The X-ray luminous galaxy cluster population at 0.9<z<~1.6 as revealed by the XMM-Newton Distant Cluster Project
R. Fassbender, H. Boehringer, A. Nastasi, R. Suhada, M. Muehlegger, A. de Hoon, J. Kohnert, G. Lamer, J.J. Mohr, D. Pierini, G.W. Pratt, H. Quintana, P. Rosati, J.S. Santos, A.D. Schwope
aa r X i v : . [ a s t r o - ph . C O ] O c t The X-ray luminous galaxy cluster population at 0 . < z < ∼ . ‡ R Fassbender , H B¨ohringer , A Nastasi , R Suhada , M M ¨uhlegger , Ade Hoon , J Kohnert , G Lamer , J J Mohr , , , D Pierini § , G W Pratt , HQuintana , P Rosati , J S Santos , A D Schwope Max-Planck-Institut f¨ur extraterrestrische Physik (MPE), Giessenbachstrasse 1, 85748Garching, Germany Leibniz-Institut f¨ur Astrophysik Potsdam (AIP), An der Sternwarte 16, 14482 Potsdam,Germany Department of Physics, Ludwigs-Maximilians Universit¨at M¨unchen, Scheinerstr. 1, 81679Munich, Germany Excellence Cluster Universe, Boltzmannstr. 2, 85748 Garching, Germany CEA Saclay, Service d’Astrophysique, L’Orme des Merisiers, Bˆat. 709, 91191Gif-sur-Yvette Cedex, France Departamento de Astronom´ıa y Astrof´ısica, Pontificia Universidad Cat´olica de Chile,Casilla 306, Santiago 22, Chile European Southern Observatory (ESO), Karl-Scharzschild-Str. 2, 85748 Garching,Germany European Space Astronomy Centre (ESAC), 7828691 Villanueva de la Canada, Madrid,SpainE-mail: [email protected]
Abstract.
We present the largest sample of spectroscopically confirmed X-ray luminoushigh-redshift galaxy clusters to date comprising 22 systems in the range 0 . < z < ∼ . Newton
Distant Cluster Project (XDCP). All systems were initially selectedas extended X-ray sources over 76.1 deg of non-contiguous deep archival XMM- Newton coverage, of which 49.4 deg are part of the core survey with a quantifiable selection functionand 17.7 deg are classified as ‘gold’ coverage as starting point for upcoming cosmologicalapplications. Distant cluster candidates were followed-up with moderately deep optical andnear-infrared imaging in at least two bands to photometrically identify the cluster galaxypopulations and obtain redshift estimates based on colors of simple stellar population models.We test and calibrate the most promising redshift estimation techniques based on the R − zand z − H colors for e ffi cient distant cluster identifications and find a good redshift accuracyperformance of the z − H color out to at least z ∼ .
5, while the redshift evolution of theR − z color leads to increasingly large uncertainties at z > ∼ .
9. Photometrically identifiedhigh- z systems are spectroscopically confirmed with VLT / FORS 2 with a minimum of threeconcordant cluster member redshifts. We present first details of two newly identified clusters, ‡ Based on observations under program IDs 079.A-0634 and 085.A-0647 collected at the European Organisationfor Astronomical Research in the Southern Hemisphere, Chile, and observations collected at the CentroAstron´omico Hispano Alem´an (CAHA) at Calar Alto, operated jointly by the Max-Planck Institut f¨urAstronomie and the Instituto de Astrof´ısica de Andaluc´ıa (CSIC). § Visiting astronomer at MPE. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP XDCP J0338.5 + z = + z = z = z clusters, of which 17 are at z ≥ . z > .
3. The median system mass of the sample isM ≃ × M ⊙ , while the probed mass range for the distant clusters spans approximately(0.7-7) × M ⊙ . The majority ( > z > . ff sets from the X-raycenter with a median value of about 50 kpc in projection and a smaller median luminosity gapto the second-ranked galaxy of ∆ m ≃ . ′ from the X-raycenter. This value suggests an increase of the fraction of very luminous cluster-associatedradio sources by about a factor of 2.5-5 relative to low- z systems. The galaxy populations in z > ∼ . z sample will allow first detailed studies ofthe cluster population during the critical cosmic epoch at lookback times of 7.3-9.5 Gyr on theaggregation and evolution of baryons in the cold and hot phases as a function of redshift and system mass. Keywords : galaxies: clusters: general – X-rays: galaxies: clusters – galaxies: evolution –cosmology: observations he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP
1. Introduction
The most extreme mass peaks in the primordial matter density field have developed intothe present day galaxy cluster population through gravitational amplification and more than13 Gyrs of hierarchical structure formation at work. As such, clusters of galaxies form thetop level of the hierarchy and are the latecomers on the stage of cosmic structures with themost extreme masses and dimensions for gravitationally bound objects. Besides their role askey tracers of the cosmic large-scale structure, clusters are also intriguing multi-componentastrophysical systems for the study of dark matter, baryons in the hot and cold phases, and amultitude of resulting interaction processes between them.However, one of the major observational challenges is to provide sizable samples ofgalaxy clusters at high redshift ( z > .
8) in order to trace the evolution of the cluster populationand their matter components back to the first half of cosmic time, corresponding to lookbacktimes of 7-10 Gyrs. Bona fide clusters of galaxies with total masses of M > ∼ M ⊙ are rareobjects, in particular at high z , which requires large survey areas (tens of square degrees)on one hand and a high observational sensitivity for the identification and investigation ofthe galaxy- and intracluster medium (ICM) components on the other hand. Examples ofsuccessful high- z galaxy cluster surveys based on optical / infrared observations of the galaxypopulations include Gonzalez et al. (2001), Gladders and Yee (2005), Olsen et al. (2007),Eisenhardt et al. (2008), Muzzin et al. (2009), Grove et al. (2009), Erben et al. (2009), Adamiet al. (2010), R¨oser et al. (2010), and Gilbank et al. (2011). X-ray selected distant clustersearches include the work of Rosati et al. (1998), Pacaud et al. (2007), ˇSuhada et al. (2010),and Mehrtens et al. (2011), while detected z > . ff ect (SZE) are reported e.g. in Marriage et al. (2010) and Williamson et al. (2011). For ageneral overview of di ff erent survey techniques and an updated status report of distant galaxycluster research we refer to the accompanying review of Rosati and Fassbender (in prep.)In this paper we provide a comprehensive overview of the XMM- Newton
Distant ClusterProject (XDCP), a serendipitous X-ray survey specifically designed for finding and studyingdistant X-ray luminous galaxy clusters at z ≥ .
8. The main aims of this article are adescription of the cluster sample construction in the XDCP and a report on the status ofthe compilation of the largest distant X-ray luminous galaxy cluster sample to date. Thepaper follows and combines a series of previous multi-wavelength studies of individual high- z clusters discovered in the XDCP k . To this end, we start with the general goals and design ofthe survey in Sect. 2, followed by an overview of the observational techniques in Sect. 3. Newresults are discussed in Sect. 4, the current sample of 22 X-ray clusters at z > . Λ CDM cosmological model with parameters( H , Ω m , Ω DE , w) = (70 km s − Mpc − , 0.3, 0.7, -1), physical quantities (e.g. R , M ) arederived for radii for which the mean total mass density of the cluster is 500 or 200 times thecritical energy density of the Universe ρ cr ( z ) at the given redshift z , and all reported magnitudes k An updated list of XDCP publications can be found at . he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP
2. The XMM-Newton Distant Cluster Project
The XMM-
Newton
Distant Cluster Project was initiated in 2003 with the main objective of asystematic search for distant X-ray luminous galaxy clusters, with a special focus on the z > z = . / . + + Newton archive o ff eredthe possibility for a new generation of serendipitous X-ray galaxy cluster surveys with anorder of magnitude better sensitivity and greatly improved resolution capabilities (e.g. Romeret al. 2001). From the very start, the XDCP focussed on the galaxy cluster population in the first halfof the present age of the Universe, i.e. at redshifts z > ∼ .
8. This specialization made thesurvey manageable in terms of the required follow-up resources and, moreover, allowed thedeployment of optimized observational techniques and instrumentation for high- z studies asdiscussed in Sect. 3. The final aim of the XDCP survey is the compilation of an X-ray selecteddistant galaxy cluster sample with a minimum of 50 test objects at z > . z >
1) to allowstatistically meaningful evolution studies of the cluster population in at least three mass andredshift bins.With such a sample numerous open questions on the formation and early evolution of themost massive bound structures in the Universe can be addressed observationally. Some of thekey areas include:(i) Galaxy evolution in the densest high- z environments(ii) Redshift evolution of the X-ray scaling relations(iii) Evolution of the thermal structure and the metal enrichment of the intracluster medium(iv) Number density evolution of massive clusters at z > . z clusters are shown in Sects. 4 & 5 and in publications on individualsystems (e.g. Santos et al. 2009; Strazzullo et al. 2010; Fassbender et al. 2011b). Combiningthe existing literature data on the scaling relations of cluster X-ray properties with recent deepX-ray observations of new distant systems from our and other projects we obtained tighterconstraints on the evolution of scaling relations with redshift as presented in Reichert et al.(2011). These results support the picture of an early energy input into the intracluster mediumas advocated in preheating models (e.g. Stanek et al. 2010; Short et al. 2010) rather than he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP
5a late energy input from interactions with a central AGN at low redshift ( z < ∼ ff ectof Dark Energy on structure growth is expected to be most pronounced in the redshift range0 < z < ∼
2. The competitive cosmological and Dark Energy constraints of Vikhlinin et al.(2009a; 2009b) based on the observed evolution of the cluster mass function with only 37moderate redshift systems (0 . < z < .
9) clearly demonstrated the high potential of distantX-ray clusters as Dark Energy probes. Therefore the XDCP survey will be ideally suited toextend this test to the next higher redshift regime soon once a sizable subsample of the surveyis completed.
The XDCP survey is based on the following four stage strategy:
X-ray source detection and candidate selection:
Deep, extragalactic ¶ XMM-
Newton archivalfields are screened for serendipitous extended
X-ray sources, which are in their vast ma-jority associated with galaxy clusters. The positions of the detected extended X-raysources are cross-correlated with available optical data and extragalactic database infor-mation to test for the existence of a detectable optical cluster counterpart. For about30% of the X-ray sources no optical counterpart could be identified. These sources areselected as distant cluster candidates for further follow-up.
Follow-up imaging and redshift estimation:
The selected distant cluster candidates aretargeted with su ffi ciently deep imaging data in at least two suitable optical or near-infrared (NIR) bands. The data allow as a first identification step to probe the existenceof an overdensity of (red) galaxies coincident with the extended X-ray source and ina second step a cluster redshift estimate based on the comparison of the color of red-ridgeline galaxies with simple stellar population (SSP) evolution models for passivegalaxies. Spectroscopic confirmation:
Photometrically identified systems at z > . Multi-wavelength follow-up of selected systems:
The most interesting and intriguing dis-tant systems are further studied in more detail in di ff erent wavelength regimes, e.g. withdeeper X-ray data or multi-band imaging observations in the optical and infrared.The first z > z = .
39 (Mulliset al. 2005), which started the ongoing era of distant cluster detections with XMM-
Newton . ¶ The extragalactic sky is defined here as the sky region with galactic latitudes | b | ≥ ◦ that avoids the largeextinction and dense stellar fields of the galactic band. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP
3. Observational techniques and reduction pipelines
The following section introduces and discusses the di ff erent relevant observational techniquesfor the first three XDCP survey stages in more detail. A full comprehensive description ofobservational aspects and reduction pipelines can be found in Fassbender (2007). The XMM-
Newton observatory currently provides by far the best capabilities for detectingthe typically faint extended X-ray sources associated with distant galaxy clusters. The mostimportant key features of XMM-
Newton for this task are (i) the large e ff ective collecting area( ∼ on-axis at 1 keV), (ii) the 30 ′ diameter field-of-view (FoV, ∼ ), and (iii) asu ffi ciently good spatial resolution of 5 ′′ -15 ′′ (FWHM) to identify distant clusters as extendedsources.The XMM- Newton data archive is a very rich resource to start a systematic searchfor distant clusters based on their characteristic X-ray signature, the extended thermal ICMemission, which clearly discriminates these sources from the point-like AGN population thatdominates the X-ray sky in extragalactic fields. For the definition of the XDCP survey fields,the public XMM-
Newton archive as of 2 November 2004 was considered, i.e. the public dataof the first 5 years of the mission. Out of the 2960 observed fields available at that timewith a combined nominal exposure time + of 72.3 Msec, 1109 fields remained after applyingthe conditions of (i) imaging mode observations of at least one of the three cameras, (ii) aminimum nominal exposure time of 10 ksec, and (iii) field positions outside the galactic plane( | b | ≥ ◦ ) and away from the Magellanic Clouds and M31 ∗ . After a further removal of(iv) major dedicated survey fields (e.g. COSMOS) and (v) constraining the area to the VLT-accessible part of the sky (DEC ≤ + ◦ ) for the follow-up program, 575 archival observationsremained as input for the survey (see Fig. 1). Out of these fields, 29 were discarded as non-usable for the survey after a visual screening of all fields.The remaining 546 XMM- Newton archival fields with a nominal total exposure timeof 17.5 Msec were processed and analyzed as detailed below. The final XDCP sample ofsuccessfully processed and analyzed fields amounts to 469 individual XMM-
Newton pointings(29 fields had corrupted data and 48 were flared) comprising 15.2 Msec of X-ray data with atotal sky coverage of 76.1 deg (see Table 1). The initial XDCP pilot study (de Hoon et al.,in prep.) for testing and qualifying the survey strategy of Sect. 2.2 was based on an earlierprocessing and candidate selection of about 20% of these fields. The task of processing several hundred XMM-
Newton archivalfields requires an e ffi cient automated X-ray reduction pipeline with minimized manualinteraction. To this end, a designated, distant cluster optimized XDCP reduction and source + The exposure time listed in the XMM-
Newton archive. ∗ The minimal angular distances for the field positions were 10.8 ◦ for the LMC, 5.3 ◦ for the SMC, and 3.2 ◦ forM31 (see e.g. Kim et al. 2004). he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 1.
Sky distribution of the 575 Southern XMM-
Newton fields considered for the XDCPsurvey, 469 of which were successfully processed and analyzed. The red squares indicatefields within the original footprint of the South Pole Telescope survey. Square symbols are notto scale. detection pipeline was developed based on the XMM Science Analysis Software † (SAS). Allselected XMM- Newton data sets were homogeneously processed with this pipeline using theversion SAS 6.5 released in August 2005.The data processing starts with the Observation Data File (ODF) for each archival field.In a first reduction step the SAS tasks cifbuild , odfingest , emchain , and epchain arerun to set up the appropriate calibration files for the field, ingest the housekeeping data, andproduce calibrated photon event files for the PN and the two MOS X-ray imaging instrumentsof XMM- Newton .In a second step, periods of increased background levels, most notably due to solarsoft proton flares, are removed from the data in a strict two level flare cleaning process(see e.g. Pratt and Arnaud 2003). This task is of crucial importance for the detectabilityof faint extended X-ray sources. Due to the flat nature of the flare spectrum, time periodswith background levels significantly higher than the quiescent count rates are in the firstcleaning stage e ffi ciently identified in the hardest energy band of 12-14 keV (10-12 keV) forthe PN (MOS) detector and removed from the data with an automated 3- σ clipping algorithm.However, residual soft flare peaks can still remain in the data, which are subsequently removedby applying a second soft-band cleaning stage to the full 0.3-10 keV band with a similarclipping procedure. The resulting cleaned photon event lists for each detector contain nowonly the selected science usable time periods, which is on average about two thirds of thenominal field exposure time, i.e. one third of the observation is typically lost due to flares andinstrumental overheads.We define the clean e ff ective exposure time as the period during which all threeinstruments in imaging operation would collect the equivalent number of soft science photonsfor the particular observation. The 48 fields with a resulting clean e ff ective exposure time † http://xmm.esac.esa.int/sas/ he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP < ff ective exposure time, with an average (median)clean field depth of 18.78 ksec (15.71 ksec).In a third step, images with a pixel scale of 4 ′′ / pixel are generated for di ff erent X-rayenergy bands from the clean event lists for each of the three instruments. The redshiftedspectra of distant clusters with ICM temperatures of 2-6 keV have their observed bulkemission in the soft X-ray band. Images are hence generated for the standard XMM bands0.3-0.5 keV, 0.5-2.0 keV, 2.0-4.5 keV, and a very broad band with 0.5-7.5 keV. Moreover, it ispossible to define a single energy band which maximizes the expected signal-to-noise ratio(SNR) for z > . eexpmap ,which contain the e ff ective local integration times associated with each detector pixel scaledto the on-axis exposure. These X-ray exposure maps, similarly to the concept of flatfields inoptical and NIR imaging, contain the calibration information on the radial vignetting function,the energy dependent detector quantum e ffi ciency, chip gaps, dead detector columns, thetransmission function of the used optical blocking filter, and the field-of-view of the detectors.Exposure corrected, i.e. flatfielded, images are obtained by dividing the photon imagesof each detector by the corresponding exposure map. The full data stack for each energy bandis obtained by combining the PN, MOS1, and MOS2 images weighted with the correspondinge ff ective collecting area of each telescope-camera system. For visual inspection purposes thecombined and exposure corrected X-ray images in each energy band are smoothed with a 4 ′′ Gaussian filter, from which logarithmically spaced X-ray flux contours are generated to beoverlaid on optical images for the source identification process (Sects. 3.1.3 & 3.2).
The X-ray source detection is run on each field individually,even in case of multiple observations of the same target or overlapping fields. In the eventof multiple detections of the same extended X-ray source in overlapping fields, the highestsignificance source is retained on the cluster candidate list, while the others are flagged asduplicate detections. The main technical reason for this field-by-field approach is that theX-ray point-spread-function (PSF) at each detector position has to be known as accuratelyas possible in order to allow a robust determination of extent likelihoods. Since the PSF ‡ ofXMM- Newton ’s telescopes varies considerably across the field, in particular as a function ofincreasing o ff -axis angle, detections in combined mosaic fields would have added significantsystematic uncertainty to the results based on the available PSF calibration and SAS status atthe time of data reduction.The main XDCP source detection method relies on a sliding box detection with theSAS task eboxdetect followed by a maximum likelihood fitting and source evaluationwith emldetect . As a preparatory step, detection masks are created with emask that ‡ The source detection relies on the tabulated energy and position dependent PSF model as provided in thecalibration database for SAS 6.5. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP esplinemap . For robustness and to avoidpossible artificial background fluctuations from spline fits with many degrees of freedom,we make use of the smooth two-component background model option, which is based onthe linear combination of a spatially constant background contribution (quiescent particleinduced background and instrumental noise) and a vignetted component (CXB and residualsof the soft proton particle background). Background maps are produced by first running eboxdetect with a local background determination around the detection cell in order toproduce a preliminary list of X-ray sources, which are subsequently excised from the fieldbefore performing the two-component fit for the global background map.The sliding box source detection is then repeated with eboxdetect using the previouslydetermined global background maps for each detector and varying detection cell sizesto account for the extended sources. This way, a list of positions of X-ray sourcecandidates is produced, which serves as input list for the subsequent detailed analysisand source characterization via maximum likelihood (ML) fitting with emldetect . Themaximum likelihood fitting for the source evaluation and parameter estimation is performedsimultaneously for all used energy bands and the three individual detectors, with theassociated global background maps, exposure maps, and detection masks provided to thetask. The maximum likelihood PSF fitting procedure applied to the photon images evaluatesthe significance for the detection ( DET ML ) and the extent (
EXT ML ) of an X-ray sourceexpressed in terms of the likelihood L = − ln p Pois (Cruddace et al. 1988), where p Pois is theprobability of a Poissonian random background fluctuation of counts in the detection cell,which would result in at least the number of observed counts. X-ray sources are flaggedas extended with core radius r c > § with a fixed β = / DET ML and the free parameters position and count rate in each band, or as extended source with extentlikelihood
EXT ML and the additional core radius parameter r c .The inherent thresholding procedure used in emldetect and the test performed forsource confusion of two PSF-like components does not allow a subsequent evaluation of theextent probabilities of all sources, but rather divides the populations into point sources with, § Radial surface brightness profile with functional form S ( r ) = S · [1 + ( r / r c ) ] − / . he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP EXT ML . This implies that the critical thresholding parametersfor the detection of extended X-ray sources have to be optimized prior to the actual detectionrun. To this end, source detection tests with various input parameter combinations wereperformed on the XMM-
Newton data set in the COSMOS field, which were compared to theactual extended X-ray source catalog of confirmed galaxy groups and clusters of Finoguenovet al. (2007).The XDCP source detection procedure follows two main objectives: (i) theconstruction of a quantifiable extended X-ray source sample ( survey sample ) with anaccurately characterizable selection function over a suitable part of the X-ray coverage(Sects. 3.1.4 & 3.1.5), and (ii) a supplementary X-ray selected cluster candidate sample( supplementary sample ) from the full XDCP sky coverage and down to the faintest feasibleX-ray flux levels that still allow the blind detection of extended sources. The scientificapplications of the first objective are statistical and cosmological studies of a well-controlledhigh- z galaxy cluster sample with quantified detection characteristics drawn from a knownsurvey volume. To this end, the final survey sample is selected from the inner parts ( Θ ≤ ′ )of the detector area (survey level 2 in Table 1 and Sect. 3.1.5) based on significant extended X-ray sources above a minimum flux cut-o ff , which is determined through extensive simulations(Sect. 3.1.4). The second objective for the compilation of the additional supplementary sample aims at an extended coverage of the accessible range of cluster parameters by considering alsosources of lower significance and at large o ff -axis angles at the expense of higher impuritylevels. Applications for this supplementary sample include (i) new rare massive clusters foundin the additional larger survey area covered by the outer parts of the detectors (survey level 1 inTable 1 and Sect. 3.1.5), (ii) the detection of lower mass and higher redshift systems at lowerflux levels, and (iii) the general exploration of the feasibility limits of the source detection andX-ray cluster surveys.The adopted XDCP source detection procedure for the construction of the survey samplerests upon the conceptually simplest detection strategy by deploying the single, distantcluster optimized, detection band for the energy range 0.35-2.4 keV. This choice is expectedto yield optimal signal-to-noise ratios for the X-ray sources associated with the targeteddistant cluster population with ICM temperatures T X > ∼ DET ML ≥ p real ≥ . > ∼ . σ ) as the minimum likelihood for the existence of a source, and EXT ML ≥ p ext ≥ . > ∼ . σ ) as lower threshold for the extent probability.For the supplementary sample, additional cluster candidates down to lower extentlikelihoods of EXT ML > p ext ≥ .
95, significance > ∼ σ ) are considered by re-running thesource detection two more times using di ff erent detection schemes. The first one is the basicXMM ‘standard scheme’ covering the energy range 0.3-4.5 keV with three input bands (0.3-0.5 keV, 0.5-2.0 keV, 2.0-4.5 keV). The second setup is an experimental ‘spectral matchedfilter scheme’ that covers the broader energy range 0.3-7.5 keV with an increased weight onthe lower energy range by using five overlapping bands (0.3-0.5 keV, 0.5-2.0 keV, 2.0-4.5 keV, he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ewavelet were additionally used as a qualitative cross-check ofdetected extended sources at low significance levels.This redundancy strategy with source detection results from di ff erent detection schemeso ff ers cross-comparison possibilities that are particularly advantageous when evaluatingflagged extended sources very close to the threshold of detectability. To this end, thesupplementary schemes add extra information to the source lists from the primary detectionband scheme, such as standard 0.5-2.0 keV flux estimates and several hardness ratios.Furthermore, the stability of the best fitting extended source model can be evaluated by cross-comparing the core radius measurements and extent likelihoods obtained with the di ff erentdetection schemes, which allows a more reliable identification of spurious sources frombackground fluctuations and spurious extent flags associated with point sources.The XDCP source detection run based on the discussed schemes and applied to the469 XMM- Newton archival survey fields resulted in about 2000 flagged extended sourcecandidates as raw input list for the combined survey and supplementary samples. Theseflagged sources are further evaluated in the three stage screening process detailed below.
A visual inspection and screening of candidate extended X-raysources detected in XMM-
Newton data is inevitable even at significance threshold levelsmuch higher than for the XDCP scheme. At the first screening stage on the X-ray level,obvious spurious detections of extended sources are removed from the source list. Variouscalibration and detection method limitations as well as instrumental artifacts can lead tospurious detections of extended sources. The most obvious false detections originate from (i)secondary detections in wings of large (partially masked out) extended sources, (ii) artifacts atthe edges of the field-of-view, and (iii) PSF residuals in the wings of very bright point sources.These ‘level 1’ spurious extended X-ray sources, totaling about 15% of the raw catalog, canbe readily and safely removed by inspecting the locations of the candidate sources in the FoVof the combined soft-band X-ray image.For identifying and removing the more subtle ‘level 2’ false detections, a second X-ray screening stage is required that is based on a close inspection and evaluation of everysource individually with complementary information on potential contaminations from opticalimaging data. For this task, a set of diagnostic images is produced to evaluate the sourceenvironment based on the X-ray flux contours, the original combined X-ray photon image,the flux distribution in the three individual detectors, and the overlaid X-ray contours onoptical imaging data. For the latter X-ray-optical overlays the online all-sky data base ofthe Second Digitized Sky Survey k (DSS 2) is queried for image cutouts in the red (DSS 2-red)and NIR (DSS 2-infrared) bands. The aim of the second X-ray screening stage is to identifyfalse detections originating from e.g. (iv) blends of three or more point sources, (v) spurioussources related to an underestimation of the local background, vi) chip boundary e ff ects,(vii) residuals from the correction of the so-called out-of-time event trails, and (viii) ‘optical k http://archive.eso.org/dss he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP .
36 2 .
84 3 .
36 4 .
00 4 .
76 5 .
64 6 .
72 8 .
00 9 .
52 11 . . . . . . . . . . . . . r c [arcsec] . . . . . . . . . . . . . . . . . . f . − . k e V [ − e r g c m − s − ] LBQS, θ = 1 . ′ p det ( f ) . . . . . r c f n ( r c ) c a b Figure 2.
Source detection simulation results.
Left:
Detection sensitivity for the central partof a deep (51.7 ksec) survey field with color coded recovery fractions according to the verticalcolor bar across the source flux versus core radius parameter plane. The dotted black lineindicates the XMM-
Newton resolution limit, whereas the dashed line follows the backgroundlimit as a function of source extent.
Right:
Illustration on how a completeness function p det ( f )is obtained as a function of flux f by weighting the detection probabilities in the source extentdirection with an assumed core-radius distribution (a) of the cluster population. Input curve (a)shows the observed local r c distribution scaled to apparent sizes at z =
1, whereas (b) and (c)are up- and down-scaled distributions by factors of 2. Plots adapted from M¨uhlegger (2010). loading’ residuals caused by bright optical sources. The conservative flagging of such ‘level2’ false detections reduces the original raw source catalog by an additional 20% resulting ina double X-ray screened input list of about 1300 extended sources with a remaining impuritylevel of 10-20% ¶ .The third and final screening stage aims to identify the optical counterparts associatedwith the extended X-ray sources based on the X-ray-optical overlays and additional queriesto the NASA Extragalactic Data Base + (NED) to check for known objects and redshiftinformation. Approximately 100 (8%) of the extended sources can be readily identified asnon-cluster objects, mostly nearby galaxies and galactic sources, e.g. supernova remnants.From the remaining list of ∼ z ∼ Newton fields, the final XDCP sample comprises 990individual galaxy cluster candidates. From this point on, the further XDCP survey e ff orts arefocussed on the identification and deeper study of the selected ∼
300 distant cluster candidates,i.e. the extended X-ray sources without optical counterpart. ¶ Based on a preliminary empirical evaluation with wide field follow-up imaging data. + he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP One of the main strengths of X-ray cluster surveys is the abilityto accurately quantify the detection process and the resulting e ff ective survey volume throughsimulations (see e.g. Pacaud et al. 2006; Burenin et al. 2007; Mantz et al. 2010; Lloyd-Davieset al. 2010). For the characterization of the XDCP survey sample a dedicated simulationpipeline was developed by M ¨uhlegger (2010) that follows the actual survey data and detectionprocedure as closely as possible.For the background limited regime of deep XMM- Newton fields, the minimum flux levels f lim required for the detection of idealized resolvable (i.e. r c > r min ) extended sources withangular core radius r c scale as f lim ( r c > r min ) ∝ r c · [ B ( Θ , φ ) / t e ff ( Θ )] / , where t e ff is the e ff ectiveexposure time at o ff -axis angle Θ and B is the total local background count rate, which canadditionally vary with azimuthal angle φ . This strong positional dependence of detectionsensitivities for a heterogeneous serendipitous XMM- Newton survey implies that the accuratereconstruction of the selection function requires a local approach for each solid angle elementof the X-ray coverage.To this end, the full XDCP survey area is characterized by analyzing the detectionperformance of 7.5 million simulated, circularly symmetric mock β -model ∗ cluster sourcesspanning a wide range of core radii (2-128 ′′ ) and net source counts (20-1280) in 25logarithmic steps each. Simulated clusters with a poissonized two-dimensional photondistribution are convolved with the local PSF and then placed directly into the observedXDCP survey fields at various o ff -axis angles and random azimuthal positions. This approachaccounts by design for all local properties at a given position in a survey field, such as localbackground, exposure time, and possible contamination from surrounding X-ray sources. Inorder to obtain su ffi cient statistics for the covered parameter space and the di ff erent positionsacross the FoV, more than 1500 field realizations are generated, each with ten additionalinserted mock clusters. These mock fields with simulated cluster sources of known flux andposition are then analyzed by the XDCP source detection pipeline for the primary detectionscheme with the optimized 0.35-2.4 keV band. The detected extended sources in each fieldrealization are subsequently matched to the simulated input catalog, from which the fractionof recovered detected cluster sources can be determined as a function of input flux, core radius,and o ff -axis angle.Figure 2 (left) shows the simulation results for the central part of one of the deepestXDCP survey fields with 51.7 ksec clean e ff ective exposure time, originally observed aspart of the Large Bright Quasar Survey (LBQS, Hewett et al. 1995). The shown detectionsensitivity as a function of total source flux versus angular core radius is representative forthe deepest part of XDCP, while the typical median survey sensitivity along the y-axis is afactor 2.5-3 higher. The figure illustrates well the XMM- Newton detection capabilities andits limitations. The ‘shark tooth’ shaped colored region of extended source detectability isconfined by two limits. The dotted black line at small core radii marks the manifestationof the XMM-
Newton resolution limit and is governed by the extent significance (
EXT ML )determination for the sources. The core radius detection threshold decreases slightly with ∗ More complicated (e.g. double- β ) models are currently not considered owing to the increasing complexity toadequately cover the model parameter space. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP DET ML ) in order toidentify the presence of a source with low central surface brightness above the backgroundlevel. This limit prevents the detection of very extended sources and closely follows theexpected scaling behavior f lim ∝ r c .The highest detection sensitivity is achieved at the tip of the ‘shark tooth’ for angular coreradii in the range 6-12 ′′ , corresponding to physical core sizes of 50-100 kpc at z > .
8. For suchrelatively compact cores, extended sources down to flux levels of ∼ − erg s − cm − in thedetection band can be identified for the deepest parts of the XDCP survey. Conversely, thisimplies a sweet spot for cluster detections at the lowest flux levels (i.e. the supplementarysample), e.g. at the highest redshifts of z > ∼ .
4, where this e ff ect introduces a detectionpreference towards clusters with compact cores. However, by applying a flux cut well abovethe known extreme tip of detectability the construction of fair and morphologically unbiasedcluster samples is still straightforward.In order to obtain detection probabilities as a function of flux p det ( f ) that provide anaverage over the cluster structures in the survey, the simulation results along the angular coreradius axis have to be weighted with the actual core radius distribution of the underlyingcluster population, which is illustrated in the right panel of Fig. 2. Ideally, one would like the z > . ′′ at z = ±
6% across our 0 . < ∼ z < ∼ . p det ( f )function. However, high- z clusters are expected to exhibit more compact cores due to thehigher critical background density at the collapse epoch. The e ff ect on the XDCP selectionfunction can be investigated by downscaling the local distribution by a factor of 2 (curve [c]),resulting in only a moderate change of the median flux limit by about 10%. The unknownstructural properties of the high- z cluster population are thus only minimally a ff ecting thesurvey sensitivity characterization, as long as the average high- z cluster core radii are notdecreasing by more than a factor of 2. A more significant decrease of the average detectionsensitivity would occur in the unexpected case of increasing core radii (curve [b]). In thefuture, we hope to recover the shape distribution function of distant clusters directly from oursurvey, once the statistics of systems with good X-ray data is su ffi ciently large.Another important result of the performed simulations is the determination of anoptimized maximal acceptable XMM- Newton o ff -axis angle for which the enclosed detectorarea is well characterizable, without compromising the detection sensitivity and reliability dueto the o ff -axis PSF characteristics and other instrumental artifacts. This optimal maximum o ff -axis angle of the well characterizable detector area was found to be Θ max = ′ , implying anenclosed solid angle of 0.126 deg per XMM- Newton field. At Θ max = ′ , the PSF FWHM he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Table 1.
Basic characteristics of the XDCP X-ray coverage for di ff erent survey levels.Properties that apply to the full area of a given survey level are indicated by a ‘yes’.Full X-ray coverage Main Survey
Gold CoverageSurvey Level SL 1
SL 2
SL 3Solid Angle [deg ] 76.1 − erg s − cm − ] ∼ ∼ | b | ≥ ◦ yes yes yesDEC ≤ ◦ yes yes yesXMM Nom. Exp. ≥
10 ksec yes yes yesXMM O ff -axis Angle ≤ ′ yes yesXMM Clean Exp. ≥
10 ksec yesN H ≤ × cm − yesLow Background / Contamination yes blurring factor is about 40% increased and the e ff ective area is decreased to 44% compared tothe on-axis characteristics of XMM- Newton .While the analysis of the full XDCP survey simulations is still ongoing, the basic globalsurvey characteristics are known and are discussed in the next section.
Table 1 provides an overview ofthe XDCP survey coverage and the basic sample properties for three di ff erent subsets of theX-ray data, called survey levels (SL). The full X-ray coverage (SL 1) comprises a total solidangle of non-overlapping area of 76.1 deg from which a combined galaxy cluster candidatesample of 990 sources was identified. This corresponds to a candidate surface density of 13.0per deg , which is comparable to the total cluster density in the XMM-LSS survey (Adamiet al. 2011). The SL 1 coverage has an average soft band sensitivity for extended sourcesof ∼ − erg s − cm − and a sample impurity of up to 20%. The main aim of this full X-ray coverage sample is to increase the area for the search of the rarest, most massive high- z systems, which requires the largest possible survey volume for these sources with flux levelsbright enough to be identified even at large detector o ff -axis angles.A complete and detailed survey characterization will be available for the main survey(SL 2) which is constrained to the X-ray coverage enclosed within the inner 12 ′ of thedetector area, comprising 49.4 deg and 752 cluster candidates. The average sensitivityis ∼ . × − erg s − cm − , which is also reflected in the increased surface density of15.2 per deg and significantly improved purity levels. The coverage of SL 2 will be themaximum solid angle for studies that require a detailed knowledge of the selection function,i.e. cosmological applications.As the starting point for the construction of a first sizable and statistically completesample of z ≥ . of X-ray data. This SL 3 has the additional he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ff ective clean exposure time of ≥
10 ksec, upper limits on the Galactichydrogen absorption column, and stricter field selection cuts concerning the background levelsand contaminating sources in the FoV. The simulation run for all 160 SL 3 fields is completed,and yielded an average soft band cluster detection sensitivity of 0 . × − erg s − cm − , witha candidate surface density of 17.5 per deg and an expected initial purity level of > coverage in the COSMOSfield (Finoguenov et al. 2007) and dedicated contiguous cluster surveys with XMM-LSS-likeexposure times (Pacaud et al. 2006).The initial impurity levels of the di ff erent subsamples are based on the selection ofextended sources down to the pursued low extent significance threshold of 2-3 σ for thesupplementary sample. With the limiting average flux levels for SL 2 or SL 3 at hand, it isstraightforward to construct statistically complete cluster subsamples with negligible impuritylevels based on subsequently applied flux cuts. As an example for the ‘gold coverage’ ofSL 3, a minimum flux cut of 1 . × − erg s − cm − imposed on the confirmed z ≥ . ∼
30% distant candidate subsample without initial optical counterpart identification,most of the initial sample impurity is part of these ∼
300 candidates selected for follow-upimaging. Hence, the fraction of spurious sources that passed the X-ray screening procedureof Sect. 3.1.3 has to be identified as false positives during the follow-up imaging campaignsdiscussed in Sect. 3.2. Preliminary results suggest that about 1 / z > ∼ . z > (0.088 deg ) for the presence of a galaxy cluster center within 30 ′′ (60 ′′ )around initially detected spurious X-ray positions. This sky area is to be compared to theexpected surface density of the objects we are looking for, which is of the order of one z ≥ . , implying a chance of <
10% in the case of allowed cluster center o ff setsof up to 1 ′ and even a factor of 4 lower for the 0.5 ′ o ff set radius. This estimate resultsin the conclusion that the odds of finding even a single ‘chance cluster’ in the full XDCPsurvey that is randomly associated with a low significance extended X-ray source is very low.However, the situation may look di ff erent in case systematic astrophysical e ff ects that canmimic extended X-ray emission are present or enhanced in high- z group and low-mass clusterenvironments, such as multiple weak AGN in clusters that could cause a systematic pointsource confusion in these systems. Answers on the presence and abundance of such systems he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Chandra
X-ray data is available for some of the potentially contaminated systems.
The automated XDCP source detection pipelineprovides approximate source parameters, such as estimates for the source flux and the coreradius. However, for a detailed characterization of the X-ray properties of spectroscopicallyconfirmed systems a more elaborate ‘post detection processing’ is required for determiningaccurate cluster luminosities and other physical parameters. To this end, we re-processthe archival data with the latest SAS version and calibration database, manually check andoptimize the quiescent time periods used for the double flare cleaning process, and checkfor potential contaminations of the source environment under study. At this stage, also thecombination of overlapping XMM-
Newton fields is considered for cases of significant signal-to-noise gain of the source. We then apply an extended version of the growth curve analysis(GCA) method of B¨ohringer et al. (2000) to the point-source excised cluster emission in orderto obtain an accurate 0.5-2 keV flux measurement of the source as a function of cluster-centricradius. Examples of the cumulative background-subtracted source flux for two clusters areshown in Sect. 4 in Fig. 8. The total cluster source flux is determined iteratively by fittinga line to the plateau level of the flux and measuring the enclosed total source flux within theplateau radius. The uncertainty of the flux measurement is determined from the Poisson errorsplus a 5% systematic uncertainty of the background estimation.The soft band restframe luminosity L . − , and the bolometric luminosity L bolX , arethen self-consistently determined within R by iterating the estimates for the cluster radiusand ICM temperature derived from the scaling relations of Pratt et al. (2009) (for details seeSuhada et al., subm.). These X-ray luminosity measurements are the physical key propertiesof the distant XDCP clusters as these are, by survey design, available for all newly detectedsystems. The application of the latest calibration of the L X -T X and L X -M scaling relationsout to high- z allows subsequent robust estimates of the other fundamental properties ICMtemperature T X and total cluster mass M (Reichert et al., 2011). Tentative first direct T X constraints from X-ray spectroscopy are feasible when several hundred source counts areavailable, which is the case for about 1 / The task for the follow-up imaging of the second XDCP survey stage is quite challenging: thephotometric identification of about 300 X-ray selected z > . ffi ciently deep to reach the highest accessible cluster redshifts and toreliably flag the unavoidable fraction of false positives. It is obvious that time and telescopee ffi cient imaging strategies are required to tackle this observational challenge.After more than 20 dedicated XDCP imaging campaigns, the data acquisition for theimaging follow-up is now close to completion. In total, we applied and tested five di ff erentimaging strategies at five telescopes using eight optical and NIR imaging instruments: (i)R + z band imaging with VLT / FORS 2, (ii) z + H imaging with OMEGA2000 at the Calar Alto3.5 m telescope , (iii) I + H imaging at NTT with SOFI, EMMI, and EFOSC 2 , (iv) g + r + i + z + H he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 3.
Simple stellar population models for the color evolution of passively evolvinggalaxies as a function of redshift.
Top left:
Color evolution diagram for a selection of colorsbased on models with solar metallicity and stellar formation redshift of z f = Top right:
The R − z, z − H, and I − H color shifted to the same origin at z = ff erent redshift regimes. Bottom:
SSP model gridsof the R − z ( left ) and z − H ( right ) color evolution for formation redshifts of three, five, and ten( di ff erent colors ) and solar ( solid lines ) and three times solar metallicity ( dashed lines ). imaging at the CTIO Blanco 4 m with MOSAIC II and ISPI , (v) and g + r + i + z + J + H + K s withthe 7-band imager GROND at the ESO / MPG 2.2 m telescope.In the following section, we will discuss cluster identification performance predictionsfor the di ff erent methods, provide an overview of a NIR data reduction pipeline developedfor the project, introduce the applied redshift estimation method, and finally evaluate andcompare the performance of the R − z and z − H colors based on our spectroscopic sample.
The minimum requirement for the reliable identification of opticalcounterparts of z > . z are variants of the red-sequence methodproposed by Gladders and Yee (2000, 2005). Based on their predictions, the optical R − z he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Table 2.
Properties of the main filters used in this section. The second column lists the centralwavelengths, columns 3-6 show the expected apparent Vega magnitudes of L* passivelyevolving galaxies with formation redshift z f = ff erent redshifts,and the last column indicates the additive positive o ff sets for the conversion to AB magnitudes.Filter Center m*( z = .
5) m*( z = .
0) m*( z = .
5) m*( z = .
0) mAB (Vega) µ m mag mag mag mag magR special Figure 4.
Absolute predicted redshift uncertainties as a function of z of di ff erent red-sequencemethods for a photometric color error assumption of σ color ≈ . · (1 + z ) mag. Errorestimates were obtained from the derivatives of the smoothed model colors in Fig. 3 using σ z ≈ dz / d (X − Y) · σ color , where (X − Y) denotes the photometric method. The black solidline illustrates the estimated redshift error for the R − z technique under the assumption of aformation redshift of z f =
5, blue shows the z − H method, and red I − H. The dotted lines use amodel formation redshift of z f = color of the cluster red-sequence was expected to yield reliable redshift estimates out to z ∼ .
4. The original XDCP follow-up imaging strategy was based upon this R − z methodusing short snapshot imaging with VLT / FORS 2 in the z
Gunn (8 min) and R
Special (16 min)broadband filters. Figure 3 displays simple stellar population (SSP) model predictions basedon PEGASE 2 (Fioc and Rocca-Volmerange 1997) for the observed redshift evolution of theR − z color in comparison to other optical / NIR colors (top), and for a model grid of three stellarformation redshifts ( z f = , ,
10) and two metallicities ( Z = Z ⊙ , Z ⊙ ).The limitation that R − z follow-up imaging of targeted high- z candidates is only e ffi cientwith the capabilities and sensitivity of VLT / FORS 2 led us to develop alternative strategiesthat are applicable to 4 m-class telescopes and, at the same time, provide a higher redshift he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP z > ∼ . ffi ciency, redshiftsensitivity, and redshift limit are z − H and I − H, which are shown in the upper right panel ofFig. 3 in comparison to R − z. Since the achievable accuracy of the red-sequence based redshiftestimate depends on the gradient of the color d (X − Y) / dz it is immediately evident that z − Hand I − H are expected to provide significantly improved redshifts at z > .
9. Furthermore, alimiting magnitude of H lim ∼
21 mag (Vega) is reachable in less than 1 h with NIR imagersat 4 m-class telescopes, corresponding to apparent magnitudes of passive galaxies of m* + + z ≃ . z ≃ σ z ≈ σ color · dz / d (X − Y),where (X − Y) denotes the photometric method, σ color the observational error of the mean red-sequence color, and σ z the resulting absolute redshift uncertainty. For a realistic photometriccolor error assumption from good quality data of σ color ≈ . · (1 + z ) mag, we obtain theexpected absolute redshift uncertainties shown in Fig. 4. As can be seen, the z − H and I − Hmethods are promising to deliver redshift estimates with uncertainties of σ z < ∼ . z ∼ .
5, while the high- z uncertainty based on the R − z color is sensitive to the assumedstellar formation redshift of the model and increases dramatically beyond z ∼ .
9, when the4000 Å break shifts out of the R
Special filter.We implemented and tested the z − H imaging technique (Fig. 5, right panels) for theidentification of high- z clusters in the year 2006 at the Calar Alto 3.5 m telescope using theNIR wide field camera OMEGA2000, with results shown throughout this work. Observationsbased on the I − H method followed from 2007 on at the 3.5 m NTT with the instrumentcombination SOFI / EMMI and SOFI / EFOSC 2. First promising I − H results at z > . / ISPI and at the ESO / MPG 2.2 m telescope with GROND allow theflexibility to make use of all of the discussed colors.
In the following we provide a brief overview of thereduction and analysis of the Calar Alto OMEGA2000 near-infrared data, which is the basisfor many results presented here, in particular for the new systems presented in Sects. 4.2 & 4.3.Details on the reduction of imaging data from other telescopes can be found in Schwope et al.(2010) for VLT / FORS 2 (z + R), in Santos et al. (2011) for NTT (I + H), in Zenteno et al. (2011)for CTIO (griz), and in Pierini et al. (subm.) for GROND data (grizJHK s ).OMEGA2000 (Bailer-Jones et al. 2000) is the wide-field NIR prime focus camera atthe Calar Alto 3.5 m telescope with a 15.4 ′ × ′ FoV and a pixel scale of 0.45 ′′ per pixel.Besides the standard NIR broadband filters J H K s , the instrument is also equipped with az-band filter in which the HAWAII-2 detector array still features a high quantum e ffi ciency he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ∼ / instrument system o ff ers an online reduction pipeline(Fassbender 2003), which allows the evaluation of the presence of a distant cluster in real-time ( + ∗ was developed for XDCP witha special focus on distant cluster applications, i.e. faint galaxies. The full data reductionprocedure can be broken up into the independent processing blocks (i) single image reduction,(ii) image summation, (iii) object mask creation, followed by a second iteration of steps(i) + (ii) with an optimized sky background modeling. For the single image reduction theindividual 40 sec (60 sec) H-band (z-band) exposures are first flatfielded and bad-pixel-corrected. A preliminary, first iteration NIR sky background model is determined from theseven dithered images taken closest to the frame of consideration (i.e. ± ff sets while identifying and rejecting cosmic ray events in theprocess. During the stacking process, individual exposures are automatically weighted toyield the optimal SNR in the final stack. Following Gabasch (2004), this optimal weightfactor in the limit of faint sources scales as T / ( B · σ ), where T is the transparency determinedfrom monitoring the fluxes of stars in each frame, B represents the background level, and σ denotes the measured seeing .The first-iteration summed image stacks in each filter are then used to create an objectmask, which flags regions with detectable object flux above the background noise. Forthe first iteration reduction, the signal of these objects was still in the images used for thesky background model, resulting in determined sky levels which are slightly biased highat these object position, which in turn translates into a slight background over-subtraction.This background bias is overcome in the second iteration of the reduction process, wherethe object fluxes in each individual exposure are masked out and replaced with the medianlevel of the surrounding unmasked detector area prior to the use of these flux-removed framesas input for the sky modeling process of the time-adjacent images. This results in the finalunbiased reduced single images, which are again stacked in sky coordinates to produce thesecond iteration final deep image stacks. These final co-added images, based on the discusseddouble-background subtraction procedure and the optimal weighting process for faint objects,now constitute the basis for the further analysis and distant cluster identification process. As a prerequisite for obtaining good distant cluster redshiftsestimates, reliable galaxy photometry and color measurements are required as part of thenext analysis block. To this end, the final image stacks in the z- and H-band are co-alignedonto identical pixel coordinate grids, i.e. the same objects in both bands have identical ∗ The full OMEGA2000 pipeline is freely available upon request. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ♯ for the source detection and for the green channel layer of RGBcolor composites. The z- and H-band stacks, in which the actual photometric measurementsare performed, are then PSF-matched to the larger on-frame measured seeing value ofthe two bands (typically 0.8-1.5 ′′ ) by applying an appropriate Gaussian smoothing kernel(i.e. σ = σ + σ ) to the frame with better seeing. As the final preparatory step for thephotometry, all frames are equipped with a proper equatorial world coordinate system (WCS)from the astrometric plate solution fit with the WCS Tools †† software package.The actual source photometry is performed with SExtractor (Bertin and Arnouts 1996)run in dual image mode, where the deep detection image is used to find the sources down tofaint magnitudes, and the photometric parameters are then extracted directly from the PSF-matched z- and H-band images at the detected source positions. The photometric calibrationis achieved in the H-band with stars from the 2-Micron All Sky Survey (2MASS, Cutriet al. 2003) directly observed within the large FoV of the science frame. The z-band isphotometrically calibrated by means of dedicated standard star observations (Smith et al.2002) throughout the night, and short photometric overlap observations of the science field inphotometric conditions.The z − H versus H color-magnitude diagram is constructed from the Galactic extinction-corrected (Schlegel et al. 1998), i.e. de-reddened, magnitudes and colors of all galaxies in theFoV, where objects in close proximity (r < ′′ / ′′ ) to the X-ray centroid of the candidatesource are highlighted (see Fig. 8). Total H-band magnitudes ( MAG AUTO ) are used alongthe x-axis since these are directly related to the model predictions. The z − H object colorsare computed from isophotal magnitudes (
MAG ISO ), which are more accurate for colordeterminations, since the object flux measurements are restricted to the connected pixelsabove the detection threshold without extrapolations.As the final step, a color-based redshift can be estimated from the analysis of the z − Hversus H color-magnitude diagram. Since the location and center of the potential distantcluster is already accurately known from the X-ray centroid, the only unknowns of thecandidate system to solve for are the redshift and richness above our limiting magnitude. Oneof the main advantages of the X-ray selected XDCP sample is its unbiased nature with respectto the galaxy populations of the systems, which we do not want to give up by requiring afully developed red-sequence, in particular at the unexplored high-redshift end. Furthermore,the number of detectable cluster galaxies at the limiting depth in z > . <
10) with even fewer accurate color measurements, especially with data taken inpoor observing conditions or pre-defined exposure times. A robust redshift estimator for ourpurposes should thus be able to work with few cluster galaxies and without requiring a highsignal-to-noise red-sequence.For the blind redshift estimation of previously unknown distant cluster candidates, the ♯ For clusters at z > ∼ . †† he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ′′ from the X-ray centroid, (2) apply a color cut of ± N within 30 ′′ in this color interval and above themagnitude limit, (3) select the central ∼
68% percentile of the N galaxies in color space (for N ≥
4, otherwise all) yielding N , (4) determine the minimum (min ) and maximum (max )color of the N galaxies from which the final best color estimate col = (min + max ) / σ = (max -min ) / ± σ is then compared to the prediction ofthe input SSP galaxy evolution model to yield the final color-based redshift estimate anduncertainty z mod ± σ z for the candidate system, with a richness estimator N and color spread σ that allow conclusions on the existence of a cluster and the presence of a red-sequencein consideration of the magnitude limit of the data set. This color estimator is much lessdemanding in terms of data depth and quality, and with respect to requiring the presence of anevolved galaxy population for the identification of a distant candidate, compared to actuallybasing the candidate evaluation on a significant discernible red-sequence in the CMD of thefollow-up data. This is of particular importance when selecting candidate clusters at z > . rely only on two assumptions: (i) the presence of ≥ ′′ fromthe X-ray centroid (i.e. the central ≃ ), whose colors are significantly redder thanthe passive member galaxies of the distant cluster candidate. The available photometry probesthe bright end of the galaxy luminosity function at the cluster redshift, where we expect theSSP models to perform reasonably well. When extending this redshift estimation procedureto lower- z calibration clusters at z < ∼ .
6, as in Sect. 3.2.4, only magnitudes brighter than m* + and coloruncertainty σ of this empirical approach are hence not necessarily equivalent to the color andscatter of the physical cluster red-sequence of confirmed early-type passive member galaxies,which are only accessible with high quality data and extensive spectroscopic information.However, in the limit of a discernible, well populated, and tight red-sequence in the CMD,col and σ will converge to the intrinsic physical parameters of the underlying red-sequence.Using the extensive spectroscopic information of XDCP (Sect. 3.3 & 5), we can nowput the most established redshift estimation techniques, R − z and z − H, to a critical test andevaluate their performance in practice based on real distant cluster follow-up imaging data. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 5.
Comparison of the observed R − z (left) and z − H (right) colors of spectroscopicallyconfirmed clusters as a function of redshift. The top panels show the measured valuescol ± σ in comparison to the solar metallicity SSP galaxy evolution models with stellarformation redshift z f = z f = z f = ff set for R − z (left) and the average model for z − H (right). The bottom panelsshow the expected achievable absolute redshift uncertainty based on these models (red dottedline) and the observed redshift o ff sets of photometric model redshifts z mod (with uncertainties)and spectroscopic redshifts z spec for each system (blue points). Blue crosses in the left panelindicate the redshift o ff sets based on the original model (red dashed line), the open symbol ofthe highest-z cluster means that no red-sequence was discernible based on the data. ffi cacy of the R − z and z − H colors.
Figure 5 (top panels) displays themeasured color col ± σ for spectroscopically confirmed clusters with available R − z (left)or z − H imaging data (right) as a function of the spectroscopic redshift. The 20 confirmedsystems with available FORS 2 (FoV 6.8 ′ × ′ ) R − z data are in their majority targeted XDCPdistant cluster candidates, whereas the larger 15.4 ′ × ′ FoV of OMEGA2000 enabled thecoverage of additional known lower-z calibration clusters at z < ∼ . − H imagingdata, with five overlapping systems present in both data sets (see Table 5, column (7)).The R − z color evolution (col ) as a function of redshift (top left panel) shows a low-scatter behavior with relatively small photometric uncertainties based on the FORS 2 datawith the targeted exposure depth of 8 min (16 min) in z (R). However, as predicted from theSSP models (dashed lines in Fig. 5 and lower left panel of Fig. 3) the R − z color flattens outat z > ∼ . z . As discussed in Sect. 3.2.1 and Fig. 4 thisflattening directly translates into significantly increasing color-based redshift uncertainties oreven a full model redshift degeneracy. The observed color-based redshift o ff sets z mod − z spec with the derived uncertainty interval are plotted in the lower left panel, together with thediscussed predicted absolute uncertainty of the stellar formation redshift z f = he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ff erence with increasing redshiftis in very good agreement with the prediction, in particular the increasing redshift degeneracyat z > ∼
1. The open symbol for cluster X2215a at z = .
457 (see Table 6) indicates that nosignature of a red-sequence (i.e. ≥ − zversus z CMD, even though the determined col from the three reddest galaxies resultedin a reasonable color. Since the physical red-sequence is present for this system (Hiltonet al. 2009), which is also seen the z − H CMD, this indicates that we have surpassed theredshift limit of the R − z observing strategy, which was originally designed to allow clusteridentifications up to z ∼ . z f = − z color o ff set of + z > . − z color. The origin of this observed R − z color o ff set of + Gunn and R
Special filters,or a systematic when calibrating magnitudes observed in the cut-on z
Gunn filter to the standardSDSS z-band system by means of SDSS standard star observations. For most of the R − zcalibration clusters in Fig. 5 we have results based on two independent reduction pipelines,yielding consistent color measurements. Our used SSP color evolution model was also cross-checked with a consistent independent model, providing support for the quality of both thereduction and the model predictions. Moreover, a physical explanation for the redder observedcolors by invoking super-solar metallicities for the average passive galaxy population (seelower left panel of Fig. 3) can be ruled out as well, since such an o ff set would then also beevident in the right panel for the z − H color.The observed z − H color evolution on the other hand is in very good agreement with theabsolute model prediction over the full probed redshift baseline 0 . < ∼ z < ∼ .
55, fully consistentwith both the z f = z f = · (1 + z ) mag color error assumed for the redshift uncertainty estimate. Even withthese larger observational uncertainties, it is evident that the z − H color clearly outperformsthe R − z approach at z > ∼ .
9, as expected from Fig. 4. Reliable z − H cluster redshift estimateshave so far been obtained out to z ∼ . ffi cient X-ray brightness withinour survey area.From the observed color evolution of the tested techniques we can confirm that R − z he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP z < ∼ .
9, given a su ffi cientlyaccurate galaxy evolution model. For clusters at z > . z < . z > ∼ . − z technique provides an e ffi cient basis up to the limiting redshiftof z ∼ .
4. At z > .
9, the newly established z − H color method provides significantly bettercolor-based redshift estimates and allows clusters identifications out to z > ∼ .
5, whereas theuncertainties at the XDCP sample separation redshift of z ≃ . − z. We provide the empirically calibrated, best fitting R − z and z − H color evolution modelsas a function of redshift in text file format as part of the supplementary material for this paper.Based on our results, a three-band follow-up approach in R + z + H for future clusteridentification projects, e.g. eROSITA (Predehl et al. 2010), can provide color-based clusterredshift estimates with uncertainties of ∆ z < ∼ . . < ∼ z < ∼ .
5. A similar performance may also be achievable with a two-band approachbased on the I − H color, which is currently still in the evaluation phase within XDCP.
The third and final stage in the XDCP distant cluster identification process is the spectroscopicconfirmation, which is an inevitable and crucial step for all subsequent studies of the z > . ffi cient spectroscopic cluster confirmations out to z > ∼ . ∼ R radii of the distant cluster candidates are in most cases < ∼ ′ and the high densitycore from where the X-ray emission was detected is typically of the order of 30 ′′ . This restrictsthe slit placement to approximately five within the region of the X-ray emission and ten within R . Moreover, the apparent magnitude of cluster galaxies of characteristic luminosity L* isclose to the spectroscopic limit for reasonable exposure times once approaching z ∼ . r < ∼ ′ ) with the detected X-ray emission, and (3) we finda minimum of three concordant redshifts of associated galaxies. Since we start with X-rayselected candidates, this strict XDCP definition for confirmed clusters is expected to yield aclean cluster sample concerning the existence of truly gravitationally bound structures. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP The spectroscopic confirmation of newly detected distantX-ray clusters is one of the prime activities for the current survey phase. In order to allowan e ffi cient and high quality reduction of the spectroscopic data for dozens of systems, wedeveloped a new spectroscopic reduction pipeline called F-VIPGI (Nastasi et al., in prep.),which is the FORS 2 adaptation of the Vimos Interactive Pipeline Graphical Interface (VIPGI,Scodeggio et al. 2005).For the spectroscopic distant cluster confirmations we make use of the 300 I grism( λ c = R =
660 and a wavelength coverage of 6 000–10 500 Å. The wavelength coverage on the blue end can be extended down to ∼ ′′ and a minimumslit length of 6 ′′ allow the placement of about 40 target slits over the 6.8 ′ × ′ FORS 2 FoV.Slits are preferentially placed on color-selected galaxies close to the expected red-sequencecolor at the estimated redshift with the highest priority assigned to objects within the detectedX-ray emission. Individual exposures are taken with net integration times of 21 min, whereasthe total number of exposures varies from 2 to 10 depending on the estimated system redshiftand the faintness of the targeted galaxies.The reduction process includes all standard reduction steps, i.e. bias subtraction,flatfielding, background subtraction, wavelength calibration, extraction of 1-D spectra, andcombination of all spectra from the individual exposures including cosmic ray event rejection.F-VIPGI performs these steps in a semi-automated way with the possibility of interactivequality checks after each process step. The wavelength calibration is achieved by means of aHelium-Argon reference line spectrum observed through the same MXU mask, which allowsan absolute calibration with typical rms errors of < ∼ The final redshifts are obtained by cross-correlating thereduced spectra with a spectral template library over a wide range of object classes using thesoftware packages EZ (Garilli et al. 2010) and IRAF/RVSAO (Kurtz and Mink 1998). The bestfitting redshift solutions are interactively checked by making use of the graphical VIPGI tools,which allow a simultaneous assessment of the observed spectrum with overplotted redshiftedline features, the corresponding sky-subtracted 2-D spectrum, and possible contaminations ofobserved features related to sky emission lines. This way, the final spectroscopic redshift foreach galaxy can be determined with typical absolute uncertainties of σ z ≃ (2-4) × − (seee.g. Table 3), corresponding to a restframe velocity uncertainty of 30-60 km / s in z ∼ ff sets of < / s from the redshift peakare considered, corresponding to ∆ z < . × (1 + z C ). The systemic cluster redshift z C canthen be robustly determined as the median of the outlier-clipped redshift distribution ofspectroscopic member galaxies. For the cases with a su ffi ciently high number of identifiedspectroscopic members ( > ∼
8) approximate cluster velocity dispersions are computed byapplying the methods of Danese et al. (1980) and Beers et al. (1990). he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 6.
Properties of the cluster XMMU J0035.8-4312 / SpARCS J003550-431224 at z = Left panel:
Color composite image (gr + iz + H) of the 2.5 ′ × ′ cluster environmentwith XMM- Newton
X-ray contours overlaid in yellow (North is up, East is to the left).Blue(red)-shifted spectroscopic cluster members with respect to the system redshift are markedby small cyan (red) circles, white circle indicate the 0.5 ′ and 1 ′ radii around the X-ray centroidposition. Right panel:
XMM-
Newton
X-ray surface brightness profile of the cluster’s extendedemission for the PN (black), MOS1 (green), and MOS2 detectors. Dashed (solid) red linesshow the best fit (PSF-corrected) single β -model profile for the PN (upper curves) and thecombined signal of the MOS instruments (lower curves).
4. New distant clusters results
In the following section we present results on two newly identified clusters at z ∼ . = The galaxy cluster SpARCS J003550-431224 was spectroscopically confirmed at a redshiftof z = Spitzer
Adaptation of the Red-sequence Cluster Survey (SpARCS) (e.g. Muzzin et al. 2009; Demarcoet al. 2010). This optically rich system was selected within SpARCS based on its red-sequencein z ′ − µ m color space and contains 10 spectroscopic members in the range 1 . < z < ∼ . ±
230 km / s was derived.Within the XDCP survey, the cluster was independently X-ray selected as the verysignificantly extended X-ray source XMMU J0035.8-4312 at an o ff -axis angle of 6.3 ′ duringthe initial source detection run (Sect. 3.1) in the XMM- Newton field with observation ID0148960101 and an e ff ective clean exposure time of 47.2 ksec. Owing to the lack of anoptical counterpart, the X-ray source was classified as a promising distant cluster candidateand followed-up at the 4 m CTIO / Blanco telescope with MOSAIC II in the g r i z bands on 11 he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP R ≃ ′′ (outer white circle)with spectroscopically confirmed members marked as either blue-shifted with respect to thesystem redshift (cyan circles) or red-shifted (red circles).Both images have the logarithmically spaced XMM- Newton
X-ray surface brightnesscontours overlaid in yellow, from which the rather peculiar and irregular X-ray sourcemorphology is evident in comparison to the full distant cluster sample shown in Fig. 9. Threedistinct local surface brightness maxima (left panel of Fig. 6) appear to be discernible in thecurrent data within the inner 30 ′′ from the X-ray centroid, which is determined as the ‘center-of-mass’ of the extended X-ray emission. This emission is characterized by the most extendedsurface brightness distribution among all clusters in the presented sample with an e ff ectivecore radius determined from the radial profile fit (right panel of Fig. 6) of r c ≃ ′′ + − ≃
280 kpc( β = + − . ), consistent with the original source detection value of r c ≃ ′′ ≃
260 kpc (forfixed β = / ffi ciently high flux level off . − , ≃ (0 . ± . × − erg s − cm − results in one of the highest extent significances(see Sect. 3.1.4) for XMMU J0035.8-4312 among the full current distant cluster sample, with EXT ML ≃
56 and a corresponding formal probability for a spurious extent of < − .The observed di ff use and distorted X-ray morphology of XMMU J0035.8-4312 can bestbe explained by an ongoing major merger scenario, where at least two main componentsare in the process of coalescence with bulk flow velocities mainly along the line-of-sight.This scenario is supported by the bimodal velocity structure of the spectroscopic membersreported in Wilson et al. (2009) with five member galaxies at z < .
33 (cyan circles in Fig. 6)and five other members centered around z ∼ .
34 (red circles). The median redshifts of thetwo spectroscopic member bins that are likely associated with di ff erent sub-components ofthe merging process di ff er by ∆ z = .
013 or a rest-frame velocity o ff set of ∆ z ≃ / s,which is typical for major mergers (e.g. Markevitch and Vikhlinin 2007). The velocity sub-structure of XMMU J0035.8-4312 is also visible in the spatial distribution of the spectroscopicmembers, where the blue-shifted galaxies, i.e. infalling from behind the cluster, are associatedwith the Southern and South-Western extensions of the X-ray emission, whereas the galaxieswith positive rest-frame velocities are all located in the Northern half of the system. Thebrightest cluster galaxy (BCG, larger cyan circle) is associated with one of the local X-raypeaks 13 ′′ (109 kpc) away from the X-ray centroid and could possibly be the former center ofthe infalling (blue-shifted) component from the SW radial direction. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 7.
Optical and X-ray properties of the clusters XDCP J0027.2 + z = + z = Top panels: ′ × ′ z + H bandcolor composite images of the clusters with XMM-
Newton
X-ray contours overlaid in yellow(North is up, East is to the left). Spectroscopic member galaxies are indicated by small circles,the two large circles mark the 0.5 ′ and 1 ′ radii around the X-ray centroid position. Centralpanels:
Same as above for a 1.5 ′ × ′ zoom on the core region with the black backgroundremapped to gray scale for contrast enhancement. Bottom panels: ′ × ′ H-band images ofthe cluster environments with X-ray contours in blue and density contours of color selectedgalaxies close to the expected red-sequence color in red, with small red circles indicating theindividual galaxies. Black circles have the same meaning as the white ones above. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP f X ( < r) [ − e r g s − c m − ] PNMOS1+MOS2r r plat F plat PNMOS1+MOS2r r plat F plat f X ( < r) [ − e r g s − c m − ] MOS1+MOS2r r plat F plat MOS1+MOS2r r plat F plat Figure 8.
Physical properties of the ICM and galaxy populations of the clusters of Fig. 7.
Top:
Growth curve of the extended X-ray emission measured for the PN (blue) and MOSdetectors (red) in the 0.5-2 keV band. Poisson errors plus 5% background uncertainties aredisplayed by the dashed lines, the vertical solid (dotted) lines depict the R (plateau level)radii.
Center: z − H versus H CMDs of the cluster fields with galaxies within 30 ′′ (60 ′′ ) fromthe X-ray centroid marked in red (green), and spectroscopic members (black squares) at r > ′ shown in blue. Black lines indicate the 50% completeness limits, blue lines the H* magnitudeat the cluster redshift, and the red dashed lines the expected color of a z f = Bottom:
Member galaxy spectra with indicated redshifted spectral features, IDs correspond toTable 3. Atmospheric absorption (top) and emission (bottom) features are overplotted in red. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP / SpARCS J003550-431224 we followed the approach of Sect. 3.1.6 and measure a soft-band luminosity of L . − , ≃ (0 . ± . × erg s − and a bolometric energy output ofL bolX , ≃ (1 . ± . × erg s − . With approximately 370 source counts in the soft bandthe cluster signal is su ffi cient to additionally allow a direct spectroscopic ICM temperaturedetermination of T X ≃ . + − keV using a local background extraction region close to the cluster(see Table 4, left column).From the measurements of L bolX , and T X we can derive total cluster mass estimatesbased on the latest M-L and M-T scaling relations (see Sect. 5.2). The luminosity-basedmass estimate M L X ≃ . + . − . × M ⊙ and the independent temperature-based one M T X ≃ . + . − . × M ⊙ are fully consistent, indicating that the observed merging process doesnot have a significant influence on the system’s location on the L-T relation comparedto relaxed clusters. We can hence establish a robust X-ray-based total mass estimate forXMMU J0035.8-4312 / SpARCS J003550-431224 of M ≃ × M ⊙ ( ± σ r ≃ (9 . ± . × M ⊙ is biased high by abouta factor of 4-5 as a result of the presented evidence for major merging activity preferentiallyalong the line-of-sight. + = We now present results of the newly confirmed cluster XDCP J0027.2 + z = Newton field with OBSID 0050140201 and an e ff ective clean exposure timeof 41.8 ksec at an o ff -axis angle of 11.1 ′ (see Tables 4 & 6). The X-ray surface brightnessdistribution of the system is more compact compared to SpARCS J003550-431224 with acore radius (for β = /
3) of r c ≃ ′′ ( ≃
110 kpc) and a mostly regular morphology featuringan elongation in the SE-NW direction as shown in Fig. 7 (left panels). The growth curveanalysis (Sect. 3.1.6) for the source yielded an unabsorbed soft-band flux of f . − , = (0 . ± . × − erg s − cm − as shown in the upper left panel of Fig. 8.Imaging follow-up observations (Sect. 3.2) in the H- (50 min) and z-band (34 min) tookplace at the Calar Alto 3.5m telescope with the OMEGA2000 NIR camera on 3 / ′′ (1.91 ′′ ) in H (z) and limiting 50% completenessVega magnitudes of H lim ≃
21 mag and z lim ≃ . + − H versus H color-magnitude diagram is displayed in the central left panel ofFig. 8 with the bright end of the observed red-sequence at the expected z − H SSP modelcolor of ≃ z f =
5, red dashed line). Towards fainter magnitudes (H > ∼
19 mag) the he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ′′ for the PSFmatched photometry resulting in a broadening of the observed cluster red-sequence. In orderto evaluate the spatial overdensity of galaxies close to the expected red-sequence color, thecolor cut 2.0 ≤ z − H ≤ z f = z f = with abackground of (3.1 ± − . The main red galaxy density concentration coincides withthe X-ray emission peak within a few arcseconds, indicating a dynamically evolved maincluster. A secondary galaxy density peak towards the East with a superimposed X-ray pointsource suggests an infalling structure on the cluster outskirts.Spectroscopic observations of the cluster environment with VLT / FORS 2 (Sect. 3.3) wereperformed on 6 September 2010 in 1 ′′ seeing conditions for a total net exposure of 1.5 h(run ID: 085.A-0647). Six secure spectroscopic cluster members with a median redshift of z = ff set of about − / srelative to the median system redshift suggests that the BCG has not yet settled down to thebottom of the cluster potential well. The member galaxies with IDs A3 and A4 at projecteddistances of 210 and 340 kpc show weak traces of [O ii ] emission. Moreover, galaxy A5 atd center ≃
520 kpc exhibits very significant [O ii ] line emission with an equivalent width of about46Å. All three galaxies (A3-A5) are close to or redder than the expected SSP model color,which could point towards dusty star formation activity (e.g. Pierini et al. 2005).Based on the spectroscopic system redshift, the cluster’s soft-band X-ray luminosity canbe determined as L . − , = (0 . ± . × erg s − or in terms of the total bolometricenergy output L bolX , ≃ (1 . ± . × erg s − . Applying the scaling relation of Sect. 5.2yields a total mass estimate for the cluster XDCP J0027.2 + L X ≃ . + . − . × M ⊙ . + = The optical and X-ray properties of the system XDCP J0338.5 + z = Newton field with OBSID 0036540101 with a clean e ff ective exposure time of 18 ksecat an o ff -axis angle of 8 ′ . The imaging follow-up in z- (53 min) and H-band (50 min) tookplace on 5 January 2006 in good but non-photometric observing conditions with 1.2 ′′ seeingduring the same campaign as for the cluster in Sect. 4.2, complemented again by photometric he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Table 3.
Spectroscopic member galaxies of XDCP J0027.2 + z = .
959 (A, top) andXDCP J0338.5 + z = .
916 (B, bottom). The table lists for each member galaxy theidentification number used in Fig. 7 (bottom), coordinates, total H-band magnitude, z − H color,projected cluster-centric distance d cen , spectroscopic redshift z spec , and its uncertainty σ z .ID RA DEC H z − H d cen z spec σ z J2000 J2000 Vega mag Vega mag arcsecA1 6.80767 17.24345 17.62 2.19 6.7 0.9485 0.0003A2 6.80770 17.24796 18.39 2.22 17.5 0.9579 0.0002A3 6.81179 17.23640 17.79 2.30 26.8 0.9509 0.0003A4 6.82149 17.24001 19.30 2.54 42.5 0.9602 0.0004A5 6.81302 17.22560 19.38 2.39 65.4 0.9600 0.0002A6 6.82069 17.26749 18.16 1.98 94.3 0.9599 0.0003B1 54.64069 0.48888 18.02 2.18 49.0 0.9166 0.0002B2 54.64089 0.48512 19.78 2.29 51.4 0.9151 0.0003B3 54.64640 0.48250 17.52 2.28 73.9 0.9147 0.0002B4 54.64936 0.48243 18.09 2.35 83.4 0.9166 0.0004B5 54.67213 0.47336 19.05 2.39 172 0.9160 0.0002B6 54.67971 0.47059 18.14 2.05 200 0.9192 0.0003B7 54.68905 0.45572 19.06 2.23 253 0.9157 0.0002 z-band calibration snapshot observations on 30 October 2006. The final image stacks reachlimiting 50% completeness Vega magnitudes of H lim ≃ . lim ≃ . / FORS 2 follow-up observations were performed on 9November 2007 (run ID: 079.A-0634) under moderate 1.5 ′′ seeing conditions for a total netexposure time of 2.2 h and yielded seven spectroscopic cluster members (Table 3, IDs B1-B7).The situation and configuration for the system XDCP J0338.5 + + + ∼ σ ) and an irregular morphologywith extensions in three directions (see Fig. 7). The X-ray centroid is located in closeproximity to a chip gap of the PN detector, which had to be discarded for this reason forthe growth-curve analysis shown in the top right panel of Figs. 8. The results based on thetwo MOS detectors yielded a flux of f . − , = (2 . ± . × − erg s − cm − , which couldbe biased high due to unresolved point source contributions within the analysis aperture. Thederived luminosities L . − , = (0 . ± . × erg s − and L bolX , ≃ (2 . ± . × erg s − are hence to be interpreted as upper limits, as is the luminosity-based mass estimate ofM L X ≃ . + . − . × M ⊙ (Tables 4 & 6).Owing to the complex multi-extension X-ray morphology, the e ff ective measured coreradius of the X-ray surface brightness distribution is quite extended ( ∼
190 kpc), with a valuein between the cases of SpARCS J003550-431224 and XDCP J0027.2 + + he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ≤ z − H ≤ and a background level of (5.6 ± − . This representation showsthat the complex morphology is also reflected in the red galaxy distribution with a maindensity extension to the East and to the North and a connecting pivot point close to the centroidof the detected extended X-ray source. Both Northern and Eastern extensions of the galaxydistribution are still within the estimated projected cluster radius of R X , ≃
940 kpc ≃ ′ based on the X-ray mass estimate. Very weak extended X-ray emission seems to be presentfor the main Eastern extension centered at an approximate distance of 1.5 ′ from the main X-ray source centroid, while the X-ray emission towards the Northern extension is dominatedby several point sources.This configuration suggests ongoing merging activity of at least three main components,similar to the discussed situation of SpARCS J003550-431224, with the di ff erence that thebulk motions of the components are along the plane of the sky rather than in the radialdirection. A merging configuration very close to the plane of the sky is also supportedby the spectroscopic members of the systems, which all exhibit small rest-frame velocityo ff sets from the median redshifts of ≤
500 km / s (Table 3, IDs B1-B7). Four of these membersare located in the Eastern extension within 90 ′′ from the main X-ray centroid, three otherswere found beyond the nominal cluster radius towards the same direction. The tentativelyidentified BCG is the galaxy with ID B3 at a projected distance of about 570 kpc from thedetermined X-ray center, which is likely part of the infalling Eastern structure. The currentlyavailable spectroscopy and confirmed cluster memberships are biased towards this Easternextension since the MXU mask was centered on this structure to also incorporate the close-bysystem XDCP J0338.7 + ff erent componentsof the cluster environment. All spectroscopic members are located close to the expectedz − H SSP model color in the CMD, although most spectra (B2-B6) show indications ofweak [O ii ] emission pointing towards some ongoing star formation activity (Fig. 8, rightcenter and bottom panels), including the tentatively identified o ff -center BCG. Future multi-wavelength studies of this system will enable a more detailed characterization of this complexbut intriguing system. Galaxy clusters observed in the first half of cosmic time can be expected to be activelyaccreting mass from the surrounding large-scale structure and to exhibit a larger fraction ofmajor merger events caught-in-the-act as part of the hierarchical structure growth process. The he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Table 4.
Properties of the galaxy clusters SpARCS J003550-431224, XDCP J0027.2 + + c and cluster radii R X , are approximate values.Property SpARCS J0035.8-43 XDCP J0027.2 +
17 XDCP J0338.5 +
00 UnitRA 00:35:50.1 00:27:14.3 03:38:30.5DEC -43:12:10.3 + + z centerBCG
13 (109) 6.7 (53) 73 (573) arcsec (kpc)
DET ML
76 54 15
EXT ML
56 24 6.6r c ( β = /
3) 31 (260) 15 (110) 24 (190) arcsec (kpc)f . − , ± .
24 0.94 ± ± − erg s − cm − L . − , ± .
22 0.40 ± ± erg s − L bolX , ± . ± ± erg s − T X + − NA NA keVR X ,
700 770 940 kpc M L X + . − . + . − . + . − . M ⊙ three systems presented in this section (see Table 4) span a wide range of X-ray morphologiesand dynamical states: (i) the multi-peaked X-ray emission of SpARCS J003550-431224owing to major merger activity along the line-of-sight (Sect. 4.1), (ii) the mostly regular X-rayproperties of XDCP J0027.2 + + z > . z > . + / irregular X-ray morphologiesamong the whole comparison sample of 22 systems presented in the next section.
5. The XDCP sample of 22 X-ray luminous galaxy clusters at z > . Combining the data presented on the three systems in Sect. 4 with previously publishedresults, we can now complete the compilation of the largest sample of spectroscopicallyconfirmed X-ray luminous galaxy clusters at z > . z cluster population. This first XDCP distant cluster sample of 22 he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Newton archive (Mehrtens et al. 2011).
Tables 5 & 6 list the optical and X-ray properties of the present XDCP galaxy cluster sampleat z > .
9. Both tables start with the cluster IDs and the system redshifts for easiercross-referencing of objects. 20 clusters have related publications in the literature (or aresubmitted / in preparation) that are listed in the last column (21) of Table 6. Five clusters (C04,C07, C08, C15, C16) were spectroscopically confirmed by other projects, from which theo ffi cial name of the first publication is listed in column (3).The stated coordinates in Table 5 (4 & 5) refer to the X-ray centroid of the detectedextended X-ray sources, from which all projected cluster-centric distances are measured,e.g. in columns (8) and (9). The given X-ray centroid position is the first moment (i.e. the‘center-of-mass’) of the extended X-ray emission as measured during the maximum likelihoodsource evaluation procedure discussed in Sect. 3.1.2. The number of spectroscopic clustermembers and the follow-up imaging technique and color (see Sect. 3.2) are given in columns(6) and (7). The cluster-centric BCG o ff sets (8) will be further analyzed in Sect. 5.4, andSect. 5.5 discusses the statistics of nearby 1.4 GHz radio sources listed in column (9). Totalmass estimates are either X-ray luminosity based (10) as discussed in Sect. 5.2, or derivedfrom other methods, where available, in column (11).Table 6 focusses on the X-ray properties of the systems starting with the XMM- Newton serendipitous source name (12), an acronym (13) used for Fig. 9, and the physical X-raysource parameters (see Sect. 3.1.6) soft-band 0.5-2 keV flux (14), bolometric luminosity (15),and X-ray temperature (16). The XMM-
Newton field observation identifier of the detectedserendipitous X-ray source is listed in (17), the e ff ective clean exposure time (ECT, seeSect. 3.1.1) of the field is given in column (18), and the o ff -axis angle of the source is statedin (19).Column (20) in Table 6 lists an overall X-ray quality (XFl), which summarizes theconfidence that the detected extended X-ray emission of the source originates predominantlyfrom thermal emission of the ICM. This flag takes into account all presently availableinformation on the source to assign a confidence class based on (i) the original sourcedetection parameters (Sect. 3.1.2), (ii) the more detailed source characterization (Sect. 3.1.6),(iii) the imaging data information to check for potentially contaminating objects (Sect. 3.2),and (iv) the optical spectra of central sources to probe for AGN signatures (Sect. 3.3). Thisevaluation yielded for 17 clusters (77%) a secure ( +++ ) X-ray quality flag, implying a highconfidence ( > h e X - r a y l u m i nou s ga l a xyc l u s t e r popu l a ti ona t . < z ≤ . a sr eve a l e db y t h e X D C P Table 5.
General properties of the 22 XDCP galaxy clusters at z > . ffi cial cluster name (3), X-ray centroid coordinates (4 + ′ . X-ray luminosity-based total mass estimates areprovided in column (10). Other mass estimates from the referenced publications in column (21) of Table 6, where available, are listed in (11) with themethod indicated (T: X-ray temperature-based, HE: hydrostatic equilibrium method, WL: weak lensing, M g : gas mass-based). The entries ‘lit.’ referto literature references of other projects listed in Table 6.ID z O ffi cial Name RA DEC Specs Follow-up d centerBCG S . M L X M J2000 J2000 cen ) 10 M ⊙ M ⊙ (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)C01 1.579 XDCP J0044.0-2033 00:44:05.2 -20:33:59.7 3 I − H 73 3.2 (0.6 ′ ) 2.9 + . − . C02 1.555 XDCP J1007.3 + + − H 36 2.2 (0.1 ′ ) 1.7 + . − . C03 1.490 XDCP J0338.8 + + − H 176 - 1.2 + . − . C04 1.457 XMMXCS J2215.9-1738 22:15:58.5 -17:38:05.8 lit. R − z, z − H 300(t) 3.3 (1.8 ′ ) 1.8 + . − . + . − . (T)C05 1.396 XDCP J2235.3-2557 22:35:20.4 -25:57:43.2 30 R − z 31 - 4.1 + . − . + . − . (HE / WL)C06 1.358 XDCP J1532.2-0837 15:32:13.2 -08:37:01.4 3 R − z 46(t) - 1.1 + . − . C07 1.335 SpARCS J0035.8-4312 00:35:50.1 -43:12:10.3 lit. grizH 109 0.2 (0.2 ′ ) 1.7 + . − . + . − . (T)C08 1.237 RDCS J1252.9-2927 12:52:54.5 -29:27:18.0 lit. lit., R − z 11 15.3 (0.8 ′ ) 3.7 + . − . + . − . (HE)C09 1.227 XDCP J2215.9-1751 22:15:56.9 -17:51:40.9 7 (5) R − z, z − H 57(t) 3.1 (0.8 ′ ) 1.0 + . − . + . − . (T)C10 1.185 XDCP J0302.1-0001 03:02:11.9 -00:01:34.3 6 z − H 47 - 2.1 + . − . C11 1.122 XDCP J2217.3 + + − H 35(t) 18.0 (0.2 ′ ) 1.8 + . − . C12 1.117 XDCP J2205.8-0159 22:05:50.3 -01:59:27.4 3 R − z, z − H 57 - 1.8 + . − . C13 1.097 XDCP J0338.7 + + − H 347(t) - 1.5 + . − . C14 1.082 XDCP J1007.8 + + − z 199 3.6 (0.8 ′ ) 1.7 + . − . C15 1.053 XLSS J0227.1-0418 02:27:09.2 -04:18:00.9 lit. z − H 113(t) - 2.0 + . − . C16 1.050 XLSS J0224.0-0413 02:24:04.1 -04:13:31.7 lit. lit. 44 0.1 (0.9 ′ ) 3.3 + . − . + . − . (HE)C17 1.000 XDCP J2215.9-1740 22:15:57.5 -17:40:25.6 10 (2) R − z, z − H 20 14.4 (0.7 ′ ) 1.1 + . − . + . − . (T)C18 0.975 XDCP J1229.4 + + − z 8(t) - 4.8 + . − . + . − . (T)C19 0.975 XDCP J1230.2 + + − z 134 1.7 (0.3 ′ ) 4.1 + . − . + . − . (T / M g / WL)C20 0.959 XDCP J0027.2 + + − H 53 3.3 (1.7 ′ ) 1.6 + . − . C21 0.947 XDCP J0104.3-0630 01:04:22.3 -06:30:03.1 7 (8) R − z, z − H 30 11.9 (0.0 ′ ) 2.1 + . − . C22 0.916 XDCP J0338.5 + + − H 573(t) - 2.6 + . − . h e X - r a y l u m i nou s ga l a xyc l u s t e r popu l a ti ona t . < z ≤ . a sr eve a l e db y t h e X D C P Table 6.
Continuation of Table 5 focused on the X-ray properties of the clusters. The XMM-
Newton source name is listed in (12), an acronym formin (13), the 0.5-2 keV soft-band X-ray flux inside the R aperture in (14), the bolometric cluster luminosity in (15), and the spectroscopic X-raytemperature in (16), where feasible. The XMM-
Newton detection field is listed in (17), the corresponding e ff ective clean time (ECT) of the field in(18), the source o ff -axis angle in (19), an overall X-ray quality flag (XFl) in (20), and relevant literature references to the cluster in (21).ID z XMM Source Name Acron. f . − , / − L bolX , / T X OBSID ECT Θ o ff XFl References a erg s − cm − erg s − keV ksec ′ (1) (2) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21)C01 1.579 XMMU J0044.0-2033 X0044 1.6 ± ± +++ Sa11C02 1.555 XMMU J1007.3 + ± ± ++ Fa11aC03 1.490 XMMU J0338.8 + ± ± + Na11C04 1.457 XMMU J2215.9-1738 X2215a 1.1 ± ± + . − . +++ St06,Hi07 / / ± ± + . − . +++ Mu05,Ro09,J09,S10C06 1.358 XMMU J1532.2-0837 X1532 0.29 ± ± + Su11C07 1.335 XMMU J0035.8-4312 X0035 0.80 ± .
24 1.8 ± . + − +++ Wi09, this workC08 1.237 XMMU J1252.9-2927 X1252 3.0 ± ± + . . +++ Ro04,De07C09 1.227 XMMU J2215.9-1751 X2215b 0.37 ± ± + . − . +++ dHo11,B10C10 1.185 XMMU J0302.1-0001 X0302 1.2 ± ± +++ Su11C11 1.122 XMMU J2217.3 + ± ± +++ Fa11cC12 1.117 XMMU J2205.8-0159 X2205 0.95 ± ± +++ Da09,Fa11cC13 1.097 XMMU J0338.7 + ± ± + Pi11C14 1.082 XMMU J1007.8 + ± ± + ∞− . +++ Sc10C15 1.053 XMMU J0227.1-0418 X0227 1.1 ± ± + . − . +++ An05,Pa07,Ad10C16 1.050 XMMU J0224.0-0413 X0224 3.1 ± ± + . − . +++ Pa07,Ma08,Ad10C17 1.000 XMMU J2215.9-1740 X2215c 0.54 ± ± + . − . +++ dHo11C18 0.975 XMMU J1229.4 + ± ± + . − . +++ Sa09C19 0.975 XMMU J1230.2 + ± ± + . − . +++ Fa11b / + ± ± +++ this workC21 0.947 XMMU J0104.3-0630 X0104 1.7 ± ± +++ Fa08C22 0.916 XMMU J0338.5 + ± ± ++ this work a Sa11: Santos et al. (2011); Fa11a: Fassbender et al. (2011a); Na11: Nastasi et al. (2011); St06: Stanford et al. (2006); Hi07 / /
10: Hilton et al.(2007, 2009, 2010); B10: Bielby et al. (2010); Mu05: Mullis et al. (2005); Ro09: Rosati et al. (2009); J09: Jee et al. (2009); S10: Strazzullo et al.(2010); Su11: ˇSuhada et al. (2011); Wi09: Wilson et al. (2009); Ro04: Rosati et al. (2004); De07:Demarco et al. (2007); dHo11: de Hoon et al. (inprep.); Fa11c: Fassbender et al. (in prep.), Da09: Dawson et al. (2009); Pi11: Pierini et al. (subm.); Sc10: Schwope et al. (2010); An05:Andreon et al.(2005); Pa07:Pacaud et al. (2007); Ad10: Adami et al. (2011); Ma08:Maughan et al. (2008); Sa09: Santos et al. (2009); Fa11b / he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP !"" X0035 z=1.335 gr iz H X2215a z=1.457 rzH
X2215c z=1.000 rzH
X2215b z=1.225 rzH
X2235 z=1.396 RzH
X1252 z=1.237 RzH
X0302 z=1.185 RzH
X2217 z=1.122 zH
X1007b z=1.082 RzH
X0227 z=1.053 rzH
X0224 z=1.050 grz
X1230 z=0.975 UB Vr iz X1229 z=0.975 RzK s X0104 z=0.947 RzH
X0338a z=1.490 zH
X1007a z=1.555 zH
X0044 z=1.579 IzH
X1532 z=1.358 VRz
X2205 z=1.117 izH
X0338b z=1.097 zH
Figure 9.
XDCP gallery of the 20 X-ray luminous galaxy clusters at z > . Newton
X-raysurface brightness contours are overlaid in yellow. Each cluster image is centered on the X-ray centroid location and has a sidelength of 1.5 ′ × ′ with the black background remappedto gray scale for contrast enhancement. The top of the panels lists the cluster acronym (seeTable 6), the system redshift, and the bands used for the shown color composite. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 10.
Redshift histogram of all currently spectroscopically confirmed XDCP galaxyclusters (black hashed) and the sample presented in this work (blue). The solid (dashed)vertical line marks the redshift z = . z = Two clusters (9%) have assigned intermediate ( ++ ) X-ray confidence flags, with a non-negligible probability of up to 20% that non-thermal emission processes may be majorcontributors ( > ∼ z = .
555 (Fassbender etal. 2011a) with an extent significance of ∼ σ hosts a radio galaxy in the center which couldemit non-thermal X-rays, and C22 at z = .
916 with its complex configuration was discussedin Sect. 4.3.A lower X-ray confidence flag ( + ) was given to three systems (14%), where theprobability of possible predominant non-thermal X-ray emission appears to be at levels > σ , i.e. very close to the detection threshold.All three systems feature a red galaxy population peaked within 30 ′′ from the X-ray centroid,of which 3-7 galaxies are spectroscopically confirmed members ( ˇSuhada et al. 2011; Nastasiet al. 2011, Pierini et al., subm.), in the case of XDCP J1532.2-0837 (C06) with signaturesof some central AGN activity. The properties of such low-mass ( M < ∼ . × M ⊙ ),high redshift ( z > ∼ .
1) systems and their X-ray point source contents are currently unexploredterritory and will require further investigations.
The histogram in Fig. 10 (blue shaded region) displays the redshift distribution of the clusterspresented in this work as well as the full current XDCP sample (black hashed) for comparison.With 17 systems at z ≥ z > .
3, this sample provides an almost homogeneousredshift coverage up to z ∼ . he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 11.
XDCP clusters in the mass versus redshift plane. The mass estimates with thelowest uncertainty for each galaxy cluster were used according to Table 5 (10 & 11). Futurestudies can investigate the properties of the galaxy cluster population in at least three redshiftbins at z > . accessible redshifts. Here we make use of the latest empirically calibrated M-L and M-Trelations by Reichert et al. (2011) with the explicit forms M L X = (1 . ± . · L bolX , erg s − . ± . · E ( z ) − . + . − . × M ⊙ and (1) M T X = (0 . ± . · (cid:18) T X (cid:19) . ± . · E ( z ) − . ± . × M ⊙ , (2)where E(z) = H(z) / H is the cosmic evolution factor of the Hubble expansion. These relationsprovide the best current constraints on the redshift evolution factors and their uncertainties,which in the case of the M-L relations is significantly slower than the self-similar modelpredictions (e.g. Kaiser 1986; B¨ohringer et al., in prep.). Since the evolution factors for therelevant redshift regime 0 . < z < ∼ . ≃ E ( z )-term dominates the error budget for luminosity-based mass estimatesat high- z together with the e ff ect of intrinsic scatter, which is currently only quantified atlow-redshifts (Pratt et al. 2009). The considered error budget hence includes this (local)intrinsic scatter, the redshift evolution uncertainty including sample bias e ff ects, the errorsin normalization and slope of the relation, and the measurement uncertainties in L X (T X ). Asa last step, M values are scaled to total mass estimates M ≃ (1 . ± . · M byassuming an NFW mass profile with concentration parameters c = . ± . ff y et al. (2008). he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP X constraints are available. In severalcases (C05, C08, C16, C19) more accurate mass estimates are available and listed in (11),which are mostly based on the standard hydrostatic equilibrium (HE) method, weak lensing(WL) measurements, or combinations thereof.Figure 11 shows the XDCP cluster mass estimates with the lowest uncertainty as afunction of the system redshift. The characteristic median mass of the sample is M ≃ × M ⊙ , with a mass range spanning approximately 0.7-7 × M ⊙ . The distributionshows a fairly homogeneous and unbiased mass sampling with indications of an increasinglower mass cut with redshift as expected. The achieved coverage of the mass-redshift planewill allow future investigations of the distant galaxy cluster population properties in at leastthree redshift (vertical dashed lines) and mass bins (horizontal dashed lines). The latter massbins allow an approximate distinction of the three classes of massive distant X-ray clusterswith M > . × M ⊙ , medium mass objects at 1 . × M ⊙ < M < ∼ . × M ⊙ , andlow mass systems with M < ∼ . × M ⊙ . Figure 9 displays an optical / NIR-X-ray gallery of all systems, in additions to the two newclusters presented in Fig. 7 (central panels). All image sizes are 1.5 ′ × ′ , corresponding tophysical length scales of ∼ z ≃ Newton
X-ray surface brightness contours areoverlaid for each system, with optimized adjusted levels for each source to allow a fairrepresentation of the underlying X-ray morphology.This X-ray surface brightness morphology is generally closely linked to the dynamicalstate of the systems (e.g. B¨ohringer et al. 2010; Mohr et al. 1993). Although the presenteddistant clusters do not constitute a representative sample and the signal strength is verylimited, a rough qualitative morphological classification can provide some first clues on thetypical high- z cluster X-ray appearance within the limitations of the XMM- Newton resolutioncapabilities. As the simplest qualitative classification, we can consider the following fourcategories: regular morphology (R), mostly regular but with a clear elongation axis (R-), intermediate states (0), and multi-peaked / irregular (M / I) morphologies. Such a schemeyields roughly 4 /
22 (18%) regular systems (C05, C08, C16, C19), 12 (55%) mostly regularmorphologies (C01, C02, C04, C06, C10, C11, C12, C14, C17, C19, C20, C21), 4 (18%)intermediate state systems (C03, C09, C13, C15), and the 2 (9%) multi-peaked / irregular X-ray morphologies (C07, C22) discussed in Sects. 4.1 & 4.3.The majority of the systems (16 /
22 or 73%) hence exhibit at least a mostly regular X-ray he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP > ∼ × M ⊙ ), feature BCGs close to the X-ray centroid, and show very evolvedgalaxy populations (e.g. Strazzullo et al. 2010; Rettura et al. 2010; Santos et al. 2009). Themore elongated X-ray structure of the mostly regular (R-) cluster category, on the other hand,may indicate the major matter accretion axis or minor merging activity (e.g. Fassbender etal. 2011b), while the multi-peaked / irregular (M / I) systems suggest ongoing major mergers(Sects. 4.1 & 4.3). ff sets and luminosity gaps Another indicator for the dynamical state of a system at low redshifts is the location of theBCG with respect to the X-ray centroid position (e.g. Haarsma et al. 2010; Smith et al. 2010;Sanderson et al. 2009). Studies of the representative REXCESS reference sample by Haarsmaet al. (2010) show that ∼
80% of the local ( z < .
2) clusters host a central dominant brightestcluster galaxy within 20 kpc of the X-ray peak, with a median o ff set for the full population of7.5 kpc (red histogram and red dashed line in the top panel of Fig. 12).The situation is clearly di ff erent at z > .
9, where a central dominant BCG coincidentwith the X-ray centroid is more an exception than the rule, as is evident from Fig. 9. The X-ray centroid position, as the first moment of the surface brightness distribution of the extendedcluster emission, is generally robustly determined with XMM-
Newton , even in the low-countregime at high- z , with an average statistical positional uncertainty of 3 ′′ . The (Gaussian)combination of this statistical error with the average systematic absolute astrometric o ff set of1 ′′ (e.g. Watson et al. 2009) leads to an average total X-ray centroid uncertainty of 25-28 kpc inthe targeted redshift regime. For this work, we conservatively assume a total positional errorradius of the X-ray centroid determination of 30 kpc (green dotted line in Fig. 12), whichwould result in an observed median BCG o ff set for the REXCESS sample of 26.8 ± ff set directions.The unambiguous identification of the BCG can be a challenging task at high- z for asignificant fraction of non-trivial cases. Owing to the standard paradigm of hierarchical built-up of BCGs (e.g. De Lucia and Blaizot 2007), the high- z progenitors of present day centrallydominant galaxies may not necessarily be the brightest galaxies at any redshift and may havemigrated long distances within the larger scale cluster environment. Related to this, threemain issues for the observational identification process arise in practice: (i) clusters may hostseveral top ranked galaxies with similar absolute magnitudes or mass (e.g. X1229, Santoset al. 2009); (ii) the brightest galaxy of the cluster environment may still be outside the formalR radius (e.g. X2217, Fassbender et al., in prep.); and (iii) o ff -center BCG candidatesmay still lack the spectroscopic membership confirmation or not all brighter galaxies atlower cluster-centric distance have been spectroscopically excluded as members (e.g. X0338c,Sect. 4.3). Such ambiguous cases (8 /
22) are flagged as tentative (t) BCG identifications incolumn (8) of Table 5, where the projected cluster-centric BCG distances are listed. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 12.
Properties of brightest cluster galaxies in the XDCP sample.
Top panel:
Comparison of the BCG o ff sets from the X-ray centroid for the low-z REXCESS sample (redhistogram) and the z > . z clusters are indicated by the blue background color. The median XDCP clustercentroid o ff sets for all BCGs (secure identifications) of 55 kpc (50 kpc) are marked by the black(blue) dashed vertical lines, whereas the green dotted line depicts the average measurementuncertainty. The median BCG o ff set for the REXCESS sample of 7.5 kpc is indicated by thered dashed line for reference. Bottom panel:
Magnitude di ff erence ∆ m between the first- andsecond-ranked cluster galaxies as a function of the BCG centroid o ff set. Clusters belongingto di ff erent redshifts bins are marked by di ff erent colors. Filled (open) symbols indicatesecure (tentative) identifications of the two top-ranked galaxies. The median magnitude gapof 0.31 mag for the full sample is marked by the horizontal black line, the green vertical linemarks the centroid measurement uncertainty as above. The black hashed histogram in Fig. 12 (top panel) shows the observed distribution ofBCG o ff sets for the full z > . he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ff set of 55 kpc (50 kpc for the secure BCGs) from the X-ray centroid,with a wing extending towards large cluster-centric distances. Considering the discussedmeasurement uncertainty, the o ff sets of only 7 /
22 systems of the sample with observedd centerBCG <
40 kpc are statistically consistent with harboring a central BCG at o ff sets of < ∼
20 kpc(see e.g. the case of X2235, Rosati et al. 2009). The determined median BCG o ff set ofd centerBCG ∼
50 kpc, on the other hand, is robust and basically una ff ected by centroid uncertaintiessince it is governed by the largest half of the distribution of cluster-centric distances.At lookback times of 7.3-9.5 Gyrs for the present sample, the observed BCG populationhas hence generally not yet reached the bottom of the cluster potential well (see e.g. the caseof X1230, Fassbender et al. 2011b), but is rather still caught in the process of inward migrationvia dynamical friction. A first hint for a further redshift evolution of the BCG o ff sets can beobtained by considering the redshift bins of Fig. 11, which yields median projected cluster-centric BCG distances of ∼
52 kpc for the first two bins at z ≤ . ∼
73 kpc for the sevenhighest-z systems at z > . z clusters, we can consider the luminosity gap ∆ m between the first- and second-rankedgalaxies. This statistic was studied by Smith et al. (2010) for a sample of massive(M ∼ M ⊙ ), low-redshift (0 . ≤ z ≤ .
3) clusters and was found to correlate tightlywith the dynamical state of the systems, e.g. large ∆ m generally imply small amounts ofsubstructure and cuspy gas density profiles. The median luminosity gap for this local referencesample is measured to be ∆ m , med ≃ .
67 mag and the fraction of clusters with very dominantBCGs with ∆ m > ∆ m of the XDCP z > . ff sets. The ∆ m measurements wereobtained in the reddest (K s , H, or z) optical / NIR band available (see column 21 in Table 6for references). The color coding groups the systems into the di ff erent redshift bins, whereasopen circles indicate tentative identifications of the first- and / or second ranked galaxies asdiscussed above. Clear trends of ∆ m with either the BCG o ff set or as a function of redshiftare not obvious in the bottom panel of Fig. 12. However, the statistics of the ∆ m distributionreveals again marked evolutionary di ff erences compared to the low- z reference sample. Themeasured median luminosity o ff set of the high- z clusters is found to be ∆ m , med ≃ .
31 mag( ∆ m , med ≃ .
28 mag for secure identifications) and the sample only contains one candidatesystem (e.g. < ∼ ∆ m > z = . z > . z < .
3. In particular, we note that our current sample based on the discussed follow-up strategy (Sects. 3.2 & 3.3) does not include any candidates that would qualify them as fossilgroups with ∆ m ≥ . < z < ∼ . z systems and their dominance with respect to second-ranked galaxies. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP The statistics of radio sources associated with high- z galaxy clusters is of prime importancefor ongoing Sunyaev-Zeldovich e ff ect (SZE) surveys (e.g. Williamson et al. 2011; Marriageet al. 2010; Planck Collaboration et al. 2011). Radio emitting sources at the cluster locationspose the main source of potential contamination for SZE selected cluster samples, since thesesources can (partially) fill in the SZE decrement signal and hence lead to an underestimationof cluster counts and mass estimates. While detailed radio source studies in clusters in thelocal Universe (e.g. Lin and Mohr 2007; Best et al. 2007; Mittal et al. 2009) and at moderateredshifts (Sommer et al. 2011) are now available, robust statistics for the z > . ′ ( ∼ ∼ ′′ FWHM resolution) andhence allows a first evaluation of the frequency of cluster-associated radio sources at brightflux densities ( > ∼ ′ / ′ / ′ of1.2% / / / /
10 NVSS radio sources at radii within 0.5 ′ / ′ / ′ from thecenters of the 22 distant XDCP clusters is to be compared to the background expectationof 0.3 / / ′ ( ∼
500 kpc) from the X-raycenter. The radio flux density of these sources spans a range of 2.2-18 mJy with a medianvalue of 3.5 mJy. A trend of the cluster-associated radio source fraction with redshift is notapparent over the three probed redshifts bins with the current statistics. In terms of clustermass bins, there is a hint that intermediate mass systems (M ∼ × M ⊙ ) may be preferredenvironments for cluster-associated radio sources with an observed fraction of approximately50%.With the assumption of a typical spectral index of α ≃ − . P . > ∼ (0 . × W Hz − for the detection of cluster-associated radio sources in the probed redshiftrange 0 . < z < ∼ .
6. A comparison to results obtained at lower redshifts is hence only possiblefor the most luminous radio source bin with P . > ∼ W Hz − , for which a low- z fraction he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP ∼
6% were determined by Lin and Mohr (2007) and Best et al.(2007), while the HIFLUGCS sample of Mittal et al. (2009) contains a fraction of ∼ z sample thus suggests an increase of the fraction ofvery luminous cluster-associated radio sources by about a factor of 2.5-5.An upper limit of P . < ∼ . × W Hz − for the potentially most luminous radiosources in the sample can be derived from the observed flux densities at the locations ofclusters C08 and C11 (see Table 5). These maximal flux densities are expected to dropby a factor of ∼
40 to < ∼ ffi ciency of massive clusters with SZE surveys for the assumed extrapolation † ,with a maximum radio source flux contribution at 150 GHz of < ∼
10% at the typical clusterdetection limit of e.g. the South Pole Telescope (Carlstrom et al. 2011). ≥ . galaxy cluster frontier As the final point to be addressed in this section, we have a closer look at the current galaxycluster redshift frontier at z > ∼ . z = .
62 in the group regime (M < M ⊙ ) by Tran et al. (2010).The three top ranked clusters from Table 5 at redshifts of 1.490, 1.555, and 1.579 arepresently the most distant, spectroscopically confirmed, X-ray luminous systems known inthe cluster regime at M > ∼ M ⊙ . The systems XDCP J0044.0-2033 (C01, Santos et al.,2011), XDCP J1007.3 + + ′′ from the X-ray centroid position are indicated in red and spectroscopic members are marked by squareboxes. Simple stellar population model predictions for stellar formation redshifts of 5 (3)are displayed by red (blue) dashed lines and green dotted lines confine the applied colorcuts for each system spanning the color range between 0.3 mag bluer than the z f = z f = ′ × ′ ( ∼ × ± − . † This extrapolation by more than a factor of 100 in frequency using α ≃ − . he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP Figure 13.
Comparison of the presently three most distant clusters in the XDCP sample:XDCP J0338.8 + z = .
490 (top), XDCP J1007.3 + z = .
555 (center), andXDCP J0044.0-2033 at z = .
579 (bottom). The left column shows the color-magnitudediagrams with red circles indicating galaxies within 40 ′′ from the X-ray center, blue symbolsrepresenting spectroscopic members at r > ′′ , and black dots all other objects in the FoV.Spectroscopic members are marked by open squares, the 50% completeness limits by thevertical dashed black lines, and the apparent characteristic H-band magnitudes H* at thecluster redshifts by the vertical dotted blue line. Horizontal blue ( z f =
3) and red ( z f = ff erent stellar formationredshifts z f , and the dotted lines confine the applied color cuts for the red galaxy densities.The corresponding 4 ′ × ′ ( ∼ × Newton
X-ray contours overlaid in blue. Large black circles indicate the60 ′′ and 30 ′′ radii around the X-ray center, small black circles mark spectroscopic members,and red circles represent color selected red objects with corresponding logarithmically spacedred density contours with levels of 3.3, 5.2, 8.0, 13, 20, 30 galaxies per arcmin covering thesignificance range of 2-29 σ above the background. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP z = .
579 (bottompanels) is a remarkable system for this cosmic epoch with its high X-ray luminosity andcorresponding high mass estimate (Tables 5 & 6). This massive structure is also reflected inthe rich galaxy population (see Fig. 9, top left panel), which marks a > ∼ σ galaxy overdensitycentered on the X-ray emission. Both the galaxy distribution and X-ray emission are elongatedalong the NS direction, which may reflect the main cluster assembly axis. The CMD iswell populated in the applied color cut region with the main noteworthy feature that thebrightest central galaxies, including the spectroscopically confirmed BCG candidate, are allsignificantly bluer than the expected red-sequence color.The second ranked system XDCP J1007.3 + z = .
555 (central panels) is anintermediate mass cluster with a central, red, radio-loud BCG. The currently available imagingdepth is 0.5 mag shallower compared to the other two fields, implying that only the brightend of the underlying galaxy population is presently accessible. Two spectroscopic membersclose to the characteristic magnitude H* and within a projected distance of 200 kpc from theX-ray centroid are blue and feature prominent [O ii ] emission lines, which provide evidencefor strong starburst activity in these massive galaxies.The seven spectroscopic members of the lower mass system XDCP J0338.8 + z = .
490 (top panels), on the other hand, show very little signs of star formation out tobeyond the nominal R radius. The BCG is a red, merging, o ff -center galaxy at the expectedSSP color, while other galaxies along the apparently well populated red-sequence seem toshow an increased spread in color, compared to lower-z clusters. Towards fainter magnitude(H ∼
21 mag) several central galaxies are just below the applied color cut, with the e ff ect thatthe otherwise central galaxy density peak (Nastasi et al. 2011) is now shifted Northward ofthe X-ray centroid position. This system features the widest spatial distribution of red galaxyoverdensities, spread over almost the full ∼ × z > ∼ . > ∼ With the recent observational advances to push the high redshift cluster frontier to z > ∼ . ff erent cluster components and their mutual interactions.Besides the aspect of reaching out to redshifts of z ∼ .
6, other key features of the he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP z < . z > ∼ . and total system mass. The presented samplewith 22 distant test objects is clearly a key step forward to achieve these goals. However,the spectroscopic follow-up of all high- z XDCP candidate clusters is still ongoing with goodprospects to double the number of the present sample over the next few years.
6. Summary and conclusions
We presented a description of the survey strategy of the XMM-
Newton
Distant Cluster Projectto detect, identify, and study X-ray luminous galaxy clusters at z > .
8. All clusters are X-rayselected as extended sources in deep archival XMM-
Newton data and are hence unbiased withrespect to their galaxy populations. We provided an overview of the X-ray data processingof the 469 survey fields and discussed the detection capabilities of XMM-
Newton concerningfaint extended X-ray sources down to soft-band flux levels of < − erg s − cm − .We discussed di ff erent imaging techniques for the e ffi cient follow-up and photometricidentification of distant cluster candidates. In particular we compared the e ffi cacy of two-bandimaging strategies based on the R − z and z − H colors using 20 spectroscopically confirmedreference clusters in each case. We applied a robust prescription to blindly measure thecharacteristic color of red cluster galaxies and its uncertainty for candidate systems ofunknown redshifts to be compared with simple stellar population galaxy evolution modelpredictions. We confirmed the general expectations on the redshift accuracy performanceof the R − z color, which yields accurate estimates at z < . . < z < ∼ . z range the z − H color provides more reliable redshift estimatesowing to its steep redshift dependence which also allows robust cluster identifications out to z > ∼ .
5. The empirically calibrated redshift evolution models for the R − z and z − H colors areprovided in table format as part of the online material.We outlined the spectroscopic cluster confirmation process with VLT / FORS 2 and ourapplied observational galaxy cluster definition based on (i) the detected extended X-rayemission, (ii) a coincident red galaxy population, and (iii) a minimum of three associatedconcordant spectroscopic member redshifts.We discussed the X-ray properties of the previously identified rich clusterSpARCS J003550-431224 at z = bolX , ≃ (1 . ± . × erg s − , an ICM temperature of T X ≃ . + − keV, and a derived consistent mass estimatefrom both measurements of about M ≃ × M ⊙ ( ± c ≃
260 kpc), multi-peaked X-ray morphology, which in conjunction with the bimodalredshift distribution provides evidence for a major merger configuration close to the line-of-sight. he X-ray luminous galaxy cluster population at . < z ≤ . as revealed by the XDCP + z = + z = + bolX , ≃ (1 . ± . × erg s − with a correspondingmass estimate of M L X ≃ . + . − . × M ⊙ . The X-ray morphology is elongated, but mostlyregular with a coincident rich red galaxy population and a central BCG with a significant rest-frame velocity o ff set of − / s. The system XDCP J0338.5 + bolX , ≃ (2 . ± . × erg s − and the mass estimate of M L X ≃ . + . − . × M ⊙ for the system are to be considered as upperlimits due to potential unresolved point source contributions to the flux measurements.These new systems together with the previously published ones constitute the largestsample of X-ray selected distant galaxy clusters to date. In total, we presented X-ray andoptical properties for 22 X-ray luminous systems at z > .
9, with an almost homogeneousredshift coverage all the way to z ∼ .
6. The sample has a median total cluster massof M ≃ × M ⊙ and spans a mass range of approximately 0.7-7 × M ⊙ . A firstqualitative (non re-presentative) assessment of X-ray morphologies of the sample showedthat the majority of the systems ( > ∼ ff sets of the brightest cluster galaxiesfrom the X-ray centroid locations. In contrast to local clusters of which ∼
80% harbor adominant BCG within 20 kpc from the X-ray center, the brightest galaxies of the majority ofthe z > . ff sets from their X-ray centers and are less dominantwith respect to the second-ranked galaxy. We find a median cluster-centric BCG o ff set forthe sample of ∼
50 kpc, with a significant tail towards large projected o ff -center distances(i.e. >
100 kpc) for about one third of the systems. The median observed luminosity gapbetween the first- and second-ranked galaxy for the high- z cluster sample is ∆ m , med ≃ . ∆ m >
1) is < ∼ ∆ m , med ≃ .
67 mag and a fraction of 37% of BCGs with ∆ m > z < . /
22 cluster locations (59%) we found the presence of a 1.4 GHz radio source within2 ′ from the X-ray centers, of which 10 /
22 (45%) are NVSS sources with flux density levelsof > ∼
30% of the systems host a cluster-associated NVSS1.4 GHz radio source with flux densities in the range of 2.2-18 mJy, predominantly at locationswithin 1 ′ (i.e. < ∼
500 kpc) from the center. With the current statistics, no change of the radio-loud cluster fraction with redshift over the probed interval is evident, while the data suggestan increase of the fraction of very luminous cluster-associated radio sources by about a factorof 2.5-5 relative to low- z systems. EFERENCES z > ∼ . > ∼ M ⊙ clusters.Although red galaxy populations close the predicted SSP model colors are already presentin these systems, drastic changes at the massive end of the galaxy populations are evidentcompared to the evolved, tight red-sequences observed in massive clusters at z < ∼ .
4. Theseobserved changes in the three most distant XDCP systems include (i) significantly bluer colorsthan the red-sequence for the brightest galaxies (C01), (ii) starburst activity for central massivegalaxies (C02), and (iii) an apparent observed increase in the red-sequence scatter (C03). Eventhough no clear picture on the evolution of the galaxy populations in these densest clusterenvironments is established yet at lookback times of > ∼ z > ∼ . z > . z > . z < . ff erent clustercomponents in the hot and cold phases. Acknowledgments / / / ff in carrying out the service observations. The XMM- Newton project is an ESAScience Mission with instruments and contributions directly funded by ESA Member Statesand the USA (NASA). This research has made use of the NASA / IPAC Extragalactic Database(NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology,under contract with the National Aeronautics and Space Administration.
References
Adami C, Durret F, Benoist C, Coupon J, Mazure A, Meneux B, Ilbert O, Blaizot J, ArnoutsS, Cappi A, Garilli B, Guennou L, Lebrun V, Lef`evre O, Maurogordato S, McCracken H J,Mellier Y, Slezak E, Tresse L and Ulmer M P 2010
Astron. Astrophys. , A81.Adami C, Mazure A, Pierre M, Sprimont P G, Libbrecht C, Pacaud F, Clerc N, SadibekovaT, Surdej J, Altieri B, Duc P A, Galaz G, Gueguen A, Guennou L, Hertling G, Ilbert O, LeF`evre J P, Quintana H, Valtchanov I, Willis J P, Akiyama M, Aussel H, Chiappetti L, Detal
EFERENCES
Astron. Astrophys. , A18.Andreon S, Valtchanov I, Jones L R, Altieri B, Bremer M, Willis J, Pierre M and Quintana H2005
Mon. Not. R. Astron. Soc. , 1250–1260.Bailer-Jones C A, Bizenberger P and Storz C 2000 in M. Iye & A. F. Moorwood, ed.,‘Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series’ Vol. 4008of
Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference pp. 1305–1316.Becker R H, Helfand D J, White R L, Gregg M D and Laurent-Muehleisen S A 2003
VizieROnline Data Catalog , 0.Beers T C, Flynn K and Gebhardt K 1990
Astron. J. , 32–46.Bertin E and Arnouts S 1996
Astron. Astrophys. Suppl. , 393–404.Best P N, von der Linden A, Kau ff mann G, Heckman T M and Kaiser C R 2007 Mon. Not. R.Astron. Soc. , 894–908.Bielby R M, Finoguenov A, Tanaka M, McCracken H J, Daddi E, Hudelot P, Ilbert O, KneibJ P, Le F`evre O, Mellier Y, Nandra K, Petitjean P, Srianand R, Stalin C S and Willott C J2010
Astron. Astrophys. , A66.B¨ohringer H, Mullis C R, Rosati P, Lamer G, Fassbender R, Schwope A and Schuecker P2005
ESO Messenger , 33.B¨ohringer H, Pratt G W, Arnaud M, Borgani S, Croston J H, Ponman T J, Ameglio S, TempleR F and Dolag K 2010
Astron. Astrophys. , A32.B¨ohringer H, Voges W, Huchra J P, McLean B, Giacconi R, Rosati P, Burg R, Mader J,Schuecker P, Simic¸ D, Komossa S, Reiprich T H, Retzla ff J and Tr¨umper J 2000
Astrophys.J. Suppl. , 435–474.Bondi M, Ciliegi P, Zamorani G, Gregorini L, Vettolani G, Parma P, de Ruiter H, Le FevreO, Arnaboldi M, Guzzo L, Maccagni D, Scaramella R, Adami C, Bardelli S, BolzonellaM, Bottini D, Cappi A, Foucaud S, Franzetti P, Garilli B, Gwyn S, Ilbert O, Iovino A, LeBrun V, Marano B, Marinoni C, McCracken H J, Meneux B, Pollo A, Pozzetti L, RadovichM, Ripepi V, Rizzo D, Scodeggio M, Tresse L, Zanichelli A and Zucca E 2003
Astron.Astrophys. , 857–867.Burenin R A, Vikhlinin A, Hornstrup A, Ebeling H, Quintana H and Mescheryakov A 2007
Astrophys. J. Suppl. , 561–582.Carlstrom J E, Ade P A R, Aird K A, Benson B A, Bleem L E, Busetti S, Chang C L,Chauvin E, Cho H M, Crawford T M, Crites A T, Dobbs M A, Halverson N W, Heimsath S,Holzapfel W L, Hrubes J D, Joy M, Keisler R, Lanting T M, Lee A T, Leitch E M, LeongJ, Lu W, Lueker M, Luong-van D, McMahon J J, Mehl J, Meyer S S, Mohr J J, MontroyT E, Padin S, Plagge T, Pryke C, Ruhl J E, Scha ff er K K, Schwan D, Shiroko ff E, SpielerH G, Staniszewski Z, Stark A A, Tucker C, Vanderlinde K, Vieira J D and Williamson R2011
PASP , 568–581.
EFERENCES
Astrophys. J. Letters , L109–L113.Condon J J, Cotton W D, Greisen E W, Yin Q F, Perley R A, Taylor G B and Broderick J J1998
Astron. J. , 1693–1716.Cruddace R G, Hasinger G R and Schmitt J H 1988 in F Murtagh and A Heck, eds,‘Astronomy from Large Databases’ pp. 177–182.Cutri R M, Skrutskie M F, van Dyk S, Beichman C A, Carpenter J M, Chester T, CambresyL, Evans T, Fowler J, Gizis J, Howard E, Huchra J, Jarrett T, Kopan E L, Kirkpatrick J D,Light R M, Marsh K A, McCallon H, Schneider S, Stiening R, Sykes M, Weinberg M,Wheaton W A, Wheelock S and Zacarias N 2003
The IRSA 2MASS All-Sky Point Source Catalog, NASA / IPAC Infrared Science Archive.Danese L, de Zotti G and di Tullio G 1980
Astron. Astrophys. , 322–327.Dawson K S, Aldering G, Amanullah R, Barbary K, Barrientos L F, Brodwin M, ConnollyN, Dey A, Doi M, Donahue M, Eisenhardt P, Ellingson E, Faccioli L, Fadeyev V, FakhouriH K, Fruchter A S, Gilbank D G, Gladders M D, Goldhaber G, Gonzalez A H, Goobar A,Gude A, Hattori T, Hoekstra H, Huang X, Ihara Y, Jannuzi B T, Johnston D, Kashikawa K,Koester B, Konishi K, Kowalski M, Lidman C, Linder E V, Lubin L, Meyers J, MorokumaT, Munshi F, Mullis C, Oda T, Panagia N, Perlmutter S, Postman M, Pritchard T, Rhodes J,Rosati P, Rubin D, Schlegel D J, Spadafora A, Stanford S A, Stanishev V, Stern D, StrovinkM, Suzuki N, Takanashi N, Tokita K, Wagner M, Wang L, Yasuda N, Yee H K C andSupernova Cosmology Project T 2009 Astron. J. , 1271–1283.De Lucia G and Blaizot J 2007
Mon. Not. R. Astron. Soc. , 2–14.Demarco R, Rosati P, Lidman C, Girardi M, Nonino M, Rettura A, Strazzullo V, van der WelA, Ford H C, Mainieri V, Holden B P, Stanford S A, Blakeslee J P, Gobat R, Postman M,Tozzi P, Overzier R A, Zirm A W, Ben´ıtez N, Homeier N L, Illingworth G D, Infante L, JeeM J, Mei S, Menanteau F, Motta V, Zheng W, Clampin M and Hartig G 2007
Astrophys. J. , 164–182.Demarco R, Wilson G, Muzzin A, Lacy M, Surace J, Yee H K C, Hoekstra H, Blindert K andGilbank D 2010
Astrophys. J. , 1185–1197.Du ff y A R, Schaye J, Kay S T and Dalla Vecchia C 2008 Mon. Not. R. Astron. Soc. , L64–L68.Eisenhardt P R M, Brodwin M, Gonzalez A H, Stanford S A, Stern D, Barmby P, Brown M J I,Dawson K, Dey A, Doi M, Galametz A, Jannuzi B T, Kochanek C S, Meyers J, MorokumaT and Moustakas L A 2008
Astrophys. J. , 905–932.Erben T, Hildebrandt H, Lerchster M, Hudelot P, Benjamin J, van Waerbeke L, SchrabbackT, Brimioulle F, Cordes O, Dietrich J P, Holhjem K, Schirmer M and Schneider P 2009
Astron. Astrophys. , 1197–1222.Fassbender R 2003 Commissioning of the near IR camera OMEGA2000 and development ofa pipeline reduction system Master’s thesis University of Heidelberg.
EFERENCES z > ∼ / Astron. Astrophys. , L73–L77.Fassbender R, B¨ohringer H, Santos J S, Pratt G W, ˇSuhada R, Kohnert J, Lerchster M, RovilosE, Pierini D, Chon G, Schwope A D, Lamer G, M ¨uhlegger M, Rosati P, Quintana H, NastasiA, de Hoon A, Seitz S and Mohr J J 2011b
Astron. Astrophys. , A78.Fassbender R, Nastasi A, B¨ohringer H, ˇSuhada R, Santos J S, Rosati P, Pierini D, M ¨uhleggerM, Quintana H, Schwope A D, Lamer G, de Hoon A, Kohnert J, Pratt G W and Mohr J J2011a
Astron. Astrophys. , L10.Finoguenov A, Guzzo L, Hasinger G, Scoville N Z, Aussel H, B¨ohringer H, Brusa M, CapakP, Cappelluti N, Comastri A, Giodini S, Gri ffi ths R E, Impey C, Koekemoer A M, KneibJ P, Leauthaud A, Le F`evre O, Lilly S, Mainieri V, Massey R, McCracken H J, Mobasher B,Murayama T, Peacock J A, Sakelliou I, Schinnerer E, Silverman J D, Smolˇci´c V, TaniguchiY, Tasca L, Taylor J E, Trump J R and Zamorani G 2007 Astrophys. J. Suppl. , 182–195.Fioc M and Rocca-Volmerange B 1997 in ‘ASSL Vol. 210: The Impact of Large Scale Near-IR Sky Surveys’ p. 257.Gabasch A 2004 Galaxy Evolution in the FORS Deep Field PhD thesis Ludwig-MaximilianUniversity Munich.Garilli B, Fumana M, Franzetti P, Paioro L, Scodeggio M, Le F`evre O, Paltani S andScaramella R 2010 PASP , 827–838.Gilbank D G, Gladders M D, Yee H K C and Hsieh B C 2011
Astron. J. , 94.Gladders M D and Yee H K C 2000
Astron. J. , 2148–2162.Gladders M D and Yee H K C 2005
Astrophys. J. Suppl. , 1–29.Gonzalez A H, Zaritsky D, Dalcanton J J and Nelson A 2001
Astrophys. J. Suppl. , 117–138.Grove L F, Benoist C and Martel F 2009
Astron. Astrophys. , 845–855.Gwyn S D J 2008
PASP , 212–223.Haarsma D B, Leisman L, Donahue M, Bruch S, B¨ohringer H, Croston J H, Pratt G W, VoitG M, Arnaud M and Pierini D 2010
Astrophys. J. , 1037–1047.Hashimoto Y, Henry J P, Hasinger G, Szokoly G and Schmidt M 2005
Astron. Astrophys. , 29–33.Hayashi M, Kodama T, Koyama Y, Tanaka I, Shimasaku K and Okamura S 2010
Mon. Not.R. Astron. Soc. , 1980–1990.Hewett P C, Foltz C B and Cha ff ee F H 1995 Astron. J. , 1498–1521.Hilton M, Collins C A, Stanford S A, Lidman C, Dawson K S, Davidson M, Kay S T, LiddleA R, Mann R G, Miller C J, Nichol R C, Romer A K, Sabirli K, Viana P T P and West M J2007
Astrophys. J. , 1000–1009.
EFERENCES
Astrophys. J. , 133–147.Hilton M, Stanford S A, Stott J P, Collins C A, Hoyle B, Davidson M, Hosmer M, Kay S T,Liddle A R, Lloyd-Davies E, Mann R G, Mehrtens N, Miller C J, Nichol R C, Romer A K,Sabirli K, Sahl´en M, Viana P T P, West M J, Barbary K, Dawson K S, Meyers J, PerlmutterS, Rubin D and Suzuki N 2009
Astrophys. J. , 436–451.Jee M J, Rosati P, Ford H C, Dawson K S, Lidman C, Perlmutter S, Demarco R, Strazzullo V,Mullis C, B¨ohringer H and Fassbender R 2009
Astrophys. J. , 672–686.Jones L R, Ponman T J, Horton A, Babul A, Ebeling H and Burke D J 2003
Mon. Not. R.Astron. Soc. , 627–638.Kaiser N 1986
Mon. Not. R. Astron. Soc. , 323–345.Kim D W, Cameron R A, Drake J J, Evans N R, Freeman P, Gaetz T J, Ghosh H, GreenP J, Harnden, Jr. F R, Karovska M, Kashyap V, Maksym P W, Ratzla ff P W, Schlegel E M,Silverman J D, Tananbaum H D, Vikhlinin A A, Wilkes B J and Grimes J P 2004
Astrophys.J. Suppl. , 19–41.Kurtz M J and Mink D J 1998
PASP , 934–977.Lerchster M, Seitz S, Brimioulle F, Fassbender R, Rovilos M, B¨ohringer H, Pierini D,Kilbinger M, Finoguenov A, Quintana H and Bender R 2011
Mon. Not. R. Astron. Soc. , 2667–2694.Lin Y T and Mohr J J 2007
Astrophys. J. Suppl. , 71–94.Lloyd-Davies E J, Romer A K, Mehrtens N, Hosmer M, Davidson M, Sabirli K, Mann R G,Hilton M, Liddle A R, Viana P T P, Campbell H C, Collins C A, Dubois E N, Freeman P,Harrison C D, Hoyle B, Kay S T, Kuwertz E, Miller C J, Nichol R C, Sahlen M, StanfordS A and Stott J P 2010 arXiv:1010.0677 .Mantz A, Allen S W, Ebeling H, Rapetti D and Drlica-Wagner A 2010
Mon. Not. R. Astron.Soc. , 1773–1795.Markevitch M and Vikhlinin A 2007
Phys. Rep. , 1–53.Marriage T A, Acquaviva V, Ade P A R, Aguirre P, Amiri M, Appel J W, Barrientos L F,Battistelli E S, Bond J R, Brown B, Burger B, Chervenak J, Das S, Devlin M J, Dicker S R,Doriese W B, Dunkley J, Dunner R, Essinger-Hileman T, Fisher R P, Fowler J W, HajianA, Halpern M, Hasselfield M, Hern’andez-Monteagudo C, Hilton G C, Hilton M, HincksA D, Hlozek R, Hu ff enberger K M, Hughes D H, Hughes J P, Infante L, Irwin K D, JuinJ B, Kaul M, Klein J, Kosowsky A, Lau J M, Limon M, Lin Y , Lupton R H, Marsden D,Martocci K, Mauskopf P, Menanteau F, Moodley K, Moseley H, Netterfield C B, NiemackM D, Nolta M R, Page L A, Parker L, Partridge B, Quintana H, Reese E D, Reid B, SehgalN, Sherwin B D, Sievers J, Spergel D N, Staggs S T, Swetz D S, Switzer E R, Thornton R,Trac H, Tucker C, Warne R, Wilson G, Wollack E and Zhao Y 2010 arXiv:1010.1065 .Maughan B J, Jones L R, Pierre M, Andreon S, Birkinshaw M, Bremer M N, Pacaud F,Ponman T J, Valtchanov I and Willis J 2008 Mon. Not. R. Astron. Soc. , 998–1006.
EFERENCES arXiv:1106.3056 .Middelberg E, Norris R P, Cornwell T J, Voronkov M A, Siana B D, Boyle B J, Ciliegi P,Jackson C A, Huynh M T, Berta S, Rubele S, Lonsdale C J, Ivison R J and Smail I 2008
Astron. J. , 1276–1290.Miley G and De Breuck C 2008
Astron. Astrophys. Rev. , 67–144.Mittal R, Hudson D S, Reiprich T H and Clarke T 2009 Astron. Astrophys. , 835–850.Mohr J J, Fabricant D G and Geller M J 1993
Astrophys. J. , 492–505.M ¨uhlegger M 2010 Simulated Observations of Galaxy Clusters for Current and Future X-raySurveys Phd thesis TU M ¨unchen.Mullis C R, Rosati P, Lamer G, B¨ohringer H, Schwope A, Schuecker P and Fassbender R2005
Astrophys. J. Letters , L85–L88.Muzzin A, Wilson G, Yee H K C, Hoekstra H, Gilbank D, Surace J, Lacy M, Blindert K,Majumdar S, Demarco R, Gardner J P, Gladders M and Lonsdale C 2009
Astrophys. J. , 1934–1942.Nastasi A, Fassbender R, B¨ohringer H, ˇSuhada R, Rosati P, Pierini D, Verdugo M, SantosJ S, Schwope A D, de Hoon A, Kohnert J, Lamer G, M ¨uhlegger M and Quintana H 2011
Astron. Astrophys. , L6.Olsen L F, Benoist C, Cappi A, Maurogordato S, Mazure A, Slezak E, Adami C, Ferrari Cand Martel F 2007
Astron. Astrophys. , 81–93.Pacaud F, Pierre M, Adami C, Altieri B, Andreon S, Chiappetti L, Detal A, Duc P, GalazG, Gueguen A, Le F`evre J, Hertling G, Libbrecht C, Melin J, Ponman T J, QuintanaH, Refregier A, Sprimont P, Surdej J, Valtchanov I, Willis J P, Alloin D, Birkinshaw M,Bremer M N, Garcet O, Jean C, Jones L R, Le F`evre O, Maccagni D, Mazure A, Proust D,R¨ottgering H J A and Trinchieri G 2007
Mon. Not. R. Astron. Soc. , 1289–1308.Pacaud F, Pierre M, Refregier A, Gueguen A, Starck J L, Valtchanov I, Read A M, Altieri B,Chiappetti L, Gandhi P, Garcet O, Gosset E, Ponman T J and Surdej J 2006
Mon. Not. R.Astron. Soc. , 578–590.Pierini D, Maraston C, Gordon K D and Witt A N 2005
Mon. Not. R. Astron. Soc. , 131–145.Planck Collaboration, Ade P A R, Aghanim N, Arnaud M, Ashdown M, Aumont J,Baccigalupi C, Balbi A, Banday A J, Barreiro R B and et al. 2011 arXiv:1101.2024 .Pratt G W and Arnaud M 2003
Astron. Astrophys. , 1–16.Pratt G W, Croston J H, Arnaud M and B¨ohringer H 2009
Astron. Astrophys. , 361–378.
EFERENCES ff ermann E, Reiprich T, Robrade J, Roh´e C, Santangelo A, Sch¨achner G, Schanz T,Schmid C, Schmitt J, Schreib R, Schrey F, Schwope A, Steinmetz M, Str¨uder L, SunyaevR, Tenzer C, Tiedemann L, Vongehr M and Wilms J 2010 in ‘Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series’ Vol. 7732 of Presented at the Societyof Photo-Optical Instrumentation Engineers (SPIE) Conference .Reichert A, B¨ohringer H, Fassbender R and M ¨uhlegger M 2011 arXiv:1109.3708 .Rettura A, Rosati P, Nonino M, Fosbury R A E, Gobat R, Menci N, Strazzullo V, Mei S,Demarco R and Ford H C 2010
Astrophys. J. , 512–524.Romer A K, Viana P T P, Liddle A R and Mann R G 2001
Astrophys. J. , 594–608.Rosati P, Borgani S, della Ceca R, Stanford A, Eisenhardt P and Lidman C 2000 in M Plionis and I Georgantopoulos, eds, ‘Large Scale Structure in the X-ray Universe,Proceedings of the 20-22 September 1999 Workshop, Santorini, Greece, eds. Plionis, M. &Georgantopoulos, I., Atlantisciences, Paris, France’ p. 13.Rosati P, della Ceca R, Norman C and Giacconi R 1998
Astrophys. J. Letters , L21.Rosati P, Stanford S A, Eisenhardt P R, Elston R, Spinrad H, Stern D and Dey A 1999
Astron.J. , 76–85.Rosati P, Tozzi P, Ettori S, Mainieri V, Demarco R, Stanford S A, Lidman C, Nonino M,Borgani S, Della Ceca R, Eisenhardt P, Holden B P and Norman C 2004
Astron. J. , 230–238.Rosati P, Tozzi P, Gobat R, Santos J S, Nonino M, Demarco R, Lidman C, Mullis C R,Strazzullo V, B¨ohringer H, Fassbender R, Dawson K, Tanaka M, Jee J, Ford H, LamerG and Schwope A 2009
Astron. Astrophys. , 583–591.R¨oser H J, Hippelein H, Wolf C, Zatloukal M and Falter S 2010
Astron. Astrophys. , A15.Sanderson A J R, Edge A C and Smith G P 2009
Mon. Not. R. Astron. Soc. , 1698–1705.Santos J S, Fassbender R, Nastasi A, B¨ohringer H, Rosati P, ˇSuhada R, Pierini D, Nonino M,M ¨uhlegger M, Quintana H, Schwope A D, Lamer G, de Hoon A and Strazzullo V 2011
Astron. Astrophys. , L15.Santos J S, Rosati P, Gobat R, Lidman C, Dawson K, Perlmutter S, B¨ohringer H, Balestra I,Mullis C R, Fassbender R, Kohnert J, Lamer G, Rettura A, Rit´e C and Schwope A 2009
Astron. Astrophys. , 49–60.Scharf C 2002
Astrophys. J. , 157–159.Schlegel D J, Finkbeiner D P and Davis M 1998
Astrophys. J. , 525.Schwope A D, Lamer G, de Hoon A, Kohnert J, B¨ohringer H, Dietrich J P, Fassbender R,Mohr J, M ¨uhlegger M, Pierini D, Pratt G W, Quintana H, Rosati P, Santos J and ˇSuhada R2010
Astron. Astrophys. , L10.
EFERENCES
PASP , 1284–1295.Short C J, Thomas P A, Young O E, Pearce F R, Jenkins A and Muanwong O 2010
Mon. Not.R. Astron. Soc. , 2213–2233.Smith G P, Khosroshahi H G, Dariush A, Sanderson A J R, Ponman T J, Stott J P, Haines C P,Egami E and Stark D P 2010
Mon. Not. R. Astron. Soc. p. 1499.Smith J A, Tucker D L, Kent S, Richmond M W, Fukugita M, Ichikawa T, Ichikawa S i,Jorgensen A M, Uomoto A, Gunn J E, Hamabe M, Watanabe M, Tolea A, Henden A,Annis J, Pier J R, McKay T A, Brinkmann J, Chen B, Holtzman J, Shimasaku K and YorkD G 2002
Astron. J. , 2121–2144.Smolˇci´c V, Finoguenov A, Zamorani G, Schinnerer E, Tanaka M, Giodini S and Scoville N2011
Mon. Not. R. Astron. Soc. p. L284.Sommer M W, Basu K, Pacaud F, Bertoldi F and Andernach H 2011
Astron. Astrophys. , A124.Stanek R, Rasia E, Evrard A E, Pearce F and Gazzola L 2010
Astrophys. J. , 1508–1523.Stanford S A, Elston R, Eisenhardt P R, Spinrad H, Stern D and Dey A 1997
Astron. J. , 2232.Stanford S A, Holden B, Rosati P, Eisenhardt P R, Stern D, Squires G and Spinrad H 2002
Astron. J. , 619–626.Stanford S A, Romer A K, Sabirli K, Davidson M, Hilton M, Viana P T P, Collins C A, KayS T, Liddle A R, Mann R G, Miller C J, Nichol R C, West M J, Conselice C J, Spinrad H,Stern D and Bundy K 2006
Astrophys. J. Letters , L13–L16.Strazzullo V, Rosati P, Pannella M, Gobat R, Santos J S, Nonino M, Demarco R, Lidman C,Tanaka M, Mullis C R, Nu˜nez C, Rettura A, Jee M J, B¨ohringer H, Bender R, Bouwens R J,Dawson K, Fassbender R, Franx M, Perlmutter S and Postman M 2010
Astron. Astrophys. , A17.Tran K V H, Papovich C, Saintonge A, Brodwin M, Dunlop J S, Farrah D, Finkelstein K D,Finkelstein S L, Lotz J, McLure R J, Momcheva I and Willmer C N A 2010
Astrophys. J.Letters , L126–L129.ˇSuhada R, Fassbender R, Nastasi A, B¨ohringer H, de Hoon A, Pierini D, Santos J S, Rosati P,M ¨uhlegger M, Quintana H, Schwope A D, Lamer G, Kohnert J and Pratt G W 2011
Astron.Astrophys. , A110.ˇSuhada R, Song J, B¨ohringer H, Benson B A, Mohr J, Fassbender R, Finoguenov A, PieriniD, Pratt G W, Andersson K, Armstrong R and Desai S 2010
Astron. Astrophys. , L3.Vikhlinin A, Burenin R A, Ebeling H, Forman W R, Hornstrup A, Jones C, Kravtsov A V,Murray S S, Nagai D, Quintana H and Voevodkin A 2009a
Astrophys. J. , 1033–1059.
EFERENCES
Astrophys. J. , 1060–1074.Vikhlinin A, McNamara B R, Forman W, Jones C, Quintana H and Hornstrup A 1998
Astrophys. J. , 558.Watson M G, Schr¨oder A C, Fyfe D, Page C G, Lamer G, Mateos S, Pye J, Sakano M, RosenS, Ballet J, Barcons X, Barret D, Boller T, Brunner H, Brusa M, Caccianiga A, CarreraF J, Ceballos M, Della Ceca R, Denby M, Denkinson G, Dupuy S, Farrell S, FraschettiF, Freyberg M J, Guillout P, Hambaryan V, Maccacaro T, Mathiesen B, McMahon R,Michel L, Motch C, Osborne J P, Page M, Pakull M W, Pietsch W, Saxton R, SchwopeA, Severgnini P, Simpson M, Sironi G, Stewart G, Stewart I M, Stobbart A M, Tedds J,Warwick R, Webb N, West R, Worrall D and Yuan W 2009
Astron. Astrophys. , 339–373.White R L, Becker R H, Helfand D J and Gregg M D 1997
Astrophys. J. , 479.Williamson R, Benson B A, High F W, Vanderlinde K, Ade P A R, Aird K A, AnderssonK, Armstrong R, Ashby M L N, Bautz M, Bazin G, Bertin E, Bleem L E, BonamenteM, Brodwin M, Carlstrom J E, Chang C L, Chapman S C, Clocchiatti A, Crawford T M,Crites A T, de Haan T, Desai S, Dobbs M A, Dudley J P, Fazio G G, Foley R J, FormanW R, Garmire G, George E M, Gladders M D, Gonzalez A H, Halverson N W, Holder G P,Holzapfel W L, Hoover S, Hrubes J D, Jones C, Joy M, Keisler R, Knox L, Lee A T, LeitchE M, Lueker M, Luong-Van D, Marrone D P, McMahon J J, Mehl J, Meyer S S, MohrJ J, Montroy T E, Murray S S, Padin S, Plagge T, Pryke C, Reichardt C L, Rest A, RuelJ, Ruhl J E, Saliwanchik B R, Saro A, Scha ff er K K, Shaw L, Shiroko ff E, Song J, SpielerH G, Stalder B, Stanford S A, Staniszewski Z, Stark A A, Story K, Stubbs C W, Vieira J D,Vikhlinin A and Zenteno A 2011
Astrophys. J. , 139.Wilson G, Muzzin A, Yee H K C, Lacy M, Surace J, Gilbank D, Blindert K, Hoekstra H,Majumdar S, Demarco R, Gardner J P, Gladders M D and Lonsdale C 2009
Astrophys. J. , 1943–1950.Zenteno A, Song J, Desai S, Armstrong R, Mohr J J, Ngeow C C, Barkhouse W A, AllamS S, Andersson K, Bazin G, Benson B A, Bertin E, Brodwin M, Buckley-Geer E J, HansenS M, High F W, Lin H, Lin Y T, Liu J, Rest A, Smith R C, Stalder B, Stark A A, TuckerD L and Yang Y 2011
Astrophys. J.734