1:
2: \documentclass[12pt,preprint]{aastex}
3: %\documentclass{emulateapj}
4:
5: \shorttitle{Two Clusters With Radio Quiet Cooling Cores}
6: \shortauthors{Donahue et al.}
7:
8:
9: %\usepackage{graphics,graphicx,epsf,natbib,emulateapj5}
10: \usepackage{graphics,graphicx,epsf,natbib}
11: \bibliographystyle{apj}
12:
13: %% Syntax:
14: %% In the graphics package:
15: %%
16: %% \begin{figure}
17: %% \includegraphics[llx,lly][urx,ury]{file}
18: %% \end{figure}
19: %%
20: %% where ll denotes 'lower left' and ur 'upper right' and the x and y
21: %% values are the coordinates of the PostScript bounding box in
22: %% points. There are 72 points in an inch.
23: %%
24: %% In the graphicx package:
25: %%
26: %% \begin{figure}
27: %% \includegraphics[key=val,key=val,...]{file}
28: %% \end{figure}
29: %%
30: %% where some of the useful keys are: angle, width, height,
31: %% keepaspectratio (='true' or 'false') and scale. Bounding box values
32: %% can be given as [bb=llx lly urx ury].
33: %%
34: \received{}
35:
36: \def\Del{\Delta}
37:
38:
39: \begin{document}
40: \title{Two Clusters of Galaxies with Radio-Quiet Cooling Cores}
41: \author{Megan Donahue and G. Mark Voit}
42: \affil{Michigan State University, Physics \& Astronomy Dept., East Lansing, MI 48824-2320}
43: \email{donahue@pa.msu.edu, voit@pa.msu.edu}
44: \author{Christopher P. O'Dea}
45: \affil{Dept. of Physics, Rochester Institute of Technology,
46: 84 Lomb Memorial Dr., Rochester, NY 14623-5603}
47: \author{Stefi A. Baum}
48: \affil{Center for Imaging Science, Rochester Institute of Technology,
49: 54 Lomb Memorial Dr., Rochester, NY 14623-5603}
50: \author{William B. Sparks}
51: \affil{Space Telescope Science Institute, 4700 San Martin Dr, Baltimore, MD 21212}
52:
53: \begin{abstract}
54: Radio lobes inflated by active galactic
55: nuclei at the centers of clusters are a promising candidate for
56: halting condensation in clusters with short central cooling times because they
57: are common in such clusters. In order to test the AGN-heating hypothesis, we obtained
58: {\em Chandra} observations of two clusters with
59: short central cooling times yet no evidence for AGN activity: Abell 1650 and Abell 2244.
60: The cores of these clusters indeed appear systematically different from cores with
61: more prominent radio emission. They do not have significant central temperature gradients,
62: and their central entropy levels are markedly higher than in clusters with stronger radio
63: emission, corresponding to central cooling times $\sim 1$~Gyr. Also, there is no
64: evidence for fossil X-ray cavities produced by
65: an earlier episode of AGN heating. We suggest that either
66: (1) the central gas has not yet cooled to the point at which feedback
67: is necessary to prevent it from condensing, possibly because it is conductively stabilized,
68: or (2) the gas experienced a major heating event $\gtrsim 1$~Gyr in the past and has
69: not required feedback since then.
70: The fact that these clusters with no evident feedback have higher central entropy
71: and therefore longer central cooling times than clusters with obvious AGN feedback
72: strongly suggests that AGNs supply the feedback necessary to suppress condensation
73: in clusters with short central cooling times.
74: \end{abstract}
75: % 6 keywords allowed
76: \keywords{galaxies:clusters:general --- galaxies:clusters:individual (A1650) ---
77: galaxies:clusters:individual (A2244) --- cooling flows}
78:
79: \section{Introduction}
80:
81:
82: The cooling-flow problem in clusters of galaxies has been one of the most
83: notorious issues in galaxy formation.
84: The cooling time ($t_c$) of gas within the central $100-200$~kpc of many
85: clusters is less than a Hubble time
86: \citep[e.g. ][]{CowieBinney77, FabianNulsen77}.
87: If there is no compensating heat source distributing
88: thermal energy over that same region, that gas ought to cool, condense, and relax
89: toward the cluster's center in a so-called ``cooling flow,''
90: but exhaustive searches in other wave bands have failed
91: to locate the $10^{12}-10^{13} \, M_\odot$ of stars or cool gas that should have
92: accumulated
93: \citep[e.g. ][]{ODea1994, Antonucci1994, McNamaraJaffe1994}.
94: Nevertheless, something unusual is happening in clusters with $t_c \ll H_0^{-1}$.
95: Significantly smaller amounts of gas have been detected in the form of
96: CO \citep{Edge2002, EdgeFrayer2003} or HI \citep{1994ApJ...436..669O},
97: vibrationally excited H$_2$ \citep{Donahue2000, Falcke1998, JaffeBremer97},
98: and evidence for star formation
99: \citep[e.g. ][]{Cardiel1998,Crawford1999, VD97, ODea2004}
100: are common in these systems, and {\em Chandra} observations have
101: shown that radio lobes sometimes carve out huge cavities in
102: the X-ray emitting gas at the centers of such clusters
103: \citep[e.g., ][]{McNamaraA2597_2001, Fabian2000_NGC1275, Blanton2003}.
104:
105: This association of star formation, line emission, and
106: relativistic plasma with cooling-flow clusters has fed
107: speculation that feedback from active galactic nuclei
108: modulates the condensation of hot gas, greatly
109: reducing the mass-cooling rates naively inferred from
110: X-ray imaging \citep[e.g., ][]{Bohringer2002, Quilis2001}.
111: However, active feedback sources
112: are not found in every cluster with $t_c \ll H_0^{-1}$.
113: For example, the nearby cooling-flow sample of
114: \citet{Peres1998}
115: consists of twenty-three clusters with
116: $\dot{M} > 100 \, M_\odot \, {\rm yr}$ inferred from {\em ROSAT}
117: imaging. Of these, thirteen have both an emission-line nebula
118: and a strong radio source, two have no emission lines but a strong
119: radio source (A2029, A3112), and three have emission lines but a
120: weak radio source (A478, A496, A2142) leaving five with no
121: emission lines and little or no radio activity (A1651, A2244,
122: A1650, A1689, A644).
123:
124: To test the idea that feedback from either an AGN, star formation,
125: or some combination of the two suppresses cooling in the cores of
126: clusters with $t_c \ll H_0^{-1}$, we observed two objects from
127: this last set of five with {\em Chandra}: A1650 ($z=0.0845$) and A2244 ($z=0.0968$).
128: These clusters are luminous X-ray
129: sources, with bolometric
130: $L_x \sim 8 \times 10^{44} h_{70}^{-2}$ erg s$^{-1}$ and
131: estimated gas $T_x$ of 5.5-7.0 keV \citep{David1993}.
132: Here we compare those clusters with an archival sample of clusters of
133: similar X-ray luminosities ($L_x=0.4-30 \times 10^{44}$ erg s$^{-1}~h_{70}^{-2}$) and temperatures
134: ($T_x=2.9-7.4$ keV), with
135: $t_c \ll H_0^{-1}$, with evidence for active feedback in the form of central
136: radio emission, and in most cases, with emission-line nebulae as well
137: \citep{Donahue2005B}. We will refer to these clusters as "active clusters."
138: All of the clusters in the \citet{Donahue2005B} sample and the two clusters
139: discussed in this paper have single, optically luminous, brightest central galaxies
140: residing at the centroid of their X-ray emission.
141: \S~2 describes the observations and calibration procedures.
142: \S~3 describes the data analysis and the extraction of entropy
143: profiles, and \S~4 discusses our results, which we summarize
144: in \S~5. For this paper we assume $H_0=70$ km s$^{-1}$ Mpc$^{-1}$
145: and a flat universe where $\Omega_M=0.3$.
146:
147:
148: \section{Observations and Calibration}
149:
150: The observation dates, flare-free exposure times, and count rates between 0.5-9.5 keV within
151: a 4' radius aperture are reported in
152: Table~\ref{log}. The back-illuminated CCD on the Chandra X-ray Observatory
153: \citep{Chandra2002}, the ACIS-S3 detector, was used for its sensitivity
154: to soft X-rays. Its field of view ($8\arcmin \times 8\arcmin$)
155: extends to about 10\% of the virial radius of each cluster, limiting our analysis
156: to the cluster cores.
157:
158: \begin{deluxetable}{lccc}
159: \tablecaption{Chandra Observations \label{log}}
160: \tablehead{
161: \colhead{Cluster} & \colhead{Observation Date} & \colhead{Exposure Time} &
162: \colhead{ACIS-S Count Rate} \\
163: \colhead{} & \colhead{} & \colhead{(s)} & \colhead{(ct s$^{-1}$)} }
164: \startdata
165: Abell 1650 & Aug 3-4, 2003 & 27,260 & 4.36 \\
166: Abell 2244 & Oct 10-11, 2003 & 56,965 & 4.35 \\
167: \enddata
168: \end{deluxetable}
169:
170: We processed these datasets using the {\em Chandra} calibration software
171: CALDB 2.29 and CIAO 3.1, released in July 2004\footnote{Chandra Interactive
172: Analysis of Observations (CIAO), http://cxc.harvard.edu/ciao/}.
173: Neither observation experienced flares.
174: We used Chandra deep background observations for our background spectra.\footnote{M. Markevitch, author of http://asc.harvard.edu/cal/ Acis/Cal\_prods/bkgrnd/acisbg/ COOKBOOK}
175: Source and background spectra were extracted using identical concentric
176: annuli containing a minimum of 20,000 counts per source spectrum.
177: Bright point sources were excluded from the event files
178: before spectral extraction. The spectra were binned to a minimum of 25 counts per
179: energy bin.
180:
181: Using XSPEC v11.3.1, we fit the projected and deprojected spectra from 0.7-7.0 keV to
182: MekaL models \citep{MekaL} with Galactic absorption attenuating the
183: soft X-rays \citep{MM1983}.
184: Since the best-fit absorption overlapped the Galactic values of $N_{\rm H}=1.56$ and $2.3 \times 10^{20}$
185: for A1650 and A2244 respectively \citep{DickeyLockman1990}, we fixed $N_{\rm H}$ at
186: those values for this analysis. The positions of the Fe-K lines were consistent with the cluster redshifts from galaxy velocities in \citet{StrubleRood1999}.
187: We computed 90\% uncertainties ($\Del \chi^2 =2.71$) for the temperature,
188: normalization, and metallicity at each
189: annulus. We constrained the metallicity to be
190: constant across 2-3 annuli. The reduced $\chi^2$ values for the fits
191: were typically 1.10-1.15. More details about our data analysis strategy and further analyses
192: are described in \citet{Donahue2005B}, where
193: we also analyze Chandra archival observations of nine other cooling-flow clusters that have central
194: radio sources and emission-line nebulae.
195:
196: Neither cluster exhibits a strong temperature gradient across the core. Abell 2244
197: is nearly isothermal with $kT = 5.5 \pm 0.5$ keV at every radius $<4'$, and Abell 1650
198: varies from $5.5 \pm 0.5$ keV in the core to $7.0 \pm 1.0$ keV at 4', statistically
199: consistent with but somewhat higher than the temperature profile over similar radii obtained with
200: XMM measurements.
201: (\citet{TakahashiYamashita2004} adopted a lower redshift ($z=0.0801$) to fit the XMM data,
202: which may indicate a calibration uncertainty.) These small inner temperature gradients contrast with those of most other cooling flow clusters,
203: which tend to be more pronounced. Both cluster cores contain a significant metallicity gradient,
204: ranging from 0.6-0.8 solar in the center to a more typical 0.2-0.3 solar outside the core,
205: consistent with \citet{TakahashiYamashita2004}. This metallicity
206: pattern is typical of the other cooling flow clusters we studied.
207:
208: \section{Data Analysis}
209:
210: The goal of our data analysis was to determine whether or not clusters without
211: obvious signatures of feedback were systematically different from those with radio sources
212: that do show signatures of feedback. Our primary results are that these two clusters
213: do not show any evidence for ghost cavities and have higher central entropy levels
214: than clusters showing evidence for feedback.
215:
216: In order to search for cavities in the intracluster medium, we adaptively smoothed
217: the X-ray data to a minimum significance of 5-sigma with both a Gaussian and a top-hat kernel.
218: We found no ``ghost bubbles." On scales larger than about 50 kpc from the center, both clusters exhibited regular, nearly round intensity contours. We also did not see evidence for
219: filaments, such as that found tracing the H$\alpha$ emission in
220: Abell 1795 \citep{2001MNRAS.321L..33F} or M~87 \citep{Sparks2004}.
221:
222: We determined the entropy profiles of these clusters by computing the
223: adiabatic constant $K = kTn_e^{-2/3}$ at each radius to quantify the
224: specific entropy. The temperature ($kT$) profiles were measured as described
225: in \S2. The electron density profiles ($n_e$) were derived by deprojecting
226: the 0.5-2.0 keV surface brightness profiles within
227: annuli having 5" widths using the technique of \citet{KCC1983}. The
228: uncertainties of the deprojected count rate profiles were estimated by bootstrapping
229: 1000 monte-carlo simulations of the original surface brightness profiles.
230: A spatially-dependent conversion of 0.5-2.0 KeV count rates to electron
231: densities was obtained from the X-ray spectroscopy. For this paper,
232: we assumed that the temperature and the count-rate conversion factor in the
233: central bin were constants.
234:
235:
236: \begin{figure}
237: %\includegraphics*[width=0.5\textwidth,angle=0]{rquiet.ps}
238: \includegraphics*[width=0.5\textwidth,angle=0]{f1.ps}
239: \caption{Entropy profiles for Abell 1650 and Abell 2244. All gas at $\sim5$ keV with entropy $\lesssim170$ keV cm$^2$ has $t_c < H_0^{-1}$. The hatched region shows the locus of
240: entropy profiles for active clusters from the sample of \cite{Donahue2005B}.
241: \label{EntropyProfiles}}
242: \end{figure}
243:
244: Figure~\ref{EntropyProfiles} shows that the entropy profiles of Abell 1650 and Abell 2244 are systematically different from the nine cooling-flow clusters in the sample of active clusters from \cite{Donahue2005B}.
245: The two radio-quiet clusters have flatter entropy profiles with larger values of central entropy.
246: To quantify this difference, we fit both a simple power law of $K = K_{100}(r/100~\rm{kpc})^{\alpha}$
247: and the same power law plus a central entropy $K =K_0 + K_{100}(r/100~\rm{kpc})^{\alpha}$ to the
248: entropy profiles, as was done for the active clusters in \citet{Donahue2005B}.
249: Table~\ref{table:entropyfits}
250: gives the best fits. We find that $\alpha \approx 0.6-0.8$ and $K_0 \approx 30-50 \,{\rm keV \, cm^2}$
251: in the radio-quiet clusters, in contrast to $\alpha \sim 1$ and $K_0 \approx 10 \,{\rm keV \, cm^2}$
252: for the active clusters. Figure~\ref{figure:EntPower} shows central entropy values
253: plotted as a function of 20~cm radio power, from the NVSS \citep{NVSS}. Abell 2244 has a weak, off-center
254: radio source that may not be associated with the cluster, plotted as an upper limit.
255:
256: \begin{deluxetable}{lccccc}
257: \tablecaption{Entropy Profile Fit Results\label{table:entropyfits}}
258: \tablehead{
259: \colhead{Cluster} & \colhead{$K_0$} & \colhead{$K_{100}$} & \colhead{$\alpha$} &
260: \colhead{Total $\chi^2$ }&\colhead{ N } \\
261: \colhead{} & \colhead{KeV cm$^2$} & \colhead{KeV cm$^2$} & \colhead{} & \colhead{} & \colhead{(d.o.f.)} }
262: \startdata
263: Abell 1650 & $27\pm5$ & $150\pm7$ & $0.80\pm0.07$ & 12 & 47 \\
264: & $=0.00$ & $177$ & $0.56\pm0.02$ & 28 & 48 \\
265: Abell 2244 & $48\pm5$ & $ 102\pm8$ & $0.97\pm0.08$ & 7 & 31 \\
266: & $=0.00$ & $162\pm3$ & $0.54\pm0.02$ & 42 & 32 \\
267: Active Sample & $8\pm4$ & $150\pm50$ & $1.2\pm0.2$ & & \\
268: & $=0.00$\tablenotemark{*} & $144\pm24$ & $0.96\pm0.15$ & & \\
269: \enddata
270: \tablenotetext{*}{The fits set to $0.00$ entropy in the cores for the sample in Donahue
271: et al. (2005B) were quite poor, except for the case of Abell 2029.}
272: \end{deluxetable}
273:
274:
275: \begin{figure}
276: %\includegraphics*[width=0.5\textwidth,angle=0]{entropy_power.eps}
277: \includegraphics*[width=0.5\textwidth,angle=0]{f2.eps}
278: \caption{ Radio power $\nu L_\nu$ from 20 cm observations \citep{NVSS}
279: (See also \citet{LedlowOwen1995} and \citet{SBO1995}, and 6
280: cm upper limits for A2244 and A1650 from \citet{Burns1990}.)
281: \label{figure:EntPower} }
282: \end{figure}
283:
284:
285:
286:
287:
288:
289: \section{Discussion}
290:
291: The significance of elevated central entropy in Abell~1650 and Abell~2244 is that
292: a larger central entropy implies a longer central cooling time compared to
293: clusters in \cite{Donahue2005B} of similar temperature. Assuming pure free-free
294: cooling, the cooling time for gas of temperature $T$ and entropy $K$ is
295: \begin{equation}
296: t_c \approx 10^8 \, {\rm yr} \left( \frac {K} {10 \, {\rm keV cm^2}} \right)^{3/2}
297: \left( \frac {kT} {5 \, {\rm keV}} \right)^{-1} \; \; .
298: \end{equation}
299: Thus, these two clusters, which show no evidence for feedback, have a central cooling
300: time $\sim 1 \,{\rm Gyr}$, while those that do show evidence for feedback have a central
301: cooling time $\sim 0.1 \, {\rm Gyr}$. According to the definition of \citet{Peres1998}, Abell~1650
302: and Abell~2244 were properly classified as cooling-flow clusters because $t_c < 5$~Gyr.
303: However, one does not expect to see significant cooling and condensation of gas
304: in these clusters for at least another $\sim 5 \times 10^8 \,{\rm yr}$. In other words, evidence
305: for feedback is seen in those clusters that can trigger it on a $\sim 10^8 \, {\rm yr}$ timescale
306: and not in clusters in which gas is not currently expected to be condensing. Here we discuss the
307: implications of this finding.
308:
309: The most straightforward interpretation of the cooling-time dichotomy between active
310: and radio-quiet clusters is that radiative cooling in cluster cores triggers AGN feedback
311: when the central gas begins to condense. \citet{Donahue2005B} find that all nine of
312: their active clusters have very similar core entropy profiles, suggesting that this set
313: of clusters has settled into a quasi-steady configuration that is episodically heated by
314: AGN outbursts on a $\sim 10^8$~year timescale. \citet{VoitDonahue2005} show that
315: outflows of $\sim 10^{45} \, {\rm erg \, s^{-1}}$ naturally maintain the
316: observed characteristics of the entropy profiles in these clusters.
317:
318: If that is the correct interpretation, then it is possible that Abell~1650 and Abell~2244 have
319: unusually long cooling times because they each experienced unusually strong AGN outbursts
320: $\gtrsim 1$~Gyr in the past. Raising the central entropy to the observed $\sim 30-50 \,
321: {\rm keV cm^2}$ levels would require an AGN outflow $\sim 10^{46} \, {\rm erg \, s^{-1}}$ \citep{VoitDonahue2005}. Such outbursts are rare but not unprecedented.
322: \citet{McNamara2005} have recently observed an outburst of this magnitude in
323: MS0735+7421, which now has a central entropy $\sim 30 \, {\rm keV \, cm^2}$.
324: The long cooling time following such an outburst would account for why we do
325: not see any sign of X-ray cavities in these two clusters.
326:
327: It is also possible that the central gas in Abell~1650 and Abell~2244 has never cooled
328: to the point at which it can trigger a strong AGN outburst. That could happen, for example, if
329: frequent merger shocks have been able to support the core entropy at the $\sim 50 \,
330: {\rm keV \, cm^2}$ level for several Gyr, or if electron thermal conduction can resupply the
331: thermal energy radiated by the central gas. One can evaluate the efficacy of thermal
332: conduction by comparing the size of a radiatively cooling system to the Field length
333: \begin{equation}
334: \lambda_{\rm F} = \left( \frac {\kappa T} {n_e^2 \Lambda} \right)^{1/2}
335: \approx 4 \, {\rm kpc} \, \left( \frac {K} {10 \,{\rm keV \, cm^2}} \right)^{3/2} f_c^{1/2}
336: \; \; ,
337: \end{equation}
338: where $\Lambda$ is the usual cooling function and $\kappa = 6 \times 10^{-7} \, f_c T^{5/2}
339: \, {\rm erg \, s^{-1} \, cm^{-1} \, K^{-7/2}}$ is the Spitzer conduction coefficient with suppression
340: factor $f_c$. The approximation assumes free-free cooling ($\Lambda \propto T^{1/2}$), which conveniently
341: makes $\lambda_{\rm F}$ a function of entropy alone. At radii $\sim 100 \, {\rm kpc}$, we find that
342: $\lambda_{\rm F} \sim r$ in all the cooling-flow systems we have studied, implying that conduction
343: can plausibly balance cooling there, as long as $f_c \sim 1$. At radii $\sim 10 \, {\rm kpc}$ in systems
344: with signs of feedback, we find $\lambda_{\rm F} < r$ even for $f_c = 1$, implying that conduction
345: cannot balance cooling at small radii, in agreement with the findings of \citet{2004MNRAS.347.1130V}.
346: At those same small radii in the two systems without signs of feedback, we find $\lambda_{\rm F}
347: \approx r$ for $f_c \approx 1$, suggesting that these systems are potentially stabilized by
348: thermal conduction, which would account for their modest temperature gradients.
349:
350: One speculation that emerges from this brief analysis of thermal conduction is that there
351: is a critical entropy profile $K(r) \approx 10 \, {\rm keV \, cm^2} \, f_c^{-1/3} (r/4 \, {\rm kpc})^{2/3}$
352: dividing conductively stabilized systems from those that require feedback. Clusters with
353: central entropy profiles below this line will continue to cool until some other heat source
354: intervenes, while conduction stabilizes those clusters above the line. One would then expect
355: the cluster population to bifurcate into systems with strong central temperature gradients
356: and feedback and those without either. Furthermore, a very powerful AGN outburst
357: could induce a transition from a feedback-stabilized state to a conductively-stabilized state by
358: raising the central entropy level to $\gtrsim 30 \, {\rm keV \, cm^2}$.
359:
360: Another potential heat source that has been suggested as a solution to the cooling-flow problem
361: is annihilation of dark matter particles such as neutralinos \citep{QinWu01,Totani2004}. In the
362: model of \citet{Totani2004}, the annihilation rate peaks in the center because of a spike in the
363: density profile owing to the central black hole. The steady heating rate in this model is not linked
364: as directly to baryon cooling as the AGN feedback model suggested here, but it is an interesting
365: alternative mechanism that could be explored further.
366:
367:
368:
369: \section{Conclusions}
370: In order to test whether AGN heating compensates for radiative cooling in the
371: cores of clusters of galaxies, we have
372: used {\em Chandra} to observe a small sample consisting of two clusters with central cooling
373: times $< H_0^{-1}$,
374: yet no evidence for prominent AGN activity: Abell 1650 and Abell 2244. The X-ray properties of the
375: cores of these clusters indeed appear systematically different from cores with
376: more prominent radio emission. While the central cooling times are shorter
377: than a Hubble time and they have strong metallicity gradients,
378: they do not have significant central temperature gradients,
379: and their central entropy levels are markedly higher than in clusters with stronger radio
380: emission, corresponding to central cooling times of a billion years. Also, there is no
381: evidence in the X-ray surface brightness maps for fossil X-ray cavities produced by
382: a relatively recent episode of AGN heating. In contrast to the central cores of the clusters
383: with stronger radio emission, these cores may be stabilized by conduction if it is
384: operating at close the Spitzer rate. We suggest that a tremendous AGN outburst, such
385: as that shocking the ICM in MS0735+74 \citep{McNamara2005} may have elevated the central
386: entropy of these clusters some $10^9$ years ago. Whether or not conduction is operative
387: in stabilizing these clusters cannot be determined, but it is energetically feasible.
388: Further theoretical development and a larger study is required to test whether the
389: timescales are consistent with entropy profiles of a larger population.
390: The fact that these clusters with no evident feedback have
391: higher central entropy than clusters with obvious feedback suggests that
392: rare but influential AGN outbursts can dramatically change the original distribution of
393: entropy in clusters of galaxies. Alternatively, the intracluster gas of these clusters
394: may have started out with higher initial
395: entropy than the ICM in the active clusters, and it has not cooled to the point of sparking strong
396: AGN feedback.
397:
398: \acknowledgements
399: Support for this work was provided by the National Aeronautics and Space Administration through Chandra Award Numbers SAO GO3-4159X and AR3-4017A issued by the Chandra X-ray Observatory Center, which is operated by the Smithsonian Astrophysical Observatory for and on behalf of the National Aeronautics Space Administration under contract NAS8-03060.
400: This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.
401:
402:
403: \bibliography{coolingflows}
404:
405:
406:
407:
408:
409:
410:
411: \end{document}
412: