1:
2: %\documentclass[referee]{aa} % for a referee version
3: %\documentclass[onecolumn]{aa} % for a paper on 1 column
4: %\documentclass[longauth]{aa} % for the long lists of affiliations
5: %\documentclass[rnote]{aa} % for the research notes
6: %
7: \documentclass[structabstract]{aa}
8: %\documentclass[tradiabstract]{aa} % for the abstract without structuration
9: % (traditional abstract)
10: %
11: \usepackage{natbib}
12: \bibpunct{(}{)}{;}{a}{}{,} % to follow the A&A style
13:
14:
15: \usepackage{graphicx}
16: \usepackage{natbib}
17: \usepackage{color}
18: \usepackage{rotating}
19: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
20: \usepackage{txfonts}
21: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
22: %
23: \begin{document}
24: \definecolor{grey}{rgb}{0.5,0.6,0.7}
25: \definecolor{darkred}{rgb}{0.5,0.0,0.0}
26: \definecolor{darkgreen}{rgb}{0.0,0.7,0.0}
27: \newcommand{\euge}[1]{{\bf\textcolor{darkred}{Euge says: #1}}}
28: \newcommand{\claudia}[1]{{\bf\textcolor{blue}{Claudia says: #1}}}
29: \newcommand{\hernan}[1]{{\bf\textcolor{darkgreen}{Hern\'an says: #1}}}
30: \newcommand{\hide}[1]{\textcolor{grey}{#1}}%
31:
32: \title{Fossil Groups in the Millennium Simulation}
33:
34: \subtitle{ Evolution of the Brightest Galaxies}
35:
36: \author{Eugenia D\'{\i}az-Gim\'enez\inst{1},
37: Hern\'an Muriel\inst{1}
38: \and
39: Claudia Mendes de Oliveira\inst{2}
40: }
41:
42: \institute{IATE (CONICET-UNC) \& OAC (UNC). Laprida 854, C\'ordoba 5000. Argentina\\
43: \email{eugeniadiazz@gmail.com, hernan@oac.uncor.edu}
44: \and
45: IAG, USP. Rua do Mat\~ao 1226, S\~ao Paulo. Brazil\\
46: \email{oliveira@astro.iag.usp.br}
47: }
48:
49: \date{Received March 11, 2008; accepted August 19, 2008 }
50:
51: % \abstract{}{}{}{}{}
52: % 5 {} token are mandatory
53:
54: \abstract
55: % context heading (optional)
56: % {} leave it empty if necessary
57: {
58: }
59: % aims heading (mandatory)
60: { %Create a catalogue of \emph{simulated} fossil groups
61: %in order to study the properties and particularly the merging
62: %histories of their first-ranked galaxies and compare
63: %those with \emph{simulated} non-fossil and \emph{observed} fossil groups.
64: We create a catalogue of \emph{simulated} fossil groups and study their properties, in particular the merging histories of their first-ranked galaxies. We compare the \emph{simulated} fossil group properties with those of both \emph{simulated} non-fossil and \emph{observed} fossil groups.
65: }
66: % methods heading (mandatory)
67: {Using simulations and a mock galaxy catalogue, we searched for massive ($>$ 5 $\times$ 10$^{13} \ h^{-1} {\cal M}_\odot$) fossil groups in
68: the Millennium Simulation Galaxy Catalogue.
69: In addition, attempted to identify observed fossil groups in the Sloan Digital Sky
70: Survey Data Release 6 using identical selection criteria.
71: %Using
72: %simulations and a mock galaxy catalogue,
73: %we investigate the expected differences between first-ranked galaxies in
74: %fossil and non-fossil systems
75: }
76: % results heading (mandatory)
77: { Our predictions on the basis of the simulation data are:
78: (a) fossil groups comprise about 5.5\% of the total population of
79: groups/clusters with masses larger than
80: 5 x 10$^{13} \ h^{-1} {\cal M}_\odot$. This fraction is consistent with
81: the fraction of fossil groups identified in the SDSS, after all observational
82: biases have been taken into account;
83: (b) about 88\%
84: of the dominant central objects in fossil groups
85: are elliptical galaxies that have a median R-band absolute magnitude of
86: $\sim -23.5-5 \ log \ h$, which is typical of the observed fossil groups known
87: in the literature;
88: (c)
89: first-ranked galaxies of systems with
90: $ {\cal M} >$ 5 x 10$^{13} \ h^{-1} {\cal M}_\odot$,
91: regardless of whether they are either fossil or
92: non-fossil,
93: are mainly formed by gas-poor mergers;
94: (d) although
95: fossil groups, in general, assembled most of their virial masses at
96: higher redshifts in comparison with non-fossil groups,
97: \emph{first-ranked
98: galaxies in fossil groups merged later,
99: i.e. at lower redshifts},
100: compared with
101: their non-fossil-group counterparts.
102: }
103: % conclusions heading (optional), leave it empty if necessary
104: {We therefore expect to observe a number of luminous
105: galaxies in the centres of fossil groups that show signs of a recent
106: major merger.}
107: % This population is still to be uncovered.}
108:
109: \keywords{ Methods:N-body simulations--
110: Methods:statistical--
111: Galaxies:clusters:general--
112: Galaxies:evolution
113: }
114:
115: \authorrunning{D\'{\i}az-Gim\'enez et al.}
116: \maketitle
117: %
118: %________________________________________________________________
119:
120: \section{Introduction}
121: \cite{Jones03} identified fossil groups as spatially extended X-ray sources
122: with an X-ray luminosity $L_X>10^{42} \ h_{50}^{-2} \ erg \ s^{-1}$ whose
123: optical counterpart was a bound system of galaxies with $\Delta M_{12}>2$ mag, where
124: $\Delta M_{12}$ was the difference in absolute magnitude in R-band between the brightest
125: and the second brightest galaxies in the system within half the projected virial
126: radius ($r_{vir}$). The dynamical masses of the systems studied so far
127: are comparable to those of
128: rich clusters ($\sim 10^{13}-10^{14} \ h^{-1} {\cal M}_\odot$) \citep{Mendes06,cyp06,Mendes08, KPJ06}.
129: Fossil groups may be of considerable importance as the place of formation of a significant
130: fraction of all giant ellipticals.
131: Beside minor differences in the definition of fossil groups,
132: their incidence rate was estimated
133: by observational, analytical, numerical, and semi-analytical analyses.
134: \cite{Vik99} and \cite{Jones03} stated that fossil groups represent (8-20)\% of observed systems in
135: the same mass range. \cite{vandenbosch07} used the 2dF Galaxy Redshift Survey data to measure
136: a fossil fraction of 6.5\% among groups with masses $(10^{13}-10^{14}) \ h^{-1} {\cal M}_\odot$.
137: \cite{M06} estimated analytically that fossil groups represent 5-40\%
138: of groups with masses in the range $\sim 10^{13}-10^{14} \ h^{-1} {\cal M}_\odot$, while the percentage decreased to 1-3\% for groups of mass larger than $10^{14} \ h^{-1} {\cal M}_\odot$, the latter result
139: having been confirmed using the photometric Sloan Digital Sky Survey Data Release 2 (SDSS DR2).
140: Numerical simulations
141: by \cite{Donghia} predicted a higher fraction of fossil systems (33\%) amongst groups of mass $\sim 10^{14} \ h^{-1} {\cal M}_\odot$. \cite{vonbenda07}, also using numerical simulations,
142: found that 24\% of groups with masses in the range $(1-5)
143: \times 10^{13} \ h^{-1}{\cal M}_\odot$ were fossil groups.
144: From semi-analytical models, \cite{sales07}
145: estimated that fossil groups represent $(8-10)\%$ of groups with masses
146: $(10^{13}-10^{15}) \ h^{-1}{\cal M}_\odot$, while \cite{Dariush07} stated that $\sim 13 \%$ of groups
147: in that mass range were fossil systems and predicted that
148: this percentage decreased to 3-4 \% for X-ray rich systems.
149:
150: A natural question is whether the large magnitude difference between the first and second ranked galaxies
151: ($\Delta$M$_{12} > $ 2 in the R-band), characteristic of these groups,
152: implies that they are a distinct class of objects or if they instead represent a tail of
153: the cluster (hereafter non-fossil group) distribution. To
154: investigate this question, \cite{Donghia} used high-resolution
155: N-body/hydrodynamical simulations to compare four
156: simulated fossil and eight non-fossil groups, all of virial masses close
157: to $1 \times 10^{14} \ h^{-1} \ {\cal M}_\odot$.
158: They found that the values of the
159: magnitude gap, $\Delta$M$_{12}$,
160: for the 12 systems, were correlated with the halo assembly time,
161: such that fossil groups assembled earlier than non-fossils. Similarly,
162: \cite{Dariush07} concluded,
163: by the study of fossil groups in the Millennium Simulation,
164: that fossils assemble a higher fraction of their masses at higher
165: redshifts than non-fossil groups.
166: The most accepted scenario for fossil groups is then that they are not a
167: distinct class but, are instead,
168: examples of groups/clusters that collapsed early.
169:
170: Although fossil groups in general do not appear differ from galaxy
171: clusters of similar masses, except for their earlier times of formation,
172: it was realised, observationally, that the first-ranked galaxies in
173: fossil and non-fossil groups differ in some respects.
174: First, their shapes differ: while fossil, first-ranked, galaxies are
175: often disky, brightest cluster galaxies are often boxy \citep{KPJ06}.
176: Second, their stellar
177: populations are dissimilar:
178: fossil group first-ranked galaxies are not as
179: old as brightest cluster galaxies \citep{delaRosa08}.
180: These findings motivated us to revisit
181: the study of \cite{Dariush07} of fossil groups in the Millennium Simulation,
182: but now focusing on the properties of the first-ranked galaxies (as
183: opposed to the complete system).
184: By searching in the Millennium Simulation, we created
185: two samples, one of fossil groups
186: with $\Delta$M$_{12} > 2$ (in the R-band)
187: and a second control sample of non-fossil groups with systems with
188: $\Delta$M$_{12} < 0.5$. Both the simulations themselves
189: and a mock catalogue were used to complete these searches.
190: In addition, we searched for fossil groups in the Sloan Digital Sky Survey Data Release 6 (SDSS DR6) \citep{AMSDSS08}
191: using the same criteria, to compare with the mock catalogue as
192: a test of the semi-analytic model and also to examine observationally the results
193: of our searching algorithm.
194:
195: The layout of this paper is as follows.
196: In Sect.~\ref{construction} we briefly describe the galaxies in the Millennium
197: Simulation, and the search for simulated fossil and non-fossil groups.
198: Section \ref{differences} contains a comparison between fossil and non-fossil groups,
199: and simulated and observed fossils. In particular, we
200: discuss the implications of
201: these results for the evolution of first-ranked galaxies in fossil groups.
202: In Sect.~\ref{samples} we describe the construction of a mock catalogue and
203: the procedures of group identification.
204: We also
205: perform an identification of fossil groups in SDSS DR6 and compare the results with
206: those obtained from the mock catalogue.
207: Finally, we summarise the paper in Sect.~\ref{conclusions}.
208:
209: \section{Construction of the Simulated Fossil Group Sample}
210: \label{construction}
211: \subsection{Dark Matter Particles and Galaxies}
212:
213:
214: The Millennium Simulation is the largest completed cosmological
215: Tree-Particle-Mesh (TPM, \cite{Xu95}) $N$-body simulation
216: \citep{Springel+05}, which evolved 10 billion ($2160^3$) dark
217: matter particles of mass $8.6 \times 10^8 \ h^{-1} {\cal M}_\odot$ within
218: a periodic box of $500 \, h^{-1}\,\rm Mpc$ on a side, using a
219: comoving, softening length of $5 \, h^{-1} \, \rm kpc$.\footnote{The
220: Millennium Simulation, developed by the Virgo Consortium, is available at
221: http://www.mpa-garching.mpg.de/millennium} The cosmological parameters
222: of this simulation correspond to a standard cosmological model
223: ($\Lambda$CDM): $\Omega_m=0.25$, $\Omega_\Lambda=0.75$, $\sigma_8=0.9$,
224: and $h=0.73$.
225: The merging history trees were stored for 60 output times separated by time intervals given by
226: $ ln(1+z_n)= n (n+35)/4200$. These are the basic
227: inputs required by the semi-analytic model.
228:
229: We use a run of \cite{deLucia07}'s semi-analytic model to extract
230: galaxies with positions, velocities, as well as absolute magnitudes (in
231: five photometric bands, BVRIK) and stellar masses, among other quantities.
232: In this model, the branches of the halo merger tree are followed forward
233: in time, and several astrophysical processes are included such as gas infall
234: and cooling, reionization of the Universe, star formation, black hole
235: growth, AGN and supernova feedback, galaxy mergers, and spectro-photometric
236: evolution. The final output at $z=0$ produced $\sim 10\times 10^6$
237: galaxies with absolute magnitudes $M_R-5 \ log \ h <-17.4$ and stellar masses larger
238: than $3 \times 10^8 \ h^{-1} {\cal M}_\odot$. This version
239: of the semi-analytic model provides an improved fit
240: to the bright end of the galaxy luminosity function, compared with former model
241: developed by \cite{Croton+06}. This sample is called Millennium Simulation
242: Galaxy Catalogue (MSGC hereafter).
243:
244: The cosmological parameter set used in the Millennium Simulation
245: was that inferred from the first-year WMAP (Wilkinson Microwave Anisotropy Probe Observations, \citealp{Spergel03}).
246: Differences in the cosmological parameter sets
247: corresponding to the $1^{st}$ and $3^{rd}$-year WMAP \citep{Spergel07}
248: (mainly in $\sigma_8$, $\Omega_m$, $n$)
249: produced a significant delay in structure formation in the WMAP3 case,
250: and the number of halos drawn for the WMAP1 cosmology to be higher.
251: However, \cite{Wang08} investigated the implications of this delay for the
252: observed properties of galaxies. They compared results obtained from a simulation
253: for the cosmological parameters of WMAP1 plus a semi-analytic model (such as MSGC)
254: with those for a simulation for the cosmological parameters of WMAP3 plus
255: a semi-analytic model. They found that the luminosity functions, correlation functions,
256: and Tully-Fisher relations were almost identical in both cases. They stated that
257: the galaxy clustering and other observable properties were far more sensitive
258: to the galaxy formation physics than to the cosmological parameters; the semi-analytic
259: parameters then may be able to compensate for the delay in structure formation producing
260: galaxy populations at z=0 that agree with observations. \cite{Wang08}
261: also concluded that substantial differences between the models appear at redshifts higher than 2, which are beyond the scope of this paper.
262:
263:
264: \subsection{FoF galaxy halos}
265: \label{halos}
266:
267: Since we were interested in studying fossil groups in the MSGC,
268: %and in the mock catalogue
269: we identified galaxy halos by using a standard
270: method. Groups of galaxies in the MSGC were identified by using
271: a Friends-of-Friends (FoF) algorithm in real space \citep{DEFW85} with
272: a linking length of 0.2 of the mean particle density, which corresponds
273: to an overdensity of 200. We checked that all galaxies given in a galaxy halo belonged
274: to the same DM halo.
275:
276: Finally, only groups of galaxies with more than $10$ members were selected
277: %in both cases.
278: We refer to these groups hereafter as FoF galaxy groups
279: (or halos).
280:
281: For all FoF galaxy groups, we compute the velocity dispersion, virial radii, and virial theorem masses. The value of the virial theorem mass was computed to be:
282: \begin{eqnarray}
283: {\cal M}_{\rm FOF} &=& { \pi \over G}\,R_h\,\sigma_{3D}^2
284: %\nonumber \\
285: %&=& 2.2 \times 10^{11} \,\left ({h \,R_h \over 10\, \hbox{kpc}} \right )\,\left
286: %({\sigma_v\over 100 \, \rm km \, s^{-1}} \right )^2\,{\cal M}_\odot \ ,
287: % \sum_{i<j} R_{ij}^{-1}} \ ,
288: \label{mviroverl}
289: \end{eqnarray}
290: where
291: $R_h = \left \langle 1/R_{ij} \right\rangle^{-1}$ is the harmonic mean
292: projected separation, given the projected separations
293: $R_{ij}$ (see eq. [10--23] of \citealp{BT87}).
294: The virial radius is $r_{vir}=\pi R_h$ and the 3-D velocity dispersions were
295: calculated in the MSGC by using the peculiar velocities
296:
297: %while in the mock catalogue the radial velocity dispersions ($\sigma_v$) are calculated using
298: %the bi-weighed estimator described by
299: %\citealp{BFG90} ($\sigma_{3D}=\sqrt{3} \sigma_v$).
300:
301:
302: \subsection{ The Sample of Fossil Groups}
303: \label{fossils}
304:
305: \begin{table}
306: \begin{center}
307: \caption{Median Properties of Fossil Groups identified in the Millennium Simulation
308: \label{medians}
309: }
310: \begin{tabular}{cccc}
311: \hline
312: \hline
313: Property & MSGC & Mock Catalogue & SDSS DR6 \\
314: \hline
315: \\
316: \# & $729 $ & $22$ & $6$ \\
317: \\
318: ${\cal M}_{FOF} \ [h^{-1}{\cal M}_\odot]$ & $7.5\times 10^{13}$ & $6.7\times 10^{13}$ & $8\times 10^{13}$ \\
319: \\
320: $\sigma_{3D} \ [ km/s ]$ & $600$ & $ 596$ & $565$ \\
321: \\
322: $r_{vir} \ [ Mpc \ h^{-1} ]$ & $0.98$ & $1.0$ & $1.1$ \\
323: \\
324: $M_R -5log(h)$ & $-23.48$ & $-23.73$ & $-22.17 *$ \\
325: \\
326: $M_* \ [h^{-1} {\cal M}_\odot] $ & $ 2.8\times 10^{11}$ & $4.2\times 10^{11}$& $-$ \\
327: \\
328: \hline
329: \end{tabular}
330: \parbox{8cm}{
331: Notes: \# number of fossil groups, ${\cal M}_{FOF}$: virial mass of the FoF galaxy group, $\sigma_{3D}$: velocity dispersion of the FoF galaxy group, $r_{vir}$ : virial radius of the FoF galaxy group, $M_R - 5 log h$: rest frame R-band absolute magnitude of the brightest galaxy within $0.5 r_{vir}$ of the FoF galaxy group, $M_* $: stellar mass of the brightest galaxy.\\
332: $*$ In this case it is in the r-band. k+e corrections are computed from \cite{Blanton+03_AJ}
333: }
334: \end{center}
335: \end{table}
336:
337: To represent fossil groups, we selected FoF galaxy groups with masses larger
338: than $5 \times 10^{13} \ h^{-1} {\cal M}_\odot $, which contained galaxies
339: with a magnitude distribution displaying a gap $\Delta$M$_{12} > $ 2 (in
340: the R-band),
341: when considering objects within a radius of
342: $0.5 r_{vir}$.
343: The adopted lower limit to the group mass ($5\times 10^{13} \ h^{-1} {\cal M}_\odot$)
344: ensured that the fossil groups chosen
345: were also X-Ray fossils, according to the
346: work of \cite{Dariush07}.
347:
348: In the MSGC, $729$ FoF galaxy halos satisfy the fossil criteria,
349: which represent $5.5\%$
350: of the FoF galaxy groups in the studied mass range.
351: The median properties of these groups are quoted in Table~\ref{medians}.
352:
353: \subsection{The Sample of Non-Fossil Groups}
354: The principal aim of this paper is to compare
355: the properties of fossil and non-fossil
356: groups. %, in particular the properties of their first-ranked galaxies.
357: \cite{Dariush07} analysed the differences between fossil and
358: non-fossil systems by concentrating on the global properties of the groups.
359: In this work, we intend to go further to not only confirm their findings but
360: also extend their analyses to the brightest galaxies in these systems.
361: We constructed
362: a control sample of non-fossil groups,
363: using the similar criteria to that used for fossil groups in the MSGC but
364: with the one difference that the magnitude gap between
365: the brightest and second
366: brightest galaxies within $0.5r _{vir}$ was less
367: than $0.5$ (for the fossil groups this had to be larger
368: than 2 magnitudes).
369: The control sample comprises $3786$ FoF galaxy halos.
370:
371: %{\bf A paragraph from here was moved to the next subsection}
372: %First, just as a check, \emph{we matched the distributions of group virial masses for
373: %fossil and non-fossil groups} and we recovered all the results described by
374: %\cite{Dariush07}. In
375: %particular, we confirmed that fossil groups, as a whole,
376: %assembled earlier than non-fossil groups, for a given range of group virial masses.
377: %However, since the main goal of this section is to study the properties of the
378: %first-ranked galaxies in groups, we thus selected different subsamples of fossil
379: %and non-fossil groups for comparisons, as described below.
380:
381: Figure \ref{masas} shows a plot of stellar masses
382: versus FoF masses for both fossil (squares) and control (crosses) samples.
383: To avoid inferring results dependent on the mass of the brightest galaxies
384: \citep{deLucia06} from the control and the fossil samples, we selected
385: subsamples with approximately equal numbers of fossil and non-fossil
386: groups, where \emph{the distribution of stellar masses of the central galaxies
387: had been matched}. This exercise provided two samples of fossil and non-fossil
388: groups each with about 680 objects.
389: We note that by insisting that the distributions of \emph{stellar masses of the
390: central galaxies} were identical for the two samples, the virial masses of the groups themselves
391: were, on average, smaller for fossil than for non-fossil groups. This effect can be seen in Fig.~\ref{masas}
392: if we match the y-axis distributions.
393: However, this procedure is justified since we were interested in
394: the properties of the first-ranked galaxies only.
395:
396: \begin{figure}
397: \centering
398: \includegraphics[width=8cm]{masvsmas.ps}
399: \caption{Scatter plot of stellar mass of the brightest galaxy and virial mass of the group. Squares correspond to
400: the sample of fossil groups, while crosses correspond to non-fossil groups.}
401: \label{masas}
402: \end{figure}
403:
404: %\section{Differences between fossil and non-fossil groups in the MSGC}
405: \section{The brightest group galaxies}
406: \label{differences}
407:
408: \begin{figure*}
409: \centering
410: \includegraphics[width=12cm]{f1.eps}
411: \caption{Cumulative distribution of assembly, formation, identity, and
412: extended identity times (from top to bottom). Thick lines correspond
413: to fossil groups, while thin lines are for non-fossil groups.}
414: \label{times}
415: \end{figure*}
416:
417: \subsection{The formation and evolution}
418: Given the availability of merger trees of galaxies in the Millennium
419: Simulation, it is possible to study the formation and evolution of
420: central galaxies in groups.
421: \cite{deLucia06} and \cite{deLucia07}
422: studied the evolution of galaxies of \emph{different} stellar mass. They defined
423: a set of particular times related to formation and evolution of
424: galaxies. Following their work, we analyse the properties of the central
425: galaxies of fossil and non-fossil groups, both samples having the \emph{same} stellar
426: mass distributions. Briefly, the different times are defined as follows:
427:
428: \begin{itemize}
429: \item Assembly time ($z_a$) is the time when $50\%$ of the final stellar mass was already
430: contained in a single galaxy.
431: \item Formation time ($z_f$) is the time when half of the mass of the stars contained in
432: the final galaxy at redshift zero have already formed.
433: \item Identity time ($z_i$) is the time when the latest major merger occurred (
434: major merger is adopted as ${\cal M}_1/{\cal M}_2<3$ ( ${\cal M}_1 > {\cal M}_2$) ).
435: \item Extended identity time ($z_e$) is the latest time when the sum of the
436: masses of all progenitors merging at that time was greater than a third of the
437: mass of the main progenitor (multiple simultaneous minor mergers).
438: \end{itemize}
439:
440: \begin{table}
441: \begin{center}
442: \caption{Median redshifts and probability values from K-S test: fossil versus non-fossils
443: \label{KSprob}
444: }
445: \begin{tabular}{cccc}
446: \hline
447: \hline
448: & Fossil & Non-Fossil & K-S\\
449: \hline
450: $z_a$ & $0.5642$ & $ 0.6871$ & $6\times 10^{-6}$ \\
451: $z_f$ & $3.5759$ & $3.5759$ & $6\times 10^{-6}$ \\
452: $z_e$ & $0.3197$ & $0.4566$ & $5\times 10^{-6}$\\
453: $z_i$ & $0.2798$ & $0.4566$ & $1\times 10^{-7}$\\
454: $z_a (FoF)$ & $0.4566$ & $0.1749$ & $0$ \\
455: \hline
456: \end{tabular}
457: \end{center}
458: \end{table}
459: We compute all of these times for the brightest galaxies in fossil and non-fossil groups, and also the assembly time of the FoF halos.
460: In Table \ref{KSprob}, we show the median of the different characteristic times and also
461: include the probabilities that both distributions, for fossil and non-fossil groups,
462: are drawn from the same distribution (Kolmogorov-Smirnov test (K-S test)).
463: \emph{K-S tests indicate significant differences in all cases}. However, given the shape of the distributions,
464: in some cases the median value is insufficient to show these differences (for instance, when the median of $z_f$ in both samples is the same), therefore, in Fig.~\ref{times}, we show the cumulative distributions of times in Fig.~\ref{times}. Thick lines correspond
465: to fossil groups, while thin lines are for non-fossil groups.
466: Brightest galaxies
467: of fossil groups have (a) assembled,
468: (b) formed their stars,
469: (c) experienced their last major merger and
470: (d) experienced multiple simultaneous mergers,
471: \emph {all at lower redshifts} than bright central galaxies in non-fossil groups. In addition, we confirm that the FoF halos of fossil groups assembled earlier than non-fossil groups.
472:
473:
474: To check if the fact that we are analysing fossil
475: and non-fossil groups with
476: different halo mass ranges could introduce any bias,
477: we repeated the process described above by
478: \emph{matching the distributions of group virial masses for fossil
479: and non-fossil groups}. We then succeeded in reproducing all of the results described by
480: \cite{Dariush07}, in particular, that fossil groups in general,
481: assembled earlier than non-fossil groups, for a given range of group virial masses.
482: We also found that central galaxies in fossil groups have assembled, and
483: experienced both their last major merger and multiple mergers
484: all at lower redshifts than the central galaxies in non-fossil groups,
485: but they have formed their stars at higher redshifts. This result resembles
486: the analysis of \cite{deLucia06} since by matching the virial masses of the
487: groups, the resulting sample of fossil groups have central galaxies with
488: stellar masses typically larger than those of non-fossil groups.
489:
490: Regardless of whether either the stellar masses of the central galaxies or the virial masses of the halos are matched, although the FoF galaxy halos of fossil systems have assembled earlier (\cite{Dariush07,Donghia}, confirmed in this work using an independent semi-analytic model), their central galaxies have, on average, assembled later and, even more importantly, first-ranked galaxies of fossil groups also continue to experience major mergers for a longer period of time than in non-fossil systems.
491:
492: %This result is new but it is not that surprising from the
493: %observational point of view.
494: %A fraction of 20\% of the fossil groups selected from the SDSS
495: %by \cite{Santos07}
496: %had a merger relic or a strongly active object as first-ranked galaxy.
497:
498: \subsection{The morphologies and morphological mix during the last major merger }
499:
500: The criteria used to select fossil groups did not include any constraints
501: on the morphological type of the first-ranked galaxy. However, it is
502: interesting to study whether the first-ranked galaxy has any tendency to
503: be an early-type galaxy.
504:
505: We must consider the shortcomings of the semi-analytic models
506: when we attempt this analysis given that,
507: historically, most semi-analytic models have failed to reproduce some trends,
508: although the new versions are fairly good at reproducing many observational results related to morphology (see \citealp{Bertone07}).
509: However, it is still interesting to observe general trends, defining morphology
510: as completed by \cite{deLucia06}. A
511: galaxy is classified as elliptical if $\Delta M < 0.4$ ($\Delta M =
512: M_{bulge}-M_{total}$ in the B-band), spiral if $\Delta M >1.56$ and S0
513: in between.
514: These authors state that this morphological type determination is
515: robust for galaxies with stellar masses larger than a few times $10^{9} \ h^{-1} {\cal M}_\odot$.
516: As can be observed from Table~\ref{medians},
517: we are dealing with galaxies beyond this limit.
518: We find that $12\%$ of central galaxies of fossil groups
519: in the MSGC are non-elliptical ($5\%$ S and $7\% $ S0's).
520: Similar percentages are found in the control sample ($3\%$ are S and $9\%$ are S0).
521:
522: Regarding the luminosities of the first-ranked galaxies in fossil groups,
523: we found that even when no specific selection criterion was used to select luminous,
524: central, group galaxies, the final sample of fossil
525: groups had quite a bright median magnitude of $\sim -23.5$.
526: We note that a search for fossil groups in the photometric SDSS \citep{Santos07},
527: using similar selection
528: criteria (but with no lower limit to the mass of the groups) produced
529: a sample of groups with a mean absolute magnitude of $M_R - 5 \ log \ h = -23.74$
530:
531: \cite{KPJ06} found that
532: central galaxies of fossil groups are
533: different in their isophotal shapes (which are often disky)
534: compared with the central galaxies of
535: non-fossil systems (which often have boxy shapes). They then suggested
536: that the central galaxies of fossil
537: groups could be the result of wet mergers (gas-rich mergers) unlike the
538: galaxies in non-fossil groups.
539: It is, therefore, interesting to study
540: the morphologies of objects that represent the pre-merging subclumps
541: that produces the central galaxy, after a major merger.
542: We call the process a wet merger if at least one of the two subclumps is
543: a spiral galaxy, a dry merger if both subclumps are elliptical galaxies
544: and mixed if one of them is elliptical and the other is an S0 or
545: both are S0s. Our results are quoted in Table~\ref{mergers}. In contrast
546: to expectations,
547: we find that the vast majority of central galaxies in
548: fossil groups in the MSGC are not produced by wet major mergers.
549: In fact, \emph{central galaxies
550: in both fossil and non-fossil groups appear to have undergone the same type
551: of mergers}, i.e. from the same morphological mix,
552: and are mostly the result of gas-poor mergers.
553: This analysis was repeated
554: for the subsample of systems that have ellipticals at their centres
555: (88\% of the total sample) and
556: similar results were found (values are also quoted in Table~\ref{mergers}).
557:
558: On the other hand, \cite{Khochfar05} investigated whether the observational
559: isophotal shape distribution of elliptical galaxies could be reproduced in
560: semi-analytic models by using both the morphology of galaxies that
561: merge in the last major merger and their mass ratio. Following their work,
562: we classified the central elliptical galaxies that have had a major merger
563: into ``boxy'' or ``disky'' according to
564: the following criteria:
565: \begin{itemize}
566: \item Ellipticals that experience last major mergers between two bulge-dominated galaxies
567: (${\cal M}_{\rm bulge} \ge 0.6 \ {\cal M}_{\rm tot}$) produce ``boxy'' remnants
568: independently of the mass ratio
569: \item Last major mergers with mass ratio $1 \le M_1 / M_2 < 2$ produce ``boxy'' ellipticals.
570: \item Last major mergers with mass ratio $2 \le M_1 / M_2 < 3$ produce ``disky'' ellipticals.
571: \end{itemize}
572: Our results for the sample with elliptical galaxies at their centres
573: that have had a major merger are quoted in Table~\ref{mergers}.
574: This result disagrees with the predictions of \cite{KPJ06} based on 7
575: elliptical galaxies in fossil groups.
576: %, but, regarding normal groups, it is in agreement with \cite{Khochfar05}
577: %and \cite{Bender92} who state about the high fraction of boxy ellipticals at z=0.
578: \emph{We found no differences between the isophotal shapes of
579: elliptical galaxies in fossil or non-fossil systems}, and
580: in both the likelihood of being boxy was higher than being disky.
581: However, the particular semi-analytic model used in this work \citep{deLucia07}
582: could produce a slightly higher fraction of boxy ellipticals than observations
583: (for instance, for non-fossil groups we found that
584: for a median B-absolute magnitude of $-21$, the ratio $N_{boxy}/N_{disky}$
585: is around $2.3$ while the observations
586: of \cite{Bender92} and the predictions of \cite{Khochfar05}
587: are close to $1.5$ (see Fig.~$3$ in \cite{Khochfar05}).
588:
589:
590:
591: \begin{table}
592: \begin{center}
593: \caption{Morphological mix of subclumps that have merged to form the brightest central galaxy
594: \label{mergers}
595: }
596: \begin{tabular}{ccc}
597: \hline
598: \hline
599: & \multicolumn{2}{c}{Full Sample}\\
600: \cline{2-3}
601: & Fossils & non-Fossils \\
602: \hline
603: DRY (E+E) & $48\%$ & $39\%$ \\
604: Mixed (E+S0 or S0+S0) & $22\%$ & $26\%$ \\
605: WET (S+E or S+S0 or S+S) & $8\%$ & $9\%$ \\
606: No major merger & $22\%$ & $26\%$ \\
607: \hline
608: \hline
609: & \multicolumn{2}{c}{ E galaxy at the centre }\\
610: \cline{2-3}
611: & Fossils & non-Fossils \\
612: \hline
613: DRY & $57\%$ & $45\%$ \\
614: Mixed & $26\%$ & $31\%$ \\
615: WET & $8$\% & $9\%$ \\
616: No major merger & $9\%$ & $15\%$ \\
617: \hline
618: boxy & $76\%$ & $75\%$ \\
619: disky & $24\%$& $25\%$ \\
620: \hline
621: \end{tabular}
622: \end{center}
623: \end{table}
624: %-----------------------------------------------------------------
625: \section{Mock galaxy catalogue and observations}
626: \label{samples}
627: Although the principal aim of this paper is to study the brightest
628: galaxies of fossil groups identified in the Millennium Simulation, a
629: direct comparison of the fossil groups identified in the both simulations
630: and observational catalogues can be used as a powerful test.
631: We have particular interest to compare the fraction of FoF systems that are fossil groups in both a mock catalogue constructed from the MSGC and the Sloan Digital Sky Survey.
632: % Of particular interest is the comparison of the fraction of FoF
633: % systems that are fossil groups in both, a mock catalogue constructed from the MSGC
634: % and the Sloan Digital Sky Survey.
635:
636: % Aiming to have results that can be directly compared with observations,
637: To derive results that can be compared directly with observations,
638: we construct a mock catalogue in redshift space using a snapshot at $z=0$ of
639: the MSGC.
640: We compute observer frame galaxy apparent magnitudes from the
641: rest-frame absolute magnitudes provided by the semi-analytical model
642: and tabulated (k+e)-corrections from \cite{Poggianti97},
643: where the corrections
644: were calculated according to an evolutionary synthesis model that reproduces the
645: integrated galaxy spectrum in the range $1000-25000 \AA$.
646: Distorted redshifts are computed using the
647: peculiar velocities of each galaxy. The mock catalogue comprises
648: $\sim 1.3\times 10^6$ galaxies with apparent magnitudes lower than $R=17.77$ within
649: the volume of one simulation box ($z_{max} \sim 0.17$, $\pi/2 \ sr$).
650:
651: Groups of galaxies in the mock catalogues (FoF galaxy halos) were identified
652: by using an algorithm similar to that developed by \cite{Huchra82},
653: which is an adaptation of the FoF algorithm that takes account of the
654: distortion caused by peculiar motions (redshift space) and the apparent
655: magnitude cut-offs. As performed in real space, groups were identified as clusters with
656: an overdensity of 200 and above. Particular care was taken in
657: estimating the group centres.
658: We used a method that computed the
659: projected centre positions by weighting appropriately the positions by the local densities and
660: luminosities \citep{Diaz05}.
661:
662: The masses, virial radii, and velocity dispersions of the FoF galaxy group were
663: computed using Equation~\ref{mviroverl}, although the radial velocity dispersions ($\sigma_v$)
664: where then calculated using
665: the biweight estimator described by
666: \citealp{BFG90} ($\sigma_{3D}=\sqrt{3} \sigma_v$). Only FoF galaxy halos with more
667: than ten members were considered.
668:
669: \subsection{Fossil Groups in a mock catalogue}
670: The criteria used to select fossil groups was described in Sect.~\ref{fossils}.
671: Besides considering only groups with masses larger
672: than $5\times 10^{13} \ h^{-1} {\cal M}_\odot$,
673: we also restricted the depth of the FoF galaxy halos by selecting groups with
674: redshifts lower than $0.1$. This restriction was based on statistical analyses
675: that revealed that the reliability of the identification algorithm in redshift space
676: increases for groups at redshifts below $z_{lim}=0.1$.
677:
678: We found a comparatively low number of fossil
679: groups: $22$ FoF galaxy halos were classified as fossils, which represented
680: $\sim 3\%$ of the FoF galaxy groups in the mock catalogue.
681: Median properties of these groups were
682: also quoted in Table~\ref{medians}. For the mock sample, we estimated the mean
683: density of fossil groups up to the median redshift of the sample,
684: which is a fairly robust measure of density. We measured $1.4\times 10 ^{-6} h^3 \ Mpc^{-3}$.
685:
686: It can be observed that the fraction of groups that satisfy the fossil criteria
687: in the mock catalogue is lower than the $\sim 5.5 \%$ found in MSGC.
688: This discrepancy occurs for the following reason.
689: In real space (MSGC) the search for galaxies whose magnitudes respect the
690: $\Delta$M$_{12} > 2$ selection criterion
691: is completed within a sphere, which cannot be replicated for
692: the mock catalogue given that we operate in redshift space, and by
693: then measuring projected distances.
694: When considering projected distances in the selection of fossil groups, bonafide
695: members in the outskirts of the groups may fail to satisfy the $\Delta$M$_{12}$
696: magnitude criterion, which then produces a smaller fraction of fossil
697: groups (a similar effect is seen when fossil groups
698: are defined to be within $1r_{vir}$
699: instead of $0.5 r_{vir}$). This hypothesis was tested in the mock catalogue by
700: reidentifying fossil groups in the mock catalogue, but, instead of using redshift
701: space information, we used the available real space information to search for galaxies
702: within a sphere of radius $0.5 r_{vir}$.
703: The percentages that we found in this simple test were far more similar to those drawn from MSGC.
704:
705: We note that in observational catalogues the entire identification procedure
706: will be affected in a similar way as that of our mock catalogue. We must therefore recall
707: % It is,therefore, worth keeping in mind
708: that percentages or number densities
709: drawn from observations will be underestimated. In particular, in our study
710: case, we derived 3\% from the mock catalogue (which should be
711: equivalent to the observations) but the real number should be closer to 5.5\%
712: for the fraction of groups that are fossils, from the entire population
713: of groups/clusters with masses larger than $5 \times 10^{13} \ h^{-1}
714: {\cal M}_\odot$.
715:
716: \subsection{Comparison with SDSS}
717: \label{SDSS}
718:
719: \begin{figure*} \centering \includegraphics[width=12cm]{aitoff.ps}
720: \caption{ Grey points are the galaxies of the main sample of
721: the SDSS.
722: Crosses are the six fossil groups presented in this work. Filled circles
723: are the fossil groups found in the \citealp{Mendes06} with $z<0.1$, and
724: empty circles are the fossil groups identified by \citealp{Santos07}
725: with $z<0.1$
726: }
727: \label{aitoff}
728: \end{figure*}
729:
730: We used the main galaxy sample with spectroscopic redshifts of the SDSS DR6
731: to identify observational fossil groups by using the same methodology
732: applied in our mock catalogue. This sample consisted of $574701$ galaxies with apparent
733: magnitudes in the r-band lower than $17.77$ and a median redshift of $0.1$.
734: Briefly, FoF galaxy halos were identified by using an algorithm similar to
735: that of \cite{Huchra82} (see \citealp{Merchan05} for further details of
736: the procedure) with an overdensity contrast of 200.
737: Then, halos with more than 10 members, masses larger than
738: $5\times 10^{13} \ h^{-1} {\cal M}_\odot $ and redshifts lower than 0.1 were selected.
739: Within this sample, we searched for fossil groups, i.e. groups with a
740: magnitude gap of larger than 2 magnitudes in the r-band
741: when considering galaxies within $0.5 r_{vir}$.
742: We found 6 fossil groups in this sample, which represented a fraction of
743: $0.55\%$ of all groups
744: with more than 10 members and masses larger than
745: $5 \times 10^{13} \ h^{-1} {\cal M}_\odot$.
746: Median properties of these fossil groups are quoted in Table~\ref{medians}.
747:
748: At first sight, this result is surprising, given that we predicted in the previous section
749: that this fraction should be close to 3\%.
750: To understand this low percentage of
751: fossil groups in the SDSS, we carefully considered the causes of incompleteness in
752: the SDSS and we then reproduce these incompleteness in our mock catalogue.
753: Two problems were quickly identified:
754: 1) the incompleteness caused by fibre collisions and 2) the incompleteness
755: caused by the fibre magnitude limits and the image deblending software.
756: The former meant that a fraction of
757: $\sim 70\%$ of galaxies that have a neighbour closer than $55''$ were missing.
758: The latter caused the spectroscopic sample to become noticeably incomplete
759: for galaxies brighter than $r=14.5$.
760: We then performed four different mock catalogues considering
761: different combinations of these incompleteness,
762: % each of these incompleteness separately and together,
763: and we report in Table~\ref{mocks} the percentages of groups considered to be
764: fossils.
765: It can be seen that the fibre collision effect introduces a small
766: bias in the percentages of fossil groups identified (which can be
767: noted by comparing columns ``with
768: close pairs'' and ``without close pairs'', in Table~\ref{mocks}) but, in contrast,
769: the lack of galaxies brighter than 14.5
770: strongly bias the results, causing the fractions to
771: be far lower than the expected 3-4\% predicted from the full mock catalogue.
772: Taking the incompleteness into account, the mean density of mock fossil
773: groups up to the median redshift of the sample is
774: $1.2 \times 10^{-7} h^3 \ Mpc^{-3}$
775:
776: To check this result, we repeated our search for groups in the
777: SDSS, completed above, but now considering only galaxies in the
778: magnitude range $14.5<r<17.77$. We found exactly what we expected from the
779: simulations: $0.27\%$ of the FoF galaxy halos were fossil systems
780: (this number can be
781: directly compared with the final line in the final column of Table \ref{mocks}).
782: For the remaining fossil groups,
783: the mean density up to the median redshift was $1.5 \times 10^{-7} h^3 \ Mpc^{-3}$,
784: which was quite similar to that obtained in the mock catalogue after all biases were
785: considered.
786:
787: The lack of bright galaxies also caused the
788: few fossil groups that we identified in the SDSS DR6 to have
789: first-ranked galaxies that were not as bright as those found in the semi-analytical model.
790:
791:
792: \begin{table}
793: \begin{center}
794: \caption{Percentage of Groups that are fossils in four different mock catalogues
795: \label{mocks}
796: }
797: \begin{tabular}{ccc}
798: \hline
799: \hline
800: & with close pairs & without close pairs\\
801: \hline
802: $R<17.77$ & $3\% $ & $4\%$ \\
803: \\
804: \hline
805: $14.5<R<17.77$ & $0.2\%$ & $0.26\%$ \\
806: \\
807: \hline
808: \end{tabular}
809: \parbox{8cm}{
810: Notes: Mock with close pairs: described in sect \ref{samples}.
811: Mock without close pairs: reproducing the fibre collision effect.
812: Both were performed with different magnitude cut-offs.
813: }
814: \end{center}
815: \end{table}
816:
817: \begin{sidewaystable*}
818: \begin{center}
819: \caption{Fossil groups in the main galaxy sample of the SDSS DR6
820: \label{SLOAN}
821: }
822: \begin{tabular}{ccccccccccccc}
823: \hline
824: \hline
825: name & $\alpha_{cm}$ & $\delta_{cm}$ & $z_{cm}$ & ${\cal M}_{vir}$
826: & $\sigma_{3D}$& $r_{vir}$& $\alpha_1$ & $\delta_1$& $z_1$ & $M_1-5log(h)$& $eClass_1$ \\
827: & [h:m:s] & [d:m:s] & & [$10 ^{13} \ h^{-1} {\cal M}_\odot $] & [$km \ s^{-1}$] &
828: [$Mpc \ h^{-1}$] & [h:m:s] & [d:m:s] & & [r-band]\\
829: \hline
830: \\
831: I & 07:35:29.04 & 39:45:21.6 & 0.0866 & 11.3& 797 & 0.86 & 07:35:36.46 & 39:45:53.57 & 0.0874 & -22.43 & -0.190478 \\
832: \\
833: \hline
834: \\
835: II & 07:58:34.32 & 37:46:12.0 & 0.0406 & 7.5 & 563 & 1.01& 07:58:28.11 & 37:47:11.87 & 0.0408 & -22.29 & -0.151681 \\
836: \\
837: \hline
838: \\
839: III& 15:19:02.64 & 04:17:49.2 & 0.0466 & 8.1 & 579 & 1.04 & 15:19:03.52 & 04:20:01.14 & 0.0468 & -22.09 & -0.177914 \\
840: \\
841: \hline
842: \\
843: IV & 07:34:33.84 & 26:53:56.4 & 0.0796 & 7.9 & 551 & 1.12 & 07:34:22.22 & 26:51:44.93 & 0.0797 & -22.35 & -0.197101 \\
844: \\
845: \hline
846: \\
847: V & 14:38:55.68 & 10:08:16.8 & 0.0552 & 7.7 & 523 & 1.21 & 14:38:47.60 & 10:07:17.26 & 0.0553 & -21.70 & -0.158786 \\
848: \\
849: \hline
850: \\
851: VI** & 08:56:37.44 & 21:52:51.6 & 0.0823 & 9.9 & 544 & 1.43 & 08:56:51.80 & 21:49:49.44 & 0.0827 & -22.17 & -0.175508 \\
852: \\
853: \hline
854: \end{tabular}
855: \parbox{20cm}{
856: Notes: $\alpha_{cm}$,$\delta_{cm}$: equatorial coordinates
857: of the centre of mass of the group; $z_{cm}$: spectroscopic redshift of the centre of mass;
858: ${\cal M}_{vir}$: virial mass of the group; $\sigma_{3D}$: 3-D velocity dispersion
859: of the group; $r_{vir}$: virial radius of the group;
860: %$\#_1$: name of the brightest galaxy in the SDSS DR6 catalogue;
861: $\alpha_1$, $\delta_1$, $z_1$ equatorial coordinates and spectroscopic redshift
862: of the brightest galaxy;
863: $M_1-5log(h)$: rest frame absolute magnitude of the brightest galaxy. k+e corrections from \citealp{Blanton+03_AJ};
864: $eClass_1$: spectral type parameter of the brightest galaxy, provided by SDSS.\\
865: $**$ It is not a fossil group (see text)
866: }
867: \end{center}
868: \end{sidewaystable*}
869:
870:
871: Despite the small number of fossil groups identified in the main spectroscopic
872: galaxy sample of SDSS, we attempted to analyse
873: their main properties.
874: In Table~\ref{SLOAN}, the properties of each of the six groups are quoted.
875: Figure \ref{aitoff} shows the angular distribution of the equatorial coordinates
876: $\alpha$ versus $\delta$ for all known fossil groups.
877: Grey points are the galaxies in the main sample of the SDSS.
878: Crosses are the six fossil groups presented in this work.
879: Filled circles are the fossil groups listed in \cite{Mendes06} with $z<0.1$ and empty circles are the fossil groups identified by \cite{Santos07} in the photometric sample of SDSS with $z<0.1$.
880: %(www.astro.iag.usp.br/~fossilgroup/fossilgroup.html).
881: There are three filled circles in the SDSS area that we failed to identify.
882: Using the SDSS DR6 Finding Chart Tool \footnote{http://cas.sdss.org/dr6/en/tools/chart/chart.asp},
883: we analysed a field about the centre of those fossil groups and
884: realised that they could not be identified because of different reasons.
885: RX J1331.5+1108 could not be identified because its main galaxy could not be observed due to possible
886: fibre collisions.
887: We failed to identify RX J1159.8+5531 as a fossil group because
888: the two-magnitude gap was not present in this group. We identified this group instead to be
889: massive and containing numerous galaxies, but the two-magnitude criterion
890: caused us to
891: reject this halo (the absolute magnitude difference in R-band
892: between the first and second ranked galaxy is $\Delta M_{12}=1.6$). Finally,
893: NGC 6034 was not detected since we did not identify a massive and numerous halo
894: in its position. Instead, we identified the central galaxy to be part of the outskirts
895: of a larger group. A closer inspection also indicated that this galaxy also
896: had a close bright neighbour in the same large group,
897: causing the magnitude gap to
898: be $\Delta M_{12}=1.07$ and not 2,
899: as the strict definition of fossil groups requires.
900:
901: It is also interesting to compare the fossil groups identified by
902: \cite{Santos07} and ourselves. Only one of six groups (number II in
903: our Table \ref{SLOAN}) was in their catalogue. One reason for these differences
904: is that \cite{Santos07} insisted
905: that all fossil groups had extended X-ray emission in the ROSAT all-sky
906: catalogue.
907: For our six groups, one has extended X-ray emission (group II)
908: and one has point source X-ray emission (as detected by XMM, group number IV)
909: and four do not have any detectable ROSAT emission (and have not been
910: observed by other satellites). Given that the sensitivity of ROSAT
911: is quite low, these groups should be studied again with deeper X-ray observations to confirm
912: their nature as fossil groups. Another important discrepancy between the
913: two works is the way in which a fossil-system is defined. Differences in the radii used to
914: test the magnitude gap criterion (virial radius or fixed radius in kpc) and the
915: way of defining the group membership may cause each study to select different systems.
916:
917: By using the NASA/IPAC Extragalactic Database
918: \footnote{http://nedwww.ipac.caltech.edu}, we also analysed whether the
919: SDSS fossil groups presented in this work had been previously
920: identified as groups or clusters of galaxies.
921: Fossils I, IV, V, and VI did not have previous identifications as systems
922: of galaxies. Fossil II corresponded to the X-ray cluster NGC 2484 GROUP
923: \citep{Popesso04} and fossil III was previously identified as the
924: cluster CAN 245 \citep{Wegner99}.
925:
926: We also tested the possibility that our fossil groups were
927: spurious identifications due to the incompleteness caused by the fibre collisions
928: or the fibre magnitude limit that affect the spectroscopic sample of the SDSS.
929: Using the SDSS DR6 Finding Chart Tool, we analysed a field of 0.5 $r_{vir}$ around
930: the centre of each fossil group and found that
931: in fossil VI the brightest galaxy of this group has a close neighbour
932: that is a spectroscopic target not observed due to the fibre collision problem.
933: The magnitude gap between these two galaxies is less than two magnitudes
934: ($\Delta M_{12} \sim 1.33$).
935: Although the redshift of the second-ranked galaxy is unknown,
936: we note that it is likely that this is not a fossil group. This is one
937: of the groups that are not detected in X-rays.
938: A similar situation applies to fossil group III, since in the area corresponding to
939: half the virial radius there is a galaxy whose (apparent) magnitude differs by less
940: than 2 from the (apparent) magnitude of the brightest galaxy,
941: but this galaxy was missing from the SDSS survey because it
942: has a very close neighbour. However, in this case, the visual inspection is insufficient to
943: confirm whether this galaxy belongs to the same group.
944:
945: In summary, the spectroscopic sample of SDSS DR6 is ,probably, not the most appropriate for
946: the of study observational fossil groups due to its inherent
947: incompleteness, which means that we are often unable to or falsely identify groups. The enormity of this galaxy data set is still however a useful guide for deeper optical observational studies or even X-ray observations, which are beyond the scope of this paper.
948: %%______________________________________________________________
949:
950: \section{Conclusions}
951: \label{conclusions}
952: To help interpret observations,
953: we have analysed the properties of the first-ranked galaxies
954: of simulated fossil groups and their
955: main differences with respect to the brightest galaxies
956: of non-fossil systems. To perform this analysis,
957: we have studied a sample of fossil groups obtained from the largest simulated galaxy
958: catalogue at present, the Millennium simulation combined with a semi-analytic model
959: of galaxy formation \citep{deLucia07}.
960: The sample of fossil groups identified was sufficiently large to enable robust
961: statistical analyses to be completed.
962: A control sample of non-fossil groups was also selected.
963: The goal of this research was twofold:
964: to predict results based on semi-analytical models about fossil groups and test the semi-analytical model of formation and evolution by comparing with observational results.
965:
966: Our analysis was performed in both the Millennium Galaxy Catalogue (MSGC) and a
967: mock catalogue. The fraction of fossils predicted by the mock catalogue was
968: compared with the fraction of fossils found in the SDSS data set by applying the same
969: algorithm of identification. If the two main sources of incompleteness
970: affecting the spectroscopic sample of SDSS were taken into account,
971: the fraction of fossil groups
972: found in the mock catalogue was consistent with that observed in the SDSS.
973: We have presented six candidate fossil systems in SDSS DR6.
974: One has been rejected after visually analysing the surroundings of the brightest galaxy.
975: The remaining groups await X-rays observations or deeper photometric studies
976: to confirm their nature as fossil groups.
977:
978: Our main results can be summarised as follows:
979: we confirm the old age feature of fossil systems and that fossil
980: groups with masses larger than $5 \times 10^{13} \ h^{-1} {\cal M}_\odot$ represent
981: $\sim 5.5\%$ of massive systems in the same mass range, in
982: the Millennium Galaxy Catalogue and a lower
983: percentage (of 3\%) of similar
984: systems in a mock galaxy catalogue. Fossils in the Mock Catalogue were identified
985: in redshift space (just as achieved in observations),
986: where galaxies from the outskirts of the groups affect
987: the $\Delta M_{12}$ selection criterion and reduce the
988: resulting number of fossils.
989: We found that $88\%$ of fossil systems have central galaxies that are ellipticals.
990: In addition, we found that
991: the first-ranked galaxies of fossil groups and non-fossil groups
992: have the same morphological mixtures, which can be considered to be the results of
993: gas poor mergers.
994: If the central galaxy of a fossil group is always elliptical, the likelihood that its progenitor galaxies merge in a dry merger is higher. Although this result disagrees with the observational studies of \cite{KPJ06} for seven elliptical galaxies in fossil groups, further observational data is required to fully resolve this issue.
995: %
996: % If the elliptical nature of the central galaxy is required in
997: % the fossil group definition,
998: % the tendency of the progenitor that merge being dry is enhanced.
999: % Despite this result is against the observational studies of \citealp{KPJ06} based on seven elliptical galaxies in fossil groups, further observational data are necessary to settle this issue.
1000: %
1001: Finally, we investigated the nature of the central galaxies by
1002: keeping track of their merging histories. On the one hand, the fossil groups in general
1003: assembled earlier than non-fossil groups, whilst their central galaxies assembled later. We also found that
1004: first-ranked galaxies in fossil groups have undergone a major merger later than
1005: their counterparts in non-fossil systems. We expect this result to be confirmed by future
1006: observational catalogues.
1007:
1008:
1009: %
1010: \begin{acknowledgements}
1011: We thank the anonymous referee for comments that helped to improve this work.
1012: We thank Dr. Laerte Sodr\'e Jr., Walter dos Santos and Dr. Ariel Zandivarez
1013: for many useful discussions in all phases of this project.
1014: This work was partially supported by the European Commission's ALFA-II programme
1015: through its funding of the Latin-American European Network for
1016: Astrophysics and Cosmology (LENAC), Consejo de Investigaciones Cient\'{\i}ficas
1017: y T\'ecnicas de la Rep\'ublica Argentina (CONICET) and Secretar\'{\i}a de Ciencia y
1018: T\'ecnica, UNC (SeCyT).
1019: CMdO acknowledges financial help from FAPESP through the
1020: thematic project 01/07342-7.
1021: The Millennium Simulation databases used in this paper and the web
1022: application providing online access to them were constructed as part of the
1023: activities of the German Astrophysical Virtual Observatory.
1024: \end{acknowledgements}
1025:
1026: \bibliographystyle{aa} % style aa.bst
1027: \bibliography{cgs} % your references Yourfile.bib
1028:
1029:
1030:
1031: %\begin{thebibliography}{}
1032: %\bibitem[Croton et al.(2006)]{Croton+06}Croton, D.J., Springel, V., White, S.D.M.,
1033: %De Lucia, G., Frenk, C.S., Gao, L., Jenkins, A., Kauffmann, G.,
1034: %Navarro, J.F., and Yoshida, N., 2006, \mnras, 365, 11.
1035: %\bibitem[Dariush et al.(2007)]{Dariush07}Dariush, A., Khosroshahi, H., Ponman, T.,
1036: %Pearce, F., Raychaudhury, S., and Hartley, W., 2007, submitted to MNRAS,
1037: %astro-ph/0708.2702.
1038: %\bibitem[Davis et al.(1985)]{DEFW85}Davis, M., Efstathiou, G., Frenk, C.S.,
1039: %and White, S.D.M., 1985, \apj, 292, 371.
1040: %\bibitem[de la Rosa et al.(2007)]{delaRosa07}de la Rosa, I., 2007, in preparation.
1041: %\bibitem[De Lucia et al.(2006)]{deLucia06}De Lucia, G., Springel, V., White, S.,
1042: %Croton, D. and Kauffmann, G., 2006, \mnras, 366, 499, astro-ph/0509725.
1043: %\bibitem[De Lucia \& Blaizot(2007)]{deLucia07}De Lucia, G. and Blaizot, J., 2007,
1044: %\mnras, 375, 2, astro-ph/0606519.
1045: %\bibitem[D'Onghia et al.(2005)]{Donghia}D'Onghia, E., Sommer-Larsen, J., Romeo, A.,
1046: %Burkert, A., Pedersen, K., Portinari, L., and Rasmussen, J., 2005, \apj, 630, L109,
1047: %astro-ph/0505544.
1048: %\bibitem[Huchra \& Geller(1982)]{Huchra82}Huchra, J.P., and Geller, M., 1982,
1049: %\apj, 257, 423.
1050: %\bibitem[Jones et al.(2003)]{Jones03} Jones, L., Ponman, T., Horton, A., Babul, A., Ebeling, H., Burke, D., 2003, \mnras, 343, 627.
1051: %\bibitem[Khosroshahi et al.(2006)]{KPJ06}Khosroshahi, H., Ponman, T. and Jones, L.,
1052: %2006, \mnras, 372, L68.
1053: %\bibitem[Santos et al.(2007)]{Santos07}Santos, W., Mendes de Oliveira, C., and
1054: %Sodr\'e Jr, L., 2007, accepted by \aj, astro-ph/0708.1945v1.
1055: %\bibitem[Springel et al.(2005)]{Springel+05}Springel, V., White, S.D.M., Jenkins,
1056: %A., Frenk, C.S., Yoshida, N., Gao, L., Navarro, J., Thacker, R.,
1057: %Croton, D., Helly, J., Peacock, J.A., Cole, S., Thomas, P.,
1058: %Couchman, H., Evrard, A., Colberg, J., Pearce, F., 2005, \nat, 435, 629, astro-ph/0504097.
1059: %\bibitem[Vikhlinin et al.(1999)]{Vik99} Vikhlinin, A., McNamara, B., Hornstrup, A., Quintana, H., Forman, W., Jones, C., Way, M., 1999, \apjs, 520, 1.
1060: %\bibitem[Xu(1995)]{Xu95}Xu, G., 1995, \apjs, 98, 355, astro-ph/9409021.
1061: %\end{thebibliography}
1062:
1063: \end{document}
1064: