Discovery and Cosmological Implications of SPT-CL J2106-5844, the Most Massive Known Cluster at z > 1

0.91 MB·English

by  R. J. Foley

Draft version January 7, 2011
PreprinttypesetusingLATEXstyleemulateapjv.8/13/10
DISCOVERY AND COSMOLOGICAL IMPLICATIONS OF SPT-CL J2106-5844, THE MOST MASSIVE KNOWN
CLUSTER AT z >1
R. J. Foley1,2, K. Andersson3,4, G. Bazin3,5, T. de Haan6, J. Ruel7, P. A. R. Ade8, K. A. Aird9, R. Armstrong10,
M. L. N. Ashby7, M. Bautz4, B. A. Benson11,12, L. E. Bleem11,13, M. Bonamente14, M. Brodwin1,
J. E. Carlstrom11,12,13,15, C. L. Chang11,12, A. Clocchiatti16, T. M. Crawford11,15, A. T. Crites11,15, S. Desai10,17,
M. A. Dobbs6, J. P. Dudley6, G. G. Fazio1, W. R. Forman1, G. Garmire18, E. M. George19, M. D. Gladders11,15,
A. H. Gonzalez20, N. W. Halverson21, F. W. High11,15, G. P. Holder6, W. L. Holzapfel19, S. Hoover11,12,
J. D. Hrubes9, C. Jones1, M. Joy14, R. Keisler11,13, L. Knox22, A. T. Lee19,23, E. M. Leitch11,15, M. Lueker19,
D. Luong-Van9, D. P. Marrone11,9, J. J. McMahon11,12,24, J. Mehl11,15, S. S. Meyer11,12,13,15, J. J. Mohr3,5,25,
T. E. Montroy26, S. S. Murray,1 S. Padin11,15,27, T. Plagge11,15, C. Pryke11,12,15, C. L. Reichardt19, A. Rest7,28,
1 J. E. Ruhl26, B. R. Saliwanchik26, A. Saro3, K. K. Schaffer11,12, L. Shaw6,29, E. Shirokoff19, J. Song24,
1 H. G. Spieler23, B. Stalder1, S. A. Stanford22, Z. Staniszewski26, A. A. Stark1, K. Story11,13, C. W. Stubbs7,1,
0 K. Vanderlinde6, J. D. Vieira11,13,27, A. Vikhlinin1, R. Williamson11,15, and A. Zenteno3,5
2 Draft version January 7, 2011

n
ABSTRACT
a
J Using the South Pole Telescope (SPT), we have discovered the most massive known galaxy cluster
at z > 1, SPT-CL J2106-5844. In addition to producing a strong Sunyaev-Zel’dovich effect signal,
6
thissystemisaluminousX-raysourceanditsnumerousconstituentgalaxiesdisplayspatialandcolor
] clustering, all indicating the presence of a massive galaxy cluster. VLT and Magellan spectroscopy
O of 18 member galaxies shows that the cluster is at z = 1.132+0.002. Chandra observations obtained
−0.003
C throughacombinedHRC-ACISGTOprogramrevealanX-rayspectrumwithanFeKlineredshifted
by z = 1.18±0.03. These redshifts are consistent with the galaxy colors found in extensive optical,
.
h near-infrared, and mid-infrared imaging. SPT-CL J2106-5844 displays extreme X-ray properties for
p a cluster, having a core-excluded temperature of kT =11.0+2.6 keV and a luminosity (within r ) of
- −1.9 500
o L (0.5−2.0 keV)=(13.9±1.0)×1044 erg s−1. The combined mass estimate from measurements of
X
r the Sunyaev-Zel’dovich effect and X-ray data is M =(1.27±0.21)×1015 h−1 M . The discovery
t 200 70 (cid:12)
s of such a massive gravitationally collapsed system at high redshift provides an interesting laboratory
a for galaxy formation and evolution, and is a probe of extreme perturbations of the primordial matter
[
density field. We discuss the latter, determining that, under the assumption of ΛCDM cosmology

1 with only Gaussian perturbations, there is only a 7% chance of finding a galaxy cluster similar to
v SPT-CL J2106-5844 in the 2500 deg2 SPT survey region, and that only one such galaxy cluster is
6 expected in the entire sky.
8 Subject headings: galaxies: clusters: individual(SPT-CLJ2106-5844)—galaxies: formation—galax-
2 ies: evolution — early universe — large-scale structure of universe
1
.
1
0
1 1Harvard-Smithsonian Center for Astrophysics, 60 Garden FlightCenter,Huntsville,AL35812
1 Street,Cambridge,MA02138 15Department of Astronomy and Astrophysics, University of
: 2ClayFellow. [email protected]. Chicago,5640SouthEllisAvenue,Chicago,IL60637
v 3Department of Physics, Ludwig-Maximilians-Universita¨t, 16Departamento de Astronoma y Astrofsica, PUC Casilla
i Scheinerstr.1,81679Mu¨nchen,Germany 306,Santiago22,Chile
X 4MIT Kavli Institute for Astrophysics and Space Research, 17DepartmentofAstronomy,UniversityofIllinois,1002West
Massachusetts Institute of Technology, 77 Massachusetts Av- GreenStreet,Urbana,IL61801
r
a enue,Cambridge,MA02139 18DepartmentofAstronomyandAstrophysics,Pennsylvania
5ExcellenceClusterUniverse,Boltzmannstr.2,85748Garch- StateUniversity,525DaveyLab,UniversityPark,PA16802
ing,Germany 19Department of Physics, University of California, Berkeley,
6Department of Physics, McGill University, 3600 Rue Uni- CA94720
versity,Montreal,QuebecH3A2T8,Canada 20Department of Astronomy, University of Florida,
7Department of Physics, Harvard University, 17 Oxford Gainesville,FL32611
Street,Cambridge,MA02138 21Department of Astrophysical and Planetary Sciences and
8Department of Physics and Astronomy, Cardiff University, Department of Physics, University of Colorado, Boulder, CO
CF243YB,UK 80309
9University of Chicago, 5640 South Ellis Avenue, Chicago, 22Department of Physics, University of California, One
IL60637 ShieldsAvenue,Davis,CA95616
10NationalCenterforSupercomputingApplications,Univer- 23Physics Division, Lawrence Berkeley National Laboratory,
sityofIllinois,1205WestClarkStreet,Urbanan,IL61801 Berkeley,CA94720
11Kavli Institute for Cosmological Physics, University of 24DepartmentofPhysics,UniversityofMichigan,450Church
Chicago,5640SouthEllisAvenue,Chicago,IL60637 Street,AnnArbor,MI,48109
12Enrico Fermi Institute, University of Chicago, 5640 South 25Max-Planck-Institutfu¨rextraterrestrischePhysik,Giessen-
EllisAvenue,Chicago,IL60637 bachstr.85748Garching,Germany
13Department of Physics, University of Chicago, 5640 South 26Physics Department, Case Western Reserve University,
EllisAvenue,Chicago,IL60637 Cleveland,OH44106
14DepartmentofSpaceScience,VP62,NASAMarshallSpace 27California Institute of Technology, 1200 E. California
2 Foley et al.
1. INTRODUCTION X-ray(e.g.,Rosatietal.2004;Mullisetal.2005;Stanford
et al. 2006; Rosati et al. 2009) or Spitzer mid-infrared
Galaxy clusters are the most massive collapsed ob-
(MIR)imaging(e.g.,Stanfordetal.2005;Brodwinetal.
jects in the Universe. Massive clusters form from the
2006; Eisenhardt et al. 2008; Muzzin et al. 2009; Wilson
most overdense perturbations on their mass scale in the
et al. 2009; Papovich et al. 2010). Only one SZ-detected
primordial matter density field and are exceedingly rare
discoveredgalaxyclusterhasbeenspectroscopicallycon-
objects. At high redshift, corresponding to a relatively
firmedatz >1priortotheresultspresentedhere(Brod-
short time after the Big Bang, they provide a power-
win et al. 2010). In principle, high-mass clusters at high
ful probe of the initial matter density field of the Uni-
redshift can provide evidence for an incomplete under-
verse. Their abundance is a particularly sensitive probe
standing of cosmology, much like the massive clusters at
ofGaussianityinthematterdensityfield(e.g.,Fry1986;
z > 0.3 discovered in the late 1990s provided evidence
Mortonson et al. 2010), and, as a result, cluster surveys
against the theoretically preferred Ω =1 cosmological
can test various models of inflation and topological de- M
modelofthetime(e.g.,Carlbergetal.1996;Henry1997;
fects (e.g., Verde et al. 2001). Observations of galaxy
Bahcall & Fan 1998; Donahue et al. 1998).
clusters, in which the constituent galaxies have similar
Duringthe2009campaignoftheSPT-SZsurvey,SPT-
formation histories, can strongly constrain hierarchical
CL J2106-5844 was detected with a signal-to-noise ratio
formation models (e.g., De Lucia & Blaizot 2007); high-
(S/N) of 18.5 in 150 GHz data alone (S/N = 22.1 in
massclustersathighredshiftareparticularlyinteresting
multi-banddata)—thelargestS/Nofanyclusterinthe
astronomical laboratories for galaxy formation and evo-
2008–2009SPT-SZsample. Initialdeepopticalfollow-up
lution.
observationsshowednoobviousclusteredgalaxiesingri,
The bulk of baryonic matter in galaxy clusters is in
butdidshowastrongdetectionofgalaxiesinthez band.
the form of fully ionized intracluster gas. The free elec-
Additionalnear-infrared andSpitzer photometryfurther
trons in this gas emit thermal Bremsstrahlung radiation
confirmed the presence of clustered galaxies. These ex-
in the X-ray, and lead to inverse Compton scattering of
tremely red galaxies are consistent with a galaxy cluster
thecosmicmicrowavebackground(CMB)photons. This
at z >1.
scatteringresultsinaspectraldistortionoftheCMBob-
Chandra observations of SPT-CL J2106-5844 revealed
servedtowardtheclusterknownasthethermalSunyaev-
a luminous and extended X-ray source. Optical spec-
Zel’dovich effect (SZ; Sunyaev & Zeldovich 1972). The
troscopy of member galaxies shows that the cluster is at
surface brightness of the SZ effect is determined solely
z =1.132. SPT-CL J2106-5844 is therefore not only one
by the thermal pressure of the intra-cluster medium in-
of the most massive SPT clusters, it is also the highest-
tegratedalongthelineofsightandisindependentofthe
redshiftspectroscopicallyconfirmedSPTcluster. Infact,
clusterredshift. ThetotalSZsignalofaclusterisamea-
it is the most massive cluster known at redshift z > 1,
sure of its total thermal energy and is therefore tightly
makingSPT-CLJ2106-5844anextremelyinterestingob-
correlated with the cluster mass. This makes the SZ
ject for detailed study.
effect a powerful technique for discovering galaxy clus-
Wepresentourinitialdetectionandfollow-upobserva-
tersathighredshiftandanefficientmeansoffindingthe
tions of SPT-CL J2106-5844 in Section 2. In Section 3,
highest-mass clusters (e.g., Carlstrom et al. 2002).
weshowthatSPT-CLJ2106-5844isa(cid:38)1015 M galaxy
With the goal of detecting galaxy clusters using the (cid:12)
clusteratz =1.132. Weexaminetheimplicationsofthe
SZ effect, the 10-m South Pole Telescope (SPT; Carl-
existence of such a massive cluster at such high redshift
strom et al. 2009) is surveying 2500 deg2 in three bands
in Section 4. We summarize and conclude in Section 5.
at millimeter wavelengths. The SPT survey is sensitive
Throughout this paper, M and M masses are de-
to galaxy clusters with a mass (cid:38) 5 × 1014 M at all 200 500
(cid:12) fined as the mass enclosed in a spherical region which
redshifts. The first three SZ-discovered clusters were
has a density 200 and 500 times the mean and critical
reported by the SPT collaboration (Staniszewski et al.
density of the Universe, respectively. Uncertainties are
2009). TheSPTcollaborationrecentlypublishedthefull
reported for 68% confidence intervals.
5-σ cluster catalog from the first 200 deg2 of the survey,
observed during the 2008 observing season (Vanderlinde 2. OBSERVATIONS,DATAREDUCTION,&INITIAL
et al. 2010). Most recently, a sample of the 26 most FINDINGS
massivegalaxyclustersextractedfromthefull2500deg2 2.1. Microwave Observations from the South Pole
SPTsurveyhasbeenusedtotestthestandardcosmolog- Telescope
ical model (Williamson et al. 2011). We refer the reader
CurrentlytheSPThasbeenusedtocompleteobserva-
to these previous SPT papers for a complete description tions of 1500 deg2 to full survey depth of 18 µK-arcmin
of the survey strategy and goals. The Atacama Cos- at 150 GHz and and additional 1000 deg2 has been sur-
mologyTelescope(ACT)hasalsorecentlyreportedtheir
veyed to a “preview” depth of 54 µK-arcmin, where the
initial catalog of SZ-discovered clusters (Marriage et al.
unitKreferstoequivalentfluctuationsintheCMBtem-
2010).
perature. By the end of the 2011 observing season, we
The discovery of z (cid:38) 1 galaxy clusters has been chal- expect the full 2500 deg2 survey region to be complete
lenging. Recently, several groups have detected galaxy
to full depth. Details of the survey strategy and data
clustersatz (cid:38)1.1,withmosteitherbeingfoundthrough
processing are described in Staniszewski et al. (2009),
Vanderlinde et al. (2010), and Williamson et al. (2011).
Blvd.,Pasadena,CA91125 The2009observationsthatyieldedthediscoveryofSPT-
28Space Telescope Science Institute, 3700 San Martin Dr.,
CL J2106-5844 included measurements at 95, 150, and
Baltimore,MD21218
29Department of Physics, Yale University, P.O. Box 208210, 220 GHz to a noise level of 42, 18, and 85 µK-arcmin.
NewHaven,CT06520-8120 SPT-CL J2106-5844 was detected in maps of 2009
Discovery and Cosmological Implications of SPT-CL J2106-5844 3
SPTdatacreatedusingtime-ordereddataprocessingand vidual exposures, resample to 0(cid:48).(cid:48)86 pixels (half the solid
map-making procedures equivalent to those described in angleofthenativeIRACpixels), andrejectcosmicrays.
Vanderlinde et al. (2010). Two different cluster extrac- Images generated from the OIR data are presented in
tion pipelines were used to detect and characterize clus- Figure 1. From the deep optical-only images (gri), it is
tersintheseSPTmaps. Bothpipelinesuseamulti-scale difficult to see any indication of a cluster. However, the
matched-filter approach similar to the one described in additionofinfrareddatashowsaclearover-abundanceof
Melinetal.(2006). Onepipeline,usedtoextractclusters veryredgalaxiesclusterednearthepeakoftheSZdecre-
from single-band 150 GHz data, is identical to the clus- ment. The over-abundance of galaxies that are spatially
ter extraction procedure described in Vanderlinde et al. coincidentandhavesimilarcolorsisaclearindicationof
(2010). Theotherpipeline, usedtoextractclustersfrom aclusterofgalaxies. Color-magnitudediagrams(CMDs)
multi-band SPT data is identical to that described in for galaxies within the imaging field of view (FOV; Fig-
Williamson et al. (2011). The final SZ observable pro- ure 2) show an overabundance of very red galaxies and
duced by both pipelines is ξ, the detection significance a probable red sequence. These data are consistent with
maximized across all filter scales. SPT-CL J2106-5844 the spectroscopic redshifts described below.
wasdetectedatξ =18.5bythesingle-bandpipelineand
2.3. Optical Spectroscopy
at ξ =22.1 by the multi-band pipeline.
An image of the filtered, multi-band SPT data — us- We obtained low-resolution spectra of several clus-
ing the filter that maximized ξ, and centered on SPT- ter members of SPT-CL J2106-5844 with the Gladders
CL J2106-5844 — is shown in the left panel of Figure 1. Image-SlicingMultislitOption(GISMO;Gladdersetal.,
Detection significance contours from this same filtered in prep.) module on IMACS mounted on Magellan
map are overlaid in Figures 1 and 4. in June 2010; however, these spectra did not yield a
definitive cluster redshift. Magellan slit masks were de-
2.2. Optical and Infrared Imaging signedusinggalaxiesselectedfromdeepopticalandMIR
imaging compatible with the red sequence expected at
In June 2009, we obtained griz imaging of SPT-
z (cid:38) 1.2, as this was the initial redshift approximation
CL J2106-5844 with the Inamori Magellan Areal Cam-
from the photometry. The selection was similar to that
era and Spectrograph (IMACS; Dressler et al. 2006) and
of Brodwin et al. (2010). This technique gives higher
Low Dispersion Survey Spectrograph (LDSS330) on the
priority to brighter galaxies with colors consistent with
twin Magellan 6.5m telescopes. We obtained additional
the red sequence. If there is a region without a suit-
I-band imaging with FORS2 (Appenzeller et al. 1998)
ablered-sequencetarget,bluegalaxieswereselectedwith
on the VLT 8m telescope in September 2010. See High
the idea that they may be emission-line cluster galax-
etal.(2010)foradescriptionofourobservationstrategy
ies. The Magellan observations were obtained using the
and reduction procedure.
f/2 camera with the the 300 l/mm “red” grism and the
SPT-CL J2106-5844 was observed at the CTIO 4m
WB6300–9500 filter. A total of sixteen 1800 s exposures
Blanco telescope using the NEWFIRM imager (Autry
were obtained on June 6 and 7.
etal.2003)on28July2010. DatawereobtainedintheJ
The VLT observations were obtained with the
and K filters under photometric conditions. All images
s GRIS 300I grism combined with the OG590 blocking fil-
were acquired using a 30 point random dither pattern.
ter. TwoVLT/FORS2masksweredesignedusinggalax-
At each dither position, six frames with 10 s exposure
ies selected from deep I band VLT pre-imaging and the
times were coadded, providing 60 s total exposure per
SpitzerIRACimagingtohavecolorscompatiblewiththe
position. The dither pattern was repeated as necessary
red sequence expected at z ≈ 1.2 to 1.3, which was ini-
toachievetotalexposuretimesof7200sinJ and4440s
tially indicated by the photometry. Each mask was pop-
in K .
s ulated with approximately 30 slitlets of typical length
NEWFIRM data were reduced using the FATBOY
10(cid:48)(cid:48). This target was originally accepted as Priority B
pipeline,originallydevelopedfortheFLAMINGOS-2in-
for VLT/FORS2 spectroscopy as part of a larger SPT
strument,andmodifiedtoworkwithNEWFIRMdatain
spectroscopy campaign. In late November 2010, Chan-
support of the Infrared Bootes Imaging Survey (Gonza-
dra X-ray spectroscopy gave an initial redshift from the
lez, in prep.). Individual processed frames are combined
Fe K line (see Section 2.4). After this result when it be-
usingSCAMPandSWARP(Bertinetal.2002),andpho-
came clear that there was a significant chance of it not
tometry is calibrated to 2MASS (Skrutskie et al. 2006).
being observed in Fall 2010, we submitted a special re-
The final images in both filters have FWHM of 1(cid:48).(cid:48)3.
questtotheESODirectortoraisethistargettoPriority
Mid-infrared Spitzer/IRAC imaging was obtained in
A. That proposal was successful, and on each of three
September 2009 as part of a larger program to follow up
nights in early December 2010 a single 2700 s exposure
clusters identified in the SPT survey. The on-target ob-
wasobtainedforatotalexposuretimeof8100sononeof
servations consisted of 8×100 s and 6×30 s dithered
the two masks. These observations were taken under ex-
exposures at 3.6 and 4.5 µm, respectively.31 The deep
cellentseeingconditions(0(cid:48).(cid:48)5to0(cid:48).(cid:48)8)butathighairmass
3.6 µm observations are sensitive to passively evolving
between 1.5 and 1.7. These data were immediately re-
cluster galaxies down to 0.1L∗ at z = 1.5. The data
ducedtorevealredshiftsof15clustermembers,enabling
werereducedfollowingthemethodofAshbyetal.(2009).
a robust cluster redshift.
Briefly,wecorrectforcolumnpulldown,mosaictheindi-
Spectroscopic reductions proceeded using standard
CCDprocessingwithIRAFandtheCOSMOSreduction
30 http://www.lco.cl/telescopes-information/magellan/
instruments-1/ldss-3-1/. package (Kelson 2003) for the VLT and Magellan data,
31 In this paper, we will refer to photometry in these bands as respectively. The data were extracted using the opti-
[3.6]and[4.5],respectively. malalgorithmofHorne(1986). Fluxcalibrationandtel-
4 Foley et al.
Fig. 1.— SPT-CL J2106-5844 at millimeter, optical, and infrared wavelengths. Left: The filtered SZ significance map derived from
multi-band SPT data. The frame subtends 12(cid:48)×12(cid:48). The negative trough surrounding the cluster is a result of the filtering of the time
ordereddataandmaps. Right: LDSS3opticalandSpitzer/IRACmid-infraredgi[3.6](correspondingtoBGRchannels)images. Theframe
subtends 4.(cid:48)8×4.(cid:48)8. The white contours correspond to the SZ significance from the left-hand panel. The circles mark spectroscopically
confirmed cluster members, where green indicates quiescent, absorption-line member galaxies and cyan indicates an active, emission-line
membergalaxy. SomespectroscopicmembergalaxiesareoutsidetheFOVforthisimage.
Horne 1988; Foley et al. 2003).
3 Three independent redshift determinations were per-
formed using a cross-correlation algorithm (IRAF
RVSAO package; Kurtz & Mink 1998), a template fit-
g) 2 ting method (SDSS early-type PCA templates), and a
a
m χ2 minimization technique by comparing to galaxy tem-
] ( 1 plate spectra. There were only minor differences in the
6 final results from the three methods. In total, we have
3. obtainedsecureredshifts,consistentwithmembershipin
[
− 0 a single cluster, for 18 galaxies. Two of these galaxies
J haveobvious[OII]emission,whiletheothershaveSEDs
consistent with passive galaxies with no signs of ongoing
−1 star formation.
18 19 20 21 A3-σclippingwasappliedaroundthepeakinredshifts
to select spectroscopic cluster members. Representative
[3.6] (mag)
spectra of cluster members and a redshift histogram of
Fig. 2.—Color-magnitudediagram(J−[3.6]vs.[3.6])forgalax- cluster members are presented in Figure 3. Redshift in-
ieswithintheIRACFOV.Suspectedred-sequenceclustermembers formationforclustermembersispresentedinTable1. A
are plotted in red. Lower-probability, but potential cluster mem-
singlegalaxywasobservedandhasasecureredshiftfrom
bers are plotted in blue. Spectroscopic members are plotted as
stars, where the red stars correspond to passive galaxies and the both Magellan and VLT. Although the VLT spectrum
blue star represents an emission-line galaxy. Additional galaxies shows clear Ca H&K absorption lines and the Magel-
in the field are plotted as black points. The size of the symbol is lanspectrumonlyshowstheD4000break, themeasured
inversely proportional to the distance to the center of the cluster
redshifts are consistent.
asdeterminedbytheclusteringofthered-sequencegalaxies. Our
5-σ limits are plotted as dotted lines. A red-sequence model cor- A robust biweight estimator was applied to the
respondingz=1.132isrepresentedasthesolidblacklineswitha spectroscopic sample to determine a mean redshift of
representativeL∗ galaxyrepresentedbytheblackdiamond. z = 1.131+0.002 and a velocity dispersion of σ =
−0.003 v
1230+270 km s−1. The uncertainty in both quantities
−180
is determined through bootstrap resampling. Since the
luriclineremovalwereperformedusingthewell-exposed
dynamics of passive and star-forming galaxies within
continua of spectrophotometric standard stars (Wade &
Discovery and Cosmological Implications of SPT-CL J2106-5844 5
Observed Wavelength (Å)
7000 7500 8000 8500 9000
4
[O II] Ca H&K
4
t 3
n
a
t 3
s
n
o r
C e
b
+ m 2
λ 2 u
f N

e
v
i
t
a 1
l 1
e
R
0 0
3400 3600 3800 4000 4200 −3 −2 −1 0 1 2
Rest Wavelength (Å) Velocity Relative to Mean (103 km s−1)
Fig. 3.— Left: Subset of optical spectra of galaxies identified to be cluster members redshifted by the cluster redshift, z =1.132. The
black vertical dashed lines show the position of [OII] λ3727 and Ca H&K λλ3933, 3968, while the grey dotted lines show the where the
spectral features would fall at ±1000 km s−1 relative to the cluster redshift. Right: Histogram of velocities for the spectroscopic cluster
members relative to the biweight mean of all members. The blue and red histograms represent objects with and without obvious [O II]
emission,respectively.
clusters are different with passive galaxies assumed to
TABLE 1
be a better tracer of the cluster potential (e.g., Fadda
Redshifts for Galaxy Members of SPT-CL J2106-5844
et al. 1996; Girardi et al. 1996; Mohr et al. 1996; Ko-
ranyi & Geller 2000), we also provide the mean red-
R.A.(◦)a Dec.(◦) zb Featurec Telescope
shift and velocity dispersion of only the passive galaxies,
z =1.132+0.002 and σ =1270+320 km s−1, respectively. 316.46581 −58.72418 1.1365(6) CaH&K VLT
−0.003 v −210 316.47601 −58.72635 1.1347(3) CaH&K VLT
Thevelocitydistributionofspectroscopicmembersrel-
316.48048 −58.69972 1.1230(3) [OII] VLT
ative to the mean velocity (Figure 3) is skewed. The 316.48086 −58.75017 1.1288(2) CaH&K VLT
bootstrap resampling produced a positive skewness in 316.49456 −58.75092 1.1148(2) CaH&K VLT
89% of the samples, which indicates that it is not very 316.50647d −58.73848 1.132 (20) D4000 Magellan
statistically significant given our sample size. However, 316.50647d −58.73848 1.1296(3) CaH&K VLT
316.51307 −58.72857 1.1312(3) CaH&K VLT
this may be an indication of substructure, consistent
316.52563 −58.75017 1.1450(6) CaH&K VLT
with the possible merger seen in the X-ray image (see 316.53273 −58.72584 1.120 (20) D4000 Magellan
Section 2.4). There is no spatial segregation of galax- 316.53346 −58.77275 1.1335(5) CaH&K VLT
ies clustered in velocity space; however, the number of 316.53434 −58.75143 1.1389(4) CaH&K VLT
316.53679 −58.75943 1.1274(6) CaH&K VLT
spectroscopic members is still small.
316.54011 −58.71250 1.1384(3) CaH&K VLT
316.54123 −58.76282 1.1384(5) CaH&K VLT
316.54297 −58.76573 1.1260(4) CaH&K Magellan
316.54292 −58.75800 1.1363(6) CaH&K VLT
316.55414 −58.72558 1.1325(3) [OII] Magellan
2.4. X-ray Observations
316.74972 −58.72926 1.1136(5) CaH&K VLT
As part of a program to follow up the most massive, a AllcoordinatesareinJ2000.
high redshift clusters detected with the SPT in 2009, b Uncertaintiesgiveninunitsof0.0001.
we have obtained a 25 ks X-ray observation of SPT- c Spectroscopicfeatureprimarilyusedindeterminingredshift.
CL J2106-5844. These data were obtained through a d ThisisthesamegalaxyobservedwithbothMagellanandVLT.
combined Chandra High Resolution Camera and Ad-
vanced CCD Imaging Spectrometer Guaranteed Time observation reveals an X-ray luminous cluster with pro-
Observation program. Data reduction was performed nounced asymmetry in the surface brightness distribu-
with the method described by Andersson et al. (2010). tion(seeFigure4),indicatingthatitmaybeinamerging
The observation resulted in 1425 source counts within state. In particular, we note that a bright concentration
r in the 0.5–7.0 keV band, corresponding to a flux, of gas is offset ∼10(cid:48)(cid:48)east of the main cluster centroid.
500
F (0.5−2.0 keV)=(2.58±0.15)×10−13 ergcm−2 s−1. X-ray observations of SPT-CL J2106-5844 were used
X
A 2(cid:48)(cid:48)×2(cid:48)(cid:48) -binned X-ray image was produced using the tomeasurethegastemperature, T , andtheX-rayana-
X
X-ray events in the 0.7–4.0 keV band (Figure 4). The log of Comptonization, Y = M ×T . The cluster
X gas X
6 Foley et al.
Fig. 4.—SPT-CLJ2106-5844atX-ray,millimeter,optical,andinfraredwavelengths. Left: ChandraX-rayimagefromthe0.7–4.0keV
band, with 4×4 binning of the original ACIS pixels (0(cid:48).(cid:48)49 on a side). The black contours correspond to the SZ significance from the
left-hand panel of Figure 1. Right: LDSS3 optical and Spitzer/IRAC mid-infrared gi[3.6] (corresponding to BRG channels) images. The
framesubtends4.(cid:48)8×4.(cid:48)8. ThewhitecontoursarethesmoothedX-raysignalfromtheleftpanel.
mass can then be inferred with the M–Y and M–T found that the cool-core fraction varies significantly be-
X X
relations of Vikhlinin et al. (2009). The r radius is tweendifferenthighredshiftX-rayselectedsamples(e.g.,
500
derived iteratively from the M–Y scaling relation as Santos et al. 2010). Future work comparing these re-
X
described in Andersson et al. (2010). See Section 3.3 sults to the cool-core fraction in SZ selected samples
for details regarding the X-ray mass estimates. A spec- should help to understand the role of selection in these
trum is extracted from the data within the derived r results. Regardless of the abundance at high redshift,
500
radius, excluding the core within 0.15 r . The spec- SPT-CL J2106-5844 is one of the few high-redshift cool-
500
trumisfit,keepingtheabsorbinghydrogencolumnfixed core galaxy clusters known.
at the galactic value n = 4.32×1020 cm−2, and the
H
metal abundance fixed at 0.3 solar. Fixing the redshift
3. MASSESTIMATESFROMCLUSTEROBSERVABLES
tothemeanspectroscopicredshift,z =1.132,resultsina
Inestimatingthemassofagalaxyclusterfromanyob-
bestfitkT =11.0+2.6 keV.Withinr (117(cid:48)(cid:48)),including
−1.9 500 servableproperty,caremustbetakentoproperlyaccount
theclustercore, SPT-CLJ2106-5844hasaluminosityof
for Eddington bias, which, in this particular case, refers
L (0.5−2.0 keV)=(13.9±1.0)×1044 erg s−1, compa-
X tothepreferentialscatteringoflower-masssystemsintoa
rable to the luminosity of the Bullet cluster (Markevitch
givenobservablebinduetoasteeplyfallingsourcepopu-
et al. 2002).
lation. ThisishandledthroughapplicationofBayes’the-
AnX-rayspectrumwasproducedfromthedatawithin
orem: we use scaling relations and noise measurements
20(cid:48)(cid:48)(0.17 r ), where the metal abundance of the gas is
500 to calculate conditional probabilities of observable O at
likely to peak. The resulting spectrum is shown in Fig-
fixed mass, P(O|M), and combine these with a mass
ure5withamodelfoldedthroughthedetectorresponse.
prior P(M) to yield a posterior estimate of the mass
The strong 6.7 keV Fe K line is present in the spec-
P(M|O). When the observable O is the same quantity
trum at approximately 3 keV. This spectrum indicates
that was used to detect the cluster, the prior probabil-
z = 1.18±0.03, a metal abundance Z = 1.44+0.66 Z ,
−0.51 (cid:12) ity of the cluster’s mass is simply the cosmological mass
and a central temperature kT = 6.5+1.7 keV. Notably, function. However, when the observable in question is
−1.1
this measurement was the first robust spectroscopic red- fromatargetedfollow-upobservation,thepriormustalso
shift measurement for this cluster. take into account the selection function of the survey in
Mergersimulationssuggestthathigh-redshiftcool-core which it was discovered. For follow-up observations tar-
clusters should be rare (Gottl¨ober et al. 2001), and geting a single, very rare or interesting object, such as
previous observations have indicated a steep decline in SPT-CL J2106-5844, the appropriate prior mass distri-
the number of cool-core clusters at high redshift (e.g., bution is strongly influenced by the selection. For this
Vikhlinin et al. 2007). However, other analyses have reason, we report masses derived from follow-up obser-
Discovery and Cosmological Implications of SPT-CL J2106-5844 7
0.01
V−1 0.02 V−1
e e
k k
s −1 0.01 s −1 5×10−3
s s
unt 5×10−3 unt
o o
c c
d d 2×10−3
e e
aliz 2×10−3 aliz
m m
nor 10−3 nor 10−3
4×10−3
2×10−3
s 2×10−3 s
al al 10−3
sidu 0 sidu 0
re−2×10−3 re −10−3
−4×10−3
1 2 5 1 2
Energy (keV) Energy (keV)
Fig. 5.—Core-excluded(circularannulusspanning0.15–1r500)andcentralregion(circularaperturewithradius0.17r500)X-rayspectra
of SPT-CL J2106-5844 (left and right panels, respectively). Overplotted (solid line) is a fit APEC spectrum redshifted to z = 1.18 and
foldedthroughthedetectorsensitivity. The6.7keVFeKlineobservedat∼3keVisbothobviousandstrongforthespectrumfromthe
centralregion.
vations with a flat mass prior only. We report both flat- lationofclusters(givenbythehalomassfunction),while
prior and mass-function-prior masses for the SZ-derived the targeted observations were conducted with a priori
mass. In Section 3.4, we combine the conditional proba- knowledge that a massive cluster was present; these two
bilities derived from different observables and apply the differentmodesofobservationdrawfromdifferentpopu-
mass-function prior to the result in order to determine lationsandthuswillexhibitdifferentlevelsofEddington
the best mass estimate based on all the available infor- bias. Toreiterate,naivelyusinganymassestimatewith-
mation. For the cosmological analyses in Section 4, we out understanding the selection of the observed popula-
use only this joint posterior mass probability density. tion can result in biased measurements.
3.1. SZ Mass Estimate
3.2. Dynamical Mass Estimate
An SZ-based cluster mass estimate is produced fol-
The dynamical mass of SPT-CL J2106-5844 can be
lowing Vanderlinde et al. (2010), using the single-band
estimated from the velocity dispersion using the M–σ
(150 GHz) detection significance ξ as a mass proxy. We v
calibration of Evrard et al. (2008). Numerical simula-
direct the reader to that work for details. Briefly, an
tions indicate that galaxies are likely to exhibit veloc-
SZ-massscalingrelation,derivedfromthesimulationsof
ity bias compared to the dark matter velocity field (e.g.,
Shawetal.(2009)andconstrainedthoughacosmological
Biviano et al. 2006; White et al. 2010). Mock observa-
analysis,isusedtoestimateasetofconditionalprobabil-
tions of simulated clusters that model the galaxy selec-
itiesP(ξ|M),takingintoaccountthemeasurementnoise
tionwithinfollowupobservationslikethosereportedhere
and the intrinsic scatter in the scaling relation. A poste-
enable these biases to be quantified and corrected (Saro
riormassestimateisthencalculatedthroughapplication
et al., in prep). Published results suggest these biases
of a mass prior. The systematic errors are estimated by
are at the level of 10% in the dispersion (White et al.
linearlyexpandingthemassestimateasafunctionofthe
2010). Here we attempt no correction to the dispersion
scaling relation parameters and using the marginalized
butaccountforthisadditionalsystematicuncertaintyin
uncertainties on these parameters.
the mass estimator.
The SZ detection of this cluster is from an untargeted
Asthedynamicalmassestimatearisesfromatargeted
survey, allowing us to take the halo mass function of
follow-up observation, the underlying mass distribution
Tinker et al. (2008) as a prior on its mass. We assume
can be well-defined (and accounted for) only in the con-
the WMAP 7-year preferred ΛCDM parameters (Lar-
text of the SZ-based selection. Such a joint estimate of
son et al. 2010) when calculating this prior; marginaliz-
mass is presented below in Section 3.4; here we report
ing over the ΛCDM cosmological parameter uncertain-
an independent (hence biased) dynamical mass estimate
ties was found to have a negligible impact on the ef-
using a flat prior on the mass.
fect of this prior. SPT-CL J2106-5844 has a single-
Using the velocity dispersion from all 18 spectroscopic
band 150 GHz significance of ξ = 18.5, which results
in an unbiased posterior SZ-derived mass estimate of clustermembersgivesMdyn,200 =1.3+−10..45×1015 h−701 M(cid:12).
M =(1.06±0.23)×1015 h−1 M . Usingthevelocitydispersionfromonlythepassivegalax-
WSZe,20a0lso report (see Section 3.470and(cid:12)Table 2) a biased ies provides a very similar result, Mdyn,200 = 1.4+−10..78 ×
SZ mass estimate, using a flat mass prior; we present it 1015 h−1 M ). In both cases, the uncertainty takes into
70 (cid:12)
here for completeness only. It is important to note that accountthestatisticaluncertaintyofthevelocitydisper-
this mass is not directly comparable to other flat-mass- sion, an additional 13% intrinsic scatter of the velocity
priorestimates(e.g.,themassesestimatedinSections3.2 dispersion due to the orientation of the velocity ellip-
and 3.3). The SZ observation was drawn from an untar- soid along the line of sight (White et al. 2010), the 4%
getedsurveyandhencefromtheentireunderlyingpopu- intrinsic scatter of the Evrard et al. (2008) M–σ rela-
v
8 Foley et al.
tion, and a 10% estimate of the systematic uncertainty
TABLE 2
in the dispersion that arises from the currently uncer-
Mass Estimates for SPT-CL J2106-5844
tainkinematicrelationshipbetweenthegalaxiesandthe
underlying dark matter.
Observable Measurement M200(1015h−701M(cid:12))
3.3. X-Ray Mass Estimate SZξ 18.5 1.06±0.23
SZξ (flatprior)18.5 1.24±0.30
As described in Section 2.4, the X-ray observations YX (14.3±3.4)×1014 M(cid:12) keV1.85±0.40
of SPT-CL J2106-5844 were used to measure the gas TX 11.0+−21..69 keV 1.83±0.76
temperature, TX, and the X-ray analog of Comptoniza- σva 1270+−321200 kms−1 1.4−+10..78
tion, Y = M × T . Following the methodology
X gas X
of Andersson et al. (2010), the data were iteratively Combined ··· 1.27±0.21
fit to determine r , T , and Y . Using the scaling Note. —UnboldedmassesindicateEddington-biasedmassestimates,
500 X X
relations of Vikhlinin et al. (2009), these values pro- calculated using flat priors on mass. Note that the SZ (untargeted)
measurementsuffersfromaconsiderablydifferentEddingtonbiasthan
vide estimates of the gas and total mass of the cluster,
the other (targeted) flat-prior estimates. The probability distribution
M = (1.27±0.07)×1014 h−1 M and M = forthecombinedmasswasobtainedbymultiplyingtheunbiased(bold)
(0.g9a3s,±5000.19)×1015 h−1 M ,respe7c0tivel(cid:12)y. EstimaYtXin,5g00the SZmassprobabilitybytheflat-priorYX massprobability(seetextfor
70 (cid:12) details). Only the unbiased SZ ξ and YX mass estimates were used to
cluster mass from the gas temperature, we find a simi- generatethecombinedmassestimate.
larlyhighmassofM =(0.92±0.37)×1015h−1M , a Asdeterminedbythebiweightestimateofthepassivegalaxies.
where we have increTaXse,5d00the mass by 17% accor7d0ing (cid:12)to biased, mass estimate from the two observables to be
the recipe in Vikhlinin et al. (2009) because of the large M =(1.27±0.21)×1015 h−1 M .
200 70 (cid:12)
degree of asymmetry in the surface brightness distribu- All mass estimates are consistent and very high for a
tion, indicative of merging activity. cluster at this redshift. Each individual mass estimate,
Forbothmassestimatesabovewehaveadoptedasys- as well as the combined estimate, indicates that SPT-
tematic uncertainty of 9% for the mass calibration of CLJ2106-5844isthemostmassiveclusteryetdiscovered
the scaling relations. The uncertainty on the self-similar at z >1. It is ∼60% more massive than SPT-CL J0546-
evolution of these relations was estimated using the sim- 5345 (Brodwin et al. 2010) at z =1.07, the second most
ulation informed estimate of 5% and 7% uncertainty on massive galaxy cluster at z >1.
the normalization at z =0.6 for the M–Y and M–T ,
X X
respectively. This uncertainty on the evolutionary term 4. LIKELIHOODANALYSIS
was extrapolated to the cluster redshift. For further dis- Since SPT-CL J2106-5844 is the highest S/N clus-
cussionontheseuncertainties,seeVikhlininetal.(2009); ter found in 2008-2009 SPT survey fields, the highest-
theappropriatemasspriortoaccountforEddingtonbias redshiftspectroscopicallyconfirmedSPTcluster,andthe
is only well-defined in the context of SZ-based selection. mostmassiveclusterdiscoveredatz >1byanymethod,
itraisesthequestion: IstheexistenceofSPT-CLJ2106-
3.4. Combined Mass Estimate 5844 consistent with a Gaussian-distributed primordial
density field in a ΛCDM universe?
The X-ray, SZ, and optical redshift data can each
Weinvestigatetheprobabilityoffindingaclusterwith
be used to produce an estimate of the mass of SPT-
the mass and redshift of SPT-CL J2106-5844 in the con-
CL J2106-5844. We list our estimates for the mass of
text of ΛCDM as follows. Based on the likelihood cal-
SPT-CLJ2106-5844fromeachoftheseobservablesinTa-
culation of Vanderlinde et al. (2010) and the CosmoMC
ble 2. Note that we have converted the M mass to
YX,500 software package (Lewis & Bridle 2002), we produce an
M assuming a Navarro-Frenk-White density pro-
YX,200 MCMC chain sampling the parameter space of spatially
file and the mass-concentration relation of Duffy et al.
flat ΛCDM cosmology, storing the halo mass function of
(2008). M is not used in the combined mass esti-
TX,500 Tinker et al. (2008) at each step in the chain. Current
matesinceitiscorrelatedwithM andhasalarger
YX,500 data sets, consisting of the WMAP 7-year data (Larson
uncertainty. Similarly, M is not used in the com-
dyn,200 et al. 2010) and the SPT 2008 cluster catalog (Vander-
bined mass estimate because of its large uncertainty.
linde et al. 2010), are used to constrain the parameter
We assume for the purposes of the combined mass es-
space explored. This yields a set of halo mass functions
timate that the uncertainties are uncorrelated between
representative of current uncertainties of ΛCDM param-
the two masses. In principle, the different observables
eter space.
could have correlated scatter in their scaling with mass,
Integrating each of these mass functions, we compute
however, this is expected to be negligible at the current
thePoissonlikelihoodoffindingaclusteratorabovethe
S/N of the observables. This simplification allows us to
mass and redshift of SPT-CL J2106-5844 in a 2500 deg2
write the posterior mass distribution as a product of the
field. For the MCMC chain described above, the ex-
two independent conditional probabilities,
istence of a cluster at or above the mass and redshift
P(M|ξ,Y )∝P(M)·P(ξ|M)·P(Y |M), (1) of SPT-CL J2106-5844 is less than 5% likely in mod-
X X
els contributing 32% of the WMAP7+V10 probability.
where P(M) is again the halo mass function of Tin- Marginalized over all WMAP7+V10 allowed parameter
ker et al. (2008), evaluated at the WMAP 7-year pre- space,wefindthatthereisa7%chanceoffindingaclus-
ferred point in ΛCDM parameter space. Functionally, ter at least as massive and at a redshift at least as high
this reduces to a product of two probability density as SPT-CL J2106-5844 in the SPT survey field. Prior to
functions, the unbiased SZ mass estimate and the flat- having observed this cluster, the expectation would be
prior Y mass estimate. We find the combined, un- that there is only one galaxy cluster at least as massive
X
Discovery and Cosmological Implications of SPT-CL J2106-5844 9
and at a redshift at least as high as SPT-CL J2106-5844 The South Pole Telescope program is supported by
in the entire sky. the National Science Foundation through grant ANT-
0638937. Partial support is also provided by the NSF
5. CONCLUSIONS PhysicsFrontierCentergrantPHY-0114422totheKavli
We present the discovery and initial characterization Institute of Cosmological Physics at the University of
ofSPT-CLJ2106-5844. Thismassiveclusterwasdiscov- Chicago, the Kavli Foundation, and the Gordon and
ered in the 2500 deg2 SPT survey as a highly significant Betty Moore Foundation. This work is based in part on
SZ decrement. Initial follow-up in the optical and in- observations obtained with the Spitzer Space Telescope
frared indicated that it was both at high redshift and (PID 60099), which is operated by the Jet Propulsion
very rich. Optical spectroscopy yielded a cluster red- Laboratory, California Institute of Technology under a
shift of z = 1.132, consistent with the redshift found contractwithNASA.Supportforthisworkwasprovided
from the X-ray spectrum of the hot intracluster gas, by NASA through an award issued by JPL/Caltech.
z = 1.18, and the broad-band SEDs of member galax- Additional data were obtained with the 6.5 m Magel-
ies. SPT-CLJ2106-5844iscurrentlythehighest-redshift lan Telescopes located at the Las Campanas Observa-
SZ-discovered galaxy cluster. tory, Chile. Support for X-ray analysis was provided
SPT-CL J2106-5844 is also extremely massive. X- by NASA through Chandra Project Numbers 12800071
ray, SZ, and velocity dispersion data provide comple- and 12800088 issued by the Chandra X-ray Observatory
mentary and consistent measurements of the mass (see Center, which is operated by the Smithsonian Astro-
Table 2). The best combined estimate of the mass is physical Observatory for and on behalf of NASA un-
M = (1.27 ± 0.21) × 1015 h−1 M , making SPT- der contract NAS8-03060. Observations from VLT pro-
CL20J02106-5844 the most massive7g0alax(cid:12)y cluster yet dis- grams 086.A-0741 and 286.A-5021 were included in this
covered at z >1. The combination of its high mass and work. We acknowledge the use of the Legacy Archive
highredshiftmakeSPT-CLJ2106-5844averyinteresting for Microwave Background Data Analysis (LAMBDA).
object in which to study galaxy formation and evolution Support for LAMBDA is provided by the NASA Office
onlyafewGyraftertheformationofthefirststars. The of Space Science. Galaxy cluster research at Harvard
X-ray data show that SPT-CL J2106-5844 has a cool is supported by NSF grant AST-1009012. Galaxy clus-
core and may be currently undergoing a merger. The ter research at SAO is supported in part by NSF grants
distribution of galaxy velocities is slightly skewed, but AST-1009649 and MRI-0723073. The McGill group ac-
is consistent with a Gaussian distribution. Additional knowledges funding from the National Sciences and En-
spectroscopy may help determine if there is significant gineeringResearchCouncilofCanada,CanadaResearch
substructure in the cluster. Chairs program, and the Canadian Institute for Ad-
ThehighmassandredshiftofSPT-CLJ2106-5844also vancedResearch. X-rayresearchattheCfAissupported
make it an interesting object from a cosmological per- through NASA Contract NAS 8-03060. The Munich
spective. The chances of finding a cluster this extreme groupacknowledgessupportfromtheExcellenceCluster
in mass and redshift in a 2500 deg2 survey is only 7%, Universe and the DFG research program TR33. R.J.F.
given current cosmological parameter constraints, with is supported by a Clay Fellowship. B.A.B is supported
only one such cluster expected in the entire sky. As the by a KICP Fellowship, support for M.Brodwin was pro-
SPTclustercataloggrows,andourunderstandingofthe videdbytheW.M.KeckFoundation,M.Bautzacknowl-
various mass-observable scaling relations improves, SZ- edges support from contract 2834-MIT-SAO-4018 from
discoveredclusterswillenableyetmorepowerfultestsof the Pennsylvania State University to the Massachusetts
the current cosmological paradigm. Institute of Technology. M.D. acknowledges support
from an Alfred P. Sloan Research Fellowship, W.F. and
C.J. acknowledge support from the Smithsonian Institu-
Facilities: Blanco (NEWFIRM), CXO (ACIS), Mag- tion, and B.S. acknowledges support from the Brinson
ellan:Baade (IMACS), Magellan:Clay (LDSS3), Spitzer Foundation.
(IRAC), South Pole Telescope, VLT:Antu (FORS2)
REFERENCES
Andersson,K.,etal.2010,ArXive-prints,1006.3068 Carlberg,R.G.,Yee,H.K.C.,Ellingson,E.,Abraham,R.,
Appenzeller,I.,etal.1998,TheMessenger,94,1 Gravel,P.,Morris,S.,&Pritchet,C.J.1996,ApJ,462,32
Ashby,M.L.N.,etal.2009,ApJ,701,428 Carlstrom,J.E.,etal.2009,ArXive-prints,0907.4445
Autry,R.G.,etal.2003,inSocietyofPhoto-Optical Carlstrom,J.E.,Holder,G.P.,&Reese,E.D.2002,ARA&A,
InstrumentationEngineers(SPIE)ConferenceSeries,Vol.4841, 40,643
SocietyofPhoto-OpticalInstrumentationEngineers(SPIE) DeLucia,G.,&Blaizot,J.2007,MNRAS,375,2
ConferenceSeries,ed.M.Iye&A.F.M.Moorwood,525–539 Donahue,M.,Voit,G.M.,Gioia,I.,Lupino,G.,Hughes,J.P.,&
Bahcall,N.A.,&Fan,X.1998,ApJ,504,1 Stocke,J.T.1998,ApJ,502,550
Bertin,E.,Mellier,Y.,Radovich,M.,Missonnier,G.,Didelon,P., Dressler,A.,Hare,T.,Bigelow,B.C.,&Osip,D.J.2006,in
&Morin,B.2002,inAstronomicalSocietyofthePacific SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)
ConferenceSeries,Vol.281,AstronomicalDataAnalysis ConferenceSeries,Vol.6269,SocietyofPhoto-Optical
SoftwareandSystemsXI,ed.D.A.Bohlender,D.Durand,& InstrumentationEngineers(SPIE)ConferenceSeries
T.H.Handley,228–+ Duffy,A.R.,Schaye,J.,Kay,S.T.,&DallaVecchia,C.2008,
Biviano,A.,Murante,G.,Borgani,S.,Diaferio,A.,Dolag,K.,& MNRAS,390,L64
Girardi,M.2006,A&A,456,23 Eisenhardt,P.R.M.,etal.2008,ApJ,684,905
Brodwin,M.,etal.2006,ApJ,651,791 Evrard,A.E.,etal.2008,ApJ,672,122
——.2010,ApJ,721,90 Fadda,D.,Girardi,M.,Giuricin,G.,Mardirossian,F.,&
Mezzetti,M.1996,ApJ,473,670
10 Foley et al.
Foley,R.J.,etal.2003,PASP,115,1220 ——.2009,A&A,508,583
Fry,J.N.1986,ApJ,308,L71 Santos,J.S.,Tozzi,P.,Rosati,P.,&Bo¨hringer,H.2010,A&A,
Girardi,M.,Fadda,D.,Giuricin,G.,Mardirossian,F.,Mezzetti, 521,A64+
M.,&Biviano,A.1996,ApJ,457,61 Shaw,L.D.,Zahn,O.,Holder,G.P.,&Dor´e,O.2009,ApJ,702,
Gottl¨ober,S.,Klypin,A.,&Kravtsov,A.V.2001,ApJ,546,223 368
Henry,J.P.1997,ApJ,489,L1+ Skrutskie,M.F.,etal.2006,AJ,131,1163
High,F.W.,etal.2010,ApJ,723,1736 Stanford,S.A.,etal.2005,ApJ,634,L129
Horne,K.1986,PASP,98,609 ——.2006,ApJ,646,L13
Kelson,D.D.2003,PASP,115,688 Staniszewski,Z.,etal.2009,ApJ,701,32
Koranyi,D.M.,&Geller,M.J.2000,AJ,119,44 Sunyaev,R.A.,&Zeldovich,Y.B.1972,Commentson
Kurtz,M.J.,&Mink,D.J.1998,PASP,110,934 AstrophysicsandSpacePhysics,4,173
Larson,D.,etal.2010,ArXive-prints,1001.4635 Tinker,J.,Kravtsov,A.V.,Klypin,A.,Abazajian,K.,Warren,
Lewis,A.,&Bridle,S.2002,Phys.Rev.D,66,103511 M.,Yepes,G.,Gottlo¨ber,S.,&Holz,D.E.2008,ApJ,688,709
Markevitch,M.,Gonzalez,A.H.,David,L.,Vikhlinin,A., Vanderlinde,K.,etal.2010,ApJ,722,1180
Murray,S.,Forman,W.,Jones,C.,&Tucker,W.2002,ApJ, Verde,L.,Jimenez,R.,Kamionkowski,M.,&Matarrese,S.2001,
567,L27 MNRAS,325,412
Marriage,T.A.,etal.2010,ArXive-prints,1010.1065 Vikhlinin,A.,Burenin,R.,Forman,W.R.,Jones,C.,Hornstrup,
Melin,J.-B.,Bartlett,J.G.,&Delabrouille,J.2006,A&A,459, A.,Murray,S.S.,&Quintana,H.2007,inHeatingversus
341 CoolinginGalaxiesandClustersofGalaxies,ed.H.B¨ohringer,
Mohr,J.J.,Geller,M.J.,Fabricant,D.G.,Wegner,G., G.W.Pratt,A.Finoguenov,&P.Schuecker,48–+
Thorstensen,J.,&Richstone,D.O.1996,ApJ,470,724 Vikhlinin,A.,etal.2009,ApJ,692,1060
Mortonson,M.J.,Hu,W.,&Huterer,D.2010,ArXive-prints, Wade,R.A.,&Horne,K.1988,ApJ,324,411
1011.0004 White,M.,Cohn,J.D.,&Smit,R.2010,MNRAS,408,1818
Mullis,C.R.,Rosati,P.,Lamer,G.,B¨ohringer,H.,Schwope,A., Williamson,R.,etal.2011,ApJ,submitted
Schuecker,P.,&Fassbender,R.2005,ApJ,623,L85 Wilson,G.,etal.2009,ApJ,698,1943
Muzzin,A.,etal.2009,ApJ,698,1934
Papovich,C.,etal.2010,ApJ,716,1503
Rosati,P.,etal.2004,AJ,127,230