1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3:
4: %%\usepackage{threeparttable}
5:
6: \slugcomment{Accepted for publication in \apj}
7:
8: \shorttitle{XMM-Newton Observations of Galaxy Groups}
9: \shortauthors{Voevodkin et al.}
10:
11: \def\head{ \vbox to 0pt{\vss \hbox to 0pt{\hskip 440pt\rm
12: LA-UR-07-0764\hss} \vskip 25pt}}
13:
14:
15: \begin{document}
16:
17: \head
18:
19: \title{X-ray Observations of Optically Selected Giant
20: Elliptical-Dominated Galaxy Groups}
21:
22:
23: \author{
24: Alexey Voevodkin,\altaffilmark{1,5}
25: Christopher J. Miller,\altaffilmark{2}
26: Konstantin Borozdin,\altaffilmark{1}
27: Katrin Heitmann,\altaffilmark{1}
28: Salman Habib,\altaffilmark{1}
29: Paul Ricker,\altaffilmark{3}
30: and Robert C. Nichol\altaffilmark{4}}
31:
32: \altaffiltext{1}{Los Alamos National Laboratory, Los Alamos, NM, USA}
33: \altaffiltext{2}{
34: NOAO/Cerro Tololo Interamerican Observatory
35: La Serena, Chile}
36: \altaffiltext{3}{Department of Physics and Astronomy, University of Illinois,
37: Urbana-Champaign, IL, USA}
38: \altaffiltext{4}{Institute of Cosmology and Gravitation,
39: University of Portsmouth,
40: Portsmouth, PO1 2EG, UK}
41: \altaffiltext{5}{Space Research Institute, Moscow, Russia}
42:
43: \begin{abstract}
44: We present a combined optical and X-ray analysis of three optically
45: selected X-ray bright groups with giant elliptical galaxies in the
46: center. These massive ellipticals were targeted for
47: \emph{XMM-Newton} X-ray observations based on their large velocity
48: dispersions and their proximity to a nearby \emph{ROSAT} X-ray
49: source. Additionally, these targets are significantly brighter in
50: the optical than their nearest neighbors. We show that one of these
51: systems meets the standard criteria for a fossil group. While the
52: other two systems have a prominent magnitude gap in the E/S0
53: ridgeline, they do not appear to have reached the fossil-like final
54: stage of group evolution.
55: \end{abstract}
56:
57: \keywords{elliptical and lenticular, cD --- galaxies:galaxies:
58: clusters --- X-rays:}
59:
60: \section{Introduction}
61:
62: X-ray bright giant elliptical galaxies have gained considerable
63: attention recently as possible end points in the evolution of galaxy
64: groups. It is often suggested that these systems form in the process
65: of merging of smaller galaxies with the associated loss of the
66: progenitors' disk component and the formation of a common, group-sized
67: dark matter halo. Details of this process are still unclear, and the
68: formation process is sensitive to both the cosmological parameters and
69: astrophysical processes involved.
70:
71: The possibility that members of compact groups could merge to form a
72: large elliptical galaxy on timescales much shorter than the Hubble
73: time was first predicted from early numerical simulations
74: ~\cite{1989Natur.338..123B}. In 1993, Ponman \& Bertram suggested
75: that a large elliptical galaxy which formed through mergers of smaller
76: galaxies could retain its diffuse halo. Such an object would exhibit a
77: high X-ray luminosity, while its optical light would be dominated by a
78: single massive galaxy. In 1994, Ponman et al. reported the discovery
79: of just such an object and termed it a ``fossil group'' (FG). Other
80: researchers have discovered similar objects and have employed terms
81: such as over-luminous elliptical galaxies (OLEGs,
82: ~\cite{1999ApJ...520L...1V}). The number of these systems reported in
83: the literature is growing. Similar to the early days of galaxy cluster
84: research, the precise object definitions used by researchers show
85: large variations ~\cite{1999ApJ...520L...1V,
86: 1999ApJ...514..133M,2000ApJS..126..209R, 2003MNRAS.343..627J,
87: 2004ApJ...612..805S,2004AdSpR..34.2525Y, 2005ApJ...624..124U}. In
88: general, the search criteria usually include a cut for both absolute
89: optical magnitude of the central object, and extended X-ray emission,
90: as well as the requirement for an optical magnitude gap between the
91: first and the second brightest member of the group.
92:
93: While both observations and numerical N-body simulations firmly
94: establish the existence of fossil-group-like objects, there are
95: several remaining questions to be addressed. The theory of
96: hierarchical structure formation predicts that more massive objects
97: form by merging of less massive objects, and consequently, that mass
98: functions for structures of different mass, such as clusters and
99: groups of galaxies, should be self-similar. While the expected
100: self-similarity is seen in N-body simulations, it seems to break down
101: observationally. The mass function of clusters is in general
102: agreement with simulations, but even the most massive groups
103: demonstrate a lack of low-mass satellites and exhibit mass functions
104: similar to the Local Group ~\cite{2007arXiv0704.2604D}. These
105: conclusions should be considered preliminary, as they have been so far
106: based on a small number of studied systems.
107:
108: Two theories have been proposed as models for the formation of giant
109: ellipticals: {\it evolutionary formation through merging} and {\it in
110: situ through accretion}. The first scenario suggests that they are
111: the end-point of group/cluster evolution through mergers and dynamical
112: friction. The second suggests that they are the massive-end point of
113: the elliptical galaxy distribution or that they formed initially with
114: a deficit of small galaxies
115: ~\cite{1999ApJ...514..133M,2003MNRAS.343..627J,2004AdSpR..34.2525Y,2006AJ....131..158M}.
116: Both of these theories have their own difficulties when trying to
117: explain the observations. If these systems have evolved from groups
118: or clusters one would expect to see cool cores, but observationally
119: cool cores are not seen at all or are smaller than expected
120: ~\cite{2004MNRAS.349.1240K,2006MNRAS.369.1211K,2004ApJ...612..805S}.
121: Mass-to-light ratios of fossil groups are unusually high (Khosroshahi
122: et al. 2007), suggesting that the star-formation history of their
123: member galaxies is different from other groups and clusters. They
124: also appear to have higher than expected concentrations as measured
125: from their density profiles.
126:
127: The fraction of fossil-like groups among the general population of
128: groups may well be a sensitive probe of structure formation. While
129: both N-body simulations and observations predict that fossil groups
130: represent a significant fraction of all groups in the mass range
131: ~$10^{13}\,M_\odot$-$10^{14}\,M_\odot$, existing observational
132: statistics do not provide tight constraints for comparison.
133: Historically, fossil groups were selected from X-ray surveys. Due to
134: to the relatively spotty or shallow sky coverage of these X-ray
135: observations, the number of identified fossil groups is low. Vikhlinin
136: et al. (1999) estimated that OLEGs comprise ~20\% of all groups and
137: clusters of comparable luminosity, but this estimate was based on only
138: 4 objects found in a $Rosat$ survey of extended objects. Jones et al.
139: (2003) found a sample of 6 fossil groups satisfying their criteria and
140: calculated their fraction to be 8-20\%. Altogether fewer than 20
141: objects have been studied so far~\cite{2006AJ....131..158M}, and
142: temperature measurements are available only for a
143: handful~\cite{2007MNRAS.377..595K}. In addition, the existing sample
144: is somewhat heterogeneous as various selection criteria were used in
145: different studies. Recently fossil group candidates were identified
146: in the \emph{Sloan Digital Sky Survey} (SDSS) data
147: \cite{2007AJ....134.1551S}. Between 6 and 34 candidates satisfying
148: their criteria were found, depending on the minimum redshift range and
149: radius of search for group members.
150:
151: Recent and planned optical surveys provide an opportunity to expand
152: the sample of fossil group candidates based on optical selection of
153: giant ellipticals. We report here the results of our attempt for such
154: an optical selection (see also Mulchaey \& Zabludoff~1999, Santos et
155: al.~ 2007). We require some evidence of X-ray emission from our
156: targets based on the $Rosat$ all-sky survey, and we are able to
157: confirm the presence of extended X-ray emission and to measure its
158: parameters after dedicated \emph{XMM-Newton} observations of the
159: selected candidates. In this paper, we present the analysis of
160: combined optical and X-ray data, including measurements of
161: temperature, metallicity, and X-ray luminosity for three X-ray bright
162: groups with giant elliptical galaxies in the center. Whenever our
163: analysis requires taking into account cosmological parameters, we
164: assume $H_0=71$ km/s Mpc$^{-1}$, $\Omega_M=0.3$, and
165: $\Omega_\Lambda=0.7$.
166:
167: \section{Data}
168:
169: In this section we describe the algorithm we used to identify X-ray
170: bright groups with a central giant elliptical galaxy in the SDSS data
171: base, and we present the \emph{XMM-Newton} data reduction process.
172:
173: \subsection{Selection Criteria}
174:
175: Fossil-like galaxy groups with a central giant elliptical galaxy were
176: most often identified based on the detection of extended hot X-ray
177: emission followed by optical observations. The SDSS data archive
178: provides the opportunity to search for such systems starting from the
179: optical data. The wealth of data from surveys like the SDSS and
180: 2dFGRS, not to mention future wide-field surveys, allows for a true
181: systematic study of properties and evolution of groups of galaxies.
182: Our selection algorithm focuses on bright ellipticals that are neither
183: completely isolated, nor in typical galaxy groups. The algorithm
184: includes the following steps.
185:
186: \begin{enumerate}
187: \item{Identify all galaxy pairs in the SDSS DR2 spectroscopic sample
188: \cite{2004AJ....128..502A} where the brighter of the two galaxies
189: is at least 2 magnitudes brighter than its counterpart and $M_r <
190: -22.5$.}
191: \item{Ensure that the bright galaxy is an early-type and has a measurable
192: velocity dispersion. Choose only the most massive ellipticals. We
193: choose those with velocity dispersions $> 200 $km s$^{-1}$.}
194: \item{Ensure that there are three galaxies within 1 Mpc, but that the
195: tenth nearest neighbor is $>$1.5 Mpc away (this is $\sim$ one
196: Abell radius). This final criteria ensures that the fossil group
197: candidate resides in a slightly overdense region, but not in a
198: generic group or cluster.}
199: \end{enumerate}
200:
201: The above selection criteria have consequences. First, the use of the
202: SDSS spectroscopic sample allows us to search a complete (at the 90\%
203: level) magnitude limited galaxy survey. However, the magnitude limits
204: and the $\delta m_{12} =2$ requirement result in a redshift limit of
205: $z \sim 0.09$ for fossil group candidates. Second, the requirement of
206: at least 3 galaxies with 1~Mpc of the bright early-type galaxy implies
207: that we will not find any Isolated Over-luminous Elliptical Galaxies
208: (IOLEGS -- as defined by Yoshioka et al.~2004).
209:
210: Applying these criteria we obtained a list of 14 candidates. We
211: visually examined all of these candidates and removed two systems that
212: had been affected by fiber collisions in the SDSS (which only observed
213: 85\% of galaxy pairs closer than 55 arcseconds). We matched the
214: remaining 12 candidates with the available X-ray information using
215: HEASARC and found seven RASS Faint Source Catalog matches. One of
216: these had been already observed (UGC00842, also known as
217: MS~$0116.3-0115$).
218:
219: Our selection criteria were not designed to create a statistical
220: sample of fossil groups. Our goal was to identify massive ellipticals
221: in slightly over-dense regions that lack bright nearest neighbors but
222: are comparatively bright in X-rays. We then targeted these
223: ellipticals to collect X-ray data and study their derived spectral
224: properties (e.g., temperatures and masses) with respect to their
225: optical properties.
226:
227: Our optical target selection was performed in 2005. We obtained
228: \emph{XMM} data for the two X-ray brightest objects from the selected
229: systems RX~J$002937.0-001218$ and RX~J$150548.7+030849$ in 2006 during
230: the XMM AO5 Cycle, and used archived \emph{XMM} data for UGC00842
231: \cite{2003xmm..prop..177M}. Everywhere below we refer to these
232: objects as rxj0029, rxj1505, and ugc00842.
233:
234: \subsection{XMM data reduction}
235:
236: %\clearpage
237: \begin{table}
238: \def\d{\phantom{1}}
239: \caption{Resulting exposures and scaling factors}\label{tab:exp_scale}
240: \centering
241: \medskip\def\arraystretch{1.15}
242: \begin{tabular}{lccccccc}
243: \hline
244: \hline
245: & \multicolumn{1}{c}{$T_{pn}$}\tablenotemark{a} & $\delta_{pn}$\tablenotemark{b} &
246: $T_{MOS1}$ & $\delta_{MOS1}$ & $T_{MOS2}$ & $\delta_{MOS2}$ \\
247: \hline
248: rxj0029 & 6.7 & 3.68 & 15.4 & 1.72 & 15.4 & 1.71 \\
249: %\hline
250: rxj1505 & ...& ... & 18.6 & 1.63 & 17.6 & 1.63 \\
251: ugc00842 & 5.2 & 1.21 & 6.3 & 0.99 & 6.1 & 0.94 \\
252: \hline
253: \end{tabular}
254: \tablenotetext{a}{$T_{xx}$ exposure time for the \emph{xx} camera in
255: kiloseconds.}
256: \tablenotetext{b}{$\delta_{xx}$ ratio of observed fluxes in $10-15$ keV band
257: outside FOV to that in ``blank'' sky data set for the \emph{xx}
258: camera.}
259: \end{table}
260: %\clearpage
261:
262:
263: We analyzed \emph{XMM} data from the EPIC/MOS and pn detectors. The
264: observations were done in Full Frame mode using the THIN optical
265: filter. Calibrated event files were generated using tasks
266: \emph{emchain} and \emph{epchain} from \emph{XMM-Newton} SAS V 7.0 and
267: the calibration database as available in July 2007 was used.
268:
269: %\clearpage
270: \begin{figure*}[t]
271: \centerline{
272: \includegraphics[width=0.33\linewidth]{f1a.ps}
273: \includegraphics[width=0.33\linewidth]{f1b.ps}
274: \includegraphics[width=0.33\linewidth]{f1c.ps}
275: }
276: \caption{X-ray images of rxj0029 (left), rxj1505 (middle), and
277: ugc00842 (right) made in the $0.5-2.0$ keV band. The smoothing scale
278: is 6". The inner circles, with radii 140'', 120'', and 130'', show
279: the areas where the X-ray spectroscopy was performed. The outer
280: circles correspond to $r_{2500}$. The bright source on the rxj1505
281: image is a quasar at $z=0.21$.}
282: \label{fig:images}
283: \end{figure*}
284:
285:
286: \begin{figure*}
287: \centerline{
288: \includegraphics[width=0.33\linewidth]{f2a.ps}
289: \includegraphics[width=0.33\linewidth]{f2b.ps}
290: \includegraphics[width=0.33\linewidth]{f2c.ps}
291: }
292: \caption{Spectra of rxj0029 (left), rxj1505 (middle) and ugc00842 (right).}
293: \label{fig:spectra}
294: \end{figure*}
295: %\clearpage
296:
297: For our analysis we follow closely the data preparation technique
298: described in~Kotov \& Vikhlinin (2005). We summarize here the most
299: important steps. We use the events corresponding to patterns 0--12 for
300: the MOS cameras and 0--4 for the pn camera. We exclude chip \#2 in the
301: MOS1 data of the rxj1505 observation since it looks brighter than the
302: others on the image compiled from the preprocessed event
303: file\footnote{See~also~\emph{XMM-Newton} helpdesk:
304: http://xmm.esac.esa.int/xmmhelp/EPICMOS?id=16999}. The data for each
305: camera is cleaned from flares using the 2--15 keV band and data from
306: the entire field of view is used, with the exception of point sources.
307: Good time intervals are generated from the periods on the lightcurves
308: for which deviations from the mean rate are less than $2\sigma$.
309: Cleaned exposure times for each camera are given in
310: Table~\ref{tab:exp_scale}. The rxj1505 data from the pn camera are
311: highly contaminated by flares, so we do not use them.
312:
313: Since the observed objects are not bright, the X-ray images contain
314: some regions which are dominated by background emission. Therefore, a
315: double-subtraction technique~\cite{2002AA...390...27A} for modeling
316: the background is used. The first step of this method is a
317: subtraction of the energetic particles induced background. In order to
318: do this, normalization coefficients between the background present in
319: the observation and background characteristics for the given camera
320: are found.
321:
322: We calculate these coefficients as a ratio of fluxes from the outside
323: field of view in the 10--15 keV band between the observation and the
324: ``template'' background file. As a ``template'' file for each camera,
325: we use a compilation of ``blank'' sky observations as described in
326: Carter \& Read~(2007). The normalization coefficients we obtain in
327: this way are listed in Table~\ref{tab:exp_scale}. For rxj0029 and
328: rxj1505 the scaling coefficients are higher than 1.5. Good values
329: should lie within 0.8--1.2 interval. In our case the data from both
330: observations are contaminated by particle background. Nevertheless, as
331: we show in our analysis, these observations are still useful for
332: imaging and spectroscopic measurements. The scaling coefficient for
333: the pn data of rxj0029 is too high, see Table~\ref{tab:exp_scale}, so
334: we did not use the pn data from this observation.
335:
336: \bigskip
337:
338: \section{X-ray Properties}
339:
340: In this section we present the spectral and imaging analysis of our
341: groups. Fig.~\ref{fig:images} shows the smoothed, particle background
342: subtracted, vignetting and exposure corrected images of rxj0029,
343: rxj1505, and ugc00842.
344:
345: \subsection{X-ray Spectral Analysis}
346:
347:
348: %\clearpage
349: \begin{table*}
350: \begin{center}
351: \def\d{\phantom{1}}
352: \caption{Spectroscopic parameters}\label{tab:spec_par}
353: \centering
354: \medskip\def\arraystretch{1.15}
355: \begin{tabular}{lcccccc}
356: \hline
357: \hline
358: & \multicolumn{1}{c}{$z$} & $F_X$\tablenotemark{a} & $L_X$\tablenotemark{b} & $T$
359: & $Z$ & $L_{X,bol}$\tablenotemark{c}\\
360: Name & & ($10^{-13}$ erg s$^{-1}$ cm$^{-2}$) & ($10^{42}$ erg
361: s$^{-1}$) &(keV)& ($Z_\odot$) & ($10^{43}$ erg s$^{-1}$)\\
362: \hline
363: rxj0029 & 0.060& $2.83\pm0.88$& $2.41\pm0.68$& $2.10\pm0.31$ & $0.66\pm0.28$ & $1.16\pm0.17$\\
364: rxj1505 & 0.042& $1.92\pm0.47$& $0.80\pm0.21$& $1.13\pm0.15$ & $0.25\pm0.10$ & $0.44\pm0.04$\\
365: ugc00842& 0.045& $5.28\pm0.62$& $2.51\pm0.25$& $1.90\pm0.30$ & $0.34\pm0.12$ & $1.63\pm0.05$\\
366: \hline
367: \end{tabular}
368:
369: \tablenotetext{a}{X-ray flux measured in the 0.5--2.0 keV band.}
370: \tablenotetext{b}{X-ray luminosity in the 0.5--2.0 keV band.}
371: \tablenotetext{c}{Bolometric luminosity (0.1--20.0 keV band) obtained by
372: extrapolation of the measured luminosity to $r_{500}$ using the
373: $\beta$-model.}
374: \end{center}
375: \end{table*}
376: %\clearpage
377:
378:
379: We extract the spectra of rxj0029, rxj1505, and ugc00842 in circles
380: with 140'', 120'', and 130'' radii, respectively (see the inner
381: circles in Fig.~\ref{fig:images}).
382:
383: The response and effective area files are generated with \emph{rmfgen}
384: and \emph{arfgen} tasks from the SAS package. The resulting spectra
385: are binned in such a way that there are at least 40 photons in every
386: bin. We use the absorbed MEKAL model~\cite{Legacy1995} as a fitting
387: model, where the Galactic absorption is fixed at a value obtained from
388: radio surveys~\cite{1990ARAA..28..215D}. However, due to the
389: contamination of the rxj0029 and rxj1505 data by the particle
390: background we have to add a power-law component without a correction
391: for the effective area. This component allows us to describe the high
392: energy part of the spectrum. We also add the same component to the
393: ugc00842 spectrum in order to describe residual contaminations by
394: flares, since any further thorough cleaning of flares, such as
395: changing $\sigma$-clipping or experiments with different cleaning
396: energy bands, only reduced the exposure time without improving the
397: spectrum.
398:
399: The data from MOS1 and MOS2, including pn in the case of ugc00842, are
400: fitted jointly in the 0.5--10 keV band, where the temperatures,
401: metalicities, and power-law slopes are held fixed, while the
402: normalizations for each spectral component and for each of the
403: detectors is kept free. The resulting spectra for the groups are shown
404: in Fig.~\ref{fig:spectra} and the derived spectral parameters are
405: given in Table~\ref{tab:spec_par}.
406:
407: Figure~\ref{fig:spectra} shows that the main feature of the group
408: spectra in the soft band is a bremsstrahlung component. The power-law
409: component in this band is several times lower. The values for
410: temperatures and abundances are primarily determined from the shapes
411: of the spectra in this band. Hence, the parameters inferred from the
412: fits are quite reliable even for the groups rxj0029 and rxj1505 which
413: are contaminated by particle induced backgrounds.
414:
415: %\clearpage
416: \begin{figure*}
417: \centerline{
418: \includegraphics[width=0.33\linewidth]{f3a.ps}
419: \includegraphics[width=0.33\linewidth]{f3b.ps}
420: \includegraphics[width=0.33\linewidth]{f3c.ps}
421: }
422: \caption{Surface brightness profiles of rxj0029 (left), rxj1505
423: (middle), and ugc00843 (right). The dotted lines show the regions
424: where temperatures and other spectroscopic parameters were measured.
425: The dashed lines show $r_{2500}$.}
426: \label{fig:profiles}
427: \end{figure*}
428: %\clearpage
429:
430: \subsection{X-ray Imaging Analysis}
431: For the imaging analysis in the 0.5--2.0 keV energy band we mask all
432: detectable point sources on the MOS1 and MOS2 detectors for three of
433: our groups. Then we add vignetting corrected images. Surface
434: brightness profiles are shown in Fig.~\ref{fig:profiles}. We use the
435: $\beta$-model ~\cite{1976AA....49..137C} to fit the data:
436: \begin{equation} S(r)=S_0 (1+ \frac{r^2}{{r_c}^2})^{-3\beta+0.5},
437: \end{equation}
438: where $r$ is the angular projected off-center distance and $r_c$ is
439: the core radius of the distribution. The best fit parameters for
440: $r_c$ and $\beta$ are given in Table~\ref{tab:prof_par} and the
441: surface brightness profiles extracted from the complete field of view
442: are shown in Fig.~\ref{fig:profiles}. The derived parameters $r_c$ and
443: $\beta$ fall in the range of values for groups of galaxies measured in
444: Osmond \& Ponman 2004. Several outer points on the brightness profiles
445: for rxj0029 and rxj1505 lie higher than one would expect, especially
446: for rxj0029. This effect is probably caused by the particle induced
447: background and vignetting corrections. The same effects but in lesser
448: degree, can also be seen in Kotov \& Vikhlinin~(2005). These last
449: points give some idea of the possible uncertainties in our results due
450: to experimental errors. As one can see, most of our points lie well
451: above this level, and our fit parameters are stable against these
452: effects.
453:
454: Assuming a spherically symmetric density distribution and hydrostatic
455: equilibrium of the intragroup gas, and using the fitted temperatures
456: and fit parameters of the $\beta$-model, we estimate the total masses
457: of the groups inside $r_{2500}$ and $r_{500}$ as
458: \begin{equation}
459: M(< r_{2500(500)}) = 1.1 \times 10^{14}M_\odot T_{keV} \beta
460: \frac{r_{2500(500)}^3}{r^2_{2500(500)}+r_c^2}
461: \end{equation}
462: The estimated values for $M_{t,2500}$ and $M_{t,500}$ (see
463: Table~\ref{tab:prof_par}) are typical for groups of galaxies.
464:
465: Knowing $r_{2500}$ and $r_{500}$ we can obtain gas mass estimations
466: inside these radii. We use the popular deprojection technique
467: \cite{1981ApJ...248...47F, 1997MNRAS.292..419W} in order to obtain
468: volume emissivity, which is then converted to the gas density and gas
469: mass. Since the brightness profiles are quite noisy, we deproject the
470: $\beta$-model fits rather than count rates. The results are shown in
471: Table~\ref{tab:prof_par}.
472:
473: The gas fractions inside $r_{2500}$ and $r_{500}$ are in good
474: agreement with the gas fractions for other groups of galaxies with the
475: same temperatures (see Fig.~4 and Fig.~6 in Sanderson et al.~2003).
476:
477: %\clearpage
478: \thispagestyle{empty}
479: %\setlength{\voffset}{15mm}
480: %{\rotate
481: \begin{table*}
482: \begin{center}
483: \caption{Parameters inferred from the X-ray data}\label{tab:prof_par}
484: \begin{tabular}{lcccccccccc}
485: \hline
486: \hline
487: & \multicolumn{1}{c}{$r_c$} & $\beta$ & $r_{2500}$\footnote{$r_{2500}$($r_{500}$) is the
488: radius inside which the mean density of the object is 2500(500) times
489: higher than the mean density of the Universe.} & $M_{g,2500}$ &
490: $M_{t,2500}$ & $f_{g,2500}$ & $r_{500}$ & $M_{g,500}$ &
491: $M_{t,500}$ & $f_{g,500}$\\
492: Name & (kpc) & & (kpc) & ($10^{11}\,M_\odot$) &
493: ($10^{13}\,M_\odot$) & ($M_g/M_t$) & (kpc) & ($10^{12}\,M_\odot$)
494: & ($10^{13}\,M_\odot$) & ($M_g/M_t$) \\
495: \hline
496: rxj0029 & $5.9\pm0.5$ &$0.40\pm0.03$ & 243 & $5.1\pm1.4$ & $2.34\pm0.40$ & ~2.2\% & 545 &
497: $2.2\pm0.6$ & $5.01\pm0.85$ & ~4.4\% \\
498: rxj1505 & $3.2\pm0.6$ &$0.37\pm0.05$ & 173 & $1.8\pm1.1$ & $0.79\pm0.15$ & ~2.3\% & 388 &
499: $0.9\pm0.5$ & $1.78\pm0.36$ & ~5.1\% \\
500: ugc00842& $5.0\pm0.5$ &$0.38\pm0.02$ & 228 & $7.5\pm2.3$ & $1.70\pm0.30$ & ~4.4\% & 509 &
501: $3.5\pm1.1$ & $4.03\pm0.69$ & ~8.8\% \\
502: \hline
503: \end{tabular}
504: \end{center}
505: \end{table*}
506: %}
507: %\clearpage
508: %\setlength{\voffset}{0mm}
509:
510:
511: \section{Optical Properties}
512:
513: %\clearpage
514: \begin{table}
515: \caption{Optical Parameters}\label{tab:optical_par}
516: \begin{tabular}{lccccccc}
517: \hline
518: \hline
519: Name\tablenotemark{a} & {$N_{spec}$}& {$\sigma_V$} & {$M_{R_c}^{BG1}$} & {$M_{R_c}^{BG2}$} & $M_{R_c}^{BG3}$&
520: $L_{opt,tot}$\\
521: {} & {} & {(km/s)} & {} & {} & & {($10^{11} L_\odot$)}\\
522: \hline
523: rxj0029 & 16 &434 & $-22.68$ & $-22.27$ & $-21.86$ & 3.09\\
524: rxj1505 & 13 &242 & $-22.55$ & $-20.86$ & $-20.79$ & 1.98\\
525: ugc00842& 16 &439 & $-23.01$ & $-20.02$ & $-19.99$ & 1.93\\
526: \hline
527: \end{tabular}
528:
529: \tablenotetext{a}{All quantities given here are calculated for galaxies inside
530: $r_{500}$.}
531:
532: \end{table}
533: %\clearpage
534:
535: The optical data are taken from the Sloan Digital Sky Survey (SDSS)
536: DR5 release~\cite{2007ApJS..172..634A}. We note that the SDSS
537: magnitudes have been found to be underestimated for bright galaxies as
538: a result of the background subtraction technique utilized by the SDSS
539: photometric pipeline~\cite{2007ApJ...662..808L, 2007AJ....133.1741B}.
540: Thus, we have corrected the petrosian magnitudes using the technique
541: described in von der Linden et al. (2007). These corrections are
542: between 0.1 and 0.2 magnitudes in the $r$ and $i$-bands. We have also
543: applied extinction corrections and k-corrections using
544: $kcorrect$\footnote{http://cosmo.nyu.edu/blanton/kcorrect/} version
545: v4\_1 \cite{2003AJ....125.2348B}. To compare to the literature we
546: transform the SDSS $r$ magnitudes to $R_{Cousins}$ using the
547: transformations given in Fukugita et al. (1996).
548:
549: We calculate velocity dispersions for our groups using galaxies with
550: available spectral information (see Table~\ref{tab:optical_par}). For
551: a consistency check we estimated $r_{500}$ using
552: \cite{1998ApJ...505...74G, 2007MNRAS.377..595K}:
553: \begin{equation}
554: r_{500} = \frac{1.2 \sigma_V}{h}.
555: \end{equation}
556: $r_{500}$ values obtained this way are within 20\% from the $r_{500}$
557: values we get from the X-ray profiles (Table~\ref{tab:prof_par}), and
558: therefore in good agreement considering the uncertainties in
559: both estimations.
560:
561: %\clearpage
562: \begin{figure*}
563: \centerline{
564: \includegraphics[width=0.33\linewidth]{f4a.ps}
565: \includegraphics[width=0.33\linewidth]{f4b.ps}
566: \includegraphics[width=0.33\linewidth]{f4c.ps}
567: }
568: \caption{The color magnitude diagram for the groups rxj0029, rxj1505,
569: and ugc00842. The open circles are photometric data from the SDSS.
570: The filled circles are those galaxies targeted for SDSS
571: spectroscopy. The squares show the galaxy data with observed SDSS
572: velocities and that are within the velocity dispersion of the group.
573: The Xs denote galaxies with observed spectra that are not group
574: members. In all figures, the region with $r_{2500}$ projected on
575: the sky is used for membership.}
576: \label{fig:cmrs}
577: \end{figure*}
578: %\clearpage
579:
580:
581: \section{Discussion and Conclusions}
582:
583: %\clearpage
584: \begin{figure*}
585: \centerline{
586: \includegraphics[width=0.33\linewidth]{f5a.ps}
587: \includegraphics[width=0.33\linewidth]{f5b.ps}
588: \includegraphics[width=0.33\linewidth]{f5c.ps}
589: }
590: \caption{X-ray contours overlayed on the SDSS images (from left to
591: right: rxj0029, rxj1505, and ugc00842).}
592: \label{fig:contour_images}
593: \end{figure*}
594:
595: \begin{figure*}
596: \plottwo{f6a.ps}{f6b.ps}
597: \plottwo{f6c.ps}{f6d.ps}
598: \caption{Comparison of the measurements done in this work (red
599: points), with the data for fossil groups taken from Khosroshahi et
600: al. (2007) (black points). Green points were taken from Osmond \&
601: Ponman (2004), blue from Helsdon \& Ponman (2003), and magenta
602: from Wu et al. (1999).}
603: \label{fig:scale_rel}
604: \end{figure*}
605: %\clearpage
606:
607: Hierarchical structure formation theory predicts the formation of
608: larger galaxies through mergers of smaller galaxies. Studies of the
609: largest galaxies provide us with a sensitive test for these
610: predictions. The merger rates depend on masses of merging galaxies and
611: on the impact parameter at first encounter. Under the usual
612: assumption of an isothermal spherical model for each galaxy and a
613: circular orbit for the smaller halo, the infall time can be estimated
614: as \cite{2005ApJ...630L.109D}:
615: \begin{equation}
616: t_{inf}=12.4~Gyr
617: \left(\frac{r_0}{100~kpc}\right)
618: \left(\frac{V_M}{700~km/s}\right)^2\left(\frac{250~km/s}{V_S}\right)^3
619: \end{equation}
620: where $r_0$ is an impact parameter, and $V_M$ and $V_S$ are
621: characteristic circular velocities for the main halo and the
622: satellite, correspondingly. One conclusion from this formula is that
623: the mergers are more efficient in groups ($V_M \lesssim$ 700~km/s),
624: than in rich clusters ($V_M \gtrsim$ 1400 km/s). Another conclusion is
625: that $L_*$ galaxies with $V_S \sim$ 250~km/s merge faster than dwarf
626: galaxies. For mergers along filaments with $r_0 \lesssim$ 10~kpc, a
627: deficit of $L_*$ galaxies is expected within the groups of mass
628: 10$^{13}$~--~10$^{14}M_\odot$ dominated by a single giant elliptical
629: galaxy. While this model explains the existence and properties of the
630: observed fossil groups and OLEGs, we would like to point out that
631: there is no clear physical justification to define a magnitude gap
632: observed in such systems at some particular level. In this dynamical
633: sense, the fossil group criteria discussed in the literature
634: \cite{2003MNRAS.343..627J} seem to be rather arbitrary. Recent
635: statistical studies of SDSS galaxy groups \cite{2007arXiv0710.5096Y}
636: confirm that the $\Delta m_{12}$ magnitude gap distribution is smooth
637: and exhibits no special features at $\Delta m_{12} \geqslant$ 2.
638: There is no clear guidance for the choice of other selection criteria,
639: like the radius of the optical search for galaxies and the acceptable
640: range of redshifts. In fact, Santos et al. (2007) applied different
641: cuts to SDSS data and came to lists of 6 to 34 candidates. Exploring
642: the parameter space may be more fruitful in this situation than by
643: setting particular cuts and limits.
644:
645:
646: The central galaxy of the group rxj1505 is much brighter than the
647: others in the group, however, the absolute difference in magnitudes
648: between the first and second brightest galaxy is 1.69, which is not
649: enough to satisfy the commonly used fossil group
650: definition~\cite{2003MNRAS.343..627J}. While this group is not
651: exceptionally bright or massive, its mass-to-light ratio and other
652: X-ray scaling relations are similar to fossil groups~\cite
653: {2007MNRAS.377..595K}. The magnitude gap for this system is close to
654: that of the group Cl~1205+44, which at z=0.59 has been identified as
655: the most distant known fossil group \cite{2005ApJ...624..124U}.
656:
657: The group rxj0029 appears to be a more unusual object, as in this case
658: the largest magnitude gap is not between the brightest and the second
659: brightest galaxy, but between the second brightest and the rest of the
660: group. While an absence of $L_*$ galaxies and a large magnitude gap
661: between the two brightest galaxies and the third brightest galaxy
662: ($\Delta m_{13}$=2.47, $\Delta m_{23}$=2.07 --- for galaxies inside
663: $r_{2500}$) assume an evolved system, the presence of the second
664: bright galaxy breaks the standard picture. In the optical the two
665: brightest galaxies look rather similar, close in their luminosities,
666: morphologies and colors (see Fig.~\ref{fig:contour_images},\
667: \ref{fig:cmrs} and Table~\ref{tab:optical_par}). However, the X-ray
668: emission is clearly centered around the brightest object, with its
669: peak at the position of the central galaxy
670: (Fig.~\ref{fig:contour_images}). One possibility is that in spite of
671: similar redshifts the second brightest galaxy is X-ray faint and
672: simply projected on the usual fossil group. However, the weighted
673: center of the X-ray emission in rxj0029 is shifted from its peak to
674: the direction of the second brightest galaxy. This may be an
675: indication that the two galaxies are interacting. One may speculate
676: that this interaction is too recent for the stars of the two galaxies
677: to merge, but still long enough, so that the gas component is stripped
678: from the second galaxy and moved toward the center of the group.
679: Continuing infall of that gas may explain an excess of X-ray emission
680: toward the second galaxy. Whether this picture is accurate or not, it
681: demonstrates again that fossil and fossil-like groups may be a more
682: heterogeneous population than is usually assumed.
683:
684: The third group, ugc00842, satisfies the standard fossil group
685: definitions. The magnitude gap between the first and second brightest
686: galaxy is greater than 2 (see Table~\ref{tab:optical_par},
687: Fig.~\ref{fig:cmrs}). Moreover, it holds for all member galaxies lying
688: inside $r_{500}$. The next bright galaxy (01:19:13.46,~$-$01:08:41.2),
689: for which $m_{12}\approx1$ is 525 kpc away from the central galaxy of
690: ugc00842. The group is very bright in the X-ray relative to its
691: optical luminosity, as is typical for fossil groups
692: (Fig.~\ref{fig:scale_rel}). Thus, this object can be added to the
693: list of known fossil groups (see~Mendes de Oliveira et al. 2006).
694:
695: Observationally, the most notable difference between fossil groups and
696: the general population of galaxy groups is that fossils are brighter
697: in the X-ray at the same optical luminosity. Our selection criteria
698: (described in Section 2.1) provided us with a range of systems
699: including a fairly average group, an evolved group, and a fossil
700: group. However, their locations on the optical/X-ray scaling-laws
701: indicate that all three appear similar to the fossil group sample
702: studied in Khosroshahi et al.~(2007).
703:
704: As a consequence, it may be that fossil groups, as classically
705: defined, do not reside in a preferred location in this parameter
706: space. Instead, it may be that the currently available group catalogs
707: fail to accurately trace the full range of parameter space in the
708: scaling-laws.
709:
710: Historically, fossil groups are characterized by their unusual X-ray
711: and optical properties. While these systems are supposed to be
712: common, less than 20 of them have been studied so far. This is
713: understandable, due to the fact that X-ray observations have so far
714: been very spotty or shallow, and as a result it has been hard to
715: collect adequate statistics for these objects. In contrast, optical
716: data provided by the SDSS contains hundreds of thousands of galaxy
717: groups suitable for this type of study. In this work we use optical
718: selection criteria to identify possible fossil group candidates, which
719: can then be further studied with dedicated X-ray observations. We
720: focused on the observation of giant elliptical galaxies, which
721: dominate a population of dwarf galaxies and are bright in X-rays
722: because of their significant masses and large amounts of stripped
723: intergroup gas. We confirmed that in all cases we detect an extended
724: X-ray emission as expected from groups in the considered mass range.
725: This allows us to conclude that the algorithm presented in this paper
726: is quite efficient for the search of groups of galaxies dominated by
727: giant elliptical galaxy. However, it should be tightened for the
728: search of fossil groups.
729:
730: \bigskip
731:
732: AV thanks Oleg Kotov for useful advice in \emph{XMM-Newton} data
733: reduction. We gratefully acknowledge helpful discussions with Alexey
734: Vikhlinin and Ann Zabludoff. We also acknowledge use of Alexey
735: Vikhlinin's $zhtools$
736: software\footnote{http://hea-www.harvard.edu/RD/zhtools/} in our X-ray
737: data analysis. We are thankful to Habib Khosroshahi for providing us
738: with data points for Fig.~\ref{fig:scale_rel}. This work was supported
739: by the LDRD program at Los Alamos National Laboratory.
740:
741: \begin{thebibliography}{1}
742:
743: \bibitem[Abazajian et al. 2004]{2004AJ....128..502A} Abazajian, K.,
744: et al.\ 2004, \aj, 128, 502
745:
746: \bibitem[Adelman-McCarthy et al. 2005]{2007ApJS..172..634A}
747: Adelman-McCarthy, J.~K., AgŸeros, M.~A., Allam, S.~S., Anderson,
748: K.~S.~J.; Anderson, S.~F. et al. 2007, ApJS, 172, 634
749:
750: \bibitem[Arnaud et al. 2002]{2002AA...390...27A} Arnaud, M., et al.
751: 2002, A\&A, 390, 27
752:
753: \bibitem[Barnes 1989] {1989Natur.338..123B} Barnes J. E., 1989,
754: Nature, 338, 123
755:
756: \bibitem[Bernardi et al. 2007]{2007AJ....133.1741B} Bernardi, M.,
757: Hyde, J.~B., Sheth, R.~K., Miller, C.~J., Nichol, R.~C., 2007, AJ,
758: 133, 1741
759:
760: \bibitem[Blanton et al. 2003]{2003AJ....125.2348B} Blanton, M.~R.,
761: Brinkmann, J., Csabai, I., Doi, M. Eisenstein, D., et al. 2003, \aj,
762: 125, 2348
763:
764: \bibitem[Carter \& Read 2007]{2007AA...464.1155C} Carter, J.~A., \& Read,
765: A.~M.\ 2007, A\&A, 464, 1155
766:
767: \bibitem[Cavaliere \& Fusco-Femiano 1976]{1976AA....49..137C}
768: Cavaliere, A., \& Fusco-Femiano, R.\ 1976, A\&A, 49, 137
769:
770: \bibitem[Dickey \& Lockman 1990]{1990ARAA..28..215D} Dickey, J.~M., \&
771: Lockman, F.~J.\ 1990, ARA\&A, 28, 215
772:
773: \bibitem[D'Onghia et al. 2005]{2005ApJ...630L.109D} D'Onghia, E.,
774: Sommer-Larsen, J., Romeo, A.~D., Burkert, A., Pedersen, K.,
775: Portinari, L., \& Rasmussen, J.\ 2005, ApJL, 630, L109
776:
777: \bibitem[D'Onghia et al. 2007]{2007arXiv0704.2604D} D'Onghia, E.,
778: Maccio', A. V., Lake, G., Stadel, J., Moore, B., 2007,
779: astro-ph/0704.2604
780:
781: \bibitem[Fabian et al. 1981]{1981ApJ...248...47F} Fabian, A.~C., Hu,
782: E.~M., Cowie, L.~L., \& Grindlay, J.\ 1981, ApJ, 248, 47
783:
784: \bibitem[Fukugita et al. 1996]{1996AJ....111.1748F} Fukugita, M.,
785: Ichikawa, T., Gunn, J.~E., Doi, M., Shimasaku, K., \& Schneider,
786: D.~P.\ 1996, \aj, 111, 1748
787:
788: \bibitem[Gigardi et al. 1998]{1998ApJ...505...74G} Girardi M.,
789: Giuricin G., Mardirossian F., Mezzetti M., Boschin W., 1998, ApJ,
790: 505, 74
791:
792: \bibitem[Helsdon \& Ponman 2003]{2003MNRAS.340..485H} Helsdon, S.~F.,
793: \& Ponman, T.~J.\ 2003, \mnras, 340, 485
794:
795: \bibitem[Jones et al. 2003]{2003MNRAS.343..627J} Jones, L.~R., Ponman,
796: T.~J., Horton, A., Babul, A., Ebeling, H., \& Burke, D.~J.\ 2003,
797: \mnras, 343, 627
798:
799: \bibitem[Khosroshahi et al. 2004]{2004MNRAS.349.1240K} Khosroshahi,
800: H.~G., Jones, L.~R., \& Ponman, T.~J.\ 2004, \mnras, 349, 1240
801:
802: \bibitem[Khosroshahi et al. 2006]{2006MNRAS.369.1211K} Khosroshahi,
803: H.~G., Maughan, B.~J., Ponman, T.~J., \& Jones, L.~R.\ 2006, \mnras,
804: 369, 1211
805:
806: \bibitem[Khosroshahi et al. 2007]{2007MNRAS.377..595K} Khosroshahi,
807: H.~G., Ponman, T.~J., \& Jones, L.~R.\ 2007, \mnras, 377, 595
808:
809: \bibitem[Kotov \& Vikhlinin 2005]{2005ApJ...633..781K} Kotov, O., \&
810: Vikhlinin, A.\ 2005, ApJ, 633, 781
811:
812: \bibitem[Lauer et al. 2007]{2007ApJ...662..808L} Lauer, T.~R., Faber,
813: S.~M., Richstone, D., Gebhardt, K., Tremaine, S., et al. 2007, ApJ,
814: 662, 808
815:
816: \bibitem[Mathews 2003]{2003xmm..prop..177M}Mathews, W. 2003
817: XMM-Newton Proposal 02033402
818:
819: \bibitem[Mendes de Oliveira et al. 2006]{2006AJ....131..158M} Mendes
820: de a Oliveira, C.~L., Cypriano, E.~S., \& Sodr{\'e}, L.~J.\ 2006,
821: AJ, 131, 158
822:
823: \bibitem[Mewe et al. 1995]{Legacy1995} Mewe, R., Caastra, J.~S., \&
824: Liedahl, D.~A. 1995, Legacy, 6, 16
825:
826: \bibitem[Mulchaey \& Zabludoff 1999]{1999ApJ...514..133M} Mulchaey,
827: J.~S., \& Zabludoff, A.~I.\ 1999, ApJ, 514, 133
828:
829: \bibitem[Osmond \& Ponman 2004]{2004MNRAS.350.1511O} Osmond,
830: J.~P.~F., \& Ponman, T.~J.\ 2004, \mnras, 350, 1511
831:
832: \bibitem[O'Sullivan et al. 2001]{2001MNRAS.328..461O} O'Sullivan, E.,
833: Forbes, D.~A., \& Ponman, T.~J.\ 2001, \mnras, 328, 461
834:
835: \bibitem[Ponman \& Bertram 1993]{1993Natur.363.51} Ponman, T.~J. \&
836: Bertram,~D. \ 1993, Nature, 363, 51
837:
838: \bibitem[Ponman et al. 1994]{1994Natur.369..462P} Ponman, T.~J.,
839: Allan, D.~J., Jones, L.~R., Merrifield, M., McHardy, I.~M., Lehto,
840: H.~J., \& Luppino, G.~A.\ 1994, Nature, 369, 462
841:
842: \bibitem[Romer et al. 2000]{2000ApJS..126..209R} Romer, A.~K., Nichol,
843: R.~C., Holden, B.~P., Ulmer, M.~P., Pildis, R.~A. et al.\ 2000,
844: ApJS, 126, 209
845:
846: \bibitem[Sanderson et al. 2003]{2003MNRAS.340..989S} Sanderson,
847: A.~J.~R., Ponman, T.~J., Finoguenov, A., Lloyd-Davies, E.~J., \&
848: Markevitch, M.\ 2003, \mnras, 340, 989
849:
850: \bibitem[Santos et al. 2007]{2007AJ....134.1551S} Santos, W.~A., Mendes de
851: Oliveira, C., \& Sodr{\'e}, L.~J.\ 2007, AJ, 134, 1551
852:
853: \bibitem[Sun et al. 2004]{2004ApJ...612..805S} Sun, M., Forman, W.,
854: Vikhlinin, A., Hornstrup, A., Jones, C., \& Murray, S.~S.\ 2004,
855: ApJ, 612, 805
856:
857: \bibitem[Vikhlinin et al. 1999]{1999ApJ...520L...1V} Vikhlinin, A.,
858: McNamara, B.~R., Hornstrup, A., Quintana, H., Forman, W., Jones, C.,
859: \& Way, M.\ 1999, ApJL, 520, L1
860:
861: \bibitem[von der Linden et al. 2007]{2007MNRAS.379..867V} von der
862: Linden, A., Best, P.~N., Kauffmann, G., White, S.~D.~M., 2007,
863: \mnras, 379, 867
864:
865: \bibitem[Ulmer et al. 2005]{2005ApJ...624..124U} Ulmer, M.~P., Adami,
866: C., Covone, G., Durret, F., Lima Neto, G. B., Sabirli, K., Holden,
867: B., Kron, R.~G., \& Romer, A.~K., 2005, ApJ, 624, 124
868:
869: \bibitem[White et al. 1997]{1997MNRAS.292..419W}White, D. A., Jones,
870: C., \& Forman, W., 1997, \mnras, 292, 419
871:
872: \bibitem[Wu et al. 1999]{1999ApJ...524...22W} Wu, X.-P., Xue, Y.-J.,
873: \& Fang, L.-Z.\ 1999, \apj, 524, 22
874:
875: \bibitem[Yang et al. 2007]{2007arXiv0710.5096Y} Yang, X., Mo, H.~J.,
876: van den Bosch, F.~C., 2007, astro-ph/0710.5096
877:
878: \bibitem[Yoshioka et al. 2004]{2004AdSpR..34.2525Y} Yoshioka, T.,
879: Furuzawa, A., Takahashi, S., Tawara, Y., Sato, S., Yamashita, K., \&
880: Kumai, Y.\ 2004, Advances in Space Research, 34, 2525
881:
882: \end{thebibliography}
883:
884: \end{document}
885: