1: %\documentclass[preprint,12pt]{aastex}
2: \documentclass[preprint2]{aastex}
3:
4: % Here's some slug-line data. The receipt and acceptance dates will be
5: % filled in by the editorial staff with the appropriate dates. Rules will
6: % appear on the title page of the manuscript until these are uncommented
7: % out by the editorial staff.
8:
9: %\received{4 February 1988}
10: %\accepted{23 September 1988}
11: %\journalid{337}{15 January 1988}
12: %\articleid{11}{14}
13:
14: % Authors may supply running head information, if they wish to do so, although
15: % this may be modified by the editorial offices. The left head contains a
16: % list of authors, usually three allowed---otherwise use et al. The right
17: % head is a modified title of up to roughly 44 characters. Running heads
18: % are not printed.
19:
20: \shorttitle{EUV Emission from Abell\ 1795 and 2199}
21: \shortauthors{Bergh\"ofer \& Bowyer}
22:
23: \begin{document}
24:
25: \title{An Analysis of BeppoSAX LECS Observations of EUV Emission in Clusters
26: of Galaxies}
27:
28: \author{Thomas W. Bergh\"ofer,\altaffilmark{1} Stuart
29: Bowyer,\altaffilmark{2}}
30:
31: \affil{Hamburger Sternwarte, Universit\"at Hamburg, Gojenbergsweg
32: 112, D-21029 Hamburg, Germany \and
33: Space Sciences Laboratory, University of California, Berkeley,
34: CA 94720-7450, USA}
35:
36: \begin{abstract}
37: Kaastra et al. (1999) have used the BeppoSAX LECS instrument to search for
38: excess EUV emission in Abell\ 2199. They claim that the results obtained
39: confirm an independent report of an excess EUV emission in this cluster
40: (Lieu et al. 1999). Using an inflight derived procedure that is better suited
41: to the analysis of extended sources and which avoids uncertainties related to
42: ground-based calibrations for the overall detector sensitivity profile, we
43: find no excess EUV emission in
44: Abell\ 2199. We also used these procedures to search for an EUV excess
45: in Abell\ 1795, but no excess was found.
46: \end{abstract}
47:
48: Subject headings: galaxies: clusters: individual (Abell 1795, Abell 2199) ---
49: techniques: image processing --- ultraviolet: general --- X-rays: general
50:
51: \section{Introduction}
52: \label{intro}
53:
54: The discovery of Extreme Ultraviolet (EUV) emission in clusters of galaxies
55: with the Extreme Ultraviolet Explorer (EUVE) has provoked considerable
56: controversy. While there is no doubt about the detection of the EUV
57: emission in excess of that produced by the well-studied X-ray emitting cluster
58: gas in Virgo (Bergh\"ofer, Bowyer, \& Korpela 2000a) and in Coma
59: (Bowyer, Bergh\"ofer, \& Korpela 1999), many clusters do not exhibit an EUV
60: excess at least at current sensitivity levels. In a series of publications
61: (Bowyer, Bergh\"ofer, \& Korpela, 1999; Bergh\"ofer, Bowyer, \& Korpela
62: 2000a,b; Bowyer, Korpela, \& Bergh\"ofer 2001) we have demonstrated that
63: the EUVE results are strongly affected by the variation of the telescope
64: sensitivity over the field of view and upon the details of the subtraction of
65: the EUV emission from the X-ray contribution.
66:
67: Kaastra et al. (1999) have analyzed BeppoSAX data obtained with the Low-Energy
68: Concentrator Spectrometer (LECS) to search for EUV emission in the Abell\ 2199
69: cluster of galaxies. Unfortunately, the telescope sensitivity profile used in
70: this work is likely to be incorrect. Quoting from their work, Kaastra et
71: al. (1999) state, "The vignetting correction for the LECS was derived from the
72: SAXDAS/LEMAT ray-trace code, assuming azimuthal symmetry around the appropriate
73: center. The correction for the support grid was also derived from that
74: package." We note that ground based simulations of the large scale
75: sensitivity of EUV and X-ray instrumentation are extraordinarily difficult to
76: construct and are notorious for being an inappropriate representation of the
77: true sensitivity functions.
78: In fact, a comparison of ground based calibration measurements with
79: ray-tracing simulations for BeppoSAX clearly demonstrated a discrepancy by a
80: factor of $\sim$1.5 at low energies (Parmar et al. 1997).
81:
82: In order to test the claim of an EUV excess in Abell\ 2199 by Kaastra et al.
83: (1999) we have reanalysed BeppoSAX LECS observations of this cluster of
84: galaxies. We also searched for excess EUV emission in Abell\ 1795 using
85: archival
86: BeppoSAX data on this cluster. In Section\ \ref{data} we describe the data and
87: reduction applied to obtain the clusters' EUV emission. Section\ \ref{results}
88: provides the results of our investigations, and a summary is presented in
89: Section\ \ref{discussion}.
90:
91: \section{BeppoSAX LECS Data and Data Reduction}
92: \label{data}
93:
94: \begin{figure}
95: \plotone{f1.eps}
96: \caption{Raw azimuthally averaged radial emission profiles of Abell\ 2199 in the
97: 0.1--0.3\,keV (solid line) and 0.5--2.2\,keV (dashed line) energy bands of the
98: BeppoSAX LECS instrument. The dotted and dashed--dotted lines are the raw
99: background profiles obtained from the standard background files,
100: respectively, in the 0.1--0.3\,keV and 0.5--2.2\,keV band.}
101: \end{figure}
102:
103: \begin{figure}
104: \plotone{f2.eps}
105: \caption{Radial emission profiles of Abell\ 1795. The labeling of the lines
106: is the same as in Figure\ 1.}
107: \end{figure}
108:
109: \begin{table}[tp]
110: \caption{\label{obslog} BeppoSAX observation log.
111: }
112: \begin{flushleft}
113: \begin{tabular}{lcc}
114: \hline
115: Target & Exposure Date & Raw Exposure Time (s) \\
116: \hline\hline
117: Abell\ 1795 & 29 Dec 1996 & 3528 \\
118: & 12 Aug 1997 & 7832 \\
119: & 28 Jan 2000 & 38166 \\
120: Abell\ 2199 & 23 Apr 1997 & 26860 \\
121:
122: \hline
123: \end{tabular}
124: \end{flushleft}
125: \end{table}
126:
127: For our investigations we have employed BeppoSAX LECS archival data of the two
128: clusters Abell\ 2199 and Abell\ 1795. Table\ \ref{obslog} provides an observation
129: log of the analysed data sets.
130: The LECS instrument aboard of BeppoSAX is a gas scintillation proportional
131: counter especially designed to perform low-energy X-ray spectroscopy. A
132: detailed description of this detector can be found in Parmar et al.
133: (1997). Here we summarize the performance of this detector relevant for
134: the study of diffuse sources in the low energy regime. The LECS has at least
135: some sensitivity in the energy range 0.1--10\,keV. Its circular field of view
136: with a radius of about 18\arcmin\ is relatively small for cluster studies.
137: At 3 keV the on-axis angular resolution is roughly 3\arcmin\ (FWHM), which
138: significantly drops to 9.7\arcmin\ (FWHM) at 0.28 keV. At these low energies
139: the resolution is dominated by the detector and is almost constant across the
140: field of view. At low energies the LECS effective area peaks near 0.2 keV.
141: For our investigations we have selected all events in the detector channels
142: 10--30 ($\sim$0.1--0.3 keV) and produced images of the clusters. The raw
143: azimuthally averaged surface profiles of Abell\ 2199 and Abell\ 1795 and their
144: statistical uncertainties are shown in Figures\ 1 and 2 (solid lines); in
145: the case of Abell\ 1795 the plot is the combined profile of the three
146: distinct LECS observations of this cluster.
147:
148: In order to determine the contribution of the X-ray emitting cluster gas
149: in the 0.1--0.3 keV band of the BeppoSAX LECS we also computed a radial
150: surface brightness profile at detector channels 50--200 ($\sim$0.5--2.2 keV).
151: Events at
152: higher energies have been ignored due to the substantial drop in effective
153: area at energies E $>$ 2.5 keV and the increase in telescope vignetting at
154: higher energies, which might affect the clusters' X-ray profiles.
155: In order to compensate for the substantially lower detector resolution at low
156: energies, we convolved the source images in the 0.5--2.2 keV band with a
157: 9.7\arcmin\ wide Gaussian. The derived
158: radial emission profiles of the two clusters in the 0.5-2.2 keV band and its
159: statistical errors are shown as dashed lines in Figures\ 1 and 2.
160:
161: For an appropriate background and sensitivity correction we obtained
162: background files from the BeppoSAX data archive. In the case of the
163: LECS detector the standard background data consist of an assemblage of blank
164: field observations with a total integration of 568562.8 seconds.
165: The dotted and dashed-dotted lines in Figures\ 1 and 2 provide the radial
166: profiles obtained from the standard background data at the detector positions
167: of Abell\ 2199 and Abell\ 1795, respectively, in the 0.1-0.3 keV and 0.5--2.2
168: keV band.
169:
170: Using procedures similar to the background subtraction method developed for
171: the EUVE Deep Survey instrument (cf. Bowyer, Bergh\"ofer, \& Korpela 1999), we
172: adopted a two-parameter profile for the background. One profile is a flat
173: background reflecting the time-dependent charged particle background. This has
174: been determined from highly obscured regions at the outer most parts of the
175: field of view. The count rates in units of $10^{-6}$cts\,s$^{-1}$\,arcmin$^2$
176: are 3.5, 5.1, and 6.3, respectively, for the background field and the science
177: exposures for Abell\ 2199 and Abell\ 1795.
178: The other profile reflecting all sensitivity changes over the
179: detector field and telescope vignetting has been constructed from the standard
180: background field. We subtracted the appropriate flat background and convolved
181: the data with a 9.7\arcmin\ wide Gaussian corresponding to the instrument
182: resolution; using the telescope point spread function from ground
183: based calibration measurements instead of a Gaussian has no significant effect.
184: Scaling factors of 4.4 and 3.9 have been applied to the derived sensitivity
185: profile to subtract this part of the background from the science exposures of
186: Abell\ 2199 and Abell\ 1795.
187: Note that this part of the background is dominated by scattered solar X-ray
188: radiation, which varies in time. Furthermore, the BeppoSAX observations used
189: to construct the standard background data were taken at the beginning of the
190: mission, whereas the cluster observations have been carried out closer to
191: solar maximum. This explains the relatively large scaling factors.
192: The given scaling factors and constant
193: background rates result in the best representation of the background in the
194: profiles. A large range of combinations of scaling factors and constant
195: background rates were explored which were consistent with acceptable fits to
196: the data, and even allowed for a possible contribution of an X-ray signal at
197: larger radii. None of these variations changed our overall findings.
198:
199: \section{Results}
200: \label{results}
201:
202: \begin{figure}
203: \plotone{f3.eps}
204: \caption{Azimuthally averaged radial emission profiles of Abell\ 2199 in the
205: 0.1--0.3\,keV (solid line) and 0.5--2.2\,keV (dashed line) band. The
206: background has been subtracted from both profiles and the fluxes in the
207: 0.5--2.2\,keV were transformed into the 0.1--0.3\,keV band to reflect the
208: contribution of the X-ray emitting cluster gas in this energy band. There
209: is no excess EUV emission in this cluster.}
210: \end{figure}
211: In Figure\ 3 we show the EUV emission in the Abell\ 2199 cluster of galaxies
212: obtained from BeppoSAX LECS data using the correct sensitivity profile
213: (solid line). The EUV emission produced by the diffuse X-ray emitting cluster
214: gas is shown by a dashed line. To transform the 0.5--2.2 keV band profile to
215: the 0.1--0.3 keV band we used XSPEC and the detector response matrix of the
216: LECS to simulate conversion factors for these two energy
217: bands of the BeppoSAX LECS. Adopting the set of parameters and models for the
218: X-ray emitting plasma and the absorption by the intervening interstellar
219: medium as described in Bowyer, Bergh\"ofer, \& Korpela (1999) we found
220: conversion factors of 10.6 and 9.9, respectively, for the 4.08 keV and 2.9 keV
221: gas components.
222:
223: Figure\ 3 demonstrates that the BeppoSAX LECS does not detect an EUV excess in
224: the cluster Abell\ 2199 when the data are analyzed correctly. We note that
225: there is a deficit of EUV emission in the central core region (R $<$ 4\arcmin)
226: of the cluster. However, with respect to the uncertainties in the cluster
227: profiles, this is a small effect and may be a result of imperfect detector
228: calibration. On the other hand, such deficits has been observed in other
229: clusters of galaxies and have been attributed in the past to the effects of
230: cooling flows; in light of current results from XMM-Newton and Chandra, this
231: effect must now be ascribed to some other cause (e.g., B\"ohringer et al. 2001,
232: Tamura et al. 2001).
233:
234: The BeppoSAX LECS results on Abell\ 1795 are shown in Figure\ 4. The solid line
235: in this figure shows the background subtracted EUV emission profile of this
236: cluster. The comparison with the expected contribution of the low energy tail
237: of the X-ray emitting gas (dashed line in this figure) demonstrates that this
238: cluster does not exhibit EUV excess emission at least at the BeppoSAX LECS
239: level of sensitivity. Using appropriate model parameters for the X-ray
240: emitting cluster gas and appropriate corrections for the interstellar
241: absorption in the direction of Abell\ 1795 (see Bowyer, Berg\"ofer, \& Korpela
242: 1999), we applied conversion factors of 12.5 and 10.5, respectively, for the
243: 6.7 keV and 2.9 keV gas components. Again, the cluster center shows a small
244: deficit in EUV radiation in the 2--4\arcmin\ radial bin when compared to
245: the expected cluster X-ray emission in the 0.1--0.3\,keV band.
246: \begin{figure}
247: \plotone{f4.eps}
248: \caption{Background subtracted radial emission profiles of Abell\ 1795 in the
249: 0.1--0.3\,keV (solid line). The dashed line provides the contribution of
250: the X-ray emitting gas as can be obtained from the LECS 0.5--2.2\,keV band.
251: There is no excess EUV emission in this cluster.}
252: \end{figure}
253:
254: \section{Discussion and Conclusion}
255: \label{discussion}
256:
257: A search for excess EUV emission has been carried out in a substantial number
258: of clusters observed with EUVE. Bowyer, Bergh\"ofer \& Korpela (1999) have
259: shown that all but two of the reported detections were the product of the use
260: of an incorrect detector sensitivity function. However, this conclusion has
261: been questioned because Kaastra et al. (1999) claimed to have found an excess
262: EUV emission in Abell 2199 using BeppoSAX LECS observations. These findings
263: appeared to support the (incorrect) finding of an excess in this cluster using
264: EUVE data. However, this result was based upon ground-based estimates of the
265: LECS detector sensitivity function. Using a procedure better suited to the
266: analysis of extended sources that avoids the known uncertainties in the
267: telescope sensitivity function, we show that there is no excess in Abell\ 2199.
268: We also searched for excess EUV emission in Abell\ 1795 using archival
269: BeppoSAX observations. No EUV excess was found.
270:
271: The results obtained here for Abell\ 2199 and Abell\ 1795 are fully consistent
272: with the results obtained on these clusters using EUVE data (Bowyer,
273: Bergh\"ofer, \& Korpela 1999). The only clusters of galaxies that have been
274: found to exhibit an excess EUV emission are the Virgo and Coma clusters
275: (Bowyer, Korpela, \& Bergh\"ofer 2001).
276: It is possible that these are the only clusters that exhibit this effect, but
277: it may be that both EUVE and BeppoSAX LECS are insufficiently sensitive to
278: detect an EUV excess in other clusters of galaxies.
279:
280: Since the only clusters with a confirmed EUV excess are Virgo and Coma, it is
281: useful to reconsider candidates for the underlying source of the EUV excess.
282: The original proposal was that this emission is thermal emission from a
283: ``warm'' (10$^6$ K) gas (Lieu et al. 1996a,b; Bowyer et al. 1996). Claims of
284: ``proof'' of this proposition have been advanced by Mittaz, Lieu, \& Lockman
285: (1998), Lieu, Bonamente, \& Mittaz (1999a), Lieu et al. (1999b), Lieu,
286: Bonamente, \& Mittaz (2000), and Bonamente, Lieu, \& Mittaz (2001a,b).
287: Lieu et al. (2000) misinterpreted small scale detector structures in the EUVE
288: data on Abell 2199 as cluster EUV emission absorbed by clumps of neutral
289: hydrogen in the cluster. A difficulty with this interpretation which is
290: independent of the data analysis problem is that the hydrogen required is
291: "... $\sim$43 times more massive than the hot ICM in this region ... (and) ...
292: implies 3 times more missing baryons than expected''(op. cite).
293:
294: The maintenance of a warm intracluster gas is quite difficult to understand
295: since gas at this temperature is at the peak of its cooling curve and would
296: cool in less than 0.5~Gyr, and on these grounds alone it was generally
297: believed that a thermal source was untenable. Observational evidence relevant
298: to this issue was obtained with the Hopkins Ultraviolet telescope (Dixon et
299: al. 1996), and FUSE (Dixon et al. 2001a,b). No Far UV line emission from gas
300: at 10$^6$ K was detected. More recently, observations of a large number of
301: clusters with XMM have been carried out. Kaastra et al. (2001) found no gas at
302: T $<$ 1 keV in Sersic 159-03, Peterson et al. (2001) found no gas at
303: T $<$ 2.7 keV in Abell 1835, and Tamura et al. (2001) found no gas at
304: T $<$ 4 keV in Abell 1795. All other clusters observed with XMM showed no
305: evidence of a cooler EUV emitting gas (Steve Kahn, private communication).
306: The sum of this evidence seems overwhelming: a thermal mechanism for the EUV
307: excess can be ruled out.
308:
309: Since the underlying source mechanism is not thermal, it must be the product
310: of some non-thermal process. Inverse Compton scattering of cosmic rays with
311: the 2.7 K background was suggested early-on as a possible source mechanism
312: (Hwang 1997, En{\ss}lin \& Biermann 1998). Sarazin \& Lieu (1998) suggested a
313: model in which a population of cosmic rays produced several Gyr ago would have
314: degraded over time and would now be unobservable as radio synchrotron emission even at very low frequencies. This population would produce an EUV flux by
315: inverse Compton scattering. Sarazin \& Lieu derived the ratio between the
316: azimuthally averaged total EUV emission and the azimuthally averaged soft
317: X-ray flux predicted by their model; this ratio increases with increasing
318: distance from the center of the cluster. Bergh\"ofer et al. (2000) derived
319: this ratio for the Virgo cluster as a test of the Sarazin \& Lieu model.
320: They found this ratio was flat with increasing distance from the center of the
321: cluster in contradiction to the prediction of the model.
322:
323: Bowyer et al. (in progress) derived this ratio for the Coma cluster
324: using data on the cluster that had been analyzed correctly. They found
325: this ratio was flat with increasing distance from the center again
326: contradicting the predictions of the Sarazin \& Lieu model.
327:
328: Despite the failure of the Sarazin \& Lieu model, the inverse Compton mechanism
329: remains as the only candidate for the source mechanism for the EUV excess.
330: However, a new difficulty for this hypothesis has recently appeared. Virtually
331: all models invoking the inverse Compton mechanism require the intracluster
332: magnetic field to be $<<$1$\mu$G. However, recent results show that cluster
333: magnetic fields are quite large. Clarke, Kronberg, \& B\"ohringer (2001)
334: studied 16 clusters with very high spatial resolution and have shown that all
335: of these clusters have B fields of 4 to 7$\mu$G. Unless this result is
336: somehow incorrect, the vast majority of models proposed for the production of
337: the EUV excess are incorrect. The only exceptions to the low field models
338: (En{\ss}lin, Lieu, \& Biermann 1999; Atoyan \& V\"olk 2000) are unlikely to be
339: appropriate (Ming, Hwang, \& Bowyer 2001).
340:
341: It is not clear whether this is a fundamental obstacle for the inverse
342: Compton scattering hypothesis or if it is simply a failure of existing
343: models. Irrespective of how widespread the occurrence of EUV excess in
344: clusters of galaxies may be, the underlying source mechanism for this emission
345: remains a mystery.
346:
347: \acknowledgments
348: This work was supported in part by a University of California grant. TWB was
349: supported in part by a Feodor-Lynen Fellowship of the
350: Alexander-von-Humboldt-Stiftung.
351:
352: \begin{thebibliography}{}
353: \bibitem{atvo00} Atoyan, A., \& V\"olk, H. 2000, \apj, 535, 45
354: \bibitem{bbk00a} Bergh\"ofer, T., Bowyer, S., \& Korpela, E. 2000a, \apj, 535,
355: 615
356: \bibitem{bbk00b} Bergh\"ofer, T., Bowyer, S., \& Korpela, E. 2000b, \apj, 545,
357: 695
358: \bibitem{boe01} B\"ohringer, H., Matsushita, K., Churazov, E., \& et al. 2001,
359: \aap, in press, astro-ph/011112
360: \bibitem{boli01a} Bonamente, M., Lieu, R., \& Mittaz, P. 2001a, \apj, 546, 805
361: \bibitem{boli01b} Bonamente, M., Lieu, R., \& Mittaz, P. 2001b, \apj, 547, L7
362: \bibitem{bola96} Bowyer, S., Lampton, M., \& Lieu, R. 1996, Science, 274, 1338
363: \bibitem{bbk99} Bowyer, S., Bergh\"ofer, T., \& Korpela, E. 1999, \apj, 526,
364: 592
365: \bibitem{bbk01} Bowyer, S., Korpela, E., \& Bergh\"ofer, T. 2001, \apj, 548,
366: 135
367: \bibitem{clkrb01} Clarke, T. E., Kronberg, P. P., \& B\"ohringer, H. 2001,
368: \apjl, 547, 111
369: \bibitem{dihu96} Dixon, W., Hurwitz, M., \& Ferguson, H. C. 1996, \apj, 469,
370: L77
371: \bibitem{disa01a} Dixon, W., Sallmen, S., Hurwitz, M., \& Lieu, R. 2001a,
372: \apj, 550, L25
373: \bibitem{disa01b} Dixon, W., Sallmen, S., Hurwitz, M. \& Lieu, R. 2001b, \apj,
374: 552, L69
375: \bibitem{ensbi98} En{\ss}lin, T. A., \& Biermann, P. L. 1998, \aap 330, 96
376: \bibitem{ensli99} En{\ss}lin, T. A., Lieu, R., \& Biermann, P. L. 1998, \aap
377: 344, 409
378: \bibitem{hwang97} Hwang, C.-Y. 1997, Science, 278, 1917
379: \bibitem{limi96a} Lieu, R., Mittaz, J., Bowyer, S., Lockman, F., Hwang, C-Y.,
380: \& Schmitt, J. H. M. M. 1996a, \apj, 458, L5
381: \bibitem{limi96b} Lieu, R., Mittaz, J., Bowyer, S., \& et al. 1996b, Science,
382: 274, 1335
383: \bibitem{libo99a} Lieu, R., Bonamente, M., \& Mittaz, J. P. D. 1999a, \apj,
384: 517, L91
385: \bibitem{libo99b} Lieu, R., Bonamente, M., Mittaz, J. P. D., \& et al. 1999b,
386: \apj, 527, L77
387: \bibitem{libo00} Lieu, R., Bonamente, M., \& Mittaz, J. P. D. 2000, \aap, 364,
388: 497
389: \bibitem{klm99} Kaastra, J., Lieu, R., Mittaz, J. P. D., \& et al. 1999, \apj,
390: 519, 119
391: \bibitem{kafeta01} Kaastra, J., Ferrigno, C., Tamura, R., \& et al. 2001,
392: \aap, 365, L99
393: \bibitem{pmb97} Parmar, A. N., Martin, D. D. E., Bavdaz, M., \& et al. 1997,
394: \aaps, 122, 309
395: \bibitem{pepa01} Peterson, J. R., Paerels, F. B. S., Kaastra, J. S., \& et al.
396: 2001, \aap, 365, L104
397: \bibitem{sali98} Sarazin, C., \& Lieu, R. 1998, \apjl, 494, 177
398: \bibitem{tam01} Tamura, T., Kaastra, J. S., Peterson, J. R. \& et al. 2001,
399: \aap, 365, L87
400: \end{thebibliography}
401:
402: \end{document}
403: