1: % aa.dem
2: % AA vers. 5.01, LaTeX class for Astronomy & Astrophysics
3: % demonstration file
4: % (c) Springer-Verlag HD
5: % revised by EDP Sciences
6: %-----------------------------------------------------------------------
7: %
8: %\documentclass[referee]{aa} % for a referee version
9: %
10: \documentclass{aa}
11: %\documentclass{aa_mio}
12: \usepackage{graphicx}
13: %
14: \begin{document}
15: %
16: \title{On the metal abundance in the core of M87}
17:
18: % \subtitle{I. Overviewing the $\kappa$-mechanism}
19:
20: \author{Silvano Molendi
21: \inst{1}
22: \and
23: Fabio Gastaldello
24: \inst{1,2}
25: %\fnmsep\thanks{Just to show the usage
26: % of the elements in the author field}
27: }
28:
29: \offprints{S.Molendi}
30:
31: \institute{
32: Istituto di Fisica Cosmica, CNR, via Bassini 15,
33: I-20133 Milano, Italy\\
34: % \email{silvano@ifctr.mi.cnr.it}
35: \and
36: Universit\`a di Milano Bicocca, Dip. di Fisica, P.za della Scienza 3
37: I-20133 Milano\\
38: % \email{gasta@ifctr.mi.cnr.it}
39: }
40:
41: % \date{Received September 15, 1996; accepted March 16, 1997}
42:
43:
44: \abstract{
45: We revisit the XMM-Newton observation of M87
46: to make a new and more detailed measurement of the
47: metal abundance profile.
48: After having verified that spectral fits
49: with single temperature models show a dramatic abundance
50: drop within 1 arcmin of the cluster core,
51: we show that more appropriate models, which include
52: a multi-temperature component to account for the strong
53: temperature gradient observed in M87, and a power-law
54: component to account for the emission of the nucleus
55: and knot A, give a substantially flat abundance profile.
56: The drastic abundance decrement found by fitting a single
57: temperature component to the data is an artifact known
58: as ``$Fe$-bias'' (Buote 2000a,b) following from the application
59: of an oversimplified spectral model to the data.
60: \keywords{X-rays: galaxies --- Galaxies: clusters--- Galaxies:
61: individual: M87
62: }
63: }
64:
65: \maketitle
66: %
67: %________________________________________________________________
68:
69: \section{Introduction}
70:
71: Our proximity to the Virgo cluster allows us to study
72: its core on physical scales unachievable in other similar systems.
73: M87, the giant elliptical galaxy at the
74: center of Virgo, is not surprisingly amongst
75: the most studied extra-galactic objects in the sky.
76: At X-ray wavelengths it has been observed by all
77: major observatories.
78: Deprojection of the X-ray images obtained with the
79: IPC instrument on Einstein (Fabian et al. 1984) showed that the gas
80: in M87 has cooling times shorter than the Hubble
81: time. Further evidence of cooling gas has come from
82: the spectrometers on-board Einstein,
83: which showed that the gas in M87 was not single-temperature
84: (e.g. Canizares et al. 1982). Subsequent observations
85: with ROSAT (Nulsen \& B\"ohringer 1995), ASCA (Matsumoto
86: et al. 1996) and BeppoSAX (Guainazzi \& Molendi 2000), showed
87: evidence of a temperature decrement with decreasing radius.
88: The ASCA and BeppoSAX data also showed evidence of a metal
89: abundance increase towards the center of M87 on scales
90: of a few arcmin, which correspond to a few tens of
91: kpc.
92:
93: A first set of results from the XMM-Newton observation
94: of M87 have been recently published in B\"ohringer et al. (2001)
95: and Belsole et al. (2001). One of the most striking
96: results presented is a
97: very marked drop in the abundance profiles of $Si$ and $Fe$
98: within 1 arcmin of the center of M87. B\"ohringer et al. (2001)
99: suggest that resonant scattering may be responsible for
100: this drop. Mathews et al. (2001),
101: using detailed radiation transfer calculations, remark that
102: the resonant scattering effect is insufficient to explain
103: the observed profile and that some sort of continuous opacity
104: is required.
105:
106: In a recent paper (Molendi \& Pizzolato 2001) we have shown
107: that the multi-phase appearance of the EPIC spectra of M87
108: within 2 arcmin from the core is not ascribable to
109: the presence of a truly multi-phase gas but is rather
110: the consequence of having, within the same line of sight,
111: emission from gas at different radii with different temperatures.
112: In this Letter we revisit the XMM-Newton observation of
113: M87; our goal is to remeasure the metal abundance profile
114: by performing a more advanced spectral analysis of the EPIC
115: data with models which take into account the projection effects
116: leading to the multi-phase appearance of the gas in M87.
117:
118: The remainder of this Letter is organized as follows.
119: In section 2 we describe the data preparation. In section
120: 3 we discuss the spectral models. In section 4
121: we present and discuss the results of our analysis.
122:
123:
124: %__________________________________________________________________
125:
126: \section{Observations and Data Preparation}
127:
128: We use XMM-Newton EPIC data from the PV observation
129: of M87/Virgo.
130: Details on the observation, as well as results from a first
131: analysis of this object, have already been
132: published in B\"ohringer et al. (2001)
133: and Belsole et al. (2001).
134:
135: We have obtained calibrated event files for the MOS1, MOS2
136: and PN cameras with SASv5.0.
137: %In this Letter we do not make use of PN data.
138: %The current SAS does not incorporate a correction for the
139: %Charge Transfer Inefficiency (CTI) of the PN detector,
140: %consequently analysis of PN spectra accumulated over large
141: %regions of the detector cannot be performed in an optimal
142: %fashion. Moreover the better spatial and spectral resolution
143: %of the MOS1 and MOS2 cameras with respect to the PN camera
144: %make them more suitable for the analysis we wish to perform.
145: Data were manually screened to remove any remaining bright
146: pixels or hot column; for the PN camera we also exclude
147: regions contaminated by out of time events.
148: Periods in which the background is
149: increased by soft-proton flares have been excluded using an
150: intensity filter; we rejected all events accumulated when
151: the count rate exceeds 15 cts/100s in the $[10-12]$
152: keV band for the MOS cameras and 25 for the PN.
153: We have accumulated spectra in 4 concentric annular regions
154: centered on the emission peak. We extend our analysis
155: out to 3 arcmin from the emission peak.
156: The bounding radii for the regions are:
157: 0$^{\prime\prime}$ and 30$^{\prime\prime}$ for region 1;
158: %15$^{\prime\prime}$ and 30$^{\prime\prime}$ for region 2;
159: 30$^{\prime\prime}$ and 1$^{\prime}$ for region 2;
160: 1$^{\prime}$ and 2$^{\prime}$ for region 3 and
161: 2$^{\prime}$ and 3$^{\prime}$ for region 4.
162: %We refrain from accumulating spectra in regions smaller
163: %than our region I because on such scales the PSF plays an important
164: %and currently unknow role.
165: We have removed point sources, and the substructures which
166: are clearly visible from the X-ray image (e.g. Belsole et al.
167: 2001) except in the innermost region. Here we have kept
168: the nucleus and knot A, the reason being that on angular scales
169: this small it is not possible to exclude completely
170: their emission. As discussed in section
171: 3 we prefer to fit the spectrum of this region
172: with a model which includes a power-law component
173: to fit the two point-like sources; note that we include only one power-law
174: component because the two sources have similar spectra
175: (B\"ohringer et al. 2001).
176:
177: Spectra have been accumulated for MOS1 and MOS2 independently.
178: The Lockman hole observations have been used for the background.
179: %Periods in which the background is
180: %increased by soft-proton flares have been excluded using an
181: %intensity filter; we rejected all events accumulated when
182: %the count rate exceeds 15 cts/100s in the $[10-12]$
183: %keV band.
184: Background spectra have been accumulated
185: from the same detector regions as the source spectra.
186: The line of sight galactic absorbing column depth towards
187: the background field is somewhat smaller,
188: $N_{\rm H} \sim 6\times 10^{19}$ cm$^{-2}$, than that found on the line of sight
189: of M87. This has a negligible impact on the soft part of our M87 spectra
190: because the $N_{\rm H} $ variation will lead to a maximum difference of
191: 10\% in the background intensity, which is in turn never more than 2\%
192: of the source intensity in the 0.5-1.0 keV energy range.
193:
194: The vignetting correction has been applied to the spectra
195: rather than to the effective areas, similarly
196: to what has been done by other authors who
197: have analyzed EPIC data (Arnaud et al. 2001).
198: Spectral fits were performed in the 0.5-4.0 keV band.
199: Data below 0.5 keV were excluded to avoid residual calibration
200: problems in the MOS response matrices at soft energies.
201: Data above 4 keV were excluded because above this energy the
202: spectra show a substantial contamination from hotter gas emitted
203: further out in the cluster, on the same line of sight.
204:
205: MOS and PN spectra are analyzed separately. As discussed in
206: Molendi (2001), there are residual cross-calibration uncertainties
207: between the two instruments which, as we shall see in the next
208: sections, sometimes lead to different measurements
209: of spectral parameters.
210:
211: %__________________________________________________________________
212:
213: \section{Spectral Modeling}
214:
215:
216:
217: All spectral fitting has been performed using version
218: 11.0.1 of the XSPEC package. All models discussed below
219: include a multiplicative component to account for the
220: galactic absorption on the line of sight of M87.
221: The column density is always fixed at a value of
222: 1.8$\times10^{20}$ cm$^{-2}$, which is derived from 21cm measurements
223: (Lieu et al. 1996).
224: We have compared our data with three different
225: spectral models.
226:
227: Model I is a single temperature model (vmekal model in XSPEC),
228: which allows to fit separately individual metals abundances.
229: This model has 13 free parameters: the temperature $T$, the
230: normalization and the abundance of $O$, $Ne$, $Na$, $Mg$,
231: $Al$, $Si$, $S$, $Ar$, $Ca$, $Fe$ and $Ni$, which are all expressed
232: in solar units.
233:
234: Model II features two single temperature components
235: (vmekal + vmekal in XSPEC)
236: This model has two free parameters more than model I:
237: the temperature and normalization of the second
238: thermal component. The metal abundance of each element of
239: the second thermal component is linked
240: to the same parameter of the first thermal component.
241: This model has
242: %been used by B\"ohringer et al. (2001)
243: %to describe the spectra from the innermost regions of M87, it has also
244: been used in the past (e.g.
245: Makishima et al. 2001 and refs. therein) as an alternative
246: to cooling-flow models.
247:
248: Model III includes a single temperature
249: plus a multi-phase component
250: (vmekal + vmcflow in XSPEC).
251: This model has two free parameters more than model I:
252: the minimum temperature, $T_{\rm min}$, and the normalization
253: of the multi-phase component.
254: The other parameters of the multi-phase
255: component are not free: the maximum temperature, $T_{\rm max}$, and the
256: metal abundance of each element are linked respectively to the
257: temperature, $T$, and the metal abundances of the same
258: elements of the single-phase component.
259: In a recent paper (Molendi \& Pizzolato 2001) we have shown
260: that such a model fits adequately the EPIC spectra of three clusters
261: which feature a temperature decrement in their core.
262: %Indeed, because of projection effects, when we look towards
263: %the core of a cluster we observe photons coming
264: %from all cluster regions, from the core to the outer regions.
265: %If, as in our case, there is a strong temperature gradient
266: %in the core, spectra accumulated from
267: %the innermost regions will be better fitted by a model
268: %characterized by a temperature distribution than by a
269: %single temperature.
270: The vmcflow component we use has originally been written
271: to fit spectra from multi-phase gas, i.e. gas that at
272: a given physical radius is characterized by a distribution of
273: temperatures rather than by a single temperature.
274: Here we are using it to describe a different physical
275: scenario where, at
276: a given physical radius the gas is all at one temperature,
277: and the multi-phase appearance of the spectrum comes
278: from having emission from many different physical radii
279: all within the same line of sight.
280:
281: For the spectrum accumulated in the innermost region
282: models II and III also include a power-law component
283: to model the emission of the nucleus and of knot A.
284:
285: % One column figure
286: %-----------------------------------------------------------
287: \begin{figure}
288: \centering
289: \includegraphics[angle=-90,width=8.0cm]{fig1.ps}
290: \caption{
291: $\Delta \chi^2$ between models I and II (filled circles),
292: and I and III (open circles), as a function of radius.
293: The solid (dashed) horizontal line indicates
294: the $\Delta \chi^2$ value for which the statistical improvement of
295: the model II (III) fit with respect to model I is significant at the
296: 99\% level according to the F-test.
297: Top panel is for MOS and bottom for PN.
298: }
299: \end{figure}
300: %
301: %______________________________________________________________
302:
303:
304: % One column figure
305: %-----------------------------------------------------------
306: \begin{figure}
307: \centering
308: \includegraphics[angle=-90,width=8.0cm]{fig2.ps}
309: \caption{
310: MOS abundance profiles for O, Si and Fe.
311: Uncertainties are at the
312: 68\% level for one interesting parameter ($\Delta \chi^2 = 1$).
313: Full symbols indicate the measurements with model I, empty squares
314: and empty triangles those for models II and III respectively.
315: }
316: \end{figure}
317: %
318: %______________________________________________________________
319:
320:
321: % One column figure
322: %-----------------------------------------------------------
323: \begin{figure}
324: \centering
325: \includegraphics[angle=-90,width=8.0cm]{fig3.ps}
326: \caption{
327: PN profiles for O, Si and Fe.
328: Symbols as in Fig. 2.
329: }
330: \end{figure}
331: %
332: %______________________________________________________________
333:
334:
335:
336:
337: \section{Results}
338:
339: In Figure 1 we plot the
340: $\Delta \chi^2$ between models I and II (filled circles)
341: and models I and III (open circle) as a function of radius.
342: The solid (dashed) horizontal line indicates
343: the $\Delta \chi^2$ value for which the statistical improvement of
344: the model II (III) fit with respect to model I is significant at the
345: 99\% level according to the F-test.
346: %Whenever a datapoint lies above this line
347: %the improvement is significant at more than the 99\% level.
348: The results plotted in the top panel refer to the MOS
349: spectra while those shown in the bottom panel are for
350: the PN spectra.
351: From Figure 1 we see that, for all bins and for both MOS and PN,
352: the improvement of models II and III with respect to model I is significant
353: at more than the 99\% level, implying that the
354: spectra are characterized by more than just one temperature.
355:
356: In the case of the innermost bin we have verified that including
357: only the second thermal component for models II and III, provides
358: a significantly better fit with respect to model I and that the
359: further inclusion of the power-law component is also statistically
360: significant. Thus for this bin we have evidence in favor of a second thermal
361: component and of a power-law component.
362:
363:
364: %Moreover from the bottom panel we see that the spanned temperature range,
365: %for the three innermost bins, becomes broader and that
366: %the contribution from cooler gas is progressively more
367: %important at smaller radii.
368:
369: In Figures 2 and 3 we report respectively the MOS and PN radial abundance
370: profiles of $O$ (top panel), $Si$ (middle panel) and $Fe$ (bottom panel),
371: while in Figures 4 and 5 we show those for
372: $Mg$ (top panel), $S$ (middle panel) and $Ar$ (bottom panel).
373: Full symbols indicate the
374: measurements with model I, empty squares and triangles those
375: with models II and III respectively.
376:
377: Figures 2, 3, 4 and 5 show that, for the two innermost
378: bins, all the metal abundances measured with model I, either with MOS or
379: PN, are systematically smaller than those measured with models II
380: or III.
381: The difference between the measurements
382: becomes larger for smaller radii as the significance of the improvement
383: of models II and III with respect to model I (see Figure 1) increases.
384:
385: We note that, while for $Si$, $Fe$, $S$ and $Ar$ the measurements obtained
386: with PN and MOS are consistent, for $O$ and $Mg$ abundances derived
387: with the MOS are systematically higher than those derived with the PN.
388: The reason lies in the current cross-calibration uncertainties
389: between the spectral response of the two instruments (Molendi 2001).
390: However, the shape of the abundance profiles
391: derived with models II and III is the same for MOS and PN: neither
392: shows evidence of a dramatic metal abundance drop such as is observed
393: with model I.
394:
395:
396:
397: % One column figure
398: %-----------------------------------------------------------
399: \begin{figure}
400: \centering
401: \includegraphics[angle=-90,width=8.0cm]{fig4.ps}
402: \caption{
403: MOS profiles for Mg, S and Ar. Symbols as in Fig. 2.
404: }
405: \end{figure}
406: %
407: %______________________________________________________________
408:
409:
410: % One column figure
411: %-----------------------------------------------------------
412: \begin{figure}
413: \centering
414: \includegraphics[angle=-90,width=8.0cm]{fig5.ps}
415: \caption{
416: PN profiles for Mg, S and Ar. Symbols as in Fig. 2.
417: }
418: \end{figure}
419: %
420: %______________________________________________________________
421:
422: Clearly single temperature models substantially
423: underestimate the metal abundance in multi-temperature spectra.
424: This is not a new result, indeed something very similar
425: has been observed in the ASCA spectra of galaxies and galaxy groups
426: (e.g. Buote 2000a,b) and goes under the name
427: of ``$Fe$-bias''.
428: The interested reader can find a detailed discussion of the
429: ``$Fe$-bia'' in Appendix A of Buote (2000a).
430: The residuals of a single temperature model applied to
431: a multi temperature spectrum around the Fe-L line complex
432: (see fig. 5 of Buote 2000b) are also quite
433: illustrative.
434:
435: Figure 1 tells us that models II and III behave very similarly
436: in terms of their improvement with respect to model I.
437: %Moreover we see that the improvement of models II and
438: %III with respect to model I becomes larger as we move inwards.
439: In the outermost radial bin, where we find the smallest
440: evidence for emission from more than a single temperature,
441: the abundances measured with models II and III
442: do not differ substantially from those derived from model I.
443: %{\it difference between MOS and PN}.
444: As we move inwards the evidence for
445: multi-temperature gas becomes
446: stronger, the $Fe$-bias becomes more important, and the abundance
447: measurements with models II and III differ more and more from those
448: obtained with model I.
449: In the innermost bin, the power-law emission from the nuclear sources
450: contributes to raising the continuum with respect to the line
451: emission thereby driving the abundances obtained with model I
452: further down.
453:
454: Since the metal abundances in the innermost region are sensitive
455: to the normalization of the power-law component
456: we have performed the following test to verify
457: that this parameter is sufficiently well determined by the
458: fitting procedure.
459: We have accumulated a spectrum for the innermost
460: bin excising the emission from the nucleus
461: and knot A, using an exclusion radius of 7$^{\prime\prime}$ around
462: the sources. We recall that for such a small exclusion radius
463: the accumulated spectrum will still be contaminated by the
464: emission of the point-sources. We have then applied models II and
465: III to this spectrum. We find that the best fitting values for
466: the metallicities are in agreement with those
467: derived by fitting the spectrum where the two sources
468: have not been eliminated.
469:
470: Close inspection of the $O$, $Si$, $Fe$ and $S$ profiles reveals a
471: small decrement in the innermost bin, the difference
472: between the measurement in this region and the one in region 2 , which
473: has the largest abundance,
474: is significant at less than 2$\sigma$ for all elements in both MOS
475: and PN with the exception of $Si$ for the PN, which is significant
476: at about 2.7$\sigma$.
477:
478: In summary, the metal abundance profiles derived by fitting single
479: temperature models are very different from the ones that are obtained
480: by fitting the more appropriate spectral models II or III. The former
481: all show a dramatic abundance decrease towards the center at radii
482: smaller than one arcmin. The latter show a flattening of the
483: abundance profile within 1 arcmin and a hint of a decrease
484: in the innermost bin.
485: In the light of these results the evidence for absorption
486: in the core of M87 is greatly reduced.
487: Observations at higher angular resolution
488: with the Chandra observatory will likely settle the issue
489: definitively.
490:
491: \begin{acknowledgements}
492: We thank the many people who have contributed to building,
493: calibrating and operating the EPIC instrument on-board XMM-Newton.
494: S. Ghizzardi is thanked for useful discussions.
495: \end{acknowledgements}
496:
497: \begin{thebibliography}{}
498:
499: % \bibitem[1966]{baker} Baker, N. 1966,
500: % in Stellar Evolution,
501: % ed.\ R. F. Stein,\& A. G. W. Cameron
502: % (Plenum, New York) 333
503:
504: \bibitem{} Arnaud, M., Neumann, D. M., Aghanim, N., Gastaud, R.,
505: Majerowicz, S., \& Hughes, J. P. 2001, A\&A, 365, L80
506:
507: \bibitem{} Belsole, E., Sauvageot, J. L., Böhringer, H., Worrall, D. M.;
508: Matsushita, K., Mushotzky, R. F., Sakelliou, I., Molendi, S.,
509: Ehle, M., Kennea, J., \& Stewart, G. 2001, A\&A, 365, L188
510:
511: \bibitem{} B\"ohringer, H., Belsole, E., Kennea, J., Matsushita, K., Molendi, S.,
512: Worrall, D. M., Mushotzky, R. F., Ehle, M., Guainazzi, M.,
513: Sakelliou, I., Stewart, G., Vestrand, W. T., Dos Santos, S.
514: 2001, A\&A, 365, L181
515:
516:
517:
518:
519: \bibitem{} Buote, D.A. 2000a, MNRAS, 311, 176
520:
521: \bibitem{} Buote, D.A. 2000b, ApJ, 539, 172
522:
523: %\bibitem{} Dickey, J.M., \& Lockman, F.J. 1990, ARA\&A, 28, 215
524:
525: \bibitem{} Fabian, A.C., Nulsen, P. E. J., Canizares, C. R. 1984 Nature, 310, 733
526:
527: \bibitem{} Guainazzi, M., \& Molendi S. A\&A, 351, L19
528:
529:
530: \bibitem{} Lieu, R., Mittaz, J. P. D., Bowyer, S., Lockman, F.J.,
531: Hwang, C.-Y.; Schmitt, J.H.M.M 1996, ApJ, 458, L5
532:
533:
534: \bibitem{} Makishima, H., Ezawa, H., Fukazawa, Y. 2001, PASP, in press
535:
536: \bibitem{} Mathews, W.G., Buote, \& D.A, Brighenti, F. 2001, ApJL in press
537: (astro-ph/0101545)
538:
539: \bibitem{} Matsumoto, H., Koyama, K., Awaki, H., Tomida, H., Tsuru, T.,
540: Mushotzky, R. \& Hatsukade, I. 1996, PASJ, 48, 201
541:
542: \bibitem{} Molendi, S. 2001, {\it Report on MOS PN cross-calibration to
543: be presented at the Leicester EPIC calibration meeting to be held
544: in June 2001}
545:
546:
547: \bibitem{} Nulsen P. E. J., \& B\"ohringer, H. 1995, MNRAS, 274, 1093
548:
549:
550: \end{thebibliography}
551:
552: \end{document}
553:
554: