1: \documentclass[apj]{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: \usepackage{amssymb,latexsym,amsmath}
4: \usepackage{apjfonts}
5: \usepackage{graphicx}
6: \usepackage{psfig}
7: %\usepackage{subeqn}
8: %\usepackage{amsmath}
9: %\usepackage[alpha,z,expert]
10: %\usepackage[hypertex]{hyperref}
11:
12: \def \th {\thinspace}
13: \def \nh {N${\rm _H}$}
14: \def \ergsec{\hbox{erg s$^{-1}$}}
15: \def \kevcmsq{\hbox{keV cm$^{2}$}}
16: \def \arcmin {\hbox{$^\prime$}}
17: \def \arcsec {\hbox{$^{\prime\prime}$}}
18: \def \chisq {$\chi ^{2}$}
19: \def\spose#1{\hbox to 0pt{#1\hss}}
20: \def\ltsim{$\mathrel{\spose{\lower 3pt\hbox{$\sim$}}
21: \raise 2.0pt\hbox{$<$}}$\thinspace}
22: \def\gtsim{$\mathrel{\spose{\lower 3pt\hbox{$\sim$}}
23: \raise 2.0pt\hbox{$>$}}$\thinspace}
24: %\def \ref {\reference{}}
25: \def \msun {${\rm M_\odot}$}
26: \def \msunyr{\hbox{{$M_{\odot}$ yr$^{-1}$}}}
27: \def \mdot{\hbox{{$\dot{M}$}}}
28: \def \mstar{${\rm M_\star}$}
29: \def \rvir{{\hbox{$r_{\rm vir}$}}}
30: \def \rs{{\hbox{$r_{\rm s}$}}}
31: \def \cvir{{\hbox{$c_{\rm vir}$}}}
32: \def \mvir{{\hbox{$M_{\rm vir}$}}}
33: \def \mgas{{\hbox{$M_{\rm gas}$}}}
34: \def \mdm{{\hbox{$M_{\rm dm}$}}}
35: \def \lcdm{{\hbox{$\Lambda$CDM}}}
36: \newcommand\solar{\hbox{{$Z_{\odot}$}}}
37: \def \degmark{^\circ}
38: \def \nh {$N_{\rm H}$}
39: \def \deg {$^\circ$}
40: \def \eg {e.g.}
41: \def \etc {etc}
42: \def \ie {i.e.}
43: \def \cf {c.f.}
44: \def \dtwentyfive {${\rm D_{25}}$}
45: \renewcommand{\labelitemi}{--}
46: \newcommand\tvir{{\hbox{$T_{\rm vir}$}}}
47: \newcommand{\source}{\mbox{NGC\,5044}}
48: \newcommand{\first}{${\rm 1^{st}}$}
49: \newcommand{\second}{${\rm 2^{nd}}$}
50: \newcommand{\third}{${\rm 3^{rd}}$}
51: \newcommand{\fourth}{${\rm 4^{th}}$}
52: \newcommand{\fifth}{${\rm 5^{th}}$}
53: \newcommand{\sixth}{${\rm 6^{th}}$}
54: \newcommand{\seventh}{${\rm 7^{th}}$}
55: \newcommand{\eighth}{${\rm 8^{th}}$}
56: \newcommand{\ninth}{${\rm 9^{th}}$}
57: \newcommand{\tenth}{${\rm 10^{th}}$}
58: \newcommand{\sersic}{S\'{e}rsic}
59: \newcommand{\apec}{APEC}
60: \newcommand{\mekal}{MEKAL}
61: \newcommand\emin{\hbox{{$E_{\rm min}$}}}
62: \newcommand\ergs{{\rm erg s$^{-1}$}}
63: \newcommand{\hnought}{${\rm H_0}$}
64: \newcommand{\rhog}{${\rho_g}$}
65: \newcommand{\zfe }{${\rm Z_{Fe}}$}
66: \newcommand{\zo }{${\rm Z_{O}}$}
67: \newcommand{\zsi }{${\rm Z_{Si}}$}
68: \newcommand{\zni }{${\rm Z_{Ni}}$}
69: \newcommand{\zs }{${\rm Z_{S}}$}
70: \newcommand{\zne }{${\rm Z_{Ne}}$}
71: \newcommand{\zmg }{${\rm Z_{Mg}}$}
72: \newcommand{\feh } {${\rm [Fe/H]}$}
73: \newcommand{\emiss } {$\Lambda(T,Z_{\rm Fe})$}
74: \newcommand{\chandra }{{\em Chandra}}
75: \newcommand{\spitzer }{{\em Spitzer}}
76: \newcommand{\xspec }{{\em Xspec}}
77: \newcommand{\acis }{{\em ACIS}}
78: \newcommand{\ciao }{{\em CIAO}}
79: \newcommand{\caldb }{{\em Caldb}}
80: \newcommand{\heasoft }{{\em Heasoft}}
81: \newcommand{\fxunits}{\mbox{ergs cm$^{-2}$ s$^{-1}$}}
82: \newcommand{\lxunits}{\mbox{ergs s$^{-1}$}}
83: \newcommand{\xmm }{{\em XMM}}
84: \newcommand{\asca }{{\em ASCA}}
85: \newcommand{\rosat }{{\em ROSAT}}
86: \newcommand{\exosat }{{\em EXOSAT}}
87: \newcommand{\einstein }{{\em Einstein}}
88: \newcommand{\sax }{{\em Beppo-SAX}}
89: \newcommand{\lx }{${\rm L_X}$}
90: \newcommand{\tx }{${\rm T_X}$}
91: \newcommand\lk{\hbox{{$L_{\rm K}$}}}
92: \newcommand\lb{\hbox{{$L_{\rm B}$}}}
93: \newcommand{\lsun }{${\rm L_\odot}$}
94: \newcommand{\leda}{{\em{LEDA}}}
95: \newcommand{\ned}{{\em{NED}}}
96: \newcommand{\heasarc }{{\em HEASARC}\footnote{ftp://legacy.gsfc.nasa.gov}}
97: \newcommand{\twomass}{{\em 2MASS}}
98: \def \dtwentyfive {${\rm D_{25}}$}
99: \newcommand{\vdag}{(v)^\dagger}
100: \newcommand{\myemail}{gasta@uci.edu}
101: \newcommand{\thin}{\thinspace}
102: \newcommand\omegam{\hbox{{$\Omega_{\rm m}$}}}
103: \newcommand\omegalambda{\hbox{{$\Omega_{\Lambda}$}}}
104: \newcommand\kmsmpc{{\rm km s$^{-1}$ Mpc$^{-1}$}}
105: \newcommand\ho{\hbox{{$H_{0}$}}}
106: \newcommand\thot{\hbox{{$T_{\rm h}$}}}
107: \newcommand\tcool{\hbox{{$T_{\rm c}$}}}
108: \slugcomment{Submitted to the Astrophysical Journal}
109: \shorttitle{The X-ray cavities, filaments and cold fronts of NGC 5044}
110: \shortauthors{Gastaldello et~al.}
111:
112: %--------------------- TITLE & ABSTRACT ----------------------------------
113: %
114:
115:
116: \begin{document}
117:
118: \title{The X-ray cavities, filaments and cold fronts in the core of the galaxy group\\ NGC 5044}
119:
120: \author {Fabio Gastaldello\altaffilmark{1,2,3},
121: David A. Buote\altaffilmark{3},
122: Pasquale Temi\altaffilmark{4,5},
123: Fabrizio Brighenti\altaffilmark{6,7},
124: William G. Mathews\altaffilmark{7}
125: \& Stefano Ettori\altaffilmark{8,9}
126: }
127: \altaffiltext{1}{IASF-Milano, INAF, via Bassini 15, Milano 20133, Italy}
128: \altaffiltext{2}{Occhialini Fellow}
129: \altaffiltext{3}{Department of Physics and Astronomy, University of
130: California at Irvine, 4129 Frederick Reines Hall, Irvine, CA 92697-4575}
131: \altaffiltext{4}{Astrophysics Branch, NASA/Ames Research Center, MS 245-6, Moffet Field, CA 94035}
132: \altaffiltext{5}{SETI Institute, Mountain View, CA 94043; and Department of physics and Astronomy, University of Western Ontario, London ON N6A, 3K7, Canada}
133: \altaffiltext{6}{Dipartimento di Astronomia, Universit\`a di Bologna, via
134: Ranzani 1, Bologna 40127, Italy}
135: \altaffiltext{7}{UCO/Lick Observatory, University of California at Santa Cruz,
136: 1156 High Street, Santa Cruz, CA 95064}
137: \altaffiltext{8}{INAF, Osservatorio Astronomico di Bologna, via
138: Ranzani 1, Bologna 40127, Italy}
139: \altaffiltext{9}{INFN, Sezione di Bologna, viale Berti Pichat 6/2, I-40127 Bologna, Italy}
140: \begin{abstract}
141: We present a two-dimensional analysis of the bright nearby galaxy group
142: NGC 5044 using the currently available \chandra\ and \xmm\ data. In the
143: inner 10 kpc a pair of cavities are evident together with a set of bright
144: X-ray filaments. If the cavities are interpreted as gas
145: displaced by relativistic
146: plasma inflated by an AGN, even in the absence of extended 1.4 GHz
147: emission, this would be consistent with a recent outburst as also indicated
148: by the extent of dust and H$\alpha$ emission. The soft X-ray filaments
149: coincident with H$\alpha$ and dust emission are cooler than the ones which do
150: not correlate with optical and infrared emission. We suggest that dust-aided
151: cooling contributes to form warm ($T \sim 10^4$ K) gas, emitting H$\alpha$
152: radiation.
153: At 31 kpc and 67 kpc a pair of cold fronts are present, indicative of
154: sloshing due to a dynamical perturbation caused by accretion of a less
155: massive group, also suggested by the peculiar velocity of the brightest
156: galaxy NGC 5044 with respect to the mean group velocity.
157: \end{abstract}
158:
159: \keywords{cooling flows --- galaxies: clusters: general --- galaxies: clusters: individual\\ (NGC 5044) --- X-rays: galaxies: clusters}
160:
161: %%%%%%%%%%%%%%%%%%%%%%%%
162: \section{Introduction}
163: \label{Introduction}
164: %%%%%%%%%%%%%%%%%%%%%%%%
165: %
166: The current X-ray observatories, \chandra\ and \xmm, have revolutionized
167: our understanding of the cores of relaxed galaxies, groups and clusters,
168: which show a highly peaked X-ray emission from a hot interstellar medium whose
169: radiative cooling time is less than 1 Gyr \citep[for recent determinations of
170: cooling times in these objects see for example][]{Voigt.ea:04,Sanderson.ea:06,Jetha.ea:07}.
171: In the absence of heating, a cooling flow is established, in which the gas
172: cools, condenses and flows toward the center, accreting onto the central
173: galaxy \citep{Fabian:94}. However the mass sink for all this supposedly
174: cooling and condensing gas has never been
175: found \citep{Donahue.ea:04}. X-ray observations with {\em Chandra} and
176: {\em XMM} have established that there is little evidence for emission from
177: gas cooling below $\sim$ \tvir/3: just when gas should be cooling most
178: rapidly it appears not to be cooling at all
179: \citep[see the recent review by][and references therein]{Peterson.ea:06}.
180:
181: A compensating heat source must therefore resupply the radiative losses, and
182: many possibilities have been proposed, including thermal conduction
183: \citep[e.g.,][]{Narayan.ea:01}, energy released by mergers
184: \citep[e.g.,][]{Motl.ea:04} or by supernovae \citep[e.g.,][]{Silk.ea:86}.
185: However, feedback from the central AGN
186: has become the most appealing solution to the problem
187: \citep[see the recent review by][and references therein]{McNamara.ea:07}.
188: There is, in fact, clear observational evidence for AGN heating as
189: the majority of brightest central galaxies of clusters and groups host a
190: radio loud AGN \citep[e.g.,][]{Burns:90,Best.ea:07} and, following the launch
191: of \chandra, in an increasing number of objects
192: such disturbances as shocks, ripples and cavities have been found in the
193: central atmospheres of clusters, groups and elliptical galaxies
194: \citep[e.g.,][]{Fabian.ea:06,Birzan.ea:04,Dunn.ea:06,Forman.ea:05,Vrtilek.ea:02,Jetha.ea:07,Croston.ea:08,Finoguenov.ea:01*2,Allen.ea:06}.
195: The cavities, which appear as X-ray
196: surface brightness depressions, have been interpreted as bubbles of
197: low density relativistic plasma inflated by radio jets, displacing the thermal
198: gas causing $PdV$ heating \citep[e.g.,][]{Churazov.ea:02}, although \citet{Mathews.ea:08}
199: pointed out that cavity formation contributes thermal energy that may
200: offset radiative cooling only by injecting ultra-hot but non relativistic gas;
201: X-ray cavities formed solely from relativistic gas have a global cooling effect.
202: In most cases, the energy introduced by
203: the AGN is more than sufficient to counteract putative cooling flows
204: \citep{McNamara.ea:07}, although the physical process of the coupling of the
205: feedback energy with the ambient medium is not well understood.
206: While many of the observed X-ray cavities are filled
207: with plasma emitting in the radio at 1.4 GHz, some are undetected at this
208: frequency and have been referred to as ``ghost cavities''.
209: These may result from
210: the aging of the relativistic particle population and be the signature of a
211: previous AGN outburst. Observations at low radio frequency of several rich
212: clusters with ghost cavities show that they are indeed filled with
213: relativistic plasma \citep[e.g., A 2597,][]{Clarke.ea:05}, although even
214: low frequency emission cannot be easily detected.
215:
216: However, some net cooling is probably occurring, as the galaxies at the
217: center of cool cores show properties not shared by typical
218: elliptical galaxies \citep{Crawford:04}:
219: 1) the presence of strong,
220: low-ionization emission line nebulae both in
221: clusters \citep[e.g.,][]{Crawford.ea:99,Conselice.ea:01} and in groups
222: \citep[e.g.][]{Macchetto.ea:96}; 2) the galaxies at
223: the center of clusters that display these nebular emission show a component
224: of excess ultraviolet/blue continuum associated with young stars
225: \citep[e.g.,][]{Cardiel.ea:98,Crawford.ea:99}. This excess blue light is again
226: extended on scales of several kpc and there is evidence that some
227: (spatially extended) star
228: formation has been triggered by interaction with the radio source
229: \citep{Mcnamara:04} 3) CO emission lines have been detected in several
230: cool core clusters at millimeter wavelengths
231: \citep[e.g.,][]{Edge:01,Salome.ea:03},
232: implying the presence of a substantial amount of warm molecular gas
233: ($10^{9-11.5}$ M$_\odot$) within a 50 kpc radius of the central
234: galaxy. Less massive warmer molecular regions have been observed in
235: H$_2$ lines, which are often spatially associated with the stellar UV
236: and H$\alpha$ emission extending over $\sim 20$ kpc \citep{Jaffe.ea:05}.
237:
238: The origin and excitation mechanism of the nebular H$\alpha$ emission
239: have been much debated and they are still poorly known. The nebulae
240: require a constant and distributed heating source
241: \citep[e.g.][]{Johnstone.ea:88} which could be stellar \citep[e.g. massive OB
242: stars,][]{Allen.ea:92} or the ICM \citep[e.g. conduction,][]{Sparks.ea:89}.
243: Deep \chandra\ X-ray imaging of Perseus \citep{Fabian.ea:03*1} and M87
244: \citep{Sparks.ea:04} have renewed interest in conduction because of the
245: spatial coincidence of filamentary H$\alpha$ and soft X-ray emission, which
246: can be due to conduction and mixing of the cold gas with the ICM.
247: The disposition of some of
248: the optical filaments in Perseus strongly suggests that the filaments
249: are due to the buoyant radio bubbles drawing out the cold gas from a
250: central reservoir \citep{Fabian.ea:03*1} and it reveals the possibility of
251: tracing the flow of the rising bubbles \citep{Hatch.ea:06}. But the
252: question is still open if galaxy-galaxy interactions are required to
253: stimulate the central optical nebulae and if the optical filaments are
254: trails of galaxies punching through a molecular hydrogen reservoir as
255: suggested by the examples in \citet{Wilman.ea:06}. The excitement caused by
256: the shift in the cooling flow paradigm is stimulating also a renewed
257: theoretical interest in the quest for the origin and excitation of
258: molecular \citep{Ferland.ea:08} and H$\alpha$ emission
259: \citep{Nipoti.ea:04,Pope.ea:08,Revaz.ea:08}.
260: In particular \citet{Nipoti.ea:04} have
261: shown how the H$\alpha$ nebulae can be stable against thermal
262: evaporation only for the conditions of temperature and pressure found
263: in cool cores, explaining their association with that type of
264: environment.
265:
266: High resolution \chandra\ observations have also revealed another interesting
267: and unexpected feature in the ICM of relaxed clusters: the presence of
268: cold fronts. Many clusters have been found to exhibit sharp arc-shaped
269: jumps in their gas density and temperature which, unlike shock fronts,
270: have the gas on the dense side cooler, so the pressure is continuous across
271: the front \citep[e.g.,][]{Markevitch.ea:00,Vikhlinin.ea:01}. In merging
272: systems they were interpreted as contact discontinuities between gases from
273: different sub-clusters \citep{Markevitch.ea:00}. However cold fronts are present
274: in the centers of many, if not most, relaxed clusters with cool cores
275: \citep[e.g.,][]{Mazzotta.ea:01,Markevitch.ea:03*1,Dupke.ea:03,Ascasibar.ea:06,Ghizzardi.ea:07}.
276: The widely accepted scenario is that cold fronts in these systems are due to
277: sloshing of the cool gas in the central gravitational potential, which is
278: set off by minor mergers/accretions; the only necessary condition is a steep
279: entropy profile as observed in relaxed clusters \citep{Ascasibar.ea:06}.
280:
281: Whereas a growing number of clusters and elliptical galaxies have deep enough
282: multi-wavelength data (X-rays, radio and optical) to study the rich
283: phenomenology of cool cores in a detailed spatially resolved fashion, only an
284: handful of groups with such coverage exists, as for example the GEMS objects
285: \citep{Osmond.ea:04} presented in \citet{Rasmussen.ea:07}, and therefore,
286: ``unfortunately, AGN heating is not as well studied in groups as in clusters''
287: \citep{McNamara.ea:07}. Examination of AGN feedback at the mass scale of groups
288: is valuable because, although the scale of outbursts in groups is less
289: energetic and often on a smaller spatial scale than in clusters, the impact can
290: be even more dramatic than in rich clusters due to
291: the shallower group potential. Statistical studies examining the impact of AGN
292: on groups are starting to address the points raised above
293: \citep{Croston.ea:05,Jetha.ea:07}.
294: Cold fronts have not been investigated in detail so far in relaxed groups of galaxies.
295:
296: In this paper we show with more detail the currently available
297: \chandra\ and \xmm\
298: data for the galaxy group NGC 5044, one of the brightest groups in X-rays:
299: it is one of the only 5 objects with ${\rm{k}}T < 2$ keV in the HIFLUGCS sample
300: \citep{Reiprich.ea:02}. Our estimate, using the new \chandra\ and \xmm\ data,
301: for the bolometric (0.1-100 keV) X-ray luminosity within
302: $r_{500}=443 h_{70}^{-1}$ kpc \citep{Gastaldello.ea:07*1} is
303: $1.05\pm0.06 \times 10^{43} h_{70}^{-2}$ \ergsec, in good agreement with the \rosat\ estimate
304: presented in \citet{Reiprich.ea:02} of $2.46 \times 10^{43} h_{50}^{-2}$ within
305: $560 h_{50}^{-1}$ kpc.
306: In our previous analysis, mainly focused on
307: radial, azimuthally averaged properties of the system \citep{Buote.ea:03,Buote.ea:03*1},
308: interesting features like holes and filaments within 10 kpc, a sharp edge,
309: resembling a cold front, at 67 kpc, and unusually low iron abundances at
310: large radii \citep{Buote.ea:04} were already pointed out.
311: An H$\alpha$ nebula is present in the core of NGC 5044 showing an extended
312: filamentary structure \citep{Goudfrooij.ea:94*1,Caon.ea:00}.
313: NGC 5044 is also remarkable because {\em Spitzer} data show extended cold dust
314: emitting at 70 $\mu m$ \citep{Temi.ea:07}. Moreover, \citet{Temi.ea:07*1} showed
315: extended 8 $\mu m$ excess (likely arising from PAH, polycyclic aromatic hydrocarbon, molecules)
316: extending out to several kpc and
317: spatially coincident with the H$\alpha$ emitting nebulosity and the
318: brightest soft X-ray emission.
319: As proposed in \citet{Temi.ea:07*1}, current evidence
320: is consistent with an internal origin of this dust, which has been buoyantly
321: transported from the galactic core out to several kpc into the hot X-ray emitting gas
322: following an AGN outburst.
323:
324: All distance-dependent quantities have been computed assuming \ho = 70 \kmsmpc,
325: \omegam = 0.3 and \omegalambda = 0.7. At the redshift of $z=0.009$ 1\arcmin\
326: corresponds to 11.1 kpc. All the errors quoted are at the 68\% confidence
327: limit.
328: %
329: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
330: \section{Observations and data preparation}
331: \label{obs}
332: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
333: %
334: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
335: \begin{figure}[th]
336: \centerline{\psfig{figure=f1.ps,height=0.3\textheight}}
337: \caption{\label{fig.source} \footnotesize \chandra\ 0.5-5 keV X-ray image of
338: the inner 30x30 kpc of \source. The image has been processed to remove point
339: sources, flat fielded with a 1.0 keV exposure map. The figure has
340: been contour binned using the algorithm of \citet{Sanders:06} with a
341: S/N of 10. Color bar units are total counts per pixel. Interesting features discussed
342: in the text are highlighted.
343: %\vspace{-0.6cm}
344: }
345: \end{figure}
346: %
347: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
348: %
349: \source\ has been observed by \chandra\ with the ACIS-S configuration on
350: September 1999 (obsID 798) for 20.7 ks and by \xmm\ on January 2001 (obsID
351: 0037950101) with the EPIC MOS and pn cameras for 23 and 20 ks, respectively
352: \citep[see also][]{Buote.ea:03,Buote.ea:03*1}.
353: We focused mainly on the \chandra\ data because
354: its PSF is much better suited for the high-spatial resolution study that
355: constitutes most of the paper. We take advantage of the \xmm\ data when we
356: discuss the properties at larger radii (see section \S\ref{subsection_edges} and
357: \S\ref{subsection_spectral.coldfronts})
358: where the resolution requirements are not so important and we can benefit from the
359: larger accessible field.
360: Here we provide just a brief description of the data preparation, more details
361: can be found in \citet{Gastaldello.ea:07*1}.
362:
363: \subsection{Chandra}
364: The data were analyzed with the X-ray analysis packages
365: \ciao\ 3.4 and \heasoft\ 6.4 in conjunction with the \chandra\ calibration
366: database (\emph{Caldb}) version 3.4.2. In order to ensure the most
367: up-to-date calibration, all data were reprocessed from the ``level 1'' events
368: files, following the standard \chandra\ data-reduction
369: threads\footnote{{http://cxc.harvard.edu/ciao/threads/index.html}}.
370: We applied the standard corrections to take account of a time-dependent
371: drift in the detector gain and charge transfer inefficiency, as implemented
372: in the \ciao\ tools. From low surface brightness regions of the active chips
373: we extracted a light-curve (5.0-10.0~keV) to identify and excise
374: periods of enhanced background. The observation was quiescent resulting in a
375: final exposure time of 20 ks.
376: Point source detection was performed using the \ciao\
377: tool {\tt wavdetect} and removed, so as not to contaminate the diffuse
378: emission, using appropriate elliptical regions containing 99\% of their flux.
379: We generated an image in the 0.5-5.0 keV and a corresponding exposure map
380: computed at an energy of 1.0 keV. A zoom of the image in the inner region
381: is presented in Fig.\ref{fig.source} and Fig.\ref{fig.cavities} and the the region
382: covered by the ACIS-S3 chip is presented in Fig.\ref{fig.innercoldfrontsb}.
383:
384: \subsection{XMM}
385: We generated calibrated event files with SAS v7.1.0 using the tasks
386: {\em emchain} and {\em epchain}. We considered only event patterns 0-12 for
387: MOS and 0 for pn, and the data were cleaned using the standard procedures for
388: bright pixels and hot column removal and pn out-of-time correction.
389: Periods of high background due to soft protons were filtered as in
390: \citet{Gastaldello.ea:07*1} resulting in a net exposure time of 22 ks for
391: MOS1 and MOS2 and 17ks for pn. For each detector, we created images in the
392: 0.5-2 keV band with point sources detected using the task {\tt ewavelet} and
393: masked using circular regions of 25 radius centered at the source position.
394: We created exposure maps for each detector and we combined the MOS images
395: into a single exposure-corrected image, smoothed on a scale of 10\arcsec,
396: shown in Fig.\ref{fig.xmmcoldfront}.
397: %
398: \section{X-ray images and surface brightness profiles}
399: \label{images}
400:
401: \subsection{The inner 10 kpc}\label{subsection.core}
402: The presence of a disturbed morphology
403: with filamentary structure is already evident from the raw \chandra\ image and
404: further confirmed by using the contour binning
405: technique of \citet{Sanders:06} (See Fig.\ref{fig.source}).
406: Two depressions in surface brightness and multiple filamentary structures, some
407: of them connected to the presence of the cavities, have been highlighted in
408: Fig.\ref{fig.source}. To highlight the presence of structure we performed an
409: ``unsharp masking'' of the \chandra\ image \citep[e.g.,][]{Fabian.ea:03}: the 0.5-5 keV
410: exposure-map-corrected image was smoothed with Gaussians of width 1 and 10
411: arcsec and the two smoothed images were then subtracted, with the resulting
412: image shown in Fig.\ref{fig.umask}. The southern cavity and the structure of filaments
413: are clearly shown in the unsharp masked image. To quantify the observed structures we
414: extracted from the exposure corrected image (the unsharp image was
415: not used in the following analysis of the surface brightness profiles)
416: an azimuthally averaged radial profile and profiles from selected angular sectors
417: as depicted in Fig.\ref{fig.umask} and listed in Table \ref{tab.profiles}
418: \citep[centered on the X-ray surface brightness peak coincident with the
419: optical center of the galaxy NGC 5044 as listed by \ned\ and by][RA 13:15:24.0 and DEC -16 23 09, J2000 coordinates]{Macchetto.ea:96}.
420:
421: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
422: \begin{figure}[t]
423: \centerline{\psfig{figure=f2.ps,height=0.3\textheight}}
424: \caption{\label{fig.umask} \footnotesize Unsharp-masked \chandra\ image created by
425: subtracting a 0.5-5 keV intensity image smoothed by a $\sigma = 10$\arcsec\
426: Gaussian and one smoothed by a $\sigma = 1$\arcsec\ Gaussian. Areas in black
427: show a deficit of counts, whereas areas in white show an excess of counts.
428: Also shown are the angular sectors used for the surface brightness profiles
429: shown in Fig.\ref{fig.profiles}, labeled by name according to the scheme of
430: Table \ref{tab.profiles}.
431: }
432: \end{figure}
433: %
434: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
435: %
436: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
437: \begin{deluxetable}{ll}
438: \tablecaption{\label{tab.profiles}}
439: \tabletypesize{\scriptsize}
440: \tablehead{\colhead{Sector} & \colhead{PA}}
441: \startdata
442: Sector I & 120-141 \\
443: Sector II & 141-182 \\
444: Sector III & 182-212 \\
445: Sector IV & 212-244 \\
446: Sector V & 244-280 \\
447: Sector VI & 310-345 \\
448: \enddata
449: \tablecomments{Position angles (measured form the N direction) for the surface brightness profiles discussed in the text in \S\ref{subsection.core}.}
450: \end{deluxetable}
451: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
452: %
453: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
454: \begin{figure*}[th]
455: %%\vspace{-0.5cm}
456: \centerline{
457: \parbox{0.5\textwidth}{
458: \psfig{figure=f3a.ps,height=0.28\textheight}}
459: \parbox{0.5\textwidth}{
460: \psfig{figure=f3b.ps,height=0.28\textheight}}
461: }
462: \centerline{
463: \parbox{0.5\textwidth}{
464: \psfig{figure=f3c.ps,height=0.28\textheight}}
465: \parbox{0.5\textwidth}{
466: \psfig{figure=f3d.ps,height=0.28\textheight}}
467: }
468: \caption{\label{fig.profiles} \footnotesize
469: \emph{Top panel:} Surface brightness profiles in selected angular sectors listed in
470: Table \ref{tab.profiles} and shown in Fig.\ref{fig.umask}. The azimuthally averaged
471: surface brightness profile is
472: plotted as a solid black line in both panels for ease of comparison.\\
473: \emph{Bottom panel:} Ratio of the surface brightness profiles of the above
474: selected angular sectors over the azimuthally averaged one.
475: }
476: \end{figure*}
477: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
478:
479: As it can be seen from Fig.\ref{fig.profiles}, there is a complex network of
480: filamentary structure of enhanced emission which highlights dramatically
481: regions of comparatively lower surface brightness. Two filaments, brighter and
482: extended (in particular compared to the azimuthally averaged profile, also
483: shown in the right panel of Fig.\ref{fig.profiles}), are evident in sector I
484: (the SE filament) and sector III (the SW filament) with an embedded cavity in
485: Sector II (also in this sector the emission is more extended compared to the average,
486: i.e. there is an excess of counts for radii greater than 1\arcmin). The filament in
487: sector III is brighter in the region 10\arcsec-30\arcsec\ but less extended,
488: i.e. dimmer at $r>1$\arcmin, than the overall emission at PA 120-182.
489: The emission in sector IV can robustly be considered filamentary, with
490: enhanced emission, even brighter than the other two filaments in the range
491: 30\arcsec-50\arcsec\ but less extended, with an embedded depression in surface brightness
492: between 10\arcsec and 30\arcsec, which stands out as a region of low emission, compared to
493: the surroundings, in Fig.\ref{fig.umask}. However this feature is at the same level of
494: surface brightness of the azimuthally averaged profile.
495: The cavity in sector VI is the real counter-part of the one in sector II.
496: Although it stands out less clearly in the unsharp-mask image, it is
497: clearly indicated in the binned images of Fig.\ref{fig.source} and in the
498: striking similarity of the surface brightness profiles within 50\arcsec\ as
499: shown in Fig.\ref{fig.profiles}. The northern cavity stands out less clearly
500: in the unsharp masked image of Fig.\ref{fig.umask} because of the dimmer regions
501: surrounding it (the northern arms are not as bright as the southern ones)
502: and because of the sharper drop in surface brightness at $r>50$\arcsec.
503: There is also evidence of enhanced emission in sector V in the
504: 20\arcsec-40\arcsec\ region, as highlighted also in the unsharp masked image.
505:
506: \subsection{The inner and outer surface brightness edges}\label{subsection_edges}
507:
508: We now zoom out to investigate the outer regions in the \chandra\
509: image, as shown in Fig.\ref{fig.innercoldfrontsb}. A sharp edge in the surface
510: brightness on the south-east (PA 120-160) is revealed in this image and
511: further confirmed by a surface brightness profile in the same sector, as
512: compared to a reference profile extracted in the PA 180-315 sector (right
513: panel of Fig.\ref{fig.innercoldfrontsb}). Whereas the profile is smooth
514: and with a continuous derivative in the reference sector, the slope of the
515: sector across the cold front is rapidly changing and the edge is
516: at $\sim 31$ kpc (170\arcsec).
517:
518: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
519: \begin{figure*}[th]
520: %%\vspace{-0.5cm}
521: \centerline{
522: \parbox{0.5\textwidth}{
523: \psfig{figure=f4a.ps,height=0.28\textheight}}
524: \parbox{0.5\textwidth}{
525: \psfig{figure=f4b.ps,height=0.28\textheight}}
526: }
527: \caption{\label{fig.innercoldfrontsb} \footnotesize
528: \emph{Left panel:} Over-plotted on the \chandra\ image are the sector used
529: to extract the surface brightness profile across the cold front and the
530: reference sector used for comparison.\\
531: \emph{Right panel:} Surface brightness profile for the cold front and the
532: reference sector. A sharp drop is visible at $\sim$ 170\arcsec\ form the
533: center, i.e. 31 kpc.
534: }
535: \end{figure*}
536: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
537:
538: Inspection of the combined MOS image shown in Fig.\ref{fig.xmmcoldfront}
539: confirms the suggestion of \citet{Buote.ea:03*1} of the presence of another
540: surface brightness edge in the north-west direction at $\sim$ 6\arcmin. Plots
541: of the surface brightness profiles across this sector and in other three
542: sectors (one including the inner edge) for comparison are shown in
543: Fig.\ref{fig.xmmprofiles}. The inner cold front is detected also in the MOS
544: image. More importantly another sharp feature (considering also the
545: \xmm\ PSF at this off-axis angle, $\sim 50$\arcsec\ 90\% encircled energy
546: fraction at 1.5 keV) is clearly revealed at $\sim$350\arcsec, i.e. 65 kpc
547: (see right panel of Fig.\ref{fig.xmmprofiles}).
548: The feature is detected also in the pn data.
549:
550: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
551: \begin{figure}[t]
552: \centerline{\psfig{figure=f5.ps,height=0.3\textheight}}
553: \caption{\label{fig.xmmcoldfront} \footnotesize Mosaic of the MOS1 and MOS2 images
554: smoothed on a 10\arcsec\ scale. The image has been divided by the summed
555: exposure maps to correct for exposure variations. Point sources have been
556: removed and filled using the \ciao\ task {\tt dmfilth}. Sectors used for the
557: extraction of surface brightness profiles discussed in the text and presented
558: in Fig.\ref{fig.xmmprofiles} are also shown.
559: }
560: \end{figure}
561: %
562: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
563: %
564: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
565: \begin{figure*}[th]
566: %%\vspace{-0.5cm}
567: \centerline{
568: \parbox{0.5\textwidth}{
569: \psfig{figure=f6a.ps,height=0.28\textheight}}
570: \parbox{0.5\textwidth}{
571: \psfig{figure=f6b.ps,height=0.28\textheight}}
572: }
573: \caption{\label{fig.xmmprofiles} \footnotesize
574: \emph{Left panel:} Surface brightness profiles in selected angular sectors,
575: with PA measured from the N direction. The inner edge can be easily seen,
576: though not as sharp as in the \chandra\ profile because of the \xmm\ larger
577: PSF.\\
578: \emph{Right panel:} Zoom over the interesting radial region for the outer
579: edge. The characteristic shape of a rapid change in the slope of the profile
580: across the edge can easily be seen in the sector PA 280-350. The profile
581: in the undisturbed sector PA 350-80 is plotted for comparison.
582: }
583: \end{figure*}
584: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
585:
586: \section{Spectral analysis and temperature maps}
587:
588: Details about the \chandra\ and \xmm\ spectral extraction and analysis
589: can be found in \citet{Humphrey.ea:06} and \citet{Gastaldello.ea:07*1}.
590: Here we briefly summarize that for \chandra\ appropriate count-weighted
591: spectral response matrices were generated for each region using the
592: standard \ciao\ tasks {\tt mkwarf} and {\tt mkacisrmf} and for \xmm\
593: using the SAS tasks {\tt rmfgen} and {\tt arfgen} in extended source mode.
594: For each spectrum, we estimated the background through local modeling
595: using the method outlined in \citet{Humphrey.ea:06} and
596: \citet{Gastaldello.ea:07*1}. The spectra were re-binned to ensure a S/N of at
597: least 3 and a minimum 20 counts bin$^{-1}$.
598:
599: \subsection{The inner 10 kpc}\label{subsection_spectral.core}
600:
601: We extracted spectra from the \chandra\ dataset in a series of regions as depicted in
602: Fig.\ref{fig.tmap}, determined using the contour binning technique \citep{Sanders:06}
603: with a S/N=50.
604: We fitted the background subtracted spectra with an APEC thermal plasma model
605: \citep{Smith.ea:01} with the absorbing column density fixed at the Galactic
606: value \citep{Dickey.ea:90}. Solar abundances are in the units of \citet{Grevesse.ea:98}.
607: To account for the undetected point-sources we
608: added a 7.3 keV bremsstrahlung component
609: \citep[this model gives a good fit to the spectrum of the detected sources in
610: nearby galaxies,][]{Irwin.ea:03} for all the regions, because they fall
611: within the twenty-fifth magnitude isophote (\dtwentyfive) of the central
612: galaxy NGC 5044.
613:
614: The temperature map thus obtained is shown in Fig.\ref{fig.tmap}. The clearest
615: feature is the cool gas present in the SW (Sector III) arm (cooler than any
616: other region in the map) which is spatially coincident with the H$\alpha$ and
617: dust filament \citep[][see Fig.\ref{fig.halpha}]{Temi.ea:07*1}.
618: The SE arm (Sector I) is hotter.
619: There is presence of cooler gas with respect to the surroundings in the
620: direction of the inner surface brightness edge (i.e. region 26 in
621: Fig.\ref{fig.tmap}).
622:
623: One possible concern about the interpretation of the spectral fits and
624: consequently of the surface brightness features is the suggestion of
625: limited multiphase gas in the temperature structure of \source, as derived
626: by the preference of two temperature (2T) models over single temperature (1T)
627: models in radial annuli \citep{Buote.ea:03}. This is further complicated by
628: projection effects, given the fact that we are dealing only with projected
629: spectra in the analysis presented here: a proper deprojection is problematic
630: due to uncertainty in the projection geometry and it is beyond the scope of the present paper.
631: However it is evident in the radial analysis
632: of \citet{Buote.ea:03} that the cool component clearly dominates in the inner
633: 10 kpc (see their Figure 6) and that a single phase description, modulo
634: projection effects, is an appropriate description of the data for these
635: inner regions. We investigated
636: fitting 2T models as in \citet{Buote.ea:03} with the addition of the
637: bremsstrahlung component for the unresolved point sources. The results
638: of the two-dimensional analysis are consistent with the radial analysis of
639: \citep{Buote.ea:03} and with the dominance in terms of emission measure of the
640: cool component over the hot one.
641:
642: To investigate in more detail the nature of the bright soft X-ray filaments
643: we extracted spectra representative of the two southern
644: filaments in two 30\arcsec$\times$45\arcsec\ rectangular regions at a distance of
645: 30\arcsec\ from the center (the blue boxes shown in Fig.\ref{fig.tmap}); the spectrum
646: taken from the SW filament has a Fe-L feature with lower
647: excitation energy lines more prominent than the spectrum from the SE filament,
648: signature of a lower temperature (see Fig.\ref{fig.comparison}). If
649: we fit the two spectra with a 1T model we obtain a temperature difference
650: which is significant at $9\sigma$ (kT = $0.65\pm0.01$ keV for the SE arm
651: and $0.78\pm0.01$ keV for the SW arm). The regions connected to the filaments are cooler
652: compared to all the other regions in the map (by $3.6\sigma$ compared to the central region \#1,
653: which has the second lowest temperature in the map, $0.70\pm0.01$).
654: If we try to fit with a 2T model
655: the spectrum of the SE arm we obtain only a small decrease in $\chi^2$
656: ($\chi^2$/dof = 57/40 for the 2T model compared to 60/42 of the 1T model).
657: If we repeat the exercise for the regions of brighter emission surrounding
658: the northern cavity (the cyan boxes shown in Fig.\ref{fig.tmap} encompassing the
659: northern filaments highlighted in Fig.\ref{fig.source}) we obtain $0.74\pm0.02$ keV for
660: the NW filament and $0.82\pm0.02$ keV for the NE filament.
661:
662: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
663: \begin{figure*}[th]
664: %%\vspace{-0.5cm}
665: \centerline{
666: \parbox{0.5\textwidth}{
667: \psfig{figure=f7a.ps,height=0.35\textheight}}
668: \parbox{0.5\textwidth}{
669: \psfig{figure=f7b.ps,angle=-90,width=0.45\textwidth}}
670: }
671: \caption{\label{fig.tmap} \footnotesize
672: \emph{Left panel:} Temperature map of the core regions of \source\ based on
673: the SN=50 contour binning. The map has been smoothed for better presentation
674: with a Gaussian kernel of 6\arcsec. Color bar units are in keV. The magenta cross marks
675: the center of NGC 5044. The spectra
676: presented in Fig.\ref{fig.comparison} and discussed in the text have been extracted from the
677: rectangular regions shown in blue and they encompass the SE and SW filament; the spectra
678: extracted from the cyan boxes and discussed in the text encompass the NE and NW filaments
679: highlighted in Fig.\ref{fig.source};.
680: \emph{Right panel:} Values and $1\sigma$ error bars for the temperature in
681: each region of the temperature map.
682: %\vspace{-0.25cm}
683: }
684: \end{figure*}
685: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
686: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
687: \begin{figure}[t]
688: \centerline{\psfig{figure=f8.ps,width=0.45\textwidth,angle=-90}}
689: \caption{\label{fig.comparison} \footnotesize The spectra from the rectangular
690: boxes taken respectively from the SW arm (black data points) and the SE arm
691: (red data points). Note the shape of the Fe-L ``hump'' and the prominence of
692: lower energy lines, signature of lower temperature, in the spectrum of the SW
693: arm.
694: }
695: \end{figure}
696: %
697: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
698: %
699: \subsection{The surface brightness edges}\label{subsection_spectral.coldfronts}
700:
701: To investigate the nature of the surface brightness edges we have to
702: determine the temperature across the surface brightness jump. We therefore
703: extracted \chandra\ spectra for the inner jump together with
704: ``control' spectra taken from regions at the same radial distance from
705: the center but at different position angles. We applied the same method to
706: the outer edge when extracting \xmm\ spectra.
707: In Fig.\ref{fig.tjumps} we plot the temperature across the edges obtained by
708: fitting a 1T model. In Table \ref{tab.coldfronts} we present the results
709: of 1T and 2T fits for selected regions inside and outside the edges together
710: with control regions at the same radial range but at different position
711: angles.
712:
713: For the inner edge a clear temperature jump is detected, with the brighter
714: regions inside the edge cooler than the outer part. 2T models are preferred
715: over 1T models which points to some degree of projection effects in the
716: temperature determination across the edge, probably caused by the cool gas
717: which is producing the edge. Other regions covering the same range of radii but
718: at different position angles do not show in-fact the same behavior, with the
719: exception to some degree of the region immediately close to the edge. We extracted
720: \xmm\ MOS and pn spectra from the same regions inside and outside the inner edge finding
721: consistent results within the 1$\sigma$ errors for both 1T and 2T models.
722:
723: For the outer edge the temperature jump is more subtle and of low statistical
724: significance ($1.6\sigma$), but, contrasted
725: with the slightly declining trend of the control regions
726: (we are in the region around the peak of the azimuthally averaged temperature
727: profile), it is consistent with the picture of cooler gas inside the edge.
728:
729: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
730: \begin{figure*}[th]
731: %%%\vspace{-0.5cm}
732: \centerline{
733: \parbox{0.5\textwidth}{
734: \psfig{figure=f9a.ps,angle=-90,height=0.25\textheight}}
735: \parbox{0.5\textwidth}{
736: \psfig{figure=f9b.ps,angle=-90,height=0.25\textheight}}
737: }
738: \caption{\label{fig.tjumps} \footnotesize
739: \emph{Left panel:} Values and $1\sigma$ error bars for the temperature across
740: the inner surface brightness edge and in control sectors.\\
741: \emph{Right panel:} Values and $1\sigma$ error bars for the temperature across
742: the outer surface brightness edge and in control sectors.
743: %\vspace{-0.25cm}
744: }
745: \end{figure*}
746: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
747: %
748: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
749: \begin{table*}[t] \footnotesize
750: \caption{Parameters from the spectral fits across the edges
751: \label{tab.coldfronts}}
752: \begin{center} \vskip -0.4cm
753: \begin{tabular}{ccccccccccc}
754: \tableline\tableline\\[-7pt]
755: & \multicolumn{4}{c}{1T} & \multicolumn{6}{c}{2T}\\
756: & & $T$ & \zfe\ & norm & & \tcool & \thot & \zfe\ & norm$_c$ & norm$_h$ \\
757: Radii PA & $\chi^2$/dof & (keV) & solar & ($10^{-4}$ cm$^{-5}$) & $\chi^2$/dof & (keV) & (keV) & solar & ($10^{-4}$ cm$^{-5}$) & ($10^{-4}$ cm$^{-5}$) \\
758: \tableline \\[-7pt]
759: {\emph{Inner edge}} \\
760: %\chandra \\
761: 100-160\arcsec\ 120-160 & 137/91 & $0.80 \pm 0.01$ & $0.59\pm0.07$ & $9.83^{+0.73}_{-0.97}$ & 88/89 & $ 0.77 \pm 0.01$ & $1.51 \pm0.16 $ & $1.20^{+0.28}_{-0.20}$ & $4.43^{+0.85}_{-0.79}$ & $1.81\pm0.25$ \\
762: 175-235\arcsec\ 120-160 & 87/66 & $1.06 \pm 0.01$ & $0.38\pm0.03$ & $5.64\pm0.31$ & 73/64 & $0.68^{+0.11}_{-0.07}$ & $1.24^{+0.07}_{-0.05}$ & $0.64^{+0.15}_{-0.10}$ & $0.62^{+0.43}_{-0.12}$ & $3.83^{+0.44}_{-0.65}$ \\
763: {\emph{Control regions}} \\
764: 100-160\arcsec\ 170-220 & 139/87 & $0.97 \pm 0.01$ & $0.77^{+0.13}_{-0.08}$ & $5.99^{+0.54}_{-0.73}$ & 118/85 & $ 0.85\pm0.01$ & $1.48^{+0.13}_{-0.61}$ & $1.52^{+0.48}_{-0.31}$ & $2.12^{+0.51}_{-0.48}$ & $1.78\pm0.30$ \\
765: 175-235\arcsec\ 170-220 & 70/62 & $1.27 \pm 0.02$ & $0.69^{+0.09}_{-0.07}$ & $3.96^{+0.23}_{-0.33}$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
766: 100-160\arcsec\ 220-270 & 90/81 & $1.07 \pm 0.01$ & $0.86^{+0.13}_{-0.10}$ & $4.71\pm0.54$ & 80/79 & $ 1.06\pm0.01$ & \nodata\ & $1.52^{+0.64}_{-0.40}$ & $2.59^{+0.94}_{-0.77}$ & $0.88^{+0.22}_{-0.28}$ \\
767: 175-235\arcsec\ 220-270 & 69/66 & $1.21 \pm 0.02$ & $0.83^{+0.15}_{-0.10}$ & $3.73\pm0.46$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
768: 100-160\arcsec\ 270-320 & 77/78 & $1.14 \pm 0.02$ & $0.88^{+0.17}_{-0.12}$ & $4.23^{+0.52}_{-0.55}$ & \nodata & \nodata & \nodata\ & \nodata & \nodata & \nodata \\
769: 175-235\arcsec\ 220-270 & 73/73 & $1.09 \pm 0.02$ & $0.61^{+0.12}_{-0.08}$ & $4.70\pm0.52$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
770: {\emph{Outer edge}} \\
771: 270-330\arcsec\ 280-350 & 214/178 & $1.15\pm 0.02$ & $0.40^{+0.07}_{-0.05}$ & $5.46^{+0.23}_{-0.26}$ & 207/176 & $ 0.66^{+0.16}_{-0.19}$ & $1.22^{+0.06}_{-0.03}$ & $0.51^{+0.28}_{-0.20}$ & $0.30^{+0.38}_{-0.13}$ & $4.71^{+0.32}_{-0.54}$ \\
772: 390-450\arcsec\ 280-350 & 83/73 & $1.23 \pm 0.05$ & $0.28^{+0.08}_{-0.05}$ & $2.59\pm0.33$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
773: {\emph{Control regions}} \\
774: 270-330\arcsec\ 350-80 & 268/235 & $1.21\pm 0.02$ & $0.58^{+0.06}_{-0.05}$ & $5.98\pm0.37$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
775: 390-450\arcsec\ 350-80 & 164/181 & $1.19\pm 0.03$ & $0.36^{+0.05}_{-0.04}$ & $5.55\pm0.41$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
776: 270-330\arcsec\ 120-190 & 186/171 & $1.31\pm 0.02$ & $0.48^{+0.05}_{-0.04}$ & $4.58^{+0.20}_{-0.30}$ & 182/169 & $ 1.06^{+0.17}_{-0.12}$ & $1.68^{+0.14}_{-0.19}$ & $0.74^{+0.19}_{-0.16}$ & $1.11^{+1.45}_{-0.58}$ & $2.73^{+0.75}_{-1.11}$ \\
777: 390-450\arcsec\ 120-190 & 126/125 & $1.27\pm 0.02$ & $0.61^{+0.12}_{-0.09}$ & $3.07\pm0.32$ & \nodata & \nodata & \nodata & \nodata & \nodata & \nodata \\
778: \tableline
779: %\\[-1.0cm]
780: \end{tabular}
781: \tablecomments{Results of 1T and 2T spectral fits for selected regions as described in the
782: text in \S \ref{subsection_spectral.coldfronts}. The first and second columns refer
783: to the radial range (in arcseconds) and azimuthal range (PA in degrees) of the fitted
784: sectors discussed in \S\ref{subsection_spectral.coldfronts}. The $norm$ parameter is the
785: emission measure of the \apec\ model as defined in \xspec:
786: $10^{-14}(\int n_en_pdV)/4\pi D^2(1+z)^2$ with units $\rm cm^{-5}$.
787: \tcool and \thot refer to the temperature of the cool and hot component of the 2T model,
788: whereas norm$_c$ and norm$_h$ refer to the correspondent emission measures.
789: No entry for 2T models
790: means that there was no improvement over a 1T model. Regions with no entry for \thot\
791: did not significantly constrain that parameter.}
792: \end{center}
793: \end{table*}
794: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
795: %
796: %=====================================================================
797: % SYSTEMATICS SYSTEMATICS SYSTEMATICS SYSTEMATICS SYSTEMATICS
798: %=====================================================================
799: %
800: \section{Systematic Errors}
801: \label{systematics}
802: This section contains an investigation of possible systematic errors
803: in the data analysis relevant for this paper. A thorough analysis of
804: many of the issues involved has been conducted in \citet{Buote.ea:03} and
805: \citet{Buote.ea:03*1}.
806:
807: \subsection{Image Binning}
808: The presence of the morphological structures discussed in the paper,
809: in particular the disturbed morphology with filamentary structure
810: of the inner regions, is already evident from the raw \chandra\ image.
811: The use of the particular binning technique of \citet{Sanders:06} does not
812: introduce any spurious feature. We also used the Weighted Voronoi Tessellation (WVT)
813: binning algorithm of \citet{Diehl.ea:06}, which is more
814: robust against the introduction of spurious features than adaptive smoothing,
815: as provided for example by the \ciao\ task {\tt csmooth} and we found the same features.
816:
817: \subsection{Plasma Codes}
818: \label{plasma}
819:
820: We compared the results obtained using the \apec\ code to those
821: obtained using the \mekal\ code \citep{Kaastra.ea:93,Liedahl.ea:95} to assess the
822: importance of different implementations of the atomic physics and
823: different emission line lists in the plasma codes.
824: We found no qualitative differences between the two codes: the fitted
825: temperatures agree to within 5\% and abundances within 10-20\% (where 1T and 2T
826: models are compared accordingly) and well within their statistical errors.
827:
828: \subsection{Bandwidth}
829:
830: We explored the sensitivity of our results to our default lower limit of
831: the bandpass, $\emin=0.5$~keV. For comparison we performed spectral fits with
832: $\emin=0.4$~keV and $\emin=0.7$~keV. The fitted temperatures are consistent
833: between models; iron abundances are better constrained, in particular for
834: 2T models, when using the larger bandwidth \citep[see][]{Buote.ea:03*1}.
835:
836: \subsection{Variable \nh}
837: \label{nh}
838: We take into account possible deviations for \nh\ from the value of
839: \citet{Dickey.ea:90} allowing the parameter to vary by $\pm25$\%.
840: We found no qualitative differences between the two cases.
841: If we leave \nh\ free to fit we derive higher values in the core, but
842: still consistent at 1$\sigma$ with the Galactic value in the majority of the
843: regions of Fig.\ref{fig.tmap}. All the other spectral parameters are basically
844: unchanged. This is not due to excess absorption but just the exploitation of
845: an additional free parameter by the fitting program to model deviations from
846: a single temperature model.
847:
848: \subsection{Background}
849:
850: Since NGC 5044 is sufficiently bright and the temperature determination stems
851: from the measurement of the shape and width of the Fe-L shell which is a
852: prominent feature in the spectra, the fitted temperature values are quite
853: insensitive to errors in the background normalization.
854:
855: %=====================================================================
856: % DISCUSSION DISCUSSION DISCUSSION DISCUSSION DISCUSSION
857: %=====================================================================
858: %
859: \section{Discussion}
860: \label{discussion}
861: \subsection{The nature of the cavities and filaments in the inner 10 kpc}
862: %
863: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
864: \begin{figure}[t]
865: \centerline{\psfig{figure=f10.ps,height=0.3\textheight}}
866: \caption{\label{fig.cavities} \footnotesize Exposure corrected 0.5-5 keV image of
867: \source. Elliptical regions in black represent the shape and size of the
868: detected X-ray cavities.
869: }
870: \end{figure}
871: %
872: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
873: %
874: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
875: \def\arraystretch{1.1}
876: \begin{deluxetable*}{cccccccc}
877: \tabletypesize{\scriptsize}
878: \tablewidth{0pt}
879: \tablecaption{Cavities time scales and derived powers. \label{tab.powers}}
880: \tablehead{
881: \colhead{$pV$\tablenotemark{a}} & \colhead{$t_{sound}$} & \colhead{$t_{buoy}$\tablenotemark{b}} & \colhead{$t_{ref}$\tablenotemark{b}} & \colhead{$P_{sound}$\tablenotemark{c}} & \colhead{$P_{buoy}$\tablenotemark{c}} & \colhead{$P_{ref}$\tablenotemark{c}} & \colhead{$L_{X}$\tablenotemark{d}}\\
882: \colhead{($10^{55}$ erg)} & \colhead{($10^7$ yr)} & \colhead{($10^{7}$ yr)} & \colhead{($10^7$ yr)} & \colhead{($10^{41}$ \ergsec)} & \colhead{($10^{41}$ \ergsec)} &
883: \colhead{($10^{41}$ \ergsec)} & \colhead{($10^{41}$ \ergsec)}
884: }
885: \startdata
886: $4.58$ & $1.34$ & $1.20$ & $3.50$ & $1.08$ & $1.21$ & $0.41$ & $32.1\pm0.09$ \\
887: & & $0.69$ & $2.02$ & & $2.10$ & $0.72$ & \\
888: \enddata
889: \tablenotetext{a}{It refers to the sum of the $pV$ energies of the two cavities}
890: \tablenotetext{b}{The first value refer to $g$ calculated
891: according to the hydrostatic equilibrium mass estimate whereas
892: the second entry refer to the estimate made using the stellar velocity dispersion of the
893: central galaxy, as described in the text.}
894: \tablenotetext{c}{The power is the $pV/t$ work only, with $\gamma/(\gamma-1)$ not accounted for.
895: For a fully relativistic plasma the values for the powers need to be multiplied by 4.}
896: \tablenotetext{d}{Bolometric (0.1-100 keV) X-ray luminosity within a radius of 27 kpc, where
897: the cooling time is 3 Gyr.}
898: \end{deluxetable*}
899: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
900:
901:
902: The X-ray analysis shown in this paper reinforces the scenario proposed
903: for \source\ by \citet{Temi.ea:07*1}, which we briefly summarize below.
904: In a recent survey of elliptical galaxies observed with the {\it Spitzer} telescope
905: \citet{Temi.ea:07} found spatially extended cold interstellar dust emitting at
906: 70 $\mu$m around many group-centered X-ray luminous elliptical galaxies.
907: The source of this dust is the dust-rich cores of the central galaxy, not from mergers
908: \citep{Mathews.ea:03,Temi.ea:07}. Since this dust is thought to be in direct contact with
909: the hot, virialized interstellar gas ($kT \sim 1$ keV), it has a short lifetime ($\sim 10^7$ yrs)
910: to sputtering destruction by thermal ions.
911: Consequently, this dust is a spatial tracer of extremely transient events that recently
912: occurred on kpc scales. \source, together with NGC 4636, has been investigated in more detail
913: in \citet{Temi.ea:07*1}. \source\ is not only extended at 70 $\mu$m but also
914: at 8 $\mu$m in a manner similar to the
915: highly asymmetric H$\alpha$ optical line emission from warm gas.
916: Current evidence is consistent with the hypothesis that dust has been buoyantly transported
917: from the galactic cores out to several kpc following a feedback heating
918: event. Furthermore, disorganized fragments of optically absorbing dusty gas
919: are visible in {\it HST} images within the central $\sim$100 parsecs,
920: providing further evidence of a recent central energy release. In particular,
921: the association of interstellar PAH emission
922: (responsible of the 8 $\mu$m emission) and
923: warm gas ($T \sim 10^4$ K) in \source\ indicates that we may be
924: viewing this galaxy at a rare moment immediately following
925: a release of energy near the central black hole.
926:
927: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
928: \begin{figure*}[th]
929: %%\vspace{-0.5cm}
930: \centerline{
931: \parbox{0.5\textwidth}{
932: \psfig{figure=f11a.ps,height=0.28\textheight}}
933: \parbox{0.5\textwidth}{
934: \psfig{figure=f11b.ps,height=0.3\textheight}}
935: }
936: \caption{\label{fig.halpha} \footnotesize
937: \emph{Left Panel:} H$\alpha$ + [N \tiny $\rm{II}$ \footnotesize] map of \source\ taken from
938: \citet{Caon.ea:00}. \emph{Right Panel:} Surface brightness contours of the
939: H$\alpha$ + [N \tiny$\rm{II}$\footnotesize] map superimposed on the contour binned image of
940: Fig.\ref{fig.source}.
941: }
942: \end{figure*}
943: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
944:
945:
946: The \chandra\ data show the impact of this energy release on the hot X-ray
947: emitting gas which is undergoing short-term gas-dynamical activity.
948: The detection of a pair of X-ray cavities allow us to give an estimate of the
949: amount of energy deposited from the AGN and the time of its release, following
950: a now standard analysis \citep[e.g.][]{Birzan.ea:04,Dunn.ea:06}.
951: For each of the two cavities a size and position was measured, assuming the
952: projected shape is an ellipse measured from the exposure-corrected,
953: un-processed image as depicted in Fig.\ref{fig.cavities}.
954: We take as a guide the cavity on the south
955: because of the higher contrast and the surface brightness profiles of
956: Fig.\ref{fig.profiles}. The two cavities are at a projected distance $R$
957: of 33\arcsec\ (6.1 kpc) from the center of the bubble to the center of
958: \source, with a projected semi-major axis $a$ of 17\arcsec\ (3.1 kpc) and a
959: projected semi-minor axis $b$ of 9\arcsec\ (1.7 kpc).
960: We do not consider here the dip in
961: surface brightness at PA 212-244 because it not a clear deviation from
962: the azimuthally averaged surface brightness profile.
963:
964: Assuming that the cavities are bubbles completely devoid of gas at the local
965: ambient temperature and taking the pressure and temperature of the gas
966: surrounding the cavities
967: to be the azimuthally averaged values at the projected radius of its center as
968: calculated in the analysis of \citet{Gastaldello.ea:07*1}, we can calculate
969: the $P{\rm{d}}V$ work done on the X-ray gas by the AGN in inflating the
970: cavities. The actual third axis $c$ of the ellipsoid is unknown: we assume a
971: prolate shape ($b=c$). In this case we obtain
972: $pV = 2.29\pm0.08 \times 10^{55}$ ergs for each cavity (the gas
973: pressure at 6.1 kpc is $2.08\pm0.07 \times 10^{-11}$ dyne cm$^{-2}$).
974: Following \citet{Birzan.ea:04} we define three different time scales and therefore potential
975: ages for the bubbles: the sonic time scale $t_{c_s} = R/c_s$, where $c_s = 445\pm3$ km/s
976: taking $kT = 0.77\pm0.01$ at 6.1 kpc; the buoyancy time scale $t_{buoy} = R/\sqrt{SC/2gV}$
977: where $V$ is the volume of the bubble, $S$ is the cross section of the bubble and $C=0.75$ is
978: the drag coefficient \citep{Churazov.ea:01}. The gravitational acceleration $g$ can be
979: calculated either as $g=GM(<R)/R^2$ using the mass profile derived in
980: \citet{Gastaldello.ea:07*1} or, as there can be reasonable doubts about the strict applicability
981: of hydrostatic equilibrium in the inner kpc of \source\ \citep[we are likely
982: underestimating mass and therefore overestimating $t_{buoy}$, as we argued
983: in][given also the failure to detect the stellar mass of the central
984: galaxy]{Gastaldello.ea:07*1}, we use the stellar velocity dispersion of the
985: central galaxy, under the approximation of an isothermal sphere and calculate
986: $g \approx 2\sigma^2/R$ \citep{Binney.ea:87}; the time required to refill the displaced volume
987: as the bubble rises upward, $t_r = 2\sqrt{r/g}$, again using for $g$ either the
988: X-ray determined mass or the stellar dispersion velocity.
989: The calculated order-of-magnitude estimates for the time scales described above
990: are shown in Table \ref{tab.powers}.
991: and they are consistent with the short
992: sputtering lifetime for the extended dust, $\sim 10^7$ yrs, discussed in
993: \citet{Temi.ea:07*1} and point towards the same episode of AGN feedback.
994:
995: We can compare estimates of the energy released by the AGN to the radiative
996: losses within the cooling region, here estimated to be the region within
997: a radius where the cooling time is less than 3 Gyr \citep[e.g.][]{Dunn.ea:06},
998: also tabulated in Table \ref{tab.powers} which corresponds to 27 kpc.
999: The task of deriving the outburst
1000: energy from the observable $pV$ is complicated by various unknowns, like
1001: the adiabatic index $\gamma$ of the material inside the bubble
1002: \citep[$\gamma=5/3$ if thermal or $\gamma=4/3$ if non-thermal, see for example the
1003: discussion in][]{Mathews.ea:08} or the correct hydrodynamic evolution of the
1004: bubble expansion, like its initial over-pressure compared to the ICM or the
1005: possibility of continuous inflation, to explain the paradoxical result of
1006: bubble energy increasing with distance from the center of the gravitational
1007: potential \citep{Diehl.ea:08}. Using the commonly adopted recipe of estimating
1008: as $4pV$, with $\gamma=4/3$, the energy input arising from the cavities alone,
1009: this energy can be at most $\sim8\times10^{41}$ \ergsec, which falls short to
1010: counteract cooling losses.
1011:
1012: It is remarkable that a pair of cavities close to the nuclear source are
1013: lacking extended high frequency radio emission. This is contrary to what is
1014: generally observed in particular in clusters of galaxies.
1015: Analysis of VLA observations
1016: at 1.56 and 4.9 GHz confirms the presence of just a point source with
1017: beam-deconvolved size $< 3$\arcsec and with a very flat ($\alpha \sim 0$)
1018: spectral index (Giacintucci et al., in preparation).
1019: Two other sources sharing the same behavior are
1020: NGC 4636 \citep[e.g.,][]{JonesC.ea:02} and NGC 4552 \citep{Machacek.ea:06}.
1021: However the cavities in these sources have been interpreted as caused by a
1022: nuclear outburst which is directly causing shock heating, also because the
1023: brighter rims surrounding the cavities are actually hotter than the
1024: surrounding medium, rather than displacement by the radio lobes
1025: \citep{JonesC.ea:02}. Or the cavities could just be related to a previous
1026: outburst and might just be ghost cavities, as
1027: suggested by \citet{OSullivan.ea:05*1} for NGC 4636 and the same reasoning
1028: could apply to NGC 5044. In that case it remains to be explained why the
1029: synchrotron emission has faded to undetectable values in these sources and
1030: not in cavities with approximately the same ages in many other sources
1031: \citep[e.g.,][]{Birzan.ea:04}. Other ghost cavity systems with a weak central radio source
1032: bearing no obvious relation to the observable cavities which are close to the central
1033: galaxy are HCG 62 \citep{Morita.ea:06} and NGC 741 \citep[][where only one cavity is present]{Jetha.ea:08}.
1034: Sensitive low-frequency radio observations will shed further light on this issue.
1035:
1036: The association of dust, H$\alpha$ and soft X-ray emission showed by
1037: \citet{Temi.ea:07*1} is also strengthened by the analysis in this paper.
1038: The presence of the N cavity and in particular of the relatively cooler
1039: emission in the NW filament (see \S\ref{subsection_spectral.core})
1040: likely explains the origin of the N H$\alpha$ filament. X-ray filaments are present at
1041: both sides of the cavities, but only
1042: the ones with the presence of dust are showing optical emission and
1043: \emph{cooler} X-ray emission (see Fig.\ref{fig.halpha} and the analysis of
1044: \S\ref{subsection_spectral.core}). \citet{Temi.ea:07*1} showed that the
1045: cospatiality of these features can be explained as the result of dust-assisted cooling
1046: in an outflowing plume of hot dusty gas: dust can cool buoyant gas to $10^4$ K, which
1047: emits the optical emission lines observed. The warm-gas phase is maintained in
1048: thermal equilibrium near $\sim10^4$ K by radiative losses and likely a combination
1049: of thermal conduction and UV heating from post-AGB stars
1050: \citep[see the calculation of Tab.5 in][ ionization from post-AGB stars
1051: could explain only 42\% of the optical line emission in \source]{Macchetto.ea:96}.
1052: It is unfortunate that the deeper H$\alpha$ observation of \citet{Caon.ea:00}, compared to the
1053: one by \citet{Goudfrooij.ea:94} used in the comparison of \citet{Temi.ea:07*1},
1054: is affected by a CCD defect in the southern region co-spatial to the X-ray
1055: cavity: it looks like the nebular emission is more extended and can cover
1056: also the cavity.
1057: Another tantalizing evidence of the association
1058: of these three components comes for example from NGC 5846
1059: \citep[][and references therein]{Trinchieri.ea:02}. Evidence of PAH emission in
1060: \spitzer\ IRS spectra has been found also in NGC 1275 \citep{Johnstone.ea:07}.
1061: It has been discussed that AGN feedback can provide
1062: some heating (negative feedback), but it can also be responsible for
1063: a positive feedback, i.e. for inducing cold gas production
1064: \citep[e.g.,][]{Pizzolato.ea:05,Revaz.ea:08}. In this scenario AGN feedback
1065: itself is responsible for the production of overdense blobs which cool
1066: rapidly producing H$\alpha$ emitting gas, molecular gas and star formation.
1067: Dust transported from the dusty disks of the central elliptical into
1068: the ICM by the episode of AGN feedback acts as a catalyst for the cooling
1069: of the gas.
1070: It would be therefore interesting for the recent renewed theoretical
1071: interest in the generation and survival of the optical line emission filaments
1072: \citep{Nipoti.ea:04,Revaz.ea:08,Pope.ea:08} to include the neglected
1073: dust-assisted cooling in the energy balance equation.
1074:
1075: \subsection{The cold fronts and the dynamical state of the group NGC 5044}
1076:
1077: The spectral analysis presented in \S\ref{subsection_spectral.coldfronts}
1078: indicated that the pair of surface brightness discontinuities detected
1079: in the \chandra\ and \xmm\ images are cold fronts.
1080: Given the large scale relaxed morphology of
1081: \source\ and its rising entropy profile \citep{Gastaldello.ea:07*2}
1082: as commonly observed for bright relaxed groups, the source nicely fits
1083: in the explanation for the emergence of these features proposed by
1084: \citet{Ascasibar.ea:06} as due to gas sloshing caused by an off-axis merger with a
1085: smaller satellite.
1086: Another tell-tale sign of this encounter, as also suggested by
1087: \citet{Ascasibar.ea:06}, is the presence of a peculiar velocity of the
1088: central galaxy, which is often exhibited by central cD galaxies in relaxed
1089: clusters \citep[e.g.,][]{Oegerle.ea:01}.
1090: Indeed NGC 5044 is known to have a $\sim$ 150 km/s peculiar velocity
1091: with respect to the mean group velocity
1092: \citep[][and references therein]{Mendel.ea:08}. A sub-clump of galaxies statistically
1093: significant to the Dressler-Schectman test \citep{Dressler.ea:88}
1094: has been detected in the north-east outskirts of the \source\ \citep{Mendel.ea:08}
1095: which might be the smaller merging sub-group responsible for the gas sloshing and
1096: the cold fronts. For a virial mass for \source\ of $3.7 \times 10^{13}$ \msun\
1097: \citep{Gastaldello.ea:07*1}, assuming a mass ratio of 5 as in the reference
1098: case of \citet{Ascasibar.ea:06}
1099: would imply a mass of $\sim 8 \times 10^{12}$ \msun\ for the satellite.
1100: If this candidate sub-clump is real, it would corroborate the proposed scenario and the
1101: mechanism proposed by \citet{Ascasibar.ea:06} for the formation of
1102: cold fronts \citep[see also the case of the cluster A 496,][]{Dupke.ea:07}
1103:
1104: Surface brightness discontinuities as cold fronts have so far been
1105: discussed in many merging elliptical galaxies and groups
1106: \citep[e.g., NGC 1404,][]{Machacek.ea:05}. An unusual discontinuity
1107: for the group NGC 507, interpreted as an
1108: abundance jump, closely related to the expansion of a radio lobe, has been discussed
1109: by \citet{Kraft.ea:04}. NGC 5044 is to our knowledge the first relaxed group for which
1110: cold fronts have been discussed in close similarity to the ubi\-qui\-tous ones
1111: detected in relaxed clusters. Further examples, given enough data quality, are likely
1112: to be discovered: suggestions have already been made for objects like MKW 4 and
1113: IC 1860 \citep{Gastaldello.ea:07*1}.
1114: %
1115:
1116: %=====================================================================
1117: % CONCLUSION CONCLUSION CONCLUSION CONCLUSION CONCLUSION
1118: %=====================================================================
1119: \section{Conclusions}
1120: \label{conclusion}
1121:
1122: We have presented results of a two-dimensional analysis of the currently available
1123: \chandra\ and \xmm\ data for the bright nearby galaxy group \source. The results
1124: can be summarized as follows:
1125:
1126: \begin{itemize}
1127: \item A pair of X-ray cavities have been detected, further confirming the recent
1128: outburst indicated by the extent and morphology of H$\alpha$ and dust emission.
1129: \item The presence of cooler filamentary X-ray emission co-spatial with H$\alpha$ and dust emission strengthens our previous suggestion that dust-aided cooling could
1130: contribute to the H$\alpha$ emission.
1131: \item The presence of a set of two cold fronts together with a peculiar velocity
1132: of the central galaxy NGC 5044 suggests a disturbance of an overall relaxed system
1133: by an off-axis merger with a smaller satellite.
1134: \end{itemize}
1135:
1136: The detection of such a rich phenomenology has been possible due to \source\
1137: being bright and nearby, much like on bigger mass scales has been possible
1138: with M87 and Perseus. As for these objects up-coming deeper X-ray observations
1139: with \chandra\ \citep{David:07} and \xmm\ \citep{Kaastra:07} will shed
1140: further light on the properties of \source\ and groups of galaxies in general.
1141: Multi-wavelength observations, i.e. in H$\alpha$ and radio bands, are needed
1142: for a deeper understanding of the physical processes in the core.
1143:
1144: %=====================================================================
1145: % ACKNOWLEDGEMENTS ACKNOWLEDGEMENTS ACKNOWLEDGEMENTS
1146: %=====================================================================
1147: \begin{acknowledgements}
1148: We would like to thank N. Caon for kindly providing the H$\alpha$ image used
1149: in Figure \ref{fig.halpha} and for useful discussions; S. Giacintucci for
1150: sharing her results prior to publication and for useful
1151: discussions; P.J. Humphrey for the use of his \chandra\ data reduction and
1152: analysis code; J. Sanders for the use of his contour binning code;
1153: L. Zappacosta for a critical reading of the manuscript; the anonymous referee
1154: for a carefulr reading and for suggestions that improved the work
1155: presented here.
1156: We made use of the WVT binning algorithm by Diehl \& Statler (2006), which is
1157: a generalization of Cappellari \& Copin's (2003) Voronoi binning algorithm.
1158: F.G., F.B. and S.E. acknowledge the financial contribution from contract
1159: ASI-INAF I/023/05/0 and I/088/06/0. D.A.B. gratefully acknowledges partial support
1160: from NASA grant NNG04GE76G, issued through the Office of Space Sciences
1161: Long-Term Space Astrophysics Program.
1162: \end{acknowledgements}
1163: %
1164:
1165: {\it Facilities:} \facility{CXO}, \facility{XMM}
1166: %
1167: %
1168: %%=====================================================================
1169: %% REFERENCES REFERENCES REFERENCES REFERENCES REFERENCES
1170: %%=====================================================================
1171: %
1172: \bibliographystyle{apj}
1173: \bibliography{gasta}
1174:
1175: \end{document}
1176: