1: \documentclass{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: %\usepackage{emulateapj5}
4:
5:
6: \usepackage{epsfig}
7: \usepackage{graphicx}
8: \def\gta{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}
9:
10:
11:
12: \shorttitle{RCS}
13: \shortauthors{Krick et al.}
14:
15:
16: \begin{document}
17: \newcommand\msun{\hbox{M$_{\odot}$}}
18: \newcommand\lsun{\hbox{L$_{\odot}$}}
19: \newcommand\magarc{mag arcsec$^{-2}$}
20: \newcommand\h{$h_{70}^{-1}$}
21:
22: %\bibliographystyle{apj}
23: \title{\bf Galaxy Clusters in the IRAC Dark Field I: \\ Growth of the red sequence}
24:
25: \author{J.E.~Krick \altaffilmark{1}, J.A.~Surace \altaffilmark{1}, D.~Thompson \altaffilmark{2}, M.L.N.~Ashby \altaffilmark{3}, J.L.~Hora \altaffilmark{3}, V.~Gorjian \altaffilmark{4}, and L.~Yan \altaffilmark{1}}
26:
27: \altaffiltext{1}{Spitzer Science Center, MS 220--6,
28: California Institute of Technology, Jet Propulsion Laboratory,
29: Pasadena, CA 91125, USA}
30: \altaffiltext {2}{Large Binocular Telescope Observatory, University of Arizona, Tucson, AZ 85721, USA}
31: \altaffiltext {3}{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge MA 02138, USA}
32: \altaffiltext {4}{Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, 91109, USA}
33: \email{jkrick@caltech.edu}
34:
35:
36:
37: \begin{abstract}
38:
39: Using three newly identified galaxy clusters at $z\sim1$
40: (photometric redshift) we measure the evolution of the galaxies
41: within clusters from high redshift to the present day by studying
42: the growth of the red cluster sequence. The clusters are located in
43: the Spitzer Infrared Array Camera (IRAC) Dark Field, an extremely
44: deep mid-infrared survey near the north ecliptic pole with
45: photometry in 18 total bands from X-ray through far-IR. Two of the
46: candidate clusters are additionally detected as extended emission in
47: matching Chandra data in the survey area allowing us to measure
48: their masses to be $M_{500}= 6.2 \pm 1.0 \times 10^{13}$ and $3.6
49: \pm 1.1 \times 10^{13}$ \msun. For all three clusters we create a
50: composite color magnitude diagram in rest-frame B-K using our deep
51: HST and Spitzer imaging. By comparing the fraction of low
52: luminosity member galaxies on the composite red sequence with the
53: corresponding population in local clusters at $z=0.1$ taken from the
54: COSMOS survey, we examine the effect of a galaxy's mass on its
55: evolution. We find a deficit of faint galaxies on the red sequence
56: in our $z\sim1$ clusters which implies that more massive galaxies
57: have evolved in clusters faster than less massive galaxies, and that
58: the less massive galaxies are still forming stars in clusters such
59: that they have not yet settled onto the red sequence.
60:
61:
62: \end{abstract}
63:
64: \keywords{galaxies: clusters --- galaxies: evolution --- galaxies:
65: photometry --- cosmology: observations}
66:
67:
68:
69:
70: \section{Introduction}
71:
72: The redshift range from $z=1$ to the present day is a particularly
73: dynamic epoch in the history of groups and clusters as evidenced by
74: the evolution of the morphology-density relation and increasing
75: fraction of blue galaxies with increasing redshift \citep{capak2007,
76: butcher1984}. Interestingly, cluster ellipticals at $z\sim1$ already
77: have a narrow distribution of red colors \citep[the red cluster
78: sequence (RCS);][]{blakeslee2003,vandokkum2001}. There is some debate
79: about the mechanism by which these cluster galaxies arrive onto the
80: red sequence. It is difficult to distinguish whether these red
81: ellipticals all formed their stars and did their merging at $z>3$,
82: then stopped forming stars when they entered the cluster environment
83: \citep{ford2004}; or if they are the product of the merging of
84: gas-poor systems which do not produce star formation
85: \citep{vandokkum2005}.
86:
87: We investigate whether the red population is still in the process of
88: forming at $z=1$ or if indeed assembly has already finished at higher
89: redshift by studying the presence of the faint end of the RCS at $z=1$
90: and comparing it to the present epoch. We measure the ratio of faint
91: to bright RCS galaxies in a sample of three $z\sim1$ clusters from the
92: IRAC Dark Field (described below). These clusters have the benefit of
93: extremely deep $3.6\,\micron$ data which allows us to study the faint
94: end of the luminosity function at rest frame near-IR, which traces the
95: peak of the spectral energy distribution in galaxies. A confirmed
96: deficit of faint galaxies on the RCS would imply that more massive
97: galaxies have evolved in clusters faster than less massive galaxies.
98: A constant fraction of faint red galaxies between $z=1$ and the
99: present would require a formation mechanism where galaxies of all
100: masses have already joined the red sequence at redshifts higher than
101: one. There is evidence that the faint end of the RCS is not
102: completely in place by z=0.8 \citep[][and references
103: therein]{delucia2007,koyama2007}, although at least some clusters at
104: these redshifts appear to have complete RCSs to M*+3.5
105: \citep{andreon2006}.
106:
107: These questions are ideally addressed with deep infrared surveys of
108: clusters at high redshift. In the last four years deep and wide area
109: surveys in the mid-infrared using the {\it Spitzer Space Telescope}
110: have substantially opened a new window on galaxy and star formation at
111: $0<z<3$. {\it Spitzer} now routinely produces imaging of
112: large fields to higher resolution and fainter depths than previously
113: possible. IRAC, the mid-infrared camera on-board Spitzer, takes
114: images at 3.6, 4.5, 5.8, and 8.0\,\micron \citep{fazio2004}. The
115: shorter wavelengths provide a direct measurement of the stellar
116: content of galaxies at redshifts as high as three. The longer
117: wavelength channels sample emission from polycyclic aromatic
118: hydrocarbons (PAHs) in low redshift galaxies, as well as direct
119: thermal emission from hot dust. The deepest such survey is the dark
120: current calibration field for the mid-infrared camera, commonly known
121: as the ``IRAC Dark Field''. These deep mid-IR data are supplemented
122: by additional 14 band photometry including Palomar u',g',r',i', {\it
123: HST/ACS} F814W, MMT $z^\prime$, Palomar J, H,\& K , Akari 11 \&
124: $15\,\micron$, Spitzer MIPS 24 \& $70\,\micron$, and Chandra ACIS-I
125: imaging. This extensive, multiwavelength dataset gives us the distinctly unique
126: opportunity to study the development of the red sequence in galaxy
127: clusters at redshift one. The second paper on galaxy clusters in this
128: series will discuss the role of star formation in the evolution of
129: clusters at $z=1$ by examining the {\it Spitzer} 24\,\micron \ data in
130: conjunction with the morphological information from the HST ACS
131: dataset.
132:
133:
134: This paper is structured in the following manner. In
135: \S\ref{observations} \& \S\ref{photz} we discuss the data and derived
136: photometric redshift determination. Details of the cluster search and
137: cluster properties are presented in \S\ref{search}. In \S\ref{cmds}
138: \& \S\ref{results} we present the color magnitude diagrams of the
139: candidate clusters and results of the faint-to bright ratios of red
140: sequence galaxies. In \S\ref{discuss} we discuss the implications for
141: the evolution of cluster galaxies. Throughout this paper we use
142: $H_0=70$km/s/Mpc, $\Omega_M$ = 0.3, $\Omega_\Lambda$ = 0.7. With this
143: cosmology, the luminosity distance at z=1 is 6607 Mpc, but the angular
144: diameter distance is a factor of $(1+z)^2$ less, or 1652 Mpc. All
145: photometry is quoted in the AB magnitude system.
146:
147:
148:
149: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%5
150:
151:
152:
153: \section{Observations \& Data Reduction}
154: \label{observations}
155:
156: \subsection{The IRAC Dark Field}
157:
158: The survey region is the IRAC Dark Field, centered at approximately
159: 17h40m +69d. The field is located a few degrees from the north
160: ecliptic pole (NEP) in a region which is darker than the actual pole
161: and is in the Spitzer continuous viewing zone so that it can be
162: observed any time IRAC is powered on for observing. These observing
163: periods are called instrument ``campaigns'', and occur roughly once
164: every three to four weeks and last for about a week. Sets of long
165: exposure frames are taken on the Dark Field at least twice during each
166: campaign totaling roughly four hours of integration time per campaign,
167: and these data are used to derive dark current/bias frames for each
168: channel. The dark frames are used by the pipeline in a manner similar
169: to ``median sky'' calibrations as taken in ground-based near-infrared
170: observing to produce the Basic Calibrated Data (BCD) for all science
171: observations. Each set of dark calibration observations collects
172: roughly two hours of integration time at the longest exposure times in
173: each channel.
174:
175:
176: The resulting observations are unique in several ways. The Dark Field
177: lies near the lowest possible region of zodiacal background, the
178: primary contributor to the infrared background at these wavelengths,
179: and as such is in the region where the greatest sensitivity can be
180: achieved in the least amount of time. The area was also chosen
181: specifically to be free of bright stars and very extended galaxies,
182: which allows clean imaging to very great depth. The observations are
183: done at many position angles (which are a function of time of
184: observation) leading to a more uniform final {\sc psf}. Finally,
185: because the calibration data are taken directly after anneals, they
186: are more free of artifacts than ordinary guest observer (GO)
187: observations. Over the course of the mission, the observations have
188: filled in roughly uniformly a region 20\arcmin \ in diameter. This
189: has created the deepest ever mid-IR survey, exceeding the depth of the
190: deepest planned regular Spitzer surveys over several times their area.
191: Furthermore, this is the only field for which a 5+year baseline of
192: mid-IR periodic observations is expected.
193:
194:
195: The IRAC data are complemented by imaging data in 14 other bands with
196: facilities including Palomar, MMT, HST, Akari, Spitzer MIPS, and
197: Chandra ACIS-I. Although the entire dark field is $> 20\arcmin$ in
198: diameter, because of spacecraft dynamics the central $\sim 15\arcmin$
199: is significantly deeper and freer of artifacts. Therefore, it is this
200: area which we have matched with the additional observations. The
201: entire dataset will be presented in detail in a future paper (Krick et
202: al, in prep). For completeness we briefly discuss here the Spitzer
203: IRAC, HST ACS, and Chandra ACIS datasets as they are the most critical
204: to this work. All space-based datasets are publicly available through
205: their respective archives.
206:
207: \subsection{Spitzer/IRAC}
208: \label{irac}
209:
210: This work is based on a preliminary combination of 75 hours of IRAC
211: imaging, which is $\approx$30\% of the expected depth not including a
212: possible warm mission. Even these 75 hours go well into the confusion
213: limit of IRAC and so the additional exposure time will not add a lot
214: of sensitivity. The Basic Calibrated Data (detector image level)
215: product produced by the Spitzer Science Center was further reduced
216: using a modified version of the pipeline developed for the SWIRE
217: survey \citep{surace2005}. This pipeline primarily corrects image
218: artifacts and forces the images onto a constant background
219: (necessitated by the continuously changing zodiacal background as seen
220: from Spitzer). The data were coadded onto a regularized 0.6\arcsec \
221: grid using the MOPEX software developed by the Spitzer Science Center.
222:
223: Experiments with DAOPHOT demonstrate that nearly all extragalactic
224: sources are marginally resolved by IRAC, particularly at the shorter
225: wavelengths, and hence point source fitting is inappropriate.
226: Instead, photometry is done using the high spatial resolution ACS data
227: as priors for determining the appropriate aperture shape for
228: extracting the Spitzer data. We do this by first running source
229: detection and photometric extraction on the coadded IRAC images using
230: a matched filter algorithm with image backgrounds determined using the
231: mesh background estimator in SExtractor (Bertin et al. 1995) . This
232: catalog is merged with the HST ACS catalog. For every object in that
233: catalog ,if the object is detected in ACS then we use the ACS shape
234: parameters to determine the elliptical aperture size for the IRAC
235: images. ACS shape parameters are determined by SExtractor on
236: isophotal object profiles after deblending, such that each ACS pixel
237: can only be assigned to one object (or the background). For objects
238: which are not detected in ACS, but which are detected in IRAC, we
239: simply use the original IRAC SExtractor photometry. Because of the
240: larger IRAC beam, we impose a minimum semi-major axis radius of
241: 2\arcsec. In all cases aperture corrections are computed individually
242: from PSF's provided by the Spitzer Science Center based on the aperture
243: sizes and shapes used for photometry.
244:
245: Final aperture photometry was performed using custom extraction
246: software written in IDL and based on the APER and MASK$\_$ELLIPSE
247: routines with the shape information from SExtractor, from either ACS
248: or IRAC as described above, using local backgrounds. Because we use
249: local backgrounds, the measured fluxes of objects near the confusion
250: limit should have a larger scatter than those non-confused objects,
251: but will on average be the correct flux. This will have the effect of
252: adding scatter to the CMD described below, but will not cause trends
253: of movement for the faint objects in color space. Additionally, this
254: will not effect the photometric redshifts, as it will likely shift all
255: IRAC points up or down, but not relative to each other. The final
256: $95\%$ completeness limit at 3.6\,\micron\ is 0.2\,$\mu$Jy or 25.7 AB
257: magnitude as calculated from a number count diagram by measuring where
258: the number of observed objects drops below 95\% of the expected number
259: of objects, where the expected numbers are calculated by fitting a
260: straight line to the brighter flux number counts.
261:
262:
263: \subsection{HST/ACS}
264: \label{acs}
265: The {\it HST} observations consist of 50 orbits with the ACS
266: comprising 25 separate pointings, all with the F814W filter (observed
267: I-band). Within each pointing eight dithered images were taken for
268: cosmic ray rejection and to cover the gap between the two ACS CCDs.
269: The ACS pipeline {\it calacs} was used for basic reduction of the
270: images. Special attention was paid to bias subtraction, image
271: registration, and mosaicing. Pipeline bias subtraction was
272: insufficient because it does not measure the bias level individually
273: from each of the four amplifiers used by ACS. We make this correction
274: ourselves by subtracting the mean value of the best fit Gaussian to
275: the background distribution in each quadrant. Due to distortions in
276: the images, registration and mosaicing was performed with a
277: combination of IRAF's {\it tweakshifts, multidrizzle,} and {\it SWarp
278: v.2.16.0} from Terapix. The actual task of mosaicing the final
279: image was complicated by the large image sizes. The single combined
280: mosaic image is 1.7GB and reading in all 200 images (160Mb each) for
281: combination is impossible for most software packages.
282:
283:
284: The final combined ACS image is $\sim 15\arcmin$ diameter coincident
285: with the deepest part of the IRAC Dark Field and is made with the
286: native $0.05\arcsec$ per pixel resolution. Photometry was performed
287: in a standard manner with SExtractor. The $3\sigma$ detection limit
288: for point sources is $I = 28.6(AB)$. The area common to both IRAC and
289: ACS contains $\sim 51,000$ detected sources.
290:
291: All cluster galaxies are detected in the mid-infrared data, and the ACS
292: data are used as priors for extraction of the mid-IR photometry.
293: Because the cluster galaxies are detected in the optical ACS images, and
294: cluster membership is derived specifically based on the optical data,
295: it is the completeness of the optical data thats sets the fundamental
296: detection limits. Thus our ability to examine the faint end of the
297: cluster (rest frame) K-band luminosity function is ultimately limited
298: by the optical data, not the infrared data.
299:
300:
301: \subsection{Chandra/ACIS-I}
302: \label{chandra}
303:
304: The field-of-view of the Chandra ACIS-I is 17\arcmin \ (for the
305: central four chips) , well-matched to the deepest central area of the
306: Dark Field. As a result, only a single pointing was required. The
307: 100ks observation was broken into three separate observations at
308: different pointing angles to reduce the effect of the gap between
309: chips. The data was reduced using the newest version of the Chandra
310: Interactive Analysis of Observations software (CIAO 4.0). Specific
311: attention was paid to destreaking, bad pixels, and background flares.
312: The task {\it merge\_all} was then used to combine event files from
313: the three observations into a final event file and image. The final
314: combined Chandra image has $0.5\arcsec$ per pixel resolution. Blind
315: pointing is expected to be 1\arcsec. Aperture photometry for extended
316: sources is done manually (see \S\ref{xray}).
317: %---------------------------------------------------------------------------------
318:
319: \section{Photometric redshifts}
320: \label{photz}
321:
322: The combined IRAC and ACS catalog contains over $50,000$ objects which
323: makes acquisition of spectroscopic redshifts impractical. Even
324: confirmation spectroscopy of red galaxies at $z=1$ in our three
325: candidate clusters will require multiple nights on 8-10m class telescopes.
326: In lieu of spectroscopy we use our extensive multi-wavelength,
327: broad-band catalog to build spectral energy distributions (SEDs) and
328: derive photometric redshifts. These SEDs are fit with template
329: spectra derived from galaxies in the Spitzer wide area infrared survey
330: survey \citep[SWIRE;][]{Polletta2007}. Since the SWIRE templates are
331: based on Spitzer observations we find them the best choice to use as
332: models for this dataset. Photometric redshifts are calculated using
333: {\it Hyperz}; a chi-squared minimization fitting program including a
334: correction for Galactic reddening \citep{bolzonella2000,
335: calzetti2000}. We emphasize that in this work photometric redshifts
336: are used only as a first step to find cluster candidates, and not to
337: determine membership within the clusters.
338:
339: %---------------------------------------------------------------------------------
340: \section{Cluster Search}
341: \label{search}
342:
343:
344: The mid-IR wavelengths of IRAC are well suited to find galaxies at
345: redshift one because the stellar peak of the spectral
346: energy distribution has red-shifted into those bands. We exploit this
347: fact in a search for clusters.
348:
349: The first step in the search is to combine the spatial information
350: from the 2-D images with photometric redshifts to visually identify
351: clusters of galaxies. We do this by plotting the locations of all
352: galaxies in a certain photometric redshift range on both the optical
353: and IR images to look for clusterings. We step through redshift space
354: in overlapping z=0.2 bins examining each for clusterings. The result
355: of this search is 14 cluster candidates at $0.8 < z < 1.4$ with
356: average areal densities of 10.3 galaxies with similar redshifts per
357: square arcminute. Next, for each candidate cluster, we examine the
358: redshift distributions of all galaxies within 0.013\degr of the center
359: ($\sim 375$ Kpc at these redshifts). We choose this radius, at around
360: one third of the virial radius, as the size at which the clustering
361: signal appears to be strongest. The candidate cluster redshift
362: distributions are compared to the average redshift distributions of 50
363: regions of the same area randomly distributed across the field. After
364: this comparison, three excellent candidate clusters remain with peaks
365: in their redshift distributions which are greater than $2\sigma$ above
366: that of the comparison fields (see Figure \ref{fig:zdist} \& Table
367: \ref{tab:clusterchar}). In addition the three best candidate clusters
368: show clear over-densities in both the ACS and IRAC images (see Figure
369: \ref{fig:clus1}). While we cannot guarantee that this search is
370: complete without extensive spectroscopy, we are confident that we have
371: found the more massive clusters at redshift one.
372:
373: In a shallow, large-area IRAC survey \citet{eisenhardt2008} finds
374: roughly two high-z clusters in an area equivalent to the IRAC dark
375: field. The DEEP2 survey finds seven spectroscopically confirmed groups
376: and clusters with $0.75 < z < 1.03$ and an upper mass limit of
377: $1\times10^{13} \msun$ in a similar area to our survey
378: \citep{fang2007}. Our finding of three candidate clusters is in agreement
379: with the number of clusters in these other surveys.
380:
381:
382: Two of the candidate clusters have centers within $1.8\arcmin$ of each
383: other. If clusters were randomly distributed, from Monte Carlo
384: simulations we would expect to find a close pair like this one 2-50\%
385: of the time depending on the areal density (here using the IRAC
386: shallow survey and DEEP2 respectively). However, from both
387: hierarchical simulations and cluster surveys we do not expect clusters
388: to be randomly distributed, instead clusters are connected by
389: filaments and are highly correlated even at redshift
390: one\citep{brodwin2007}. If our two clusters are truly close in all
391: three dimensions then they must lie along a filament. Our photometric
392: redshifts do not allow for the precision needed to know if they indeed
393: are 3-D neighbors.
394:
395:
396:
397: \subsection{X-ray Properties}
398: \label{xray}
399:
400: Two of the three candidate clusters show extended emission in our
401: 100ks Chandra ACIS-I image. This both confirms that they are bound
402: clusters because at these detected luminosities they can only be
403: clusters, and allows us to determine their masses (see Figure
404: \ref{fig:chandra}). We are also able to place a limit on the mass of
405: the third, Chandra undetected, cluster candidate.
406:
407: X-ray luminosities are obtained by doing aperture photometry on the
408: merged ACIS-I image with the CIAO task {\it dmextract}. Aperture
409: sizes were chosen to be 0.9\arcmin (0.4Mpc) which fully encompasses
410: all of the excess flux coming from the clusters but avoids neighboring
411: point sources on the Chandra image. Point sources prevent us from
412: using background regions directly adjacent to the clusters, so instead
413: we use the average of 20 background regions taken from empty regions
414: all over the frame. Net counts are 172 $\pm 31$ and 72 $\pm 29$
415: (Poisson errors) for the clusters respectively. It is possible that
416: the brighter cluster (on the right in Figure \ref{fig:chandra}) is
417: contaminated by a point source which is effecting this count rate,
418: however with such low number of photons we have not attempted to
419: separate this possible point source from the cluster and background.
420:
421: Taking the count rate from the image, we use the Portable Interactive
422: Multi-Mission Simulator (PIMMS v 3.9d) to derive a flux assuming a
423: thermal bremsstrahlung model with $4\times10^{20}$cm$^{-2}$ of Galactic
424: NH (measured for the location of this field), and a temperature of
425: 5KeV. This temperature is chosen randomly and is calculated more
426: accurately in the following iterative process. Once we have
427: calculated a flux and luminosity of the clusters, we use a combination
428: of the \citet{maughan2006} and \citet{vikhlinin2002} $L_x - T $
429: relation based on Chandra and XMM data on 22 clusters with $z > 0.4$
430: to derive a more realistic temperature. We then go back and use this
431: new temperature to re-calculate the X-ray flux and luminosity. The
432: two clusters with detections have luminosities of $3.6\pm 0.6\times
433: 10^{43}$ and $1.6 \pm0.7 \times 10^{43}$ erg/s in the 0.5-2 Kev band.
434:
435: \subsection{Mass}
436: \label{mass}
437:
438:
439: To derive the mass from the X-ray luminosity for the two detected
440: clusters, we use the \citet{maughan2007} $L_x - M_{500}$ relation
441: based on 34 clusters with $0.5<z<1.3$. Derived masses are $M_{500}=
442: 6.2 \pm 1.4 \times 10^{13}$ and $3.6 \pm 1.4 \times 10^{13}$ \msun.
443: Since the second cluster is a very low count detection, we take this
444: mass to be the upper limit to the mass of the third, Chandra
445: undetected, cluster. Quoted errors only represent
446: error in the measurements, including uncertainties in redshift, but do
447: not take into account error in the models and should therefore be
448: taken as optimistic.
449:
450: There is some evidence that using X-ray luminosities to estimate mass
451: is unreliable at high redshift in the sense that higher redshift
452: clusters have lower x-ray luminosities for a given mass
453: \citep{lubin2004, fang2007}. The DEEP2 survey of the aforementioned
454: seven clusters at $0.75< z< 1.0 $ detect none of their clusters in a
455: 200ks Chandra image \citep{fang2007}. A possible reason for this is
456: that clusters are not virialized at higher redshifts, an assumption
457: which is necessary to use the cluster hot gas to measure mass.
458: Alternatively, \citet{andreon2008} have proposed that there are no
459: underluminous X-ray clusters and that previous work has either not
460: properly measured cluster mass via velocity dispersion, or have not
461: correctly interpreted their X-ray observations. If high-z clusters
462: are underluminous in X-rays then there are two relevant implications:
463: 1) our high-z clusters have true masses which are higher than measured
464: from their Chandra luminosities, and 2) this leaves open the
465: possibility that the Chandra undetected cluster is a true cluster with
466: mass larger than the limit afforded by the current non-detection. The
467: resolution of this issue is beyond the scope of this paper.
468:
469: As a check of cluster mass we have considered using the SDSS relation
470: of optical richness to the weak lensing mass of clusters at low
471: redshifts \citep{johnston2007}. This relies on multiple assumptions
472: and relations such as that between the number of red sequence galaxies
473: inside of $1h^{-1}$Mpc with luminosities greater than $0.4L^*$ and
474: $r_{200}$, the radius at which the cluster is 200 times the critical
475: density of the universe \citep{hansen2007} . Furthermore one has to
476: then use the derived value of $r_{200}$ with the weak-lensing relation
477: to arrive at $M_{200}$. To our knowledge none of these relations have
478: been tested at high redshifts and we therefore choose not to make this
479: calculation.
480:
481: The measured mass of these clusters is consistent with the masses of
482: the, to date, roughly 35 published groups and clusters with confirmed
483: redshifts above 0.9 \citep{stanford1997, ebeling2001, stanford2001,
484: stanford2002,eblanton2003, rosati2004, margoniner2005, mullis2005,
485: siemiginowska2005,elston2006, stanford2006, fang2007,
486: finoguenov2007, eisenhardt2008}. About two thirds of these clusters
487: are detected in multi-wavelength surveys such as DEEP2, COSMOS, and
488: the IRAC shallow survey and have an average mass of
489: $3.8\times10^{13}\msun$. The rest are detected in serendipitous X-ray
490: imaging or are targeted for their location around radio galaxies.
491: These other clusters all have higher masses ($>1E14\msun$) which is
492: consistent with it being easier to detect high-mass clusters in the
493: X-ray and high mass clusters around radio galaxies. Overall we expect
494: from hierarchical formation that clusters at $z=1$ should be less
495: massive than clusters at $z=0$.
496: %---------------------------------------------------------------------------------
497:
498:
499: \section{Color Magnitude Diagrams}
500: \label{cmds}
501:
502: Most galaxies in clusters are ellipticals that have approximately the
503: same red color regardless of magnitude and therefore form a red
504: cluster sequence (RCS) in a color magnitude diagram (CMD), as long as
505: the color is chosen to span the 4000 \AA\ break. Past studies with
506: HST have shown that the RCS is in place at least by redshift one, if
507: not before \citep{blakeslee2003,vandokkum2001}. We choose to make
508: CMDs of the dark field clusters with ACS F814W and Spitzer
509: $3.6\,\micron$. At $z\sim1$ this corresponds to rest-frame B-K which
510: does span the Balmer break. We generate a composite color magnitude
511: diagram for all three clusters including all galaxies within
512: 0.017\degr ($\sim 500$ Kpc) of the candidate cluster centers (Figure
513: \ref{fig:cmd}). The upper axis is plotted as absolute K-band
514: magnitude in order to compare to a low redshift sample. Because
515: $3.6\,\micron$ at $z=1$ corresponds to K-band at $z\simeq0.1$, we do not
516: have to apply a K-correction to the data, which avoids a large set of
517: uncertainties. We do make a correction for the redshifting of the
518: bandpass and a correction for the luminosity evolution of galaxies
519: over time taken from the stellar evolution models of
520: \citet{bruzual2003}. The color of RCS is consistent with the
521: predicted colors from simple stellar populations with a single burst
522: of star formation above z=3 and a Salpeter initial mass function
523: \citep{bruzual2003}.
524:
525:
526: Without benefit of spectroscopy we determine membership based on
527: location in the CMD. Pinpointing the location of the RCS becomes
528: easier when using a composite CMD from three clusters. To determine
529: the location of the RCS, the bright end of the composite red sequence
530: ($[3.6] < 22.5$) is fit with a biweight function \citep{beers1990}
531: where the slope is fixed to -0.12 (to match the low redshift slope
532: below). The faint end is left out of the fit because of the larger
533: amount of contamination by field galaxies. Members are taken to be
534: those galaxies which lie within $\pm 0.5$ mag of the RCS. We determine
535: the significance of the existence of a red sequence in our CMD by
536: doing a monte carlo calculation with 100 realizations of the CMDs of
537: randomly selected galaxies in the dataset. From this calculation we
538: find that the number of galaxies on the true RCS is significant
539: compared to random galaxies at the $6.7\sigma$ level. The histogram
540: of member galaxies in the composite cluster is shown in Figure
541: \ref{fig:cmd}.
542:
543:
544: Although the RCS is mainly composed of cluster galaxies at a
545: particular redshift, there will be some fore- and back-ground galaxies
546: that will contaminate the membership count. We statistically subtract
547: fore-and back-ground galaxies within the measured RCS. The number of
548: contaminating galaxies is calculated by counting the average number of
549: galaxies from 50 random regions of the same size as the clusters which
550: have colors consistent with the measured RCS. The average number of
551: contaminating galaxies within a cluster area (0.017\degr) and their
552: standard deviation are $40.3\pm7.3$ and $16.9\pm8.0$ respectively for
553: the magnitude bins used below ( $-20 < M_k < -18$ and $-22 < M_k <
554: -20$). After this subtraction the number of member galaxies remaining
555: in each cluster is a few to none in the faint bin, and about twenty
556: each in the bright bin.
557:
558:
559:
560: We create CMDs for a comparison sample of low redshift clusters using
561: the same process on publically available B- and K- band data from the
562: Cosmic Evolution Survey \citep[COSMOS;][]{capak2007}. From COSMOS we
563: use four X-ray confirmed clusters with average $0.09 < z < 0.12$
564: \citep[][Id's 42, 58, 113, 140]{finoguenov2007}. These clusters have
565: an average mass of $M_{500} = 1.1\times 10^{13} \msun$ as determined
566: from the X-ray temperature. Figure \ref{fig:cosmoscmd} shows both the
567: CMD of the composite COSMOS data from four clusters and the K-band
568: distribution of RCS member galaxies including a statistical correction
569: for fore- and back-ground galaxies taken from neighboring COSMOS
570: regions.
571:
572: This comparison sample is not ideal since these low redshift clusters
573: are roughly the same mass as the high-z clusters, and we expect
574: clusters to hierarchically gain mass over time through the infall of
575: other groups and clusters. Unfortunately, to our knowledge, there is
576: no sufficiently deep near-IR imaging on local $\sim 10^\{14\}$\msun
577: clusters. While it is possible that more massive clusters will have
578: fewer faint red galaxies than a less massive cluster due to merging,
579: we do not expect this to be a large effect. Deep, wide-field, near-IR
580: imaging of local clusters is required to study this effect in detail.
581:
582: The CMD's from the two samples have clear differences. Despite our
583: ultradeep data, the distribution of the $z=0.1$ galaxies extends to
584: fainter magnitudes than that for the clusters at $z=1$ (thanks to the
585: approximately four magnitudes of surface brightness extinction). We
586: plot the deeper low redshift data to $M_K =-14$, whereas for the high
587: redshift dark field clusters we can only plot a smaller range of
588: magnitudes extending to $M_k=-18$. Everything fainter than that is
589: below our $5\sigma$ detection threshold, so we are unable to compare
590: that magnitude range with the lower redshift clusters.
591: %---------------------------------------------------------------------------------
592:
593: \section{Results}
594: \label{results}
595:
596: We investigate the evolution of the fraction of faint galaxies on the
597: red sequence from redshift one to the present. To make the comparison
598: between the two different redshift samples, we choose a rest-frame
599: K-band absolute magnitude range which is both brighter than the $5
600: \sigma$ detection threshold for both datasets, and not too far towards
601: the bright end of the luminosity function that extremely small number
602: statistics would effect the measurement. We divide the resulting
603: magnitude range into two bins which are two magnitudes wide; $-20 <
604: M_k < -18$ (faint bin), and $-22 < M_k < -20$ (bright bin).
605:
606: The quantity we want to measure is the amount of faint galaxies on the
607: RCS at $z=1$ compared to the present epoch as an evolutionary measure
608: of the build-up of the red sequence. This is calculated as the ratio
609: of faint RCS galaxies to bright RCS galaxies. This faint-to-bright
610: ratio is shown explicitly for both cluster samples in Figure
611: \ref{fig:ratio} and can also be seen by comparing Figures
612: \ref{fig:cmd} \& \ref{fig:cosmoscmd}. We find that the relative
613: number of faint galaxies on average in clusters at redshift one is
614: less than in the average cluster at redshift 0.1 at the $3\sigma$
615: level.
616:
617:
618: The major sources of error in this work are the background measurement
619: of field galaxies, the area over which cluster membership is
620: determined, the possibility that the third cluster is a chance
621: projection, and the size of the magnitude bins. Recall that
622: completeness in the IRAC data is not a limiting factor to this
623: measurement because we use ACS source locations and shapes as a prior
624: for doing IRAC photometry. To quantify the error originating from the
625: field galaxy subtraction, error bars on Figure \ref{fig:ratio} show
626: one standard deviation on the background measurement propagated to the
627: ratio of faint to bright number of galaxies.
628:
629: To estimate the effect of changing the area over which cluster members
630: are counted we re-calculate the faint-to-bright ratio using a radius
631: which is 50\% larger than the original radius (2.25 times the area).
632: The original radius was chosen as a compromise between the relative
633: physical sizes of the implied area around the low redshift and high
634: redshift clusters. For reference the original 500\,Kpc radius circle
635: is shown on the IRAC images in Figure \ref{fig:clus1}. The new,
636: larger radius, ratios are shown as stars in Figure \ref{fig:ratio}.
637: The effect of changing radius is negligible.
638:
639: We examine the possibility that the third cluster is a chance
640: projection, not a cluster, and therefore contaminating this
641: measurement. Spectroscopy is the best way to confirm that it is a
642: cluster. Without spectroscopy, we re-calculate the ratios without the
643: third cluster, using just the two X-ray detected clusters. The same
644: trend of the bright galaxies far outnumbering the faint galaxies on
645: the red sequence at z=1 is recovered, albeit with a larger noise
646: contribution. Additionally, we test that any two clusters are
647: dominating the signal by removing one cluster from the measurement in
648: turn. When we recalculate ratios each time the significance of the
649: signal goes down. From this we conclude that all three clusters are
650: contributing to the signal.
651:
652: Lastly, we consider the possibility that the four magnitude range used
653: is too large in that it goes too close to the very uncertain bright
654: end of the luminosity function of the z=0.1 clusters, and too far into
655: a regime where photometry is confused at the faint end of the
656: luminosity function in the z=1.0 clusters. To test our results, we
657: re-calculate the ratios with 1.5 magnitude wide bins centered at $M_K
658: =-20$. We find, again, the same trend of faint galaxies disappearing
659: from the red sequence population at high redshift.
660:
661: There is a clear deficit of faint galaxies on the red sequence at z=1
662: compared to the current epoch when taking into account possible
663: sources of error.
664:
665:
666: %---------------------------------------------------------------------------------
667:
668: \section{ Discussion}
669: \label{discuss}
670:
671:
672: We find that there are fewer faint red cluster galaxies at high
673: redshift than low redshift in comparison to the number of bright red
674: galaxies. There are two possible explanations for this. There could be
675: overall less faint cluster galaxies at z=1 at all colors than the
676: present epoch, or the faint end of the red sequence is not yet in
677: place, and instead those galaxies which will fill the faint end at
678: z=0, are still forming stars at higher redshifts.
679:
680: The first explanation, that there are overall, at all colors, fewer
681: faint galaxies at z=1 than at z=0, goes directly against the theory of
682: hierarchical formation. We expect that clusters at higher redshifts
683: will have more faint galaxies than today, and that over time the faint
684: galaxies will merge into brighter, more massive galaxies. This will
685: have the effect of clusters at higher redshifts having steeper faint
686: end slopes of the luminosity function than today's flatter slopes, or
687: exactly the opposite trend of what would be suggested if overall there
688: were less faint galaxies at $z=1$. While we do not think this is a
689: likely solution, it is interesting that in a simulation of cluster
690: luminosity functions, \citet{khochfar2007} find a slight trend that
691: the faint end slope of clusters does steepen from redshift 1 to 0,
692: whereas overall they find the hierarchical formation trend of
693: flattening slope from z=6 to 0. This is probably due to noise in
694: their calculation. It would be nice to know how the overall faint end
695: slope of clusters is evolving from z=1 to 0, however we do not have a
696: large enough sample to attack this problem.
697:
698: It is more likely that the deficit of faint red galaxies at high z is
699: due to the fact that those faint galaxies are still forming stars, and
700: so have bluer colors at those redshifts. This implies that a) bright,
701: massive galaxies have already shut off their star formation in
702: clusters by z=1, and b) faint, less massive galaxies are still forming
703: stars in clusters at z=1. We know that clusters
704: exhibit red sequences at higher z, so it is not new that massive
705: galaxies have already undergone some process, excited by the cluster
706: environment, which keeps them from forming stars, and lands them on
707: the red sequence. This is further evidence for the popular theory of
708: 'downsizing', in which the more massive galaxies evolve first. However, all
709: galaxies in clusters do not follow the same evolutionary processes,
710: instead evolution from the blue cloud to the red sequence seems to be
711: mass dependent. Whatever process stops star formation in clusters
712: appears has not yet happened at z=1 to the less massive galaxies.
713: Those galaxies are still being allowed to form enough stars to stay
714: off of the red sequence. The red population of clusters is not yet
715: fully in place by z=1.
716:
717:
718: We do not compare the specific values of faint-to-bright ratios of this
719: work with those in the literature as different wavelengths, areas, and
720: definitions of faint and bright are used. However the trend for a
721: deficit of faint galaxies at high redshifts is in agreement with the
722: works of \citet{delucia2007, gilbank2008, stott2007} and
723: \citet{koyama2007}, but in contrast to \citet{andreon2007}. The cause
724: of the difference is unclear, but we note that we use a much redder,
725: wider filter set at high redshifts (rest-frame B-K) than
726: \citet{andreon2007} (rest-frame u-g). We also compare with the same
727: rest-frame color at low redshift without applying a k-correction.
728:
729:
730: Future work in this field requires a large enough sample of confirmed
731: clusters at high-z with consistent observations well below $M^*$ to be
732: able to split the sample on cluster properties. There are hints that
733: the faint-to-bright ratios are cluster mass or richness dependent, but
734: the literature shows contradictory trends for these effects likely due
735: to the small samples used to date \citep{gilbank2008, delucia2007}.
736: With this evidence for ongoing star formation in clusters at z=1, in
737: paper two of this series we will examine the star formation rates and
738: morphologies of the cluster galaxies using deep SPITZER MIPS
739: $24\,\micron$ and HST ACS F814W data. Additionally as the community continues to
740: build a larger sample of high redshift clusters we will be able to
741: study their properties, in particular their suitability for dark
742: energy number count surveys \citep{wang2004}.
743: %---------------------------------------------------------------------------------
744:
745: \acknowledgments
746:
747: We acknowledge E. Rykoff for help with the mass measurement. We thank
748: the anonymous referee for useful suggestions on the manuscript. This
749: research has made use of data from the Two Micron All Sky Survey,
750: which is a joint project of the University of Massachusetts and the
751: Infrared Processing and Analysis Center/California Institute of
752: Technology, funded by the National Aeronautics and Space
753: Administration and the National Science Foundation. This work was
754: based on observations obtained with the Hale Telescope, Palomar
755: Observatory as part of a continuing collaboration between the
756: California Institute of Technology, NASA/JPL, and Cornell University,
757: the Spitzer Space Telescope, which is operated by the Jet Propulsion
758: Laboratory, California Institute of Technology under a contract with
759: NASA, the MMT Observatory, a joint facility of the Smithsonian
760: Institution and the University of Arizona, and the NASA/ESA Hubble
761: Space Telescope, obtained at the Space Telescope Science Institute,
762: which is operated by the Association of Universities for Research in
763: Astronomy, Inc., under NASA contract NAS 5-26555. These observations
764: are associated with program \#10521. Support for program \#10521 was
765: provided by NASA through a grant from the Space Telescope Science
766: Institute, which is operated by the Association of Universities for
767: Research in Astronomy, Inc., under NASA contract NAS 5-26555.
768:
769:
770: {\it Facilities:} \facility{Palomar (LFC, WIRC)}, \facility{MMT
771: (Megacam)}, \facility{HST (ACS)}, \facility{Spitzer (IRAC, MIPS)},
772: \facility{Akari}, \facility{CXO (ACIS)}
773:
774: %------------------------------------------------------------------------
775: %\clearpage
776: %\bibliography{ms.bbl}
777:
778: \begin{thebibliography}{43}
779: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
780:
781: \bibitem[{{Andreon}(2006)}]{andreon2006}
782: {Andreon}, S. 2006, \aap, 448, 447
783:
784: \bibitem[{{Andreon}(2007)}]{andreon2007}
785: ---. 2007, ArXiv e-prints, 710
786:
787: \bibitem[{{Andreon} {et~al.}(2008){Andreon}, {de Propris}, {Puddu}, {Giordano},
788: \& {Quintana}}]{andreon2008}
789: {Andreon}, S., {de Propris}, R., {Puddu}, E., {Giordano}, L., \& {Quintana}, H.
790: 2008, \mnras, 383, 102
791:
792: \bibitem[{{Beers} {et~al.}(1990){Beers}, {Flynn}, \& {Gebhardt}}]{beers1990}
793: {Beers}, T.~C., {Flynn}, K., \& {Gebhardt}, K. 1990, \aj, 100, 32
794:
795: \bibitem[{{Blakeslee} {et~al.}(2003){Blakeslee}, {Franx}, {Postman}, {Rosati},
796: {Holden}, {Illingworth}, {Ford}, {Cross}, {Gronwall}, {Ben{\'{\i}}tez},
797: {Bouwens}, {Broadhurst}, {Clampin}, {Demarco}, {Golimowski}, {Hartig},
798: {Infante}, {Martel}, {Miley}, {Menanteau}, {Meurer}, {Sirianni}, \&
799: {White}}]{blakeslee2003}
800: {Blakeslee}, J.~P. {et~al.} 2003, \apjl, 596, L143
801:
802: \bibitem[{{Blanton} {et~al.}(2003){Blanton}, {Gregg}, {Helfand}, {Becker}, \&
803: {White}}]{eblanton2003}
804: {Blanton}, E.~L., {Gregg}, M.~D., {Helfand}, D.~J., {Becker}, R.~H., \&
805: {White}, R.~L. 2003, \aj, 125, 1635
806:
807: \bibitem[{{Bolzonella} {et~al.}(2000){Bolzonella}, {Miralles}, \&
808: {Pell{\'o}}}]{bolzonella2000}
809: {Bolzonella}, M., {Miralles}, J.-M., \& {Pell{\'o}}, R. 2000, \aap, 363, 476
810:
811: \bibitem[{{Brodwin} {et~al.}(2007){Brodwin}, {Gonzalez}, {Moustakas},
812: {Eisenhardt}, {Stanford}, {Stern}, \& {Brown}}]{brodwin2007}
813: {Brodwin}, M., {Gonzalez}, A.~H., {Moustakas}, L.~A., {Eisenhardt}, P.~R.,
814: {Stanford}, S.~A., {Stern}, D., \& {Brown}, M.~J.~I. 2007, \apjl, 671, L93
815:
816: \bibitem[{{Bruzual} \& {Charlot}(2003)}]{bruzual2003}
817: {Bruzual}, G., \& {Charlot}, S. 2003, \mnras, 344, 1000
818:
819: \bibitem[{{Butcher} \& {Oemler}(1984)}]{butcher1984}
820: {Butcher}, H., \& {Oemler}, A. 1984, \apj, 285, 426
821:
822: \bibitem[{{Calzetti} {et~al.}(2000){Calzetti}, {Armus}, {Bohlin}, {Kinney},
823: {Koornneef}, \& {Storchi-Bergmann}}]{calzetti2000}
824: {Calzetti}, D., {Armus}, L., {Bohlin}, R.~C., {Kinney}, A.~L., {Koornneef}, J.,
825: \& {Storchi-Bergmann}, T. 2000, \apj, 533, 682
826:
827: \bibitem[{{Capak} {et~al.}(2007){Capak}, {Aussel}, {Ajiki}, {McCracken},
828: {Mobasher}, {Scoville}, {Shopbell}, {Taniguchi}, {Thompson}, {Tribiano},
829: {Sasaki}, {Blain}, {Brusa}, {Carilli}, {Comastri}, {Carollo}, {Cassata},
830: {Colbert}, {Ellis}, {Elvis}, {Giavalisco}, {Green}, {Guzzo}, {Hasinger},
831: {Ilbert}, {Impey}, {Jahnke}, {Kartaltepe}, {Kneib}, {Koda}, {Koekemoer},
832: {Komiyama}, {Leauthaud}, {Lefevre}, {Lilly}, {Liu}, {Massey}, {Miyazaki},
833: {Murayama}, {Nagao}, {Peacock}, {Pickles}, {Porciani}, {Renzini}, {Rhodes},
834: {Rich}, {Salvato}, {Sanders}, {Scarlata}, {Schiminovich}, {Schinnerer},
835: {Scodeggio}, {Sheth}, {Shioya}, {Tasca}, {Taylor}, {Yan}, \&
836: {Zamorani}}]{capak2007}
837: {Capak}, P. {et~al.} 2007, \apjs, 172, 99
838:
839: \bibitem[{{De Lucia} {et~al.}(2007){De Lucia}, {Poggianti},
840: {Arag{\'o}n-Salamanca}, {White}, {Zaritsky}, {Clowe}, {Halliday}, {Jablonka},
841: {von der Linden}, {Milvang-Jensen}, {Pell{\'o}}, {Rudnick}, {Saglia}, \&
842: {Simard}}]{delucia2007}
843: {De Lucia}, G. {et~al.} 2007, \mnras, 374, 809
844:
845: \bibitem[{{Ebeling} {et~al.}(2001){Ebeling}, {Jones}, {Fairley}, {Perlman},
846: {Scharf}, \& {Horner}}]{ebeling2001}
847: {Ebeling}, H., {Jones}, L.~R., {Fairley}, B.~W., {Perlman}, E., {Scharf}, C.,
848: \& {Horner}, D. 2001, \apjl, 548, L23
849:
850: \bibitem[{{Eisenhardt} {et~al.}(2008){Eisenhardt}, {Brodwin}, {Gonzalez},
851: {Stanford}, {Stern}, {Barmby}, {Brown}, {Dawson}, {Dey}, {Doi}, {Galametz},
852: {Jannuzi}, {Kochanek}, {Meyers}, {Morokuma}, \& {Moustakas}}]{eisenhardt2008}
853: {Eisenhardt}, P.~R.~M. {et~al.} 2008, ArXiv e-prints, 804
854:
855: \bibitem[{{Elston} {et~al.}(2006){Elston}, {Gonzalez}, {McKenzie}, {Brodwin},
856: {Brown}, {Cardona}, {Dey}, {Dickinson}, {Eisenhardt}, {Jannuzi}, {Lin},
857: {Mohr}, {Raines}, {Stanford}, \& {Stern}}]{elston2006}
858: {Elston}, R.~J. {et~al.} 2006, \apj, 639, 816
859:
860: \bibitem[{{Fang} {et~al.}(2007){Fang}, {Gerke}, {Davis}, {Newman}, {Davis},
861: {Nandra}, {Laird}, {Koo}, {Coil}, {Cooper}, {Croton}, \& {Yan}}]{fang2007}
862: {Fang}, T. {et~al.} 2007, \apjl, 660, L27
863:
864: \bibitem[{{Fazio} {et~al.}(2004){Fazio}, {Hora}, {Allen}, {Ashby}, {Barmby},
865: {Deutsch}, {Huang}, {Kleiner}, {Marengo}, {Megeath}, {Melnick}, {Pahre},
866: {Patten}, {Polizotti}, {Smith}, {Taylor}, {Wang}, {Willner}, {Hoffmann},
867: {Pipher}, {Forrest}, {McMurty}, {McCreight}, {McKelvey}, {McMurray}, {Koch},
868: {Moseley}, {Arendt}, {Mentzell}, {Marx}, {Losch}, {Mayman}, {Eichhorn},
869: {Krebs}, {Jhabvala}, {Gezari}, {Fixsen}, {Flores}, {Shakoorzadeh}, {Jungo},
870: {Hakun}, {Workman}, {Karpati}, {Kichak}, {Whitley}, {Mann}, {Tollestrup},
871: {Eisenhardt}, {Stern}, {Gorjian}, {Bhattacharya}, {Carey}, {Nelson},
872: {Glaccum}, {Lacy}, {Lowrance}, {Laine}, {Reach}, {Stauffer}, {Surace},
873: {Wilson}, {Wright}, {Hoffman}, {Domingo}, \& {Cohen}}]{fazio2004}
874: {Fazio}, G.~G. {et~al.} 2004, \apjs, 154, 10
875:
876: \bibitem[{{Finoguenov} {et~al.}(2007){Finoguenov}, {Guzzo}, {Hasinger},
877: {Scoville}, {Aussel}, {B{\"o}hringer}, {Brusa}, {Capak}, {Cappelluti},
878: {Comastri}, {Giodini}, {Griffiths}, {Impey}, {Koekemoer}, {Kneib},
879: {Leauthaud}, {Le F{\`e}vre}, {Lilly}, {Mainieri}, {Massey}, {McCracken},
880: {Mobasher}, {Murayama}, {Peacock}, {Sakelliou}, {Schinnerer}, {Silverman},
881: {Smol{\v c}i{\'c}}, {Taniguchi}, {Tasca}, {Taylor}, {Trump}, \&
882: {Zamorani}}]{finoguenov2007}
883: {Finoguenov}, A. {et~al.} 2007, \apjs, 172, 182
884:
885: \bibitem[{{Ford} {et~al.}(2004){Ford}, {Postman}, {Blakeslee}, {Demarco},
886: {Jee}, {Rosati}, {Holden}, {Homeier}, {Illingworth}, \& {White}}]{ford2004}
887: {Ford}, H. {et~al.} 2004, in Astrophysics and Space Science Library, Vol. 319,
888: Penetrating Bars Through Masks of Cosmic Dust, ed. D.~L. {Block},
889: I.~{Puerari}, K.~C. {Freeman}, R.~{Groess}, \& E.~K. {Block}, 459--+
890:
891: \bibitem[{{Gilbank} {et~al.}(2008){Gilbank}, {Yee}, {Ellingson}, {Gladders},
892: {Loh}, {Barrientos}, \& {Barkhouse}}]{gilbank2008}
893: {Gilbank}, D.~G., {Yee}, H.~K.~C., {Ellingson}, E., {Gladders}, M.~D., {Loh},
894: Y.-S., {Barrientos}, L.~F., \& {Barkhouse}, W.~A. 2008, \apj, 673, 742
895:
896: \bibitem[{{Hansen} {et~al.}(2007){Hansen}, {Sheldon}, {Wechsler}, \&
897: {Koester}}]{hansen2007}
898: {Hansen}, S.~M., {Sheldon}, E.~S., {Wechsler}, R.~H., \& {Koester}, B.~P. 2007,
899: ArXiv e-prints, 710
900:
901: \bibitem[{{Johnston} {et~al.}(2007){Johnston}, {Sheldon}, {Wechsler}, {Rozo},
902: {Koester}, {Frieman}, {McKay}, {Evrard}, {Becker}, \& {Annis}}]{johnston2007}
903: {Johnston}, D.~E. {et~al.} 2007, ArXiv e-prints, 709
904:
905: \bibitem[{{Khochfar} {et~al.}(2007){Khochfar}, {Silk}, {Windhorst}, \&
906: {Ryan}}]{khochfar2007}
907: {Khochfar}, S., {Silk}, J., {Windhorst}, R.~A., \& {Ryan}, Jr., R.~E. 2007,
908: \apjl, 668, L115
909:
910: \bibitem[{{Koyama} {et~al.}(2007){Koyama}, {Kodama}, {Tanaka}, {Shimasaku}, \&
911: {Okamura}}]{koyama2007}
912: {Koyama}, Y., {Kodama}, T., {Tanaka}, M., {Shimasaku}, K., \& {Okamura}, S.
913: 2007, \mnras, 382, 1719
914:
915: \bibitem[{{Lubin} {et~al.}(2004){Lubin}, {Mulchaey}, \& {Postman}}]{lubin2004}
916: {Lubin}, L.~M., {Mulchaey}, J.~S., \& {Postman}, M. 2004, \apjl, 601, L9
917:
918: \bibitem[{{Margoniner} {et~al.}(2005){Margoniner}, {Lubin}, {Wittman}, \&
919: {Squires}}]{margoniner2005}
920: {Margoniner}, V.~E., {Lubin}, L.~M., {Wittman}, D.~M., \& {Squires}, G.~K.
921: 2005, \aj, 129, 20
922:
923: \bibitem[{{Maughan}(2007)}]{maughan2007}
924: {Maughan}, B.~J. 2007, \apj, 668, 772
925:
926: \bibitem[{{Maughan} {et~al.}(2006){Maughan}, {Jones}, {Ebeling}, \&
927: {Scharf}}]{maughan2006}
928: {Maughan}, B.~J., {Jones}, L.~R., {Ebeling}, H., \& {Scharf}, C. 2006, \mnras,
929: 365, 509
930:
931: \bibitem[{{Mullis} {et~al.}(2005){Mullis}, {Rosati}, {Lamer}, {B{\"o}hringer},
932: {Schwope}, {Schuecker}, \& {Fassbender}}]{mullis2005}
933: {Mullis}, C.~R., {Rosati}, P., {Lamer}, G., {B{\"o}hringer}, H., {Schwope}, A.,
934: {Schuecker}, P., \& {Fassbender}, R. 2005, \apjl, 623, L85
935:
936: \bibitem[{{Polletta} {et~al.}(2007){Polletta}, {Tajer}, {Maraschi},
937: {Trinchieri}, {Lonsdale}, {Chiappetti}, {Andreon}, {Pierre}, {Le F{\`e}vre},
938: {Zamorani}, {Maccagni}, {Garcet}, {Surdej}, {Franceschini}, {Alloin},
939: {Shupe}, {Surace}, {Fang}, {Rowan-Robinson}, {Smith}, \&
940: {Tresse}}]{Polletta2007}
941: {Polletta}, M. {et~al.} 2007, \apj, 663, 81
942:
943: \bibitem[{{Rosati} {et~al.}(2004){Rosati}, {Tozzi}, {Ettori}, {Mainieri},
944: {Demarco}, {Stanford}, {Lidman}, {Nonino}, {Borgani}, {Della Ceca},
945: {Eisenhardt}, {Holden}, \& {Norman}}]{rosati2004}
946: {Rosati}, P. {et~al.} 2004, \aj, 127, 230
947:
948: \bibitem[{{Siemiginowska} {et~al.}(2005){Siemiginowska}, {Cheung}, {LaMassa},
949: {Burke}, {Aldcroft}, {Bechtold}, {Elvis}, \& {Worrall}}]{siemiginowska2005}
950: {Siemiginowska}, A., {Cheung}, C.~C., {LaMassa}, S., {Burke}, D.~J.,
951: {Aldcroft}, T.~L., {Bechtold}, J., {Elvis}, M., \& {Worrall}, D.~M. 2005,
952: \apj, 632, 110
953:
954: \bibitem[{{Stanford} {et~al.}(1997){Stanford}, {Elston}, {Eisenhardt},
955: {Spinrad}, {Stern}, \& {Dey}}]{stanford1997}
956: {Stanford}, S.~A., {Elston}, R., {Eisenhardt}, P.~R., {Spinrad}, H., {Stern},
957: D., \& {Dey}, A. 1997, \aj, 114, 2232
958:
959: \bibitem[{{Stanford} {et~al.}(2002){Stanford}, {Holden}, {Rosati},
960: {Eisenhardt}, {Stern}, {Squires}, \& {Spinrad}}]{stanford2002}
961: {Stanford}, S.~A., {Holden}, B., {Rosati}, P., {Eisenhardt}, P.~R., {Stern},
962: D., {Squires}, G., \& {Spinrad}, H. 2002, \aj, 123, 619
963:
964: \bibitem[{{Stanford} {et~al.}(2001){Stanford}, {Holden}, {Rosati}, {Tozzi},
965: {Borgani}, {Eisenhardt}, \& {Spinrad}}]{stanford2001}
966: {Stanford}, S.~A., {Holden}, B., {Rosati}, P., {Tozzi}, P., {Borgani}, S.,
967: {Eisenhardt}, P.~R., \& {Spinrad}, H. 2001, \apj, 552, 504
968:
969: \bibitem[{{Stanford} {et~al.}(2006){Stanford}, {Romer}, {Sabirli}, {Davidson},
970: {Hilton}, {Viana}, {Collins}, {Kay}, {Liddle}, {Mann}, {Miller}, {Nichol},
971: {West}, {Conselice}, {Spinrad}, {Stern}, \& {Bundy}}]{stanford2006}
972: {Stanford}, S.~A. {et~al.} 2006, \apjl, 646, L13
973:
974: \bibitem[{{Stott} {et~al.}(2007){Stott}, {Smail}, {Edge}, {Ebeling}, {Smith},
975: {Kneib}, \& {Pimbblet}}]{stott2007}
976: {Stott}, J.~P., {Smail}, I., {Edge}, A.~C., {Ebeling}, H., {Smith}, G.~P.,
977: {Kneib}, J.-P., \& {Pimbblet}, K.~A. 2007, \apj, 661, 95
978:
979: \bibitem[{{Surace} {et~al.}(2005){Surace}, {Shupe}, {Fang}, {Evans}, {Alexov},
980: {Frayer}, {Lonsdale}, \& {SWIRE Team}}]{surace2005}
981: {Surace}, J.~A., {Shupe}, D.~L., {Fang}, F., {Evans}, T., {Alexov}, A.,
982: {Frayer}, D., {Lonsdale}, C.~J., \& {SWIRE Team}. 2005, in Bulletin of the
983: American Astronomical Society, Vol.~37, Bulletin of the American Astronomical
984: Society, 1246--+
985:
986: \bibitem[{{van Dokkum}(2005)}]{vandokkum2005}
987: {van Dokkum}, P.~G. 2005, \aj, 130, 2647
988:
989: \bibitem[{{van Dokkum} {et~al.}(2001){van Dokkum}, {Stanford}, {Holden},
990: {Eisenhardt}, {Dickinson}, \& {Elston}}]{vandokkum2001}
991: {van Dokkum}, P.~G., {Stanford}, S.~A., {Holden}, B.~P., {Eisenhardt}, P.~R.,
992: {Dickinson}, M., \& {Elston}, R. 2001, \apjl, 552, L101
993:
994: \bibitem[{{Vikhlinin} {et~al.}(2002){Vikhlinin}, {VanSpeybroeck}, {Markevitch},
995: {Forman}, \& {Grego}}]{vikhlinin2002}
996: {Vikhlinin}, A., {VanSpeybroeck}, L., {Markevitch}, M., {Forman}, W.~R., \&
997: {Grego}, L. 2002, \apjl, 578, L107
998:
999: \bibitem[{{Wang} {et~al.}(2004){Wang}, {Khoury}, {Haiman}, \& {May}}]{wang2004}
1000: {Wang}, S., {Khoury}, J., {Haiman}, Z., \& {May}, M. 2004, \prd, 70, 123008
1001:
1002: \end{thebibliography}
1003:
1004:
1005: %\bibliography{jkrick}
1006: %------------------------------------------------------------------------
1007:
1008: % Table generated by Excel2LaTeX from sheet 'Sheet1'
1009: \begin{deluxetable}{ccccccc}
1010: \tablewidth{0pc}
1011: \tablecolumns{7}
1012: \tablecaption{Cluster Characteristics \label{tab:clusterchar}}
1013:
1014: \tablehead{
1015: \colhead{Cluster} &
1016: \colhead{ra} &
1017: \colhead{dec} &
1018: \colhead{$N_{gals}$} &
1019: \colhead{z$_{peak}\tablenotemark{a}$} &
1020: \colhead{L$_{x}$ (0.5-2.0 Kev)} &
1021: \colhead{M$_{500}$}
1022: \\
1023: \colhead{ } &
1024: \colhead{J2000 (deg)} &
1025: \colhead{J2000 (deg)} &
1026: \colhead{$r < 500 Kpc$} &
1027: \colhead{ } &
1028: \colhead{$1\times10^{43}$erg/s} &
1029: \colhead{$1\times10^{13}\msun$}
1030: }
1031:
1032: \startdata
1033: 1 & 264.68160 & 69.04481 & 215 & $1.0\pm0.1$ & $3.6\pm0.6$ & $6.2\pm1.4$ \\
1034:
1035: 2 & 264.89228 & 69.06851 & 255 & $1.0\pm0.1$ & $1.6\pm0.7$ & $3.6\pm1.4$ \\
1036:
1037: 3 & 264.83102 & 69.09031 & 241 & $1.0\pm0.2$ & $\le1.6\pm0.7$ & $\le3.6\pm1.1$ \\
1038:
1039: \enddata
1040: \tablenotetext{a}{Redshift peak and one sigma uncertainty are measured from a gaussian fit to the redshift distribution.}
1041: \end{deluxetable}
1042: %------------------------------------------------------------------------
1043:
1044: \clearpage
1045:
1046:
1047: \begin{figure}
1048: \epsscale{0.3}
1049: \plotone{f1a.ps}
1050: \plotone{f1b.ps}
1051: \plotone{f1c.ps}
1052: \caption[Redshift distribution]{Redshift distributions of the three
1053: candidate clusters. The solid histogram shows the redshifts for all
1054: galaxies within $\sim 375$\,Kpc of the cluster centers. The dotted
1055: line shows the average redshift distribution of the field galaxy
1056: population measured in regions with the same area. These
1057: distributions are based on photometric redshifts. }
1058: \label{fig:zdist}
1059: \epsscale{1}
1060: \end{figure}
1061: %-----------------------------------------------------------------
1062:
1063: \begin{figure}
1064: \epsscale{0.4}
1065: \plotone{f2a.eps}
1066: \plotone{f2b.eps}
1067: \plotone{f2c.eps}
1068: %\plotone{acs_1.ps}
1069: \caption[Cluster 1 ]{Color composite images of candidate clusters from
1070: ACS F814 (blue), MMT $z^\prime$ (green), and IRAC $3.6\,\micron$
1071: (red) data all smoothed to the resolution of the IRAC data. The
1072: circle shows the 0.017\degr (500Kpc) radius used to count the number
1073: of galaxies on the red sequence. }
1074: \label{fig:clus1}
1075: \epsscale{1}
1076: \end{figure}
1077: %-----------------------------------------------------------------
1078:
1079: \begin{figure}
1080: \epsscale{0.5}
1081: \plotone{f3.eps}
1082: \caption[chandra]{Chandra ACIS-I 0.3 - 7kev image of extended emission
1083: in the field. The two detected clusters are encircled with a 500Kpc
1084: radius circle. The third, undetected cluster location is shown with
1085: a dashed line circle. The detected cluster on the right has a
1086: higher count rate and therefore a higher mass. Specifically the
1087: extended emission is associated with the center of the cluster and
1088: not with the stars in the bottom left of Figure \ref{fig:clus1}.The
1089: extended emission just below the cluster on the right corresponds to
1090: a cluster at z$_{phot}$=0.25, and is not discussed in this paper. }
1091: \label{fig:chandra}
1092: \epsscale{1}
1093: \end{figure}
1094:
1095: %-----------------------------------------------------------------
1096: \begin{figure}
1097: \epsscale{0.4}
1098: \plotone{f4a.ps}
1099: \plotone{f4b.ps}
1100: \caption[cmd]{Left: Color magnitude diagram of all 711 galaxies within
1101: 500\,Kpc of the centers of the three clusters at z=1 from the IRAC
1102: dark field. Greyscale indicates the density of galaxies within each
1103: bin. Lines represent the RCS $\pm 0.5$mag. The top scale is
1104: absolute rest frame K-band magnitude. The dashed line shows the
1105: $5\sigma$ detection limit of the ACS data which does not come close
1106: to our fitted RCS. The CMD is plotted only to magnitudes well
1107: brighter than the measured 95\% completeness limit of the IRAC
1108: data. Right: Distribution of galaxies along the red sequence for the
1109: composite cluster at z=1. Dashed lines delineate the faint and
1110: bright magnitude bins used in the analysis.}
1111: \label{fig:cmd}
1112: \epsscale{1}
1113: \end{figure}
1114:
1115: %-----------------------------------------------------------------
1116: \begin{figure}
1117: \epsscale{0.4}
1118: \plotone{f5a.ps}
1119: \plotone{f5b.ps}
1120: \caption[cosmoscmd]{Color magnitude diagram and K-band distribution
1121: for the composite COSMOS cluster at z=0.1. Lines have the same
1122: meaning as Figure \ref{fig:cmd}. For clarity this CMD only includes
1123: those galaxies with photometric redshifts in the range of the
1124: clusters, however the actual measurement is made in the exact same
1125: way as for the clusters at $z=1$ which is to use the RCS minus a
1126: statistical background for membership information.}
1127: \label{fig:cosmoscmd}
1128: \epsscale{1}
1129: \end{figure}
1130: %-----------------------------------------------------------------
1131:
1132: \begin{figure}
1133: \epsscale{0.5}
1134: \plotone{f6.ps}
1135: \caption[ratio]{Ratio of faint-to-bright galaxies on the red sequence
1136: in both the samples at high and low z. Squares represent background
1137: subtracted numbers of member galaxies on the red sequence measured
1138: inside of a 500 KPC radius. Stars show the same measurement for a
1139: 750 Kpc radius. Error bars are only shown on the 500 Kpc points and
1140: represent error in the background measurement. }
1141: \label{fig:ratio}
1142: \epsscale{1}
1143: \end{figure}
1144:
1145: %-----------------------------------------------------------------
1146:
1147:
1148: \end{document}
1149:
1150:
1151: % LocalWords: mag arcsec IRAC Krick Surace Spitzer CA redshift IR FLAMEX PAHs
1152: % LocalWords: timescales arcsecond redshifts polycyclic AGN NEP ultradeep psf
1153: % LocalWords: HST ACS MMT Akari MIPS Chandra ACIS ch capak blakeslee vandokkum
1154: % LocalWords: ellipticals delucia RCSs andreon Ks publically XXX CCDs calacs
1155: % LocalWords: mosaicing IRAF's tweakshifts multidrizzle SWarp Terapix refs AB
1156: % LocalWords: SExtractor SEDs SWIRE Polletta SED Hyperz chi bolzonella gini cd
1157: % LocalWords: clumpings Eisenhardt overdense mips CMD CMDs Balmer biweight Gyr
1158: % LocalWords: Salpeter bruzual Flamingos Extragalactic Elston finoguenov Mk ms
1159: % LocalWords: Mpc gilbank stott koyama cdm downsizing gonzalez holmberg IPAC
1160: % LocalWords: NED bbl jkrick ks destreaking PIMMS NH KeV maughan vikhlinin XMM
1161:
1162: % LocalWords: erg Kev lubin SDSS johnston hansen Rykoff acs chandra cmd dist
1163: % LocalWords: cosmoscmd cosmosdist Lx Kpc virialized overdensity sfr zdist ps
1164: % LocalWords: rcsratio overdensities pointings Mb redshifting khochfar ESA NAS
1165: % LocalWords: kev Gorjian Yan Caltech RCS dataset datasets multi Calzetti wang
1166: % LocalWords: calzetti clusterings bremsstrahlung underluminous CMD's JPL eps
1167: % LocalWords: Greyscale stanford ebeling eblanton rosati margoniner mullis AZ
1168: % LocalWords: siemiginowska elston fazio multiwavelength al surace coadded IDL
1169: % LocalWords: regularized MOPEX DAOPHOT extragalactic Bertin isophotal PSF's
1170: % LocalWords: deblending Jy areal arcminute virial eisenhardt Monte Carlo LFC
1171: % LocalWords: brodwin dmextract monte carlo infall WIRC Megacam CXO natexlab
1172: % LocalWords: ArXiv de Propris Puddu Giordano Quintana Gebhardt Franx Gronwall
1173: % LocalWords: Illingworth tez Bouwens Broadhurst Clampin Demarco Golimowski Le
1174: % LocalWords: Hartig Infante Martel Miley Menanteau Meurer Sirianni Helfand pc
1175: % LocalWords: Becker Miralles Pell Moustakas Charlot Oemler Armus Bohlin Ajiki
1176: % LocalWords: Kinney Koornneef Storchi Bergmann Aussel Mobasher Scoville Blain
1177: % LocalWords: Shopbell Taniguchi Tribiano Sasaki Brusa Carilli Comastri Guzzo
1178: % LocalWords: Carollo Cassata Colbert Giavalisco Hasinger Ilbert Impey Jahnke
1179: % LocalWords: Kartaltepe Kneib Koda Koekemoer Komiyama Leauthaud Lefevre Lilly
1180: % LocalWords: Liu Miyazaki Murayama Nagao Porciani Renzini Salvato Scarlata ra
1181: % LocalWords: Schiminovich Schinnerer Scodeggio Sheth Shioya Tasca Zamorani
1182: % LocalWords: Poggianti Arag Salamanca Zaritsky Clowe Halliday Jablonka von
1183: % LocalWords: der Milvang Jensen Rudnick Saglia Simard Fairley Perlman Scharf
1184: % LocalWords: Horner Barmby Dey Doi Galametz Jannuzi Kochanek Morokuma Cardona
1185: % LocalWords: Lin Mohr Raines Gerke Nandra Laird Koo Croton Hora Ashby Huang
1186: % LocalWords: Kleiner Marengo Megeath Melnick Pahre Patten Polizotti Willner
1187: % LocalWords: Hoffmann Pipher Forrest McMurty McCreight McKelvey McMurray Mann
1188: % LocalWords: Moseley Arendt Mentzell Losch Mayman Eichhorn Krebs Jhabvala vre
1189: % LocalWords: Gezari Fixsen Flores Shakoorzadeh Jungo Hakun Karpati Kichak Jee
1190: % LocalWords: Whitley Tollestrup Bhattacharya Carey Glaccum Lacy Lowrance Smol
1191: % LocalWords: Laine hringer Cappelluti Giodini Griffiths Mainieri Sakelliou
1192: % LocalWords: Homeier Puerari Groess Yee Ellingson Gladders Loh Barrientos dec
1193: % LocalWords: Barkhouse Wechsler Koester Rozo Frieman Evrard Annis Windhorst
1194: % LocalWords: Kodama Tanaka Shimasaku Okamura Mulchaey Wittman Schwope Tajer
1195: % LocalWords: Schuecker Fassbender Maraschi Trinchieri Lonsdale Chiappetti deg
1196: % LocalWords: Maccagni Garcet Surdej Franceschini Alloin Shupe Rowan Tresse
1197: % LocalWords: Tozzi Ettori Lidman Nonino Borgani Della Ceca Cheung LaMassa
1198: % LocalWords: Burke Aldcroft Bechtold Worrall Spinrad Romer Sabirli Viana
1199: % LocalWords: Collins Liddle Nichol Conselice Smail Pimbblet Alexov Frayer
1200: % LocalWords: Dokkum VanSpeybroeck Markevitch Forman Grego Khoury Haiman
1201: % LocalWords: ccccccc gaussian
1202: