1: \documentclass[apj]{emulateapj}
2: \usepackage{apjfonts}
3: \citestyle{aa}
4:
5: \slugcomment{Accepted by ApJ Letters}
6: \shortauthors{Tran et al.}
7: \shorttitle{Late Assembly of Massive Cluster Galaxies}
8:
9: \begin{document}
10:
11: \newcommand{\kms}{~km~s$^{-1}$}
12: \newcommand{\logh}{$+5\log h$}
13: \newcommand{\mipsmu}{$24\mu$m}
14:
15:
16: \title{The Late Stellar Assembly of Massive Cluster Galaxies Via Major Merging}
17:
18: \author{Kim-Vy H. Tran$^1$, John Moustakas$^2$, Anthony
19: H. Gonzalez$^3$, Lei Bai$^4$, Dennis Zaritsky$^4$, \& Stefan
20: J. Kautsch$^3$}
21:
22: \footnotetext[1]{Institute for Theoretical Physics, University of
23: Z\"urich, CH-8057 Z\"urich, Switzerland}
24: \footnotetext[2]{Center for Cosmology and Particle Physics, 4
25: Washington Place, New York University, New York, NY 10003}
26: \footnotetext[3]{Department of Astronomy, University of Florida,
27: Gainesville, FL 32611}
28: \footnotetext[4]{Steward Observatory, University of Arizona, 933
29: North Cherry Avenue, Tucson, AZ 85721}
30:
31:
32: \begin{abstract}
33:
34: We present multi-wavelength observations of the brightest galaxies in
35: four X-ray luminous groups at $z\sim0.37$ that will merge to form a
36: cluster comparable in mass to Coma. Ordered by increasing stellar
37: mass, the four brightest group galaxies (BGGs) present a time sequence
38: where BGG-1, 2, \& 3 are in merging systems and BGG-4 is a massive
39: remnant ($M_{\ast}=6.7\times10^{11}M_{\odot}$). BGG-1 \& 2 have
40: bright, gravitationally bound companions and BGG-3 has two nuclei
41: separated by only 2.5~kpc, thus merging at $z<0.5$ increases the
42: BGG mass by $\gtrsim40$\% ($t_{MGR}<2$ Gyr) and $V$-band luminosity by
43: $\sim0.4$ mag. The BGGs' rest-frame $(B-V)$ colors correspond to
44: stellar ages of $>3$ Gyr, and their tight scatter in $(B-V)$ color
45: ($\sigma_{BV}=0.032$) confirms they formed the bulk of their stars at
46: $z>0.9$. Optical spectroscopy shows no signs of recent ($<1.5$ Gyr)
47: or ongoing star formation. Only two BGGs are weakly detected at
48: \mipsmu, and X-ray and optical data indicate the emission in BGG-2 is
49: due to an AGN. All four BGGs and their companions are early-type
50: (bulge-dominated) galaxies, and they are embedded in diffuse stellar
51: envelopes up to $\sim140$~kpc across. The four BGG systems must
52: evolve into the massive, red, early-type galaxies dominating local
53: clusters. Our results show that: 1) massive galaxies in groups and
54: clusters form via dissipationless merging; and 2) the group
55: environment is critical for this process.
56:
57:
58: \end{abstract}
59:
60: \keywords{galaxies: clusters: general -- galaxies: elliptical and
61: lenticular, cD -- galaxies: evolution}
62:
63:
64: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
65: \section{Introduction}
66:
67: Whether massive cluster galaxies at $z\sim0$ build up a significant
68: fraction of their stellar masses at $z<1$ via merging continues to be
69: an intensely debated question. In the paradigm of hierarchical
70: formation, galaxies continue to grow via accretion of smaller
71: satellites \citep{peebles:70}, and recent models predict that
72: brightest cluster galaxies assemble half their stellar masses at
73: $z<0.5$ \citep{delucia:07b}. However, direct merging is not favored
74: in virialized clusters due to their high velocity dispersions
75: ($\sigma_{1D}\sim1000$~\kms).
76:
77: Brightest cluster galaxies (BCG) are particularly compelling tracers
78: of mass growth because they are the most massive galaxies in the
79: universe, can be observed to high redshifts \citep{gunn:75}, and
80: should have a rich merger history. A number of BCGs in the local
81: universe have multiple nuclei \citep{hoessel:85,lauer:86,laine:03}
82: which suggests recent merging. However, studies of BCGs in the more
83: distant universe disagree as to whether BCGs have grown significantly
84: in stellar mass since redshift $z\sim1$ \citep{brough:02,whiley:08}.
85:
86: While merging of red, luminous members is observed in groups and
87: dynamically young clusters \citep{vandokkum:99,tran:05b,mcintosh:07},
88: the merging frequency and thus its importance remains uncertain. The
89: key may be environment: merging is rare in established clusters but
90: more frequent in groups that are accreted onto clusters. Because
91: groups have a higher fraction of passive early-type galaxies relative
92: to the field \citep{wilman:05,jeltema:07}, merging between such
93: systems will form massive, passive galaxies.
94:
95: Our study focuses on the brightest galaxies in four X-ray luminous
96: galaxy groups at $z\sim0.37$ discovered in the {\it Las Campanas
97: Distant Cluster Survey}\footnote{The LCDCS was designed to find galaxy
98: clusters at $z\sim0.35-0.9$ and was sensitive to groups only at the
99: lowest redshifts.} \citep{gonzalez:01}. These four galaxy groups are
100: gravitationally bound to each other and form an extended structure
101: that will assemble into a galaxy cluster comparable in mass to Coma by
102: the present epoch \citep[hereafter G05]{gonzalez:05}. By studying
103: galaxies in the group environment prior to cluster assembly, we
104: directly test whether galaxy-galaxy merging on group scales can drive
105: stellar growth in the massive galaxies observed in local clusters, and
106: whether a large fraction of new stars is formed in the process.
107: Throughout the paper, we use $H_0=70$~km~s$^{-1}$~Mpc$^{-1}$,
108: $\Omega_M=0.3$, and $\Omega_{\Lambda}=0.7$; at $z=0.37$, this
109: corresponds to a scale of 5.12 kpc per arcsec and a lookback time of 4
110: Gyr.
111:
112:
113: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
114: \section{Observations}
115:
116: Follow-up $Chandra$ ACIS-I imaging and optical spectroscopy of the
117: unique super-group 1120 field confirmed the existence of four X-ray
118: luminous groups at $z\sim0.37$ (hereafter SG1120). These groups lie
119: within a projected region of $\sim3$~Mpc and have X-ray temperatures
120: of $T_X=1.7-3.0$ keV (G05). Our study combines imaging and
121: spectroscopy from six different observatories\footnote{This work is
122: based on observations made with 1) the NASA/ESA {\it Hubble Space
123: Telescope}, obtained at the Space Telescope Science Institute, which
124: is operated by AURA, Inc., under NASA contract NAS 5-26555; 2) the
125: {\it Spitzer Space Telescope}, which is operated by the Jet Propulsion
126: Laboratory, California Institute of Technology under a contract with
127: NASA; 3) the {\it Chandra X-ray Telescope}, obtained at the Chandra
128: X-ray Observatory Center, which is operated by the Smithsonian
129: Astrophysics Observatory for and on behalf of NASA under contract
130: NAS8-03060; 4) the 4 meter Mayall telescope at Kitt Peak National
131: Observatory; 5) the 6.5 meter Magellan Telescopes at Las Campanas
132: Observatory, Chile; and 6) the ESO Telescopes at the Paranal
133: Observatories (072.A-0367, 076.B-0362, and 078.B-0409).}.
134:
135: % KPNO, HST, Chandra, Spitzer, VLT, Magellen
136:
137:
138: \subsection{Optical \& Near-Infrared Imaging}\label{kcorr}
139:
140: The optical photometry is measured from VLT/VIMOS imaging in $BVR$ for
141: a nearly contiguous $18'\times20'$ region, and from a contiguous
142: mosaic ($11'\times18'$) composed of 10 pointings taken with the
143: HST/ACS in F814W. The high resolution ACS imaging is critical for
144: identifying close galaxy pairs and merger signatures. Two pointings
145: with KPNO/FLAMINGOS also provide a contiguous $K_s$ mosaic
146: ($16'\times19'$).
147:
148: Using the HST/ACS mosaic as the master detection image, line-matched
149: catalogs were generated using SExtractor v2.5.0 \citep{bertin:96}.
150: Rest-frame absolute magnitudes, K-corrections, and stellar masses were
151: determined with {\it k-correct} v4.1 \citep{blanton:07} which uses
152: templates based on models from \citet[hereafter BC03]{bruzual:03}. We
153: used MAG$_{-}$AUTO photometry from the $BVRK_s$ imaging and assumed
154: minimum photometric uncertainties in each bandpass of 0.05 mag. The
155: photometry have also been corrected for foreground Galactic extinction
156: using the \citet{schlegel:98} dust maps and the \citet{odonnell:94}
157: Milky Way extinction curve, assuming $R_V=3.1$.
158:
159:
160:
161: \subsection{Optical Spectroscopy}
162:
163: The initial survey using VLT/VIMOS confirmed that four of the five
164: X-ray regions are galaxy groups at $z\sim0.37$ while the fifth is a
165: cluster at $z=0.48$ (G05). Targets were selected from a
166: magnitude-limited catalog ($R\leq22.5$ mag), and the spectra reduced
167: using a combination of IRAF\footnote{IRAF is distributed by the
168: National Optical Astronomy Observatories.} routines and custom
169: software; see \citet{tran:05a} for details on the reduction and
170: redshift determination.
171:
172: Further spectroscopy was obtained using Magellan/LDSS3 and VLT/FORS2.
173: In both cases, targets were selected from $K_s$ catalogs ($K_s\leq20$
174: mag); for reference, the brightest supergroup galaxy has $K_s=14.62$
175: mag. Guided by the redshift distribution, we define the supergroup's
176: redshift range to be $0.32\leq z\leq0.39$ where the brightest group
177: galaxies lie at $0.354\leq z\leq0.372$ (Table~\ref{tab:bgg}). These
178: spectra provided 97 new members for a total of 198 confirmed
179: supergroup galaxies across the approximately $20'\times20'$ region.
180:
181: The medium resolution spectroscopy corresponds to 0.7\AA~pix$^{-1}$
182: (LDSS3), 1.65\AA~pix$^{-1}$ (FORS2), and 2.5\AA~pix$^{-1}$ (VIMOS).
183: The spectral range for all three instruments covers [OII]$\lambda3727$
184: to [OIII]$\lambda5007$ for most supergroup members.
185:
186: \begin{deluxetable}{lrrrrrr}
187: \tablecaption{Properties of Brightest Group Galaxies\tablenotemark{a}}
188: \tablewidth{0pt}
189: \tablehead{
190: \colhead{BGG} & \colhead{$z$} &
191: \colhead{$V$} & \colhead{$(B-V)$} &
192: \colhead{$M_{\ast}$} &
193: \colhead{$L_{IR}$\tablenotemark{b}} &
194: \colhead{$\sigma_{1D}$ \tablenotemark{c}} \\
195: & & & & \colhead{($M_{\odot}$)} & \colhead{(erg~s$^{-1}$)} &
196: \colhead{(km~s$^{-1}$)}
197: }
198: \startdata
199: 1a & 0.3540 & $-22.46$ & 0.86 & $2.9\times10^{11}$ & \nodata & $329\pm66$ \\
200: 1b & 0.3532 & $-21.49$ & 0.85 & $1.2\times10^{11}$ & \nodata & $329\pm66$ \\
201: 2a & 0.3706 & $-22.97$ & 0.94 & $5.0\times10^{11}$ & $1.5\times10^{44}$ & $421\pm83$ \\
202: 2b & 0.3704 & $-22.04$ & 0.92 & $2.1\times10^{11}$ & \nodata & $421\pm83$ \\
203: 3 & 0.3713 & $-23.20$ & 0.94 & $5.4\times10^{11}$ & $8.0\times10^{43}$ & $588\pm90$ \\
204: 4 & 0.3720 & $-23.38$ & 0.91 & $6.7\times10^{11}$ & \nodata & $565\pm130$ \\
205: \enddata
206: \tablenotetext{a}{Positions for the individual galaxy groups are in G05.}
207: \tablenotetext{b}{Infrared luminosities determined from SST/MIPS 24$\mu$m
208: fluxes.}
209: \tablenotetext{c}{Line-of-sight velocity dispersions ($\sigma_{1D}$)
210: are measured using all members within $500$~kpc of the BGGs (13-36
211: galaxies for each group) and using the biweight estimator
212: \citep{beers:90}. }
213: \label{tab:bgg}
214: \end{deluxetable}
215:
216:
217: \subsection{\mipsmu~Imaging}
218:
219: Wide-field MIPS imaging ($0.36^{\circ}\times1^{\circ}$) was taken in
220: slow scan mode, and the data processed with the MIPS Data Analysis
221: Tool \citep[DAT version 3.02;][]{gordon:05} where array-averaged
222: background subtraction was applied to improve the flat-fielding. At
223: \mipsmu, the exposure time is $\sim1000$~s~pix$^{-1}$ and the
224: 3$\sigma$ point source detection limit $\sim100\mu$Jy. We use DAOPHOT
225: II \citep{stetson:87} to detect sources and measure their fluxes, and
226: we follow the same strategy as described in \citet{papovich:04}. The
227: spectroscopically confirmed members are correlated with the
228: \mipsmu~catalog using a $2''$ search radius. Total IR luminosities
229: are determined using star-forming galaxy templates developed by
230: \citet{dale:02}, and $L_{IR}$ converted to star formation rates using
231: Eq. 3 in \citet{kennicutt:98b}.
232:
233:
234:
235: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
236: \section{Results}
237:
238: \subsection{Brightest Group Galaxies}
239:
240: Images of the brightest group galaxy (BGG) within each of the four
241: X-ray luminous groups at $z\sim0.37$ are shown in
242: Fig.~\ref{fig:tnails}. The BGGs are ordered in increasing stellar
243: mass (see \S\ref{kcorr}) and range from
244: $2.9-6.7\times10^{11}M_{\odot}$ (Table~\ref{tab:bgg}).
245:
246: \begin{figure}
247: \epsscale{1.15}
248: \plotone{f1.eps}
249: \caption{{\it Top:} VLT/VIMOS $BVR$ images ($20''\times20''$) of
250: the four brightest group galaxies ordered in
251: increasing stellar mass, $i.e.$ a rough time sequence. BGG-1 \& 2
252: both have bright, gravitationally bound companions (marked with
253: crosses). All four BGGs have extended, diffuse stellar envelopes that
254: are up to 140~kpc across (BGG-1). {\it Bottom:} ACS F814W imaging
255: of the same BGGs reveals the spectacular double nucleus in BGG-3 (image
256: inset; projected separation of 2.5~kpc).
257: \label{fig:tnails}}
258: \end{figure}
259:
260: All of the BGGs are early-type galaxies with no visual signs of
261: ongoing star formation. These BGGs are in groups with masses less
262: than that of galaxy clusters ($\sim10^{14}M_{\odot}$
263: vs. $\sim10^{15}M_{\odot}$) at $\sim1/3$ the lookback time, yet they
264: already morphologically resemble the massive early-type galaxies that
265: are common in $z\sim0$ clusters, $e.g.$ Coma \citep{davis:76}. The
266: BGGs are even comparable in stellar mass to the most massive brightest
267: {\it cluster} galaxies in the local universe
268: \citep[$M_{\ast}\sim5\times10^{11}M_{\odot}$;][]{vonderlinden:07}
269:
270: Spectroscopy uncovers the second striking result: the two lowest mass
271: BGGs are part of merging systems (Fig.~\ref{fig:tnails}). BGG-1 has a
272: companion galaxy ($D_{proj}=34$~kpc, $\delta$v=170\kms) where both
273: are embedded in a common, extended envelope that has a semi-major axis
274: of $\sim140$~kpc; this stellar envelope is especially prominent in
275: the $R$-band imaging and is reminiscent of the stellar plume
276: associated with a major merger forming a brightest cluster galaxy at
277: $z=0.39$ \citep{rines:07}. BGG-2 also has a companion galaxy
278: ($D_{proj}=48$~kpc, $\delta$v=53\kms; note
279: $\delta$v$\approx\sigma_{cz}$) within a faint, extended stellar halo.
280:
281: In the ground-based imaging (PSF$\sim0.7''$), BGG-3 looks like a
282: normal early-type galaxy but the high resolution ACS imaging ($0.05''$
283: pix$^{-1}$) reveals its spectacular double nucleus
284: (Fig.~\ref{fig:tnails}). As part of our spectroscopic survey, we
285: aligned a slit along the two nuclei and confirm that both are part of
286: the same galaxy. The nuclei are separated by only 2.5~kpc. BGG-3
287: also has a diffuse stellar halo.
288:
289: BGG-4 is the most massive galaxy in SG1120 and lies in the group with
290: the highest X-ray temperature (G05). Other than a diffuse stellar
291: halo, BGG-4 has no obvious signs of recent ($t<1$ Gyr) merging, $e.g.$
292: multiple nuclei. However, recent simulations indicate that massive
293: ellipticals ($M_{\ast}>6\times10^{11}M_{\odot}$) like BGG-4 form via
294: dissipationless merging of two spheroids \citep{naab:06}.
295:
296: Taken as a whole, these four BGGs present a time sequence where
297: luminous ($L>L^{\ast}$) galaxies continue to merge at late times, as
298: predicted by hierarchical formation \citep{peebles:70}. The merging
299: galaxies are spheroids that have no visible signs of current or recent
300: star formation, but do have extended stellar envelopes. During the
301: merging process, stars from these spheroids are gravitationally
302: disrupted and likely to contribute to the intracluster light
303: \citep{mihos:03}. The resulting merger remnant will be a massive,
304: early-type galaxy embedded in an extended, diffuse stellar envelope,
305: $e.g.$ BGG-4.
306:
307:
308: \subsection{Color-Magnitude Diagram \& Stellar Mass}
309:
310: Figure~\ref{fig:cmd} shows the rest-frame $(B-V)$ color versus
311: absolute $V$-band magnitude for the 146 supergroup members that fall
312: on the HST/ACS mosaic. Passive members (N=88; $\sim60$\%), are
313: well-fit using the color-magnitude (CM) relation measured in
314: CL~1358+62, a massive galaxy cluster at $z=0.33$ \citep[hereafter
315: vD98]{vandokkum:98a}. The CM relation is normalized to the passive
316: members where outliers have been iteratively removed using the median
317: absolute deviation. We note that $\sim60$\% of the members are E/S0
318: galaxies.
319:
320: \begin{figure}
321: \epsscale{1.0}
322: \plotone{f2.eps}
323: \caption{Rest-frame color-magnitude diagram for the 146 members
324: ($0.32\leq z\leq0.39$) on the ACS mosaic (small dots). The large
325: filled dots are passive members ([OII] EW$<5$\AA) and the large open
326: circles denote the BGGs and merging companions (connected with the
327: dotted lines). The slope of the CM relation is taken from CL~1358+62
328: \citep[$z=0.33$;][]{vandokkum:98a}, and the fit is normalized to the
329: passive members. After merging, the two faintest BGGs will increase
330: their $V$-band luminosities by $\sim0.4$ mag.
331: \label{fig:cmd}}
332: \end{figure}
333:
334:
335: All of the BGGs lie within $\Delta (B-V)>-0.1$ of the CM relation and
336: are thus red. While they all lie blue-ward of the CM relation,
337: assuming the BGGs evolve passively their $(B-V)$ colors will increase
338: such that the BGGs will be tightly clustered around the CM relation by
339: $z\sim0$. Their $(B-V)$ colors already correspond to single stellar
340: populations that are at least 3 Gyr old (solar metallicity; BC03).
341:
342: Included in Fig.~\ref{fig:cmd} and Table~\ref{tab:bgg} are
343: the companion galaxies to BGG-1 and BGG-2. Once these companions
344: merge with their BGGs via, $e.g.$ dynamical friction, the resulting
345: BGGs will be $\sim0.4$ magnitudes brighter.
346:
347: The stellar masses determined from the photometry provide further
348: evidence of BGG build-up. The companion galaxies to BGG-1 and BGG-2
349: are each $\sim40$\% of their respective BGG's mass, and the two nuclei
350: in BGG-3 have a luminosity ratio of about 1:2. Our data indicate that
351: the BGGs grow in mass by $\gtrsim40-50$\% during the most recent
352: merger. We estimate the merging timescale by setting it equal to the
353: dynamical friction timescale \citep{binney:87} and find that
354: $t_{MGR}\lesssim2$ Gyr, consistent with merging timescales measured
355: for luminous red galaxies \citep{conroy:07} and for close pairs in
356: cosmological simulations \citep{kitzbichler:08}.
357:
358:
359: %\subsection{Rest-frame $(B-V)$ Scatter}
360:
361: A powerful feature of a color-magnitude analysis is using the scatter
362: in color to determine the relative stellar ages of different galaxy
363: populations \citep{bower:98}. The $(B-V)$ scatter (RMS) for the six
364: members in the BGG systems is $0.032$; this is significantly smaller
365: than the color scatter of the passive members ($\sigma_{BV}=0.116$;
366: Fig.~\ref{fig:cmd}), $i.e.$ the BGGs are more homogeneous in stellar
367: age than the passive population.
368:
369: Because $(B-V)$ scatter decreases as stellar populations age, we also
370: can constrain the formation redshift of the BGGs. For simplicity, we
371: model the BGGs as single stellar populations (SSPs) and allow SSPs to
372: form any time up to $t$ where $t_0$ is the epoch at $z=0.37$. Using
373: Eq.~A3 in vD98, we estimate the $(B-V)$ scatter from 5000 SSPs
374: generated at random times and find that to have $\sigma_{BV}=0.032$,
375: all of the SSPs must form by $t=0.65t_0$. In the current cosmology,
376: this corresponds to all of the stars in the BGGs forming at $z>0.9$.
377:
378: Assuming the BGGs continue to evolve passively, their $(B-V)$ scatter
379: decreases to $\sigma_{BV}\sim0.020$ by $z=0$. This is even smaller
380: than the intrinsic scatter of $\sigma_{BV}=0.028\pm0.004$ measured
381: using the 104 brightest $H$-band early-types in Coma at $z=0.02$
382: \citep{eisenhardt:07}. Note that $\sim60-70$\% of the supergroup
383: galaxies are already passive and/or red (Fig.~\ref{fig:cmd}), thus
384: later merging events are unlikely to significantly change the massive
385: BGGs' mean stellar ages.
386:
387:
388: \subsection{Composite Optical Spectrum}
389:
390: Figure~\ref{fig:bgg1d} shows the composite rest-frame optical spectrum
391: of the six galaxies in the BGG systems. The composite BGG spectrum
392: has the strong H\&K absorption features associated with an old
393: ($>10^9$ Gyr) stellar population and has no emission features nor
394: Balmer absorption. \citet{jeltema:07} also find that
395: intermediate-redshift ($0.2<z<0.6$) BGGs lack [OII]$\lambda3727$
396: emission.
397:
398:
399: \begin{figure}
400: \epsscale{1.0}
401: \plotone{f3.eps}
402: \caption{Composite rest-frame spectrum of the six members making up the
403: merging BGG systems; each spectrum was normalized to its flux level
404: at $6400<\lambda<6600$\AA~before summing. The
405: high signal-to-noise spectrum shows no signs of ongoing or recent
406: star formation (within $\sim1.5$ Gyr).
407: \label{fig:bgg1d}}
408: \end{figure}
409:
410:
411: Optical emission lines are physically driven by ongoing star formation
412: while strong Balmer absorption, H$\delta$ in particular, provides a
413: fossil record of any star formation that ended within the last
414: $\sim1.5$ Gyr \citep[$e.g.$][]{couch:87}. The absence of these
415: features in the composite BGG spectrum indicates that these members
416: have not formed any stars recently, and the spectrum's $D_N(4000)$
417: index of 1.72 corresponds to a mean stellar age of $\sim1.7$ Gyr
418: (solar metallicity; BC03).
419:
420: \subsection{\mipsmu~Detections}
421:
422: Only BGG-2 \& 3 are detected at \mipsmu, and assuming the mid-infrared
423: emission is due to dusty star formation leads to rates of 6.7 and 3.6
424: $M_{\odot}$~yr$^{-1}$, respectively. However, dusty star-forming
425: red-sequence galaxies detected at \mipsmu~tend to show weak
426: [OII]$\lambda3727$ emission in their optical spectra \citep{wolf:05},
427: and neither BGG-2 nor 3 do (Fig.~\ref{fig:bgg1d}). While
428: \mipsmu~emission is often attributed to ongoing star formation, it can
429: also be due to an active galactic nucleus (AGN). BGG-2 is an X-ray
430: point source, thus both its X-ray and \mipsmu~emission are likely due
431: to an AGN. BGG-3's weaker \mipsmu~detection makes it difficult to
432: discriminate between a faint AGN and star formation.
433:
434: BGG-2's multi-wavelength properties are consistent with an old stellar
435: population ($t_{\ast}\gtrsim2$ Gyr) and AGN. Current models favor the
436: formation of luminous red galaxies via merging of progenitors with
437: supermassive black holes; the gravitational perturbations bring fresh
438: gas to the small central region and trigger an AGN phase
439: \citep[$e.g.$][]{springel:05a}. BGG-2, and possibly BGG-3, are
440: compatible with such a model.
441:
442:
443: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
444: \section{Discussion \& Conclusions}
445:
446: These four BGG snapshots provide unprecedented observational evidence
447: for the late ($z<0.5$) stellar assembly of massive spheroidal galaxies
448: in clusters via dissipationless merging of massive progenitors. Two
449: of the four BGGs are part of merging pairs and another has a double
450: nucleus with a separation of only 2.5~kpc. When ordered by
451: increasing stellar mass ($2.9-6.7\times10^{11}M_{\odot}$), the BGGs
452: present a time sequence where galaxy-galaxy merging increases the BGG
453: mass by $\gtrsim40$\% ($t_{MGR}\lesssim2$ Gyr) and $V$-band luminosity
454: by $\sim0.4$ mag. Our results are remarkably consistent with recent
455: claims from semi-analytic models that brightest cluster galaxies tend
456: to assemble at $z<0.5$ \citep{delucia:07b}; such a model also naturally
457: explains the observed number of brightest cluster galaxies in the
458: local universe with multiple nuclei
459: \citep[$\sim20-40$\%;][]{hoessel:85,laine:03}.
460:
461: Despite the merger activity, the BGGs are dominated by old ($>10^9$
462: Gyr) stellar populations and show little sign of ongoing or recent
463: star formation. The BGGs have rest-frame $(B-V)$ colors corresponding
464: to mean stellar ages of $>3$ Gyr, and their tight scatter in color
465: ($\sigma_{BV}=0.032$) confirms that the BGGs formed their stars at
466: $z>0.9$. If the BGGs continue to age passively, their color scatter
467: by $z\sim0$ will be even smaller than that of the $\sim100$ brightest
468: early-type galaxies in Coma ($\sigma_{BV}=0.020$ vs. 0.028). Their
469: composite optical spectrum shows no sign of star formation in the last
470: $\sim1.5$ Gyr. While BGG-2 \& 3 are weakly detected at \mipsmu, the
471: X-ray and optical data indicate that BGG-2's \mipsmu~emission is due
472: to AGN rather than star formation, and possibly for BGG-3 as well.
473:
474: SG1120 thus provides unique and powerful confirmation of hierarchical
475: formation on both galaxy and cluster scales because 1) SG1120 has
476: massive galaxies assembling from galaxy-galaxy merging, and 2) it is a
477: proto-cluster assembling from the merging of galaxy groups. Our
478: findings show that the most massive galaxies in groups and clusters
479: form via dissipationless merging, and that the galaxy group
480: environment is critical to forming the massive galaxies dominating
481: local clusters.
482:
483:
484:
485: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
486:
487: \acknowledgments
488:
489: K.T. acknowledges support from the Swiss National Science Foundation
490: (grant PP002-110576) and thanks J. Blakeslee for help during the
491: initial ACS reduction. J.M. acknowledges support from
492: NASA-06-GALEX06-0030 and Spitzer G05-AR-50443, and L.B. from NASA
493: Spitzer programs through JPL subcontracts \#1255094 and \#1256318.
494: Support was also provided by NASA HST G0-10499, JPL/Caltech SST
495: GO-20683, and Chandra GO2-3183X3.
496:
497:
498: \bibliographystyle{/Users/vy/aastex/apj}
499: \bibliography{/Users/vy/aastex/tran}
500:
501:
502:
503: \end{document}
504:
505:
506: