)$, particle density, $N(s)$,
221: and metal abundances, $A(s)$.
222: %
223: The algorithm works from shell to shell inwards,
224: fitting the plasma parameters of the shell under consideration
225: and using for the outer shells the best fit parameters determined before.
226: The expression minimized by the fit algorithm for shell $s$ (counting from
227: the cluster center outwards) reads:
228: \[
229: \chi^2(T(s),N(s),A(s);s)
230: )
231: \,\,=\,\,\rule{2.5cm}{0.0cm}
232: \]
233: \[
234: \sum_{r' \in s}\,
235: \sum_{I'}\,
236: \left(\!C(r',I')-
237: \,
238: \frac{c(r')}{4\pi D_L\,^2}
239: \sum_{s'=s}^{11}\!
240: \int_0^\infty\!
241: V(s',r')\;
242: \times
243: \rule{0cm}{0.55cm}
244: \right.
245: \]
246: \[
247: \kappa(N_{\rm H}(r');E')\,\,
248: j_{\rm RS}\left(T(s'),N(s'),A(s');E'\right)\,\times
249: \]
250: \begin{equation}
251: \left.
252: D_{\rm r'}(I',E')\,dE'\rule{0cm}{0.55cm}\right)^2
253: \end{equation}
254: %
255: The first sum of the right term runs over the regions $r'$ which have their
256: centers within the shell $s$ under consideration;
257: the second sum runs over the energy bins $I'$ of each energy spectrum;
258: in the third sum $s'$ runs over the shell under consideration $s$ and all
259: outer shells.
260: %
261: $C(r',I')$ is the number of counts found in energy bin $I'$ of
262: region $r'$. The constants $c(r')\,\approx\,1$ allow for a slight deviation
263: of the emissivity from the value expected from the assumed
264: ellipsoidal geometry.
265: We adjust these constants after fitting the flux normalization of a shell.
266: %
267: %
268: The value $D_{\rm L}$ denotes the luminosity distance
269: and $\kappa$ and $j_{\rm RM}$ describe
270: the photo-electric absorption and the emissivity of the
271: Raymond-Smith plasma as function of photon energy $E'$, respectively.
272: Finally, $D_{\rm r'}(I',E')$ is the detector response matrix
273: which gives the detection area times the probability with which an X-ray
274: photon of energy $E'$ will be reconstructed within the energy bin $I'$
275: (averaged over the detector coordinates of region $r'$).
276: %
277: Due to the location of the cluster near the galactic plane we anticipated
278: substantial variation of the neutral hydrogen column density across the
279: field of view. We therefore performed radio observations with the Dominion
280: Astrophysical Observatory to obtain $N_{\rm H}$-values with
281: arcminute resolution (principal investigator: T.\ Willis).
282: Although the $N_{\rm H}$-values change by 25\% over the
283: 1.25$^\circ$$\times$2.5$^\circ$ field of view,
284: the Chandra observations lie in a region of rather constant values
285: of between 0.84$\times$$10^{22}$~cm$^{-2}$ and 0.9$\times$$10^{22}$~cm$^{-2}$.
286: We use in the following the $N_{\rm H}$-values
287: from the radio observations. Independent determination of the mean
288: $N_{\rm H}$-value from Chandra data gives consistent results.
289: Note that our inferred hydrogen column densities deviate from those
290: inferred by Leahy et al. (2000)
291: ($N_{\rm H}$ = (0.57$\pm$0.03)$\times$$10^{22}$~cm$^{-2}$)
292: from fitting the combined ROSAT and EXOSAT data with a one-temperature
293: Raymond-Smith model.
294: For each shell $s$ we quote as radial distance $R_s$ the mean
295: distances of all analysis regions contributing to the plasma fit of this shell.
296: We define the distance of a point from the cluster center
297: as the half-diameter of the major axis of the associated ellipse.
298:
299: Technically the deprojection-analysis is performed with the
300: \verb+CIAO+-tool \verb+sherpa+ running on a single
301: script which contains all the information about the
302: geometry of the shells and analysis regions.
303: %
304: %
305: All errors are quoted on 90\% confidence level.
306: We estimate the increase of the statistical error on the parameters
307: of the inner shells due to the statistical uncertainty of the fit-parameters
308: of the outer shells with a simple Monte Carlo simulation and
309: accordingly scale the errors on the fit-parameters.
310: The uncertainty of the background normalization is estimated to be 10\%
311: \cite{Vikh:01}. Determining the associated error in the parameters by
312: repeating the full deprojection analysis with a background scaled
313: up and down by 10\%, we add these errors in quadrature to the
314: statistical ones.
315: The results of the outermost ICM shell depend on the assumed
316: shell thickness. Since this is an inherent limitation of the deprojection
317: method we do not show the fit results of the 11th ICM shell in the
318: following.
319: %
320: The fits give satisfactory $\chi^2$-values for 95 out of the 104
321: analysis regions. For 9 regions large $\chi^2$-values indicate a deviation
322: of the plasma properties from the ellipsoidal shell or contamination
323: by a field source, and we exclude these regions from the overall fit.
324: %
325: \section{Results}
326: \label{results}
327: %
328: %
329: %
330: The Chandra image shows statistically significant evidence for X-ray emission
331: out to 19 arcmin from the cluster core. We do not find evidence for
332: abrupt changes in the ICM surface brightness, so we have no direct evidence
333: for shocks or contact discontinuities.
334:
335: Figure \ref{f2} shows the results of the deprojection analysis.
336: Averaged over all shells, the mean cluster temperature is 5.1$\pm$0.3~keV.
337: We have weak evidence for a hotter cluster core with
338: $k_{\rm B}T\,=$ 7.9$^{+4.4}_{-2.4}$~keV at $R_1=$ 1.4~arcmin and
339: a temperature decrease at the outermost 10th shell with
340: $k_{\rm B}T\,=$ 3.5$^{+1.2}_{-0.9}$~keV $R_{10}=$ 18.1~arcmin
341: ($k_{\rm B}$ is the Boltzmann constant).
342: Various estimates show that the low temperature of the outermost shell
343: might be an artifact due to the soft diffuse galactic X-ray emission
344: which begins to substantially contaminate the cluster emission at these
345: distances from the cluster center.
346: %
347: The dotted line of Fig.~\ref{f2}(a) shows the projected temperatures
348: determined by fits of one component Raymond-Smith models.
349: Naturally, the deprojected temperature varies more than the projected
350: temperature which averages over several ICM shells.
351:
352: The particle density (Fig.~\ref{f2}(b)) decreases monotonically
353: from 3.1$\times10^{-3}$ cm$^{-2}$ at the cluster center
354: to 3.9$\times10^{-4}$ cm$^{-3}$ at the cluster periphery.
355: The errors on the metal abundances of individual ICM shells are rather large
356: (typically about 0.4 solar abundances) and we do not show them here.
357: The mean value averaged over all shells is 0.2$\pm0.1$ solar abundance.
358: %
359: The thermal gas pressure, defined as $p\,$= $N\,k_{\rm B}\,T$
360: is shown in Fig.~\ref{f2}(c).
361: The entropy per particle relative to the mean value of all shells
362: $s_0$ is shown in Fig.~\ref{f2}(d) and has been computed
363: from $s=$ $\frac{3}{2}$ $k_{\rm B}$ $\ln{(T\,\rho^{-2/3})}$ -
364: $s_0$ with $\rho\approx$ $0.6$ $m_{\rm P}$ $N$ and $m_{\rm P}$
365: the proton mass. The entropy profile will be discussed in more detail in
366: the next section.
367: %
368: Finally, the radiative cooling times are shown in
369: Fig.~\ref{f2}(e).
370: The cooling times have been computed according to
371: $t_{\rm cool}\,=$ $\frac{3}{2}$ $N$ $k_{\rm B}$ $T$
372: $(\Lambda(N,T))^{-1}$ with the cooling function $\Lambda$ which denotes the
373: plasma emissivity including thermal bremsstrahlung and
374: line emission processes \cite{Suth:93}.
375: All radiative cooling times exceed the Hubble time, the
376: shortest cooling time being 30$^{+20}_{-9}$~Gyr.
377: %
378: %
379: %
380: %
381: \begin{figure}[th]
382: \begin{center}
383: \resizebox{2.9in}{!}{\plotone{figure02.eps}}
384: \vspace*{-0.5cm}
385: \end{center}
386: \caption{\label{f2} \small Results of the deprojection
387: analysis showing from top to bottom:
388: (a) temperature $T$, (b) particle density $N$,
389: (c) gas pressure, (d) entropy, and (e) radiative cooling times.
390: The dotted line in panel (a) shows the projected temperature,
391: and the dashed line in panel (d) is for illustrative purposes only
392: (see text).}
393: \vspace*{-0.5cm}
394: \end{figure}
395: %
396: \section{Discussion}
397: \label{discussion}
398: %
399: Based on a deprojection analysis of ROSAT surface brightness data
400: Leahy \& Yin (2000) inferred a cluster temperature decreasing from
401: $\simeq$6~keV at the cluster periphery (at $\simeq$20 arcmin from the
402: cluster core) to about $\simeq$3~keV at the cluster center.
403: The Chandra imaging spectroscopy data however, does not confirm this
404: temperature decrease. On the contrary, both, the projected and deprojected
405: temperature profiles indicate a hot cluster core.
406: %
407: The discrepancy might be explained by one or more of the numerous assumptions
408: entering the deprojection of surface brightness images mentioned in the
409: introduction, as e.g.\ a true dark matter core radius substantially deviating
410: from the assumed one.
411:
412: A hot cluster core might be the result of shock heating.
413: This possibility can be studied with the entropy per particle which
414: is able to distinguish between ICM heating by adiabatic compression or
415: by shocks. By definition, adiabatic compression does not increase the
416: ICM entropy, while shock heating, converting the energy of bulk plasma
417: motion into heat, does increase it.
418: %
419: A possible interpretation of the entropy profile is based on the dashed line
420: in Fig.~\ref{f2}(d) which shows a fit to the entropy of the second to
421: tenth shell and suggests that the entropy increases with distance
422: from the cluster center. Indeed, in the framework of hierarchical clustering
423: scenarios one expects that cluster formation through mass accretion
424: produces a radially increasing ICM entropy as consequence of the
425: growing strength of the cluster accretion shock as the total virialized
426: cluster mass accumulates (see e.g.\ Tozzi et al.\ 2001, and references
427: therein).
428: %
429: The large entropy of the cluster core might be the result of shock heating
430: by a jet or wind from the central radio galaxy 3C~129.1.
431: The required power is modest: $\simeq 10^{43}$
432: $\frac{\Delta (k_{\rm B}\,T)}{\rm 5\,keV}$
433: $\frac{t_{\rm 3C129.1}}{\rm 100~Myr}$ ergs s$^{-1}$
434: where $\Delta (k_{\rm B}\,T)$ is the temperature increase at the
435: cluster center and $t_{\rm 3C129.1}$ denotes the lifetime of the source.
436: However, at the cluster center we do not find structure in the ICM surface
437: brightness in direct support of this hypothesis.
438: %
439:
440: The analysis presented here is plagued by large statistical
441: and systematic errors. The first stem from the modest integration time
442: and the low surface brightness of 3C~129.
443: The latter mainly results from the asymmetric shape of the ICM and possible
444: contribution of the diffuse galactic X-ray emission at larger cluster core
445: distances.
446: For brighter, more symmetric clusters the spectroscopic deprojection technique
447: opens the possibility to determine the shape of the dark matter potential.
448: The only uncertainties of such an analysis derive from the uncertain
449: ICM geometry and from undetected ICM pressure components as kinetic pressure,
450: magnetic field pressure and Cosmic Ray
451: pressure.
452: \\[2ex]
453: %
454: {\it Acknowledgments:} The author would like to thank H.\ V\"olk and
455: D.\ Harris for fruitful discussions on X-ray emission from galaxy cluster
456: and AGN jets. W.M.~Lane kindly provided the 330~MHz VLA map shown in
457: Fig.~\ref{f1}, and T.~Willis the $N_{\rm H}$-data used for the deprojection
458: analysis. Very helpful comments by an anonymous referee are gratefully
459: acknowledged.
460: This research has been supported by the NASA (NAS8-39073 and GO 0-1169X).
461: %
462: \begin{thebibliography}{}
463: {\small
464: \bibitem[Allen et al.\ 2001]{Alle:01}{Allen, S.W., Ettori S., Fabian A.C.\
465: 2001, MNRAS , 324, 877}
466: \bibitem[Arnaud et al.\ 2000]{Arna:00}{Arnaud, M., Aghanim, N., Gastaud, R.,
467: et al.\ 2000, A\&A, 365, L67}
468: \bibitem[Edge \& Stewart 1991]{Edge:91}{Edge, A.C., Stewart, G.C.\ 1991,
469: MNRAS, 252, 414}
470: \bibitem[Evrard 1990]{Evra:90}{Evrard, A.E.\ 1990, ApJ, 363, 349}
471: \bibitem[Eilek 1999]{Eile:99}{Eilek, J.\ 1999, MPE Report 271, 71}
472: \bibitem[Fabian 1994]{Fabi:94}
473: {Fabian, A.\ C.\ ARAA, 1994, 32, 277}
474: \bibitem[Honda et al.\ 1996]{Hond:96}{Honda, H., et al. 1996, ApJ, 473, L71}
475: \bibitem[Kaastra et al.\ 2001]{Kaas:01}{Kaastra, J.S., Ferrigno, C.,
476: Tamura, T., et al.\ 2001, A\&A, 365, L99}
477: \bibitem[Kikuchi et al. 1999]{Kiku:99}{Kikuchi, K., Furusho, T., Ezawa, H.,
478: Yamasaki, N., Ohashi, T., Fukazawa, Y., Ikebe, Y.\ 1999, PASJ, 51, 301}
479: \bibitem[Lane et al.\ 2001]{Lane:01}{
480: Lane, W.M., Harris, D.E., Ensslin, T.A., Kassim, N.E., Perley, R.A.\
481: 2001, AAS, 199, 9814}
482: \bibitem[Leahy \& Yin 2000]{Leah:00}{Leahy, D.A., Yin, D.\ 2000,
483: MNRAS, 313, 617}
484: %
485: \bibitem[Markevitch et al.\ 1998]{Mark:98} {Markevitch, M.,
486: Forman, W.R., Sarazin, C.L., Vikhlinin, A.\ 1998, ApJ, 503, 77}
487: \bibitem[Markevitch et al.\ 2000]{Mark:00a} {Markevitch, M., Ponman,
488: T.J., Nulsen, P.E., et al.\ 2000, ApJ, 541, 542}
489: \bibitem[Metzler \& Evrard 1994]{Metz:94}
490: {Metzler, C.A., Evrard, A.E.\ 1994, ApJ, 437, 564}
491: \bibitem[McNamara et al.\ 2000]{McNa:00}
492: {McNamara, B.R., Wise, M., Nulsen, P.E.J., et al.\ 2000, ApJ, 534, L135}
493: \bibitem[Narayan \& Medvedev 2001]{Nara:01}{
494: Narayan, R., Medvedev, M.V., ApJ, 562, L129}
495: \bibitem[Roettiger et al.\ 1996]{Roet:96}
496: {Roettiger, K., Burns, J.O., Loken, C.\ 1996, ApJ, 473, 651}
497: \bibitem[Schmidt et al.\ 2001]{Schm:01}{Schmidt, R.W., Allen S.W., Fabian A.C.\
498: 2001, MNRAS, 327, 1057}
499: \bibitem[Sutherland \& Dopita 1993]{Suth:93} {Sutherland, R.S., Dopita, M.A.\
500: 1993, ApJS, 88, 235}
501: \bibitem[Tamura et al.\ 2001]{Tamu:01}
502: {Tamura, T., Kaastra, J.S., Peterson, J.R., et al.\ 2001, A\&A, 365, L87}
503: \bibitem[Tozzi \& Norman 2001]{Tozz:01}
504: {Tozzi, P., Norman, C.\ 2001, ApJ, 546, 63}}
505: \bibitem[Vikhlinin et al.\ 2001]{Vikh:01}{Vikhlinin, A., Markevitch, M., Murray, S.S.\ 2001, ApJ, 551, 160}
506: \bibitem[White 2000]{Whit:2000}{White, D. A. 2000, MNRAS, 312, 663}
507: \end{thebibliography}
508: \end{document}
509: