Large homogeneous samples of galaxy clusters, spanning a wide range of redshift, are potentially powerful tools to study the evolution of the large scale structure in the Universe. Unfortunately, finding clusters at cosmologically interesting lookback times, say z > 0.5, let alone defining a complete sample, is a time consuming and difficult task. As a result, the much-needed observational constraints for theories of structure formation have not been forthcoming.
In an attempt to remedy this situation, Piero Rosati, in collaboration with Colin Norman and Roberto Della Ceca (all at Johns Hopkins), have embarked on a project, the ROSAT Deep Cluster Survey (RDCS), aimed at comstructing a large homogeneous sample of distant galaxy clusters selected solely on the basis of their X-ray properties. The X-ray selection offers two main advantages over optical selection: clusters are high contrast objects in the X-ray sky and the selection function can be modeled in a relatively straightforward way, being essentially that of a flux-limited sample. Cluster candidates are selected from a serendipitous search for extended X-ray sources in deep pointed observations drawn from the ROSAT-PSPC archive. A wavelet-based technique is used to detect and characterize low surface brightness X-ray sources. The completeness flux limit of the survey, 1 x 10-14 erg/cm2/s, is determined by the flux level at which extended and point-like emission can be reliably distinguished. The ROSAT-PSPC with its high sensitivity, low background, and good angular resolution (~ 30" FWHM), allows fluxes an order of magnitude fainter than those in previous X-ray cluster surveys to be reached. This selection technique yielded 150 candidates over an area of ~ 50 square degrees, drawn from 180 X-ray fields scattered across the two galactic caps.
To identify these candidates, Rosati and collaborators have undertaken a large optical follow-up program, consisting of deep imaging using the KPNO 4-m and 2.1-m, the CTIO 4-m and 1.5-m, multislit spectroscopy carried out with the CryoCam Spectrograph at the KPNO 4-m for the clusters in the North, and with the ESO 3.6m for those in the South. The imaging survey in I and V bands, now nearing completion, has shown a high success rate of identification, with about 100 new clusters confirmed to date (see Figure 1). These findings imply a surface of density of ~ 10 clusters/deg2 at the survey flux limit. The spectroscopic follow-up work has secured 75 cluster redshifts so far, spanning the range 0.1-0.8 (40 in the North). A significant fraction of the newly discovered clusters lie at high redshift: 28 at z > 0.4, 18 at z > 0.5 (Figure 2). The rapid build up of such a sample of spectroscopically confirmed distant clusters underscores the efficiency and the validity of the X-ray selection. This large fraction of distant clusters also implies that there is no dramatic dearth of X-ray clusters at high redshifts or negative strong evolution as suggested by previous shallower X-ray surveys; this finding is in keeping with the results of optical surveys (e.g. the Palomar Deep Cluster Survey by Postman and collaborators). A detailed investigation of the issue of cluster evolution will soon be possible when the redshift survey is complete and the Xray luminosity function (XLF) is constructed at different redshifts.
The depth of the RDCS allows the faint end of the XLF to be probed at moderate-to-high redshifts for the first time. In addition to constraining cosmological models, this opens up the possibility to study galaxy evolution in systems with X-ray luminosities equal to and well below the local L* (roughly the Coma cluster), which span a variety of rich environments and constitute the bulk of the cluster population in the Universe.