1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{emulateapj5}
3: \usepackage{epsfig}
4: \newcommand{\chandra}{{\it{Chandra}}}
5: \newcommand{\rosat}{{\it ROSAT}}
6: \newcommand{\asca}{{\it ASCA}}
7: \newcommand{\xmm}{{\it XMM}}
8: \newcommand{\xmmn}{{\it XMM-Newton}}
9: \newcommand{\beppo}{{\it BeppoSAX}}
10: \newcommand{\mdot}{\dot{M}}
11: \newcommand{\msunyr}{{\rm M}_{\sun}~{\rm yr}^{-1}}
12: \newcommand{\fx}{ergs s$^{-1}$ cm$^{-2}$}
13: \newcommand{\lxh}{$h_{70}^{-2}$ ergs s$^{-1}$}
14: \newcommand{\hkpc}{$h_{70}^{-1}$~kpc}
15: \newcommand{\hmpc}{$h_{70}^{-1}$~Mpc}
16: \newcommand{\zsun}{$Z_{\sun}$}
17: %%%modified emulateapj for figures...
18: %\newcommand{\myfigure}{figure} % for normal
19: \newcommand{\myfigure}{figurehere} % for emulate apj
20: \makeatletter
21: %\newenvironment{tablehere}
22: % {\def\@captype{table}}
23: % {}
24: \newenvironment{figurehere}
25: {\def\@captype{figure}}
26: {}
27: \makeatother
28: %
29: \submitted{Received 2002 January 24; Accepted 2002 May 22}
30: %\journalinfo{Vol. XXX, No. X}
31: %
32: %\slugcomment{Submitted to the Astrophysical Journal Letters, 05/21/02}
33:
34: \shorttitle{\chandra{} observations of A2029}
35: \shortauthors{Lewis et al.}
36:
37: \begin{document}
38:
39: \title{\chandra{} Observations of Abell 2029: No Cooling Flow and a
40: Steep Abundance Gradient}
41:
42: \author{Aaron D. Lewis\altaffilmark{1,2}, John T.
43: Stocke\altaffilmark{2}, David A. Buote\altaffilmark{1}}
44: \email{lewisa@uci.edu, stocke@casa.colorado.edu, buote@uci.edu}
45:
46: \altaffiltext{1}{University of California, Irvine, Department of Physics
47: and Astronomy, 4171 Frederick Reines Hall, Irvine, CA, 92697-4575}
48: \altaffiltext{2}{Center for Astrophysics and Space Astronomy, University
49: of Colorado, 389 UCB, Boulder, CO 80309}
50:
51: \received{}
52:
53: \accepted{}
54:
55: \journalid{}{}
56:
57: \articleid{}{}
58:
59: \begin{abstract} We have obtained high spatial resolution temperature
60: and abundance profiles for the galaxy cluster Abell 2029 with the
61: \chandra{} ACIS-S instrument. Our observations reveal that the spectra
62: are well-fit by a single-phase gas in each annulus. While the
63: temperature of the intracluster medium drops from $\sim9$~keV at
64: 3\arcmin{} (260~\hkpc) to $3$~keV in the central $5\arcsec$ (5~\hkpc) of
65: the cluster, there is no evidence for gas emitting at temperatures below
66: 3~keV. The addition of a cooling flow component does not improve the
67: fits, despite previous claims for a massive cooling flow. There is also
68: no evidence for excess absorption above the Galactic N$_H$ value. We
69: also observe a steep Fe abundance gradient, such that $Z_{Fe} \gtrsim
70: 2$~\zsun (assuming meteoritic solar Fe abundance) in the core,
71: consistent with significant enrichment from Fe-rich Type Ia supernovae
72: in the cD galaxy. The Fe abundance drops to $\approx 0.5$~\zsun{} at
73: 3\arcmin{} (260~\hkpc), in good agreement with previous \beppo{}
74: measurements. The high resolution image reveals neither a strong central
75: point source, nor any filamentary structure related to a cooling flow or
76: a merger. The absence of a strong merger signature argues against the
77: creation of the wide-angle-tail radio source morphology in a merger
78: event. \end{abstract}
79:
80: \keywords{galaxies:clusters:individual (A2029) --- cooling flows ---
81: intergalactic medium --- X-ray:galaxies --- galaxies:abundances}
82:
83: \section{Introduction\label{sec_intro}}
84: %\epsscale{0.5}
85:
86: Abell 2029 (A2029 hereafter) is a richness class II, Bautz-Morgan type I
87: nearby cluster of galaxies at a redshift of $z=0.0767$ \citep{abe89}. It
88: has a very large cD galaxy (UGC 9752 = IC 1101) with a stellar envelope
89: extending well beyond 600~\hkpc{} \citep{uso91}, and is one of the most
90: optically regular rich clusters known \citep{dre78}. It has been well
91: studied optically \citep[see e.g.,][]{dre81,joh87}, and can be
92: characterized as a compact, relaxed, cD-galaxy dominated system, with no
93: emission-line emitting galaxies within 600~\hkpc{} of its core
94: \citep{dal00}. Throughout this Letter, we assume a cosmology of
95: H$_0=70$~$h_{70}$~km~s$^{-1}$~Mpc$^{-1}$, $\Omega_{matter}=0.3$, and
96: $\Lambda=0.7$, implying a luminosity distance to A2029 of 347~\hmpc{}
97: and a scale of 1.45 kpc~arcsec$^{-1}$.
98:
99: A2029 has been previously studied in the X-rays using various
100: instruments: the \rosat{} HRI \citep{sar92}, \rosat{} PSPC and \asca{}
101: \citep{buo96,sar98,whi00}, and \beppo{} \citep{mol99,irw00}. It is
102: one of the brightest X-ray clusters in the sky ($F_X = 7.5 \times
103: 10^{-11}$\fx{} in the 2-10 keV band, \citealt{dav93}), corresponding to
104: $L_X (2-10$keV)$ =1.1 \times 10^{45}$~\lxh{}. From \asca{} spectra,
105: \citet{mol99} find an intracluster medium (ICM) temperature in the
106: central 2\arcmin{} of $T_X = 8.3\pm0.2$~keV declining to $\sim5$~keV at
107: a radius of 10\arcmin{} (972~\hkpc).
108:
109: Previous imaging and spectral studies have inferred a large cooling flow
110: with $\mdot\simeq 400-600$~$\msunyr$ \citep{edg92,fabi94,arn88,per98},
111: \citep[but c.f.,][]{whi00}. However, evidence that it is not a cooling
112: flow includes no detection of H$\alpha$ or \ion{O}{2} emission in the cD
113: galaxy, and no blue stellar colors in the core \citep{mcn89}. The cD
114: galaxy also hosts a wide-angle-tail (WAT) radio source (PKS 1509+49 =
115: 4C+06.53), which is highly unusual in a cooling flow system
116: \citep{bur90}. Simulations show that WATs may be the signatures of
117: ongoing cluster mergers; these simulated systems also exhibit hot shock
118: fronts \citep{lok95,roe96}.
119:
120: Obviously the relaxed cooling flow and merger/hot shock scenarios are
121: mutually exclusive, and for this reason we have obtained high spatial
122: resolution imaging spectroscopy with the \chandra{} observatory in order
123: to
124: \begin{\myfigure}
125: \centerline{\epsfig{file=f1.eps, scale=0.50}}
126: %\epsscale{0.5}
127: %\plotone{f1.eps}
128: \caption{\chandra{} ACIS-S image of Abell 2029. The image is 4$\arcmin$
129: (348 \hkpc) on a side and has been smoothed with a Gaussian of $\sigma =
130: 2\arcsec$. North is up and East is left. Logarithmically spaced contours
131: are overlaid to indicate morphology.}
132: \label{fig_image}
133: \end{\myfigure}
134: \smallskip
135: \noindent
136: investigate this enigmatic object. In this Letter we describe the
137: global temperature structure, resolved metal abundances, as well as
138: highly spatially-resolved spectral properties of the central regions. A
139: future publication will measure the dark matter profile.
140:
141: We note that the authors of most X-ray spectral analyses to date have
142: used ``photospheric'' Fe abundances. Following the suggestion of
143: \citet{ish97}, we use the abundance measurements of \citet{and89} for
144: spectral fitting, except for Fe where we use the correct ``meteoritic''
145: value, Fe/H $ = 3.24 \times 10^{-5}$ by number. This results in Fe
146: abundance values a factor 1.44 times larger than the ``photospheric''
147: values. We will note this factor when comparing to other work.
148:
149: \section{Observations and Data Reduction\label{sec_obs}}
150:
151: A2029 was observed with the \chandra{} observatory on 2000 12 April for
152: a deadtime-corrected exposure of 19.8 ksec. The cluster was centered on
153: the backside-illuminated S3 chip, operating at a temperature of -120 C.
154: The data were re-processed using CIAO 2.2.1 and version 2.7 of the
155: \chandra{} calibration database (CALDB). We mitigated the effects of
156: charge transfer inefficiency using the ACISCtiCorrector.1.37
157: software\footnote{Available from the \chandra{} contributed software
158: page at \url{http://asc.harvard.edu/cont-soft/soft-exchange.html}}
159: \citep{tow00}. The data were screened for energy (0.3 to 8.0 keV), grade
160: (ASCA 02346), and status (0). We examined the lightcurve for possible
161: flaring of the background, and removed 365 seconds of suspect exposure.
162: We then scaled and subtracted the available source-free extragalactic
163: sky background maps\footnote{The background maps have been CTI-corrected
164: in exactly the same manner as the source data.} using the make\_acisbg
165: software created by Maxim Markevitch \citep[see, e.g.][]{mar00}.
166: %\footnote{Available at
167: %\url{http://cxc.harvard.edu/contrib/maxim/acisbg/}}.
168:
169: In Figure \ref{fig_image} we present the smoothed, background-subtracted ACIS-S
170: image of A2029 in the energy range $0.3-8.0$ keV. The image reveals A2029 to be
171: quite regular; although there are some position-angle dependent variations
172: within a $1\arcmin$ radius, we find no evidence of bright shock fronts,
173: filaments, or bright point sources. The X-ray emission is somewhat elliptical,
174: with a position angle of $\sim25\arcdeg$ (North to East), consistent with the
175: \rosat{} PSPC image analysis of \citet{buo96}.
176:
177:
178: \section{Spatially Resolved Spectral Properties \label{sec_space}}
179:
180: To analyze the spectrum of the hot ICM, we extracted spectra from eleven
181: concentric annuli centered on the peak of the X-ray emission (coincident
182: with the BCG at RA = $15^h 10^m 56.^s2$, Decl. = $+ 05\arcdeg 44\arcmin
183: 44\arcsec$, J2000). The annuli were of varying widths providing
184: approximately 10,000 counts per annulus in the energy range $0.3-8.0$
185: keV, except that to investigate the core with maximum spatial
186: resolution, we chose a central region $5\arcsec$ in radius, containing
187: approximately 5,000 counts.
188:
189: \subsection{Radial Temperature Profile\label{subsec_radt}}
190:
191: Using {\sc xspec}, we fit the extracted spectra with the {\sc apec}
192: plasma emission model\footnote{We have also performed fits with the {\sc
193: mekal} model. We find no statistical differences in temperature or
194: abundance measurements.} absorbed by neutral hydrogen. We adopt the
195: weighted average Galactic value of N$_H = 3.14 \times 10^{20}$~cm$^{-2}$
196: obtained from the W3N$_H$ HEASARC tool.
197: The Fe abundance was allowed to
198: be a free parameter, with all other elements tied to Fe in their solar
199: ratios.
200:
201: \begin{\myfigure}
202: \centerline{\epsfig{file=f2.eps, scale=0.33}}
203: %\epsscale{0.33}
204: %\plotone{f2.eps}
205: \caption{\chandra{} radial temperature profile of A2029. Solid lines (and filled
206: circles) are the results of the 3D spectral deprojection. Dotted lines (and
207: open squares) are the 2D analysis. Horizontal axis shows both arcsec ({\it lower
208: axis}) and \hkpc ({\it upper axis}) units.}
209: \label{fig_t}
210: \end{\myfigure}
211: \smallskip
212: %\noindent
213:
214: To properly recover the three-dimensional properties of the
215: X-ray emitting ICM, we have performed a spectral deprojection analysis
216: using the {\sc xdeproj} code of \citet{buo00c}. We start at the outside
217: working our way to the core in an ``onion-peeling'' method which
218: accounts for the cumulative projected emission from the outer annuli,
219: obtaining temperatures, densities, abundances, and any other desired
220: parameters. For details of our deprojection technique see \citet{buo00c}
221: and \citet{P7}.
222: To estimate the uncertainties on the fitted parameters we simulated
223: spectra for each annulus using the best-fitting models and fit the
224: simulated spectra in exactly the same manner as the actual data. From
225: 100 Monte Carlo simulations we compute the standard deviation for each
226: free parameter which we quote as the ``1$\sigma$'' error.
227: The reference model we have used is {\sc apec} with deprojection,
228: denoted {\sc apec} (3D).
229: %$chi^2$/dof = 134.02/111.
230: For comparison, we have also performed a standard two-dimensional
231: analysis with no deprojection, denoted {\sc apec} (2D).
232: % $chi^2$/dof = 138.17/111.
233:
234: %%\begin{tablehere}
235: %\begin{center}
236: %\begin{deluxetable}{lc}
237: %\tablecaption{Models and Fit Quality in the Core\label{tab_fits}}
238: %%\begin{tabular}{lc}
239: %%\tableline
240: %%\tableline
241: %\tablehead{
242: %\colhead{Model} & \colhead{($\chi^2$/dof)} \\
243: %}
244: %\startdata
245: %%\tableline
246: %{\sc apec} (3D) & 134.02/111 \\
247: %{\sc apec} (2D) & 138.17/111 \\
248: %{\sc apec} (3D-0.7) & 87.10/91 \\
249: %{\sc apec+apec (3D)} & 132.60/108 \\
250: %{\sc apec}+CF (3D) & 133.29/109 \\
251: %{\sc apec}+POW & 133.62/109 \\
252: %\enddata
253: %%\tableline
254: %%\end{tabular}
255: %%\smallskip
256: %%{\sc Note.}--
257: %\tablecomments{Table models are defined in \S\S \ref{subsec_radt}
258: %\& \ref{subsec_alt}.}
259: %\end{deluxetable}
260: %\end{center}
261: %%\end{tablehere}
262: %
263: In Figure \ref{fig_t} we show the
264: radial temperature profile of A2029 derived from the 3D analysis ({\it
265: solid lines} and filled circles) and the 2D analysis ({\it dotted lines}
266: and open squares). Figure \ref{fig_t} shows an obvious drop in
267: temperature in the core, with increasing values at larger radii (our
268: final bin is between 160-186 arcsec, $2.7-3.1\arcmin$, or
269: 232-270~\hkpc). The 3D model shows a more dramatic drop in the center,
270: the signal of which is masked in the 2D fits as hotter gas from larger
271: radii is observed in projection over the cooler core. We observe the gas
272: temperature to be approximately 8-10 keV at 100-200 arcseconds. The gas
273: drops to a minimum temperature of $3.1\pm0.4$~keV in the inner
274: $5\arcsec$. The data suggest a continuing increase in temperature,
275: though the profile is possibly flattening at the limit of our
276: measurements. We note that the sharp bin-to-bin fluctuations seen for
277: the 3D model in the four outermost annuli are due to the nature of the
278: 3D deprojection algorithm, and should not be considered a physical
279: feature. This effect can be avoided by regularizing (in effect
280: smoothing) the parameters, which we do not perform for this Letter (but
281: see \citealt{buo00c} for details of the procedure). The 3D fluctuations
282: in the outer bins vary about the 2D temperature values in these bins
283: without significant bias. This is expected because at large radii
284: the gas is nearly isothermal and also suffers the least from projection
285: effects, such that the 2D profile can be used to verify the 3D results in these
286: annuli. The quality of the fits is only marginally improved using the 3D vs. 2D
287: method, with the most improvement occurring in the inner 2 annuli where we also
288: observe the largest change in temperature.
289:
290: The formal quality of the spectral fits is quite good ($\chi^2$/dof =
291: 134.02/111 in the central annulus, for the reference model), but there exist
292: significant (10-15\%) residuals below 0.5 keV, where the data are in excess of
293: the model. At these energies we have either under-subtracted the background or
294: there are calibration issues outstanding. In the outer annuli where the
295: observations are background dominated (especially at high energy), the
296: background estimate seems secure and we are inclined to doubt the calibration
297: at E $< 0.5$~keV. Between 0.65 and 0.75 keV, the data fall below the model in
298: the inner three annuli, but this `absorption feature' is much sharper than the
299: spectral resolution of the S3 chip and its nature is not clear. These features
300: can be see in Fig. \ref{fig_spec}, wherein we present the spectrum and
301: residuals from the fit in Annulus 2 (between 5 and 17\arcsec).
302:
303: Allowing the neutral hydrogen column density to be a free parameter in
304: the fits results in N$_H$ values less than half the Galactic value in
305: the outer regions which climb to slightly higher than Galactic in the
306: central bin.
307: %, with $\sim20\%$ errors in that bin.
308: As a test, we have performed an additional set of fits eliminating all
309: energies below 0.7 keV ({\sc apec} (3D-0.7) $\chi^2$/dof = 87.10/91). We
310: have repeated this exercise with all the other models presented in this
311: Letter. We find that the removal of low energy residuals does not change
312: either the temperature or abundance measurements beyond their errors.
313: The rise in N$_H$ in the center would appear to accommodate the
314: `absorption feature' near 0.6 keV in the central bin, but elimination of
315: energies below 0.7 keV merely weakens the constraint on N$_H$ without
316: lowering the fitted value. In either case, there is almost no
317: improvement in the fit quality in the inner 2 annuli. Freeing N$_H$ does
318: result in a significant improvement in the fit quality for annuli 3-11.
319: As expected, eliminating energies below 0.7 keV in the fits obtains
320: higher (but still sub-Galactic) N$_H$ values. Although it is possible
321: there is a decrement in actual Galactic N$_H$ along the line of sight to
322: A2029, this would not be consistent with any previous observations (at a
323: variety of wavelengths).
324: %We suggest that the
325: %sub-Galactic measurements of N$_H$ are due to systematic errors, a
326: %hypothesis which will be examined in the light of forthcoming data on
327: %the low energy calibration of the ACIS-S instrument.
328: %
329: %,
330: %though it has some affect on the N$_H$ values when these are free
331: %parameters (see below). We will further investigate these residuals, and
332: %their possible causes in future work.
333:
334: \subsection{Fe abundance profile}
335:
336: In Figure \ref{fig_z} we show the radial metal abundance profile of
337: A2029 from the {\sc apec} (3D) model fits. Ratio of Fe to its solar
338: (meteoritic) value is shown. We observe subsolar (approximately 0.5
339: \zsun) abundances outside of 20\arcsec, and a very sharp increase in the
340: core to $\sim2$ \zsun. The abundances of other elements (S, Si, Mg and
341: O) are not constrained to better than $\sim50\%$ except in a few central
342: bins where the emission lines are more pronounced relative to the lower
343:
344: \begin{\myfigure}
345: \centerline{\epsfig{file=f3_newxv.ps, scale=0.30}}
346: %\epsscale{0.4}
347: %\plotone{f3.eps}
348: \caption{Spectrum in Annulus 2 (5-17\arcsec) between 0.3 and 2.0 keV.
349: Fractional residuals shown at bottom. Model is {\sc apec} (3D), with Galactic
350: N$_H$ absorption. Note that we do not see these residual features at larger radii.}
351: \label{fig_spec}
352: \end{\myfigure}
353: \smallskip
354: \begin{\myfigure}
355: \centerline{\epsfig{file=f4.eps, scale=0.30}}
356: %\epsscale{0.35}
357: %\plotone{f4.eps}
358: \caption{\chandra{} deprojected radial Fe abundance profile of A2029.
359: Horizontal axis shows both arcsec ({\it lower axis}) and kpc ({\it upper
360: axis}) units.}
361: \label{fig_z}
362: \end{\myfigure}
363: \smallskip
364: \noindent
365: temperature ICM. As we are unable to place significant constraints on
366: any interesting abundance ratios we do not present these data. We note
367: that allowing additional element abundances to vary does not effect the
368: Fe abundances beyond the errors shown in Figure \ref{fig_z}.
369:
370: Our measurements are entirely consistent with the $\sim$arcmin spatial
371: resolution \beppo{} results of \citet{mol99} and \citet{irw01},
372: who measured $Z_{Fe} \sim0.5$~\zsun{} (0.72~\zsun{} in meteoritic units)
373: within the central 2\arcmin, declining to $\sim0.2$ (0.29) at larger
374: radii ($\sim10\arcmin$). An emission-weighted average of our
375: best-fitting \chandra{} Fe abundance measurements in bins 1-9
376: ($0-2.2\arcmin$) yields $Z_{Fe} = 0.77$~\zsun{} (meteoritic). We also
377: see a decline to values consistent with 0.35~\zsun{} (within our errors)
378: for our outermost data points which extend to 3\arcmin.
379: %\subsection{Intrinsic Absorption}
380: %
381: %Allowing the neutral hydrogen column density to be a free
382: %parameter in the fits results in N$_H$ values
383: %less than half the Galactic value in the outer regions which climb to
384: %slightly higher than Galactic in the central bin, with $\sim20\%$ errors in that bin.
385: % (though consistent with Galactic
386: %within the error on N$_H$).
387: %
388: %Similarly, if allowed to be free, the Oxygen abundance drops to zero at all radii to
389: %accommodate the 0.6 keV ``absorption'' (there is a bright O line at 0.55 keV which
390: %must be eliminated to improve the fit). A complete lack of oxygen is incompatible with
391: %the Fe abundances measurements (which are securely constrained by both the 6.7 keV Fe
392: %line, and the Fe-L complex near 1 keV at high and low temperatures, respectively.)
393: %
394: %Neither free N$_H$ nor free O abundance affects either the ICM temperature or Fe
395: %abundance measurements beyond the errors shown in Figures \ref{fig_t} and \ref{fig_z}.
396: %
397: %Although some previous fits indicated intrinsic absorption \citep{whi91,all97},
398: %specifically related to an inferred cooling flow, subsequent X-ray analysis showed no
399: %need for intrinsic absorption, and fits consistent with the Galactic N$_H$ column
400: %density \citep{sar98,whi00}. The \citet{sar98} analysis allows for a final degeneracy
401: %wherein significant absorption exists within a central cooling flow region, the
402: %signature of which is masked by the cooling flow emission itself. We address the
403: %possibility of a cooling flow in the next section.
404:
405: \section{Analysis of the Core\label{sec_core}}
406:
407: The unsurpassed spatial resolution of \chandra{} allows us to
408: investigate the core of A2029 with high S/N in the central 5\arcsec{}
409: region. The observation of A2029 was pointed 1.3\arcmin{} from the
410: ACIS-S aimpoint (the cluster center was placed 1.2\arcmin{} south, and
411: 30\arcsec{} east of the aimpoint to place more extended emission on the
412: S3 chip), which should result in an effective PSF of $\lesssim
413: 1.5\arcsec$ FWHM (it is within the region of the detector where the
414: encircled energy within 1\arcsec{} is $\geq 50\%$\footnote{see the
415: \chandra{} Proposers' Observatory Guide Rev 3.0.}).
416:
417: \subsection{Alternative Spectral Models\label{subsec_alt}}
418:
419: From our single-temperature fit we obtain $T_X = 3.1\pm0.4$~keV, and a
420: good fit to several strong emission features across the entire spectrum,
421: except for the low energy absorption feature near 0.6 keV. Although the
422: quality of the fit in this annulus is acceptable (134.02/111
423: $\chi^2$/dof), we have investigated additional or alternative spectral
424: models.
425:
426: We have added to the first three annuli in turn: a second temperature
427: component ({\sc apec+apec} (3D) $\chi^2$/dof = 132.60/108), a constant
428: pressure cooling flow, with the material cooling from the temperature of
429: the {\sc apec} model ({\sc apec}+CF (3D) $\chi^2$/dof = 133.29/109), and
430: a power law model ({\sc apec}+POW (3D) $\chi^2$/dof = 133.62/109). Each
431: model was absorbed by N$_H$, which was fixed to the Galactic value for
432: all models except CF. None of these models provided a significant
433: improvement to the fits. The cooling flow fit yields a cooling rate of
434: $0.0\pm0.7\msunyr$, similar to other recent XMM-Newton analyses \citep[see,
435: e.g.,][]{kaa01,boh02}.
436: %in disagreement with the analysis of \citet{per98},
437: %who found $\mdot=554\msunyr$.
438:
439: \subsection{Constraints on a Central Point Source}
440:
441: In Figure \ref{fig_sb} we show the azimuthally averaged surface
442: brightness profile of A2029 obtained from counts in the $0.3-8.0$~kev
443: band. We extracted counts from the background-subtracted image in annuli
444: 1.5\arcsec{} wide out to a radius of 186\arcsec{}. We then subtracted an
445: in-field background taken from a region outside the extraction area
446: still on the S3 chip (the southwest corner).
447: %We have not performed an
448: %exposure correction due to the current limitations on the accuracy of a
449: %wide-band {\it and} wide-field exposure map.
450: For the present analysis of
451: the core, the impact of an exposure map is negligible, and we have not
452: performed such a correction.
453: \begin{\myfigure}
454: \centerline{\epsfig{file=f5.eps, scale=0.5}}
455: %\epsscale{0.5}
456: %\plotone{f5.eps}
457: \caption{\chandra{} surface brightness profile of A2029 in the
458: $0.3-8.0$~keV band ({\it open squares}). Overlaid are two Gaussians of
459: FWHM 1.5 ({\it dashed curve}) and 5\arcsec ({\it dotted curve}),
460: respectively, as well as a $\beta$-model fit ({\it solid curve}), with
461: $\beta=0.60$, $r_{core}=42\arcsec$.}
462: \label{fig_sb}\end{\myfigure}
463: \smallskip
464: %\noindent
465:
466:
467: The profile is smooth, and well constrained to the limit of our
468: observations. There is a significant peak in the profile at $\lesssim
469: 20\arcsec$. This peak is wider than a Gaussian of 5\arcsec{} FWHM
470: (overlaid as a dotted line), and thus unambiguously resolved (for
471: reference we have also overlaid a Gaussian of 1.5\arcsec{} FWHM [{\it
472: dashed line}] representing an upper limit to the ACIS PSF). We also show
473: a standard $\beta$-model
474: fit to the data omitting the inner 20\arcsec{}
475: to obtain an acceptable fit. Because we see no spectral evidence for
476: either a power-law source, or a cooling flow, we suggest that the peak
477: may represent a cusp in the gravitational potential due to the mass of
478: the cD galaxy.
479: %This high surface brightness feature will be discussed
480: %further elsewhere.
481: %Although the best-fit power law model had a very low normalization,
482: %it is possible for the spectrum to accommodate (in a poorer fit) a source with
483: %X-ray luminosity up to $\sim10^{42}$\lxh{}, which is consistent
484: %with other FR1 radio sources observed in X-rays \citep{can99}.
485:
486: \section{Discussion} We have analyzed high spatial resolution \chandra{}
487: spectra of the A2029 cluster of galaxies. We observe both strong temperature
488: and abundance gradients within the central arcmin, with $T_X$ dropping to
489: $\sim3$~keV, and $Z_{Fe}$ rising to $\approx2$~\zsun. We find no evidence for
490: excess absorption in the core, though there are some remaining residuals in the
491: low energy ($\le0.7$~keV) region of the spectrum which are not yet understood.
492: Additional model components (including a cooling flow) are not required, and do
493: not improve the fits. Thus, A2029 bears no evidence of multi-phase gas,
494: refuting earlier claims of $\mdot \sim 400-600 \msunyr$ and thus is similar to,
495: although more extreme than, many other recent \chandra{} and \xmm{}
496: observations \citep[see][for a review]{mcn02}. From the image, we find no
497: evidence of a strong shock or major cluster merger, and thus a lack of support
498: for the \citet{lok95} model for producing WAT radio sources in cluster merger
499: events. There is a small peak in the inner 10\arcsec{} of the surface brightness
500: profile, indicating a second component in excess of a standard $\beta$-model.
501:
502: The super-solar Fe abundance measurements in the core of A2029 are
503: suggestive of enrichment by metals created by Fe-rich SN Ia and ejected
504: from the cD and other galaxies in the core. Since the gaseous system is
505: relaxed and fairly undisturbed (no recent mergers), this gradient could
506: have been maintained to the present epoch from an earlier time. This
507: also agrees with the analysis of \citet{deg01} (although we now resolve
508: the enrichment at $<2\arcmin$), who found that the abundance gradient
509: matched the stellar light profile of the galaxies in the cluster core.
510:
511: \acknowledgments
512:
513: This work was supported by \chandra{} grant G00-1021X. ADL thanks Beth
514: Lewis for her ongoing support.
515:
516: \input{ms.bbl}
517: %\bibliographystyle{apj}
518: %\bibliography{adlrefs}
519:
520: \end{document}
521:
522: