astro-ph0107557/ms.tex
1: %\documentclass[12pt, preprint]{aastex}
2: \documentclass{aastex}
3: 
4: 
5: 
6: \usepackage{graphicx}
7: 
8: \usepackage{emulateapj5}
9: \usepackage{times}
10: 
11: 
12: \makeatletter
13: \newenvironment{inlinetable}{%
14: \def\@captype{table}%
15: \noindent\begin{minipage}{0.999\linewidth}\begin{center}\footnotesize}
16: {\end{center}\end{minipage}\smallskip}
17: 
18: \newenvironment{inlinefigure}{%
19: \def\@captype{figure}%
20: \noindent\begin{minipage}{0.999\linewidth}\begin{center}}
21: {\end{center}\end{minipage}\smallskip}
22: \makeatother
23: 
24: \def\keV{ke\kern-0.05emV}
25: \newcommand{\chandra}{\emph{Chandra}}
26: \newcommand{\chandraintitle}{{\itshape\footnotesize CHANDRA}\/}
27: \newcommand{\ROSAT}{\textsl{ROSAT}}
28: \newcommand{\NGC}{NGC\,}
29: \renewcommand{\arcsec}{\ensuremath{''}}
30: \renewcommand{\mp}{m_{\!p}}
31: 
32: 
33: 
34: \def\asca       {{\em ASCA}\/}
35: \def\chandra    {{\em Chandra}\/}
36: \def\rxj1720    {{ RXJ1720.1+2638}\/}
37: \def\ginga    	{{\em Ginga}\/}
38: \def\xmm        {{\em XMM}\/}
39: \def\rosat      {{\em ROSAT}\/}
40: \def\am         {$^\prime$}
41: \def\as         {$^{\prime\prime}$}
42: \def\cmsq       {~cm$^{-2}$}
43: \def\kms        {~km$\;$s$^{-1}$}
44: \def\kmsmpc     {~km$\;$s$^{-1}\,$Mpc$^{-1}$}
45: \def\msunyr     {$M_{\odot}\;$yr$^{-1}$}
46: \def\degd       {$^{\circ}\!$}
47: \def\second      {{\prime\prime}}
48: 
49: 
50: 
51: \slugcomment{Accepted for publication in ApJ Letters}
52: 
53: \begin{document}
54: 
55: 
56: \title{Evidence for a heated gas bubble inside the ``cooling flow'' region of MKW3s}
57: 
58: \author{P.\ Mazzotta\altaffilmark{1,2}, J.S. Kaastra\altaffilmark{3}, 
59: F.B. Paerels\altaffilmark{4}, C. Ferrigno\altaffilmark{3},
60: S. Colafrancesco\altaffilmark{5}, R. Mewe\altaffilmark{3}, and
61:  W.R. Forman\altaffilmark{1}}
62: \altaffiltext{1}{Harvard-Smithsonian Center for Astrophysics, 60 Garden St.,
63: Cambridge, MA 02138; mazzotta@cfa.harvard.edu}
64: \altaffiltext{2}{ESA Fellow}
65: \altaffiltext{3}{SRON Laboratory for Space Research Sorbonnelaan
66: 2,3584 CA Utrecht, The Netherlands}
67: \altaffiltext{4}{Astrophysics Laboratory, Columbia University, 550
68: West 120th Street, New York, NY 100027, USA}
69: \altaffiltext{5}{Osservatorio Astronomico di Roma, Via Frascati, 33,
70: 00040, Monteporzio 
71: Italy}
72: 
73: \shorttitle{Evidence for a heated gas bubble inside the cooling flow region of MKW3s}
74: \shortauthors{MAZZOTTA ET AL.}
75: 
76: 
77: %
78: %
79: %
80: \begin{abstract}
81: We report on the  deep Chandra observation of central $r=200$~kpc region of
82: the cluster of galaxies MKW3s which was previously identified as a
83: moderate cooling flow cluster. The Chandra image reveals two striking
84: features -- a $100$~kpc long and $21$~kpc wide filament, extending from the 
85: center to the south-west and
86: a nearly circular, $50$~kpc diameter depression 90 kpc south of the X-ray
87: peak.  The temperature map shows that the filamentary structure is colder
88: while the surface brightness depression is hotter than the average
89: cluster temperature at any radius. 
90: The hot and the cold regions indicate that both cooling and
91: heating processes are taking place in the center of MKW3s.
92: We argue that the surface 
93: brightness depression is produced by 
94: a heated, low-density gas bubble along the line of sight.
95: We suggest that the heated bubble is produced by
96: short-lived nuclear outbursts from the central galaxy.
97: \end{abstract}
98: 
99: \keywords{galaxies: clusters: general --- galaxies: clusters: individual
100:   (MKW3s) --- X-rays: galaxies --- cooling flows}
101: 
102: %
103: %
104: %
105: \section{Introduction}
106: It has been argued that 
107: cooling flows (see Fabian 1994 for a review) are the 
108: natural state of cluster cores and that more than $40\%$ of  
109: clusters have flows depositing more than $100$~h$_{50}^{-2}$~M$_\odot$~yr$^{-1}$
110: (see e.g. Peres et al. 1998).
111: However, recent   \chandra ~ and XMM results  have not confirmed
112: cooling flow rates calculated  from
113: simple models. 
114: The clearest discrepancy with the ``standard'' cooling flow model
115: is obtained with the
116: XMM-RGS:
117: the spectra from  cooling flow regions in clusters  show a 
118: remarkable lack of emission lines from gas with $T\le 1$~keV 
119:  (Peterson et al. 2001, Kaastra et al. 2001, Tamura et al. 2001).
120: To explain the observed disagreement  with ``simple'' cooling flow
121: expectations, the above  authors propose a number of possible physical
122: mechanisms (see also Fabian et al. 2001). 
123: One mechanism invoked to reduce the cooling rate
124: is  intracluster medium (ICM) heating by powerful radio sources (mainly AGN) 
125: in the nucleus of the central galaxy.  
126: Among the several difficulties of 
127: such a scenario is that, to date, no cluster shows the ICM heating 
128: taking place.
129: 
130: We present the temperature map of the central $r=200$~kpc
131: region of MKW3s derived from a deep \chandra ~ observation.
132: Even though MKW3s was previously identified as 
133: a moderate cooling flow cluster  ($\dot M\approx 170$\msunyr, 
134: Peres et al. 1998), we
135: argue  that it shows evidence for 
136:  hot gas bubbles which could be heated 
137: by AGN activity of the central galaxy.
138: We also show evidence for a cold filament similar to that seen  
139: by \chandra ~ in A1795 (Fabian et al. 2000) but we  defer
140: to a later publication (Mazzotta et al. 2001, in preparation) for 
141: a discussion.
142: 
143: 
144: We use $H_0=50$~km~s$^{-1}$~kpc$^{-1}$, which implies a linear scale of
145: 1.21~kpc per arcsec at the distance of MKW3s ($z=0.045$).
146: 
147: %
148: % 
149: %
150: 
151: 
152: \section{Cluster Observations}
153: 
154: 
155: %
156: %
157: %
158: \subsection{X-ray Imaging Analysis}
159: 
160: MKW3s was observed  on April 2000 in  ACIS-I with an exposure of
161: $\approx 57$~ksec.
162: We cleaned the data as in Mazzotta et al. (2001).
163: 
164: 
165: %%%%%%%%%%%%%%%%%%%%%
166: % FIGURE 1
167: %%%%%%%%%%%%%%%%%%%%%
168: 
169: \begin{inlinefigure}
170: \centerline{\includegraphics[width=0.95\linewidth]{f1.eps}}
171: \caption{\chandra ~ image of the central $r=200$~kpc region of 
172:  MKW3s in the $0.5-8$~keV energy band.
173: Each pixel corresponds to $4^\second \times 4^\second$. 
174: The arrows indicate the most  prominent features:
175: the filamentary structure  and the
176:  surface brightness depression.}\label{fig:x-ray}
177: \end{inlinefigure}
178: 
179: A background subtracted, vignetting corrected  \chandra ~ image of the 
180: X-ray emission in the $0.5-8$~keV energy band, from MKW3s, is 
181: shown in Fig.~\ref{fig:x-ray}. 
182: The image reveals  
183: a  surface brightness filament 
184: extending from the cluster center to the south-west and a 
185: circular-like depression  south of the X-ray peak.
186: Both features are evident  in the soft  band ($0.5-2$~keV) as
187: well as in the hard band ($2-8$~keV) 
188: indicating that they are produced by a significant  gas density
189: deviation rather than by  excess  absorption  and/or 
190:  gas temperature variations.
191: The surface brightness depression  lies at $\approx 90$~kpc 
192: ($\approx 75^\second$) from the X-ray peak
193: and exhibits a circular shape with radius $r\approx 25$~kpc 
194: ($r\approx 20^\second)$.
195: To check the statistical significance of the observed  depression, we
196: extracted the count rates 
197: from a circular region inscribed in  the depression  
198: ($r=20^\second$; hereafter region S; see
199: Fig.~\ref{fig:tmap}) and the  count rates in  two similar 
200: regions to the east and to the  west of the depression.
201: The absorbed flux from region S is 
202: $\approx 14\%$ lower than the flux from the neighboring regions
203: at a $4.7\sigma$ confidence level.
204: 
205: 
206: %%%%%%%%%%%%%%%%%%%%%
207: % FIGURE 2
208: %%%%%%%%%%%%%%%%%%%%%
209: 
210: \begin{inlinefigure}
211: \centerline{\includegraphics[width=0.95\linewidth]{f2.eps}}
212: \caption{
213: Temperature map  with overlaid ACIS-I X-ray
214: surface brightness contours (spaced by a factor of $\sqrt{2}$) in the
215: 0.5-8~keV energy band after adaptive smoothing.   The black cut-out regions identify
216:   the point sources that were masked out. The  statistical
217:   error in the temperature map is $<\pm 0.4$~keV at 68\% significance
218:   level ( $<\pm 0.8$~keV at 90\%). 
219: The actual temperatures in the regions 
220: indicated by the cardinal numbers from 1 to 9 are reported in  
221: Fig.~\ref{fig:regions}.
222: All regions numbered 1-8 are circles with $r=20^\second$ ($\approx 25$~kpc).} 
223:   \label{fig:tmap}
224: \end{inlinefigure}
225: 
226: 
227: 
228: %
229: %
230: %
231: %%%%%%%%%%%%%%%%%%%%%
232: % FIGURE 3
233: %%%%%%%%%%%%%%%%%%%%%
234: 
235: \begin{inlinefigure}
236: \centerline{\includegraphics[width=0.95\linewidth]{f3.eps}}
237: \caption{Projected emission-weighted temperature in the
238:  corresponding regions defined in  Fig.~\ref{fig:tmap} 
239: (error bars $90\%$ confidence level).}
240: \label{fig:regions}
241: \end{inlinefigure}
242: 
243: \subsection{Temperature structure}
244: 
245: 
246: All spectral fits were made in the  $0.8-8$~keV energy band.
247: Moreover, we multiplied the ARFs
248: by a constant factor 0.93 for $E<2$~keV (Vikhlinin 2000). 
249: Each spectrum was grouped to have a minimum of 25 counts per bin and 
250: fitted 
251:  with an absorbed 
252: single-temperature thermal (1-T) model (MEKAL).
253:  We fixed $N_H$ 
254: to the Galactic value ($N_H=3.04\times 10^{20}$~cm$^{-2}$) and we leave,
255: as free parameters, the  temperature, the metal abundance, and the emission
256: measure.
257: 
258: Our technique for temperature map determination is the same as that in
259: Vikhlinin, Markevitch, \& Murray (2000).
260: We extracted images in 7 energy bands
261: ($0.80-0.99-2.48-3.50-4.50-6.00-7.01-8.50$~keV)
262: and smoothed the images with a Gaussian filter
263: whose width increased  smoothly from
264: $\sigma=4^\second$ at $r=0^\second$
265: to  $\sigma=20^\second$ at $r=200^\second$.
266: We determined the
267: temperature by fixing
268: $N_H$ at the galactic value and 
269: $Z=0.6\ Z_\odot$ (the best-fit  average of the $r=100$~kpc
270: region).
271: In Fig.~\ref{fig:tmap} we overlay the  
272: X-ray contours   
273: on the  temperature map.
274: To determine the accuracy of the temperature map, we fit the  
275: X-ray spectra in 
276: 9 selected regions (indicated in Fig.~\ref{fig:tmap} by 
277: numbers 1 to 9) 
278: with a 1-T model as shown in
279: Fig.~\ref{fig:regions}.
280: 
281: We  derived  the radial temperature profile of the entire cluster 
282: in twenty elliptical regions centered on the X-ray peak as shown
283: in  Fig.~\ref{fig:profile}. 
284: For comparison we report the temperature 
285: profile derived from an XMM  
286: and an \asca ~ observation of MKW3s
287: (Ferrigno et al. 2001, in preparation, Markevitch et al. 1998);  
288: they are in excellent agreement.
289: 
290: 
291: 
292: 
293: %%%%%%%%%%%%%%%%%%%%%
294: % FIGURE 4
295: %%%%%%%%%%%%%%%%%%%%%
296: 
297: \begin{inlinefigure}
298: \centerline{\includegraphics[width=0.95\linewidth]{f4.eps}}
299:   \caption{\chandra ~ and XMM temperature profiles.
300: Temperatures are extracted from ellipsoidal regions
301:   centered on the X-ray peak and oriented as the cluster surface
302:   brightness (error bars $90\%$ confidence level).  The  
303: ratio of the minor and major axes  is  $\approx 0.7$.
304:  Red crosses: 
305: temperature profile obtained from the \chandra ~ observation.
306: Black circles: temperature profile obtained from the XMM observation.
307: Blue crosses: temperature measurements obtained from the \asca ~ observation.
308: Solid diamond and triangle indicate the projected gas temperature of
309:   the region S (also 7) and 2, respectively (error bars $90\%$ confidence level). }
310:   \label{fig:profile}
311: \end{inlinefigure}
312: 
313: 
314: %
315: %
316: %
317: \subsubsection{Temperature Structure of region S}
318: 
319: First, we performed a variety of spectral fits to clarify the
320: nature of the surface brightness depression. By freeing $N_H$  
321: in the spectral fit for the region S 
322: we find that it is consistent with the galactic value. Hence, the
323:   depression is not produced by excess absorption 
324: but rather  suggests the presence of  a low gas density bubble 
325:  along the line of sight.
326: Fig.~\ref{fig:profile} shows that the projected temperature of region S  
327: (solid diamond)  is higher than
328: the average temperature at any other radius.
329: 
330: The observed spectrum from region S may  indicate either that
331: the gas in the bubble  is hotter than the ambient gas or
332: that the emission from the bubble is dominated by a non-thermal
333: component (e.g. a power-law).  
334: To test the origin of the emission from the depression, 
335: we fit the spectrum using two  simple spectral models:
336: an absorbed  two-temperature (2-T) model and  an absorbed
337: 1-T plus a power-law model.  
338: For the first thermal component, we fixed the normalization
339: to the estimated projected emission measure of the main cluster gas
340: excluding a sphere of $r_s=25$~kpc located in the plane of the sky
341: at 90~kpc from the cluster center.
342: We also fixed $N_H$ to the galactic value and the temperature of the first
343: thermal component to $T=3.5$~keV, the temperature of the  regions
344: adjacent to the depression.
345: For the second component of the 2-T model, we find 
346: $T=7.5^{+2.5}_{-1.5}$ with
347: $\chi^2/(d.o.f)= 53.8/(55).$
348: For the power-law model we find $\Gamma=1.6^{+0.2}_{-0.2}$  $k=1.8\times 10^{-5}$~mJy  with
349: $\chi^2/(d.o.f) = 56.4/(55)$ (here $S_x=k(E/1$~keV$)^{-\Gamma}$).
350: Both models fit the observed spectrum very well.
351: 
352: %%%%%%%%%%%%%%%%%%%%%
353: % FIGURE 5
354: %%%%%%%%%%%%%%%%%%%%%
355: 
356: \begin{inlinefigure}
357: \centerline{\includegraphics[width=0.95\linewidth]{f5.eps}}
358: \caption{Digitized Sky Survey image with overlaid X-ray
359: surface brightness contours as in Fig.~\ref{fig:tmap} .  
360: The circle identifies the region S inscribed in the X-ray surface
361: brightness depression of Fig.~\ref{fig:x-ray}.} 
362:   \label{fig:dss}
363: \end{inlinefigure}
364: 
365: %
366: %
367: %
368: \subsection{Optical and Radio Images}
369: 
370: In Fig.~\ref{fig:dss}  we  overlay the  X-ray
371: contours  
372:  on the 
373: Digital Sky Survey (DSS) image. 
374: The X-ray brightness peak coincides with the  
375: optical center of the cluster cD galaxy.
376: The optical image shows no galaxy concentration
377: coincident with either the filament or the  X-ray surface
378: brightness depression.
379: 
380: In  Fig.~\ref{fig:radio} we show
381: the temperature map  with overlaid  
382:  VLA radio (1.4~GHz) contours. 
383: The radio image shows only two bright structures:
384: the northern radio  source is associated with the  dominant cluster 
385: galaxy, the origin of the external source to the south is unclear.
386: This source
387: has a steep radio spectrum ($\Gamma=3.10$; De Breuck et al. 2000),  
388: but  steep spectra are common  both  in cluster halos 
389: (see e.g. Giovannini, Tordi, \& Feretti 1999) as well as in high redshift 
390: galaxies (see e.g. De Breuck et al.  2000). 
391: If it is related to the cluster,
392: it does not coincide with either the X-ray depression or with the
393: X-ray filament.  
394: 
395: 
396: %
397: %
398: %
399: \section{Nature of the surface brightness depression}
400: 
401: In this paper we concentrate our  discussion on
402:  the  nature of the X-ray surface 
403: brightness depression.
404: In $\S~2.2.1$ we showed that the
405:  1-T fit to the spectrum from region S gives a
406: temperature higher than the radially averaged gas temperature 
407: at any radius.
408: This higher temperature may   indicate either that
409: the gas in the bubble  is hotter than the ambient gas or 
410: that the bubble is dominated by non-thermal emission.
411: Unfortunately, simple spectral analysis does not allow us to 
412: distinguish between these possibilities since  both thermal and
413: non-thermal
414:  models 
415: give acceptable fits to the observed spectrum.
416: Thus, below we  separately discuss the two possibilities and conclude
417:  that thermal emission is more likely. 
418: 
419: %
420: %
421: %
422: 
423: %%%%%%%%%%%%%%%%%%%%%
424: % FIGURE 6
425: %%%%%%%%%%%%%%%%%%%%%
426: 
427: \begin{inlinefigure}
428: \centerline{\includegraphics[width=0.95\linewidth]{f6.eps}}
429:   \caption{X-ray temperature map  with  VLA radio (1.4~GHz) contours
430:   obtained from the First Survey Catalog (contours spaced by a factor of $\sqrt{2}$). The
431:   circle indicates the  region S.}   
432:   \label{fig:radio}
433: \end{inlinefigure}
434: 
435: 
436: \subsection{Non-thermal  emission}
437: 
438: If the emission from the bubble is dominated by a power-law component,
439: then it may be produced 
440: by  inverse-Compton (IC) scattering of a power-law distribution of
441: relativistic electrons with the cosmic microwave background (CMB) 
442: photons (see e.g. 
443: Harris \& Grindlay 1979).
444: If this is the case we should also expect 
445:  synchrotron radio emission.
446: The radio and the X-ray fluxes provide an estimate
447: of  the magnetic field strength in the bubble (Harris \& Grindlay 1979). 
448:  Since the FIRST image of MKW3s  shows no 
449: significant radio emission associated with the bubble,  at first
450: we assume that the radio flux $S_r$ is equal to the survey detection
451: limit: $S_r(1.4$~GHz$)= 1$~mJy. In $\S~2.2.1$ we showed
452:  that the X-ray spectral fit  
453: gives $S_x(1$~keV$)=1.8\times 10^{-5}$~mJy and $\Gamma=1.6$.
454: Using these values we find that  $B_{IC}\approx 8\times 10^{-10}$~Gauss.
455: Assuming the same radio flux, we estimate the magnetic field strength
456: using the minimum energy condition (Miley 1980) and we find
457:  $B_{me}\approx  7\times 10^{-5}$~Gauss.  
458: 
459: The two magnetic field estimates show a great discrepancy as
460: $B_{me}/B_{IC}\approx 8\times 10^{4}$.
461: Because $B_{me}\propto S_r^{0.286}$ and $B_{IC}\propto S_r^{0.625}$, 
462: the above discrepancy is even worse if the actual radio flux is $S_r<1$~mJy.
463: If we believe  that the minimum  energy condition is
464: valid (see e.g. Readhead 1994) then it appears unlikely
465: that the observed X-ray emission from the  bubble can be produced by
466: IC. Moreover, from the  upper limit on the radio flux, we estimate that the 
467: synchrotron energy density in the bubble 
468: is  $<10^{-4}$ times the energy density of
469: the CMB. Thus, we also can exclude the possibility that the observed
470: X-ray 
471: emission is
472: produced by the
473: synchrotron self Compton process.
474: 
475: %
476: %
477: %
478: \subsection{Thermal emission} 
479: 
480: The above considerations suggest that the emission from the low 
481: density bubble is more likely thermal. 
482: The exact determination of the gas temperature and density of the bubble
483: depend on the actual shape and on the position of the bubble with
484: respect to the plane of the sky. 
485: Let us assume that the bubble is a sphere and is
486: located close to the plane of the sky. 
487: In $\S~2.2.1$ we showed that in this case the deprojected gas temperature is
488: $T_b=7.5_{-1.5}^{+2.5}$~keV.
489: Using the difference between the flux 
490: in region S and the adjacent regions, we estimate  the ratio 
491: $\gamma$ between the  bubble gas density  $n_b$ and the ambient
492:  gas  density  $n_a$ to be $\gamma=0.71$. 
493: If the bubble  is at an angle $\theta$ with respect to the plane of
494: the sky, then the contribution of the 
495: bubble to the total flux must decrease.
496: Thus,  to obtain  the
497: observed depression, the ratio $n_b/n_a$ must be smaller
498: ($\gamma<0.71$).
499: Consequently the deprojected bubble gas 
500: temperature should be $T_b>7.5$~keV. 
501: By requiring  $n_b>0$ ($\gamma>0$) 
502: we can constrain the position of the bubble to $|\theta|<42$\degd .   
503: Given the density of the bubble as a function of $\theta$, we estimated
504: the corresponding deprojected temperature and, thus, the pressure.
505: By  assuming that the temperature $T_a$ of the ambient gas outside the bubble
506: is equal to  $T_a\approx 3.5-4$~keV
507: (see also Fig.~\ref{fig:profile}), we find that, for
508: $|\theta|<30$\degd ,
509: the gas pressure  $P_b$ in the  bubble is 1-2 times larger than
510: the pressure $P_a$ of the surrounding gas.
511: Unfortunately, for 
512:  $|\theta|>30$\degd ~ it is not possible to derive an accurate estimate of the 
513: bubble pressure as its gas  density goes rapidly to zero and the resulting
514: deprojected temperature is not constrained.
515: The above considerations strongly
516: suggest that 
517: the
518: gas in the bubble was  heated,  
519: started to expand  (or is still expanding) and, 
520: is now  (or soon will  be) in pressure equilibrium with the 
521: surrounding gas. 
522: 
523: \section{Discussion}
524: 
525: It is not clear how the gas inside the bubble 
526: can have  been heated to such a high temperature.
527: Nevertheless, we may expect that the bubble 
528: formed recently.
529: If we assume that the bubble expands 
530: close to the sound speed,  the bubble
531: age is $t\approx 2.7\times 10^{7}$~yr, 
532: 
533:  much shorter  than the expected age of the cluster.
534: 
535:  The complex temperature structure and the non trivial ellipticity of the
536: cluster core,  suggest that the gas is not in hydrostatic
537: equilibrium and/or MKW3s is undergoing a major merger. 
538: This last interpretation, however,
539: is not supported by either the optical data
540: or the X-ray surface brightness analysis which
541: indicate that MKW3s is a relaxed system. 
542: Girardi et al. (1997) 
543: show that MKW3s exhibits no substructure and that
544: the galaxy velocity distribution is Gaussian and unimodal. 
545: Moreover the velocity of the cD galaxy is consistent with the cluster mean
546: and the
547: velocity dispersion is  consistent with the cluster mean
548: X-ray temperature. 
549: Furthermore, the X-ray image does not show evident substructures.
550: As shown by Buote \& Tsai (1996) the power ratio $P_2/P_0$  of the
551: multipole terms of order 2 and 0, respectively, represent a critical
552: indicator of the cluster dynamical status:
553: for MKW3s, this value  is consistent with those of many relaxed systems 
554: and significantly lower of those of well known merger clusters.
555: 
556: 
557: One possible mechanism for heating gas in the bubble
558:  is energy injection arising from  activity in the 
559: nucleus of the cD galaxy. This scenario,  however, seems to be
560: inconsistent with the  radio map that shows no sign of
561: strong nuclear activity (see $\S~3$).
562: However, 
563: we may be observing a cluster in which 
564: the black hole  
565: in the central galaxy undergoes short  intervals ($10^{7}-10^{8}$~yr) of
566: strong activity followed by periods ($10^{9}-5\times 10^{9}$~yr) 
567: of relative quiescence (e.g. Binney \& Tabor 1995, Soker et al. 2000).
568: Such a scenario suggests that the bubble was 
569: heated during the last radio burst.
570: If this interpretation is correct and the
571: gas was heated by radio jets, then the hot regions 
572: on the north side of the cD galaxy (labeled as regions 2 and 3 in
573: Fig.~\ref{fig:tmap})  may have been heated
574: by a northern jet component.
575: The fact that  we do not detect a significant
576: corresponding surface brightness depression suggests that the
577: outburst was  strongly asymmetric. 
578: Moreover, intermittent energy injection can
579: contribute to form the observed temperature asymmetry in the core of MKW3s 
580: (see Churazov et al. 2000 for a discussion of the evolution of buoyant
581: bubbles in cluster atmospheres).
582: 
583: Finally, the discovery of this heated low-density gas  bubble  
584: assumes a particular importance as it clearly
585: shows heating taking place in the cluster core region (previously 
586: identified as a
587: cooling flow region) of an apparently ``relaxed'' cluster.
588: If widespread, such gas heating mechanisms may contribute to
589: substantially reduce  the cooling rate  and may explain the absence  of 
590: emission lines from cool ($1$~keV or below) gas in recent cluster 
591: observations. Furthermore,  we should expect to find them in other
592: relaxed clusters which would clearly differentiate among the
593: possibilities for their origin.
594: 
595: 
596: \acknowledgments
597: We thank the referee D. Buote, for useful comments and suggestions.
598: P.M. thanks D. Harris and M. Markevitch for useful discussions.
599: P.M. is particularly grateful to A. Vikhlinin for sharing his
600: software tools that simplified the analysis of this \chandra~ observation. 
601: P.M. acknowledges an ESA fellowship
602:  and thanks the Center for
603: Astrophysics for the hospitality. 
604: Support for this study was provided
605: by NASA contract NAS8-39073, grant NAG5-3064, and by the Smithsonian
606: Institution. 
607: The Laboratory for Space Research Utrecht is supported
608: financially by NWO, the Netherlands Organization for Scientific Research.
609: 
610: %
611: %
612: %
613: \begin{references}
614: 
615: %\reference{}  Allen, S.\ W.,  Taylor, G.B.,  Nulsen, P.E.J.,
616: % Johnstone,  R.M.,  David, L.P.,  Ettori, S., Fabian,  A.C.,  Forman, W.,
617: %  Jones, C., \& McNamara, B.  2000, \mnras, in press (astro-ph/0101162)
618: 
619: 
620: \reference{} Binney, J.\ \& Tabor, G.\ 1995, \mnras, 276, 663 
621: 
622: \reference{} Buote, D.\ A.\ \& Tsai, J.\ C.\ 1996, \apj, 458, 27 
623: 
624: \reference{} Churazov, E.,  Br\"uggen, M., Kaiser, C.R., B\"ohringer,
625: H.,\& Forman, W. 2000, ApJ in press (astro-ph/0008215)
626: 
627: 
628: %\reference{} David, L.\ P., Nulsen, P.\ E.\ J., McNamara,  B.\ R.,  Forman, W., Jones,
629: %C.,  Ponman, T.,  Robertson, B., \& Wise, M. 2000, ApJ, 
630: %in press (astro-ph/0010224)
631: 
632: \reference{} De Breuck, C., van Breugel, W., 
633: R{\"o}ttgering, H.\ J.\ A., \& Miley, G.\ 2000, \aaps, 143, 303 
634: 
635: 
636: %\reference{} Ekers, R.\ D.\ \& Simkin, S.\ M.\ 1983, \apj, 265, 85 
637: 
638: \reference{} Fabian, A.\ C.\ 1994, \araa, 32, 277 
639: 
640: \reference {}  Fabian, A.\ C.\,  Sanders, J.S.,  Ettori, S., Taylor,
641:  G.\ B.,
642:  Allen, S.\ W.,  Crawford,  C.\ S., Iwasawa, K., \& Johnstone,  R.\ M. 
643: 2000, \mnras, in press (astro-ph/0011547)
644: 
645: \reference{} Fabian, A.\ C., Mushotzky, R.\ F., Nulsen, P.\ E.\ J., \& 
646: Peterson, J.\ R.\ 2001, \mnras, 321, L20 
647: 
648: %\reference{} Ferrigno et al. 2001, in preparation
649: 
650: \reference{} Giovannini, G., Tordi, M., \& Feretti, L.\ 1999, New 
651: Astronomy, 4, 141 
652: 
653: \reference{} Girardi, M., Escalera, E., Fadda, D., Giuricin, G., 
654: Mardirossian, F., \& Mezzetti, M.\ 1997, \apj, 482, 41 
655: 
656: 
657: 
658: \reference{} Harris, D.\ E.\ \& Grindlay, J.\ E.\ 1979, \mnras, 188, 25 
659: 
660: \reference{} Kaastra, J.\ S., Ferrigno, C., Tamura, T., Paerels, F.\ B.\ 
661: S., Peterson, J.\ R., \& Mittaz, J.\ P.\ D.\ 2001, \aap, 365, L99 
662: 
663: %\reference{} McNamara, B.\ R., Wise, M.,
664: % Nulsen, P.\ E.\ J., David, L. P., Sarazin, C. L.,
665: % Bautz, M., Markevitch, M., Vikhlinin, A.,
666: % Forman, W.\ R., Jones, C., \& Harris, D.\ E.  2000, \apjl, 534, L135 
667: 
668: \reference{}   Markevitch, M., Forman, W. R.
669: Sarazin,  C.\ L. \&  Vikhlinin A. ApJ 503, 77, 1998
670: 
671: \reference{} Markevich, M. et al. 2000, CXC memo
672: (http://asc.harvard.edu/cal ``ACIS'', ``ACIS Background'')
673: 
674: \reference{}  Mazzotta, P.,   Markevitch, M.,
675:  Vikhlinin, A.,  Forman, W.\ R.,   David, L.\ P., \&
676:  VanSpeybroeck, L., 2001, ApJ, in press (astro-ph/0102291)
677: 
678: %\reference{} Mazzotta, P. et al., 2001b, in preparation
679: 
680: \reference{} Miley, G.\ 1980, \araa, 18, 165 
681: 
682: \reference{} Peres, C.\ B., Fabian, 
683: A.\ C., Edge, A.\ C., Allen, S.\ W., Johnstone, R.\ M., \& White, D.\ A.\ 
684: 1998, \mnras, 298, 416 
685: 
686: %\reference{} Peterson, J.\ R., Paerels, F.\ B.\ S., Kaastra, J.\ S., 
687: %Arnaud, M.,
688: % Reiprich, T.\ H., Fabian, A.\ C., Mushotzky, R.\ F.,
689: % Jernigan, J.\ G., \& Sakelliou, I.  2001, \aap, 365, L104 
690: 
691: \reference{} Peterson, J.\ R., et al.
692:  2001, \aap, 365, L104 
693: 
694: \reference{} Readhead, A.\ C.\ S.\ 1994, \apj, 426, 51 
695: 
696: 
697: \reference{} Soker, N., White, R.\ E., David, L.\ P., \& McNamara, B.\ R.\ 
698: 2001, \apj, 549, 832 
699: 
700: \reference{} Tamura, T., et al.
701:  2001, \aap, 365, L87 
702: 
703: 
704: \reference{} Vikhlinin, A., 2000, \chandra ~ calibration memo
705: 
706: \reference{} Vikhlinin, A., Markevitch, M., \& Murray, S.\ S. 2000, \apj,
707: in press (asro-ph/0008496)
708: 
709: \end{references}
710: 
711: \end{document}
712: \newpage
713: 
714: 
715: %%%%%%%%%%%%%%%%%%%%%
716: % FIGURE CAPTIONS
717: %%%%%%%%%%%%%%%%%%%%%
718: 
719: 
720: %%%%%%%%%%%%%%%%%%%%%
721: % FIGURE 1
722: %%%%%%%%%%%%%%%%%%%%%
723: 
724: \begin{figure}
725: \caption{\chandra ~ image of the central $r=200$~kpc region of 
726:  MKW3s in the $0.5-8$~keV energy band.
727: Each pixel corresponds to $4^\second \times 4^\second$. 
728: The arrows indicate the most  prominent features:
729: the filamentary structure  and the
730:  surface brightness depression.}\label{fig:x-ray}
731: \end{figure}
732: 
733: %%%%%%%%%%%%%%%%%%%%%
734: % FIGURE 2
735: %%%%%%%%%%%%%%%%%%%%%
736: 
737: \begin{figure}
738: \caption{
739: Temperature map  with overlaid ACIS-I X-ray
740: surface brightness contours (spaced by a factor of $\sqrt{2}$) in the
741: 0.5-8~keV energy band after adaptive smoothing.   The black cut-out regions identify
742:   the point sources that were masked out. The  statistical
743:   error in the temperature map is $<\pm 0.4$~keV at 68\% significance
744:   level ( $<\pm 0.8$~keV at 90\%). 
745: The actual temperatures in the regions 
746: indicated by the cardinal numbers from 1 to 9 are reported in  
747: Fig.~\ref{fig:regions}.
748: All regions numbered 1-8 are circles with $r=20^\second$ ($\approx 25$~kpc).} 
749:   \label{fig:tmap}
750: \end{figure}
751: 
752: 
753: 
754: %%%%%%%%%%%%%%%%%%%%%
755: % FIGURE 3
756: %%%%%%%%%%%%%%%%%%%%%
757: 
758: \begin{figure}
759: \caption{Projected emission-weighted temperature in the
760:  corresponding regions defined in  Fig.~\ref{fig:tmap} 
761: ($90\%$ confidence level error bars).}
762: \label{fig:regions}
763: \end{figure}
764: 
765: %%%%%%%%%%%%%%%%%%%%%
766: % FIGURE 4
767: %%%%%%%%%%%%%%%%%%%%%
768: 
769: \begin{figure}
770:   \caption{\chandra ~ and XMM temperature profiles.
771: Temperatures are extracted from ellipsoidal regions
772:   centered on the X-ray peak and oriented as the cluster surface
773:   brightness (error bars $90\%$ confidence level).  The  
774: ratio of the minor and major axes  is  $\approx 0.7$.
775:  Red crosses: 
776: temperature profile obtained from the \chandra ~ observation.
777: Black circles: temperature profile obtained from the XMM observation.
778: Blue crosses: temperature measurements obtained from the \asca ~ observation.
779: Solid diamond and triangle indicate the projected gas temperature of
780:   the region S (also 7) and 2, respectively (error bars $90\%$ confidence level). }
781:   \label{fig:profile}
782: \end{figure}
783: 
784: 
785: %%%%%%%%%%%%%%%%%%%%%
786: % FIGURE 5
787: %%%%%%%%%%%%%%%%%%%%%
788: 
789: \begin{figure}
790: \caption{Digitized Sky Survey image with overlaid X-ray
791: surface brightness contours as in Fig.~\ref{fig:tmap} .  
792: The circle identifies the region S inscribed in the X-ray surface
793: brightness depression of Fig.~\ref{fig:x-ray}.} 
794:   \label{fig:dss}
795: \end{figure}
796: 
797: 
798: %%%%%%%%%%%%%%%%%%%%%
799: % FIGURE 6
800: %%%%%%%%%%%%%%%%%%%%%
801: 
802: \begin{figure}
803:   \caption{X-ray temperature map  with  VLA radio (1.4~GHz) contours
804:   obtained from the First Survey Catalog (contours spaced by a factor of $\sqrt{2}$). The
805:   circle indicates the  region S.}   
806:   \label{fig:radio}
807: \end{figure}
808: 
809: 
810: 
811: \end{document}
812: 
813: 
814: 
815: