Distant cluster hunting. II. A comparison of X-ray and optical cluster detection techniques and catalogs from the ROSAT optical X-ray survey

We present and analyze the optical and X-ray catalogs of moderate-redshift cluster candidates from the ROSAT Optical X-Ray Survey, or ROXS. The survey covers the sky area contained in the fields of view of 23 deep archival ROSAT PSPC pointings, 4.8 square degrees. The cross-correlated cluster catalogs were constructed by comparing two independent catalogs extracted from the optical and X-ray bandpasses, using a matched-filter technique for the optical data and a wavelet technique for the X-ray data. We cross-identified cluster candidates in each catalog. As reported in Paper I, the matched-filter technique found optical counterparts for at least 60% (26 out of 43) of the X-ray cluster candidates; the estimated redshifts from the matched filter algorithm agree with at least 7 of 11 spectroscopic confirmations (Deltaz less than or similar to 0.10). The matched filter technique, with an imaging sensitivity of m(I) similar to 23, identified approximately 3 times the number of candidates (155 candidates, 142 with a detection confidence > 3 sigma) found in the X-ray survey of nearly the same area. There are 57 X-ray candidates, 43 of which are unobscured by scattered light or bright stars in the optical images. Twenty-six of these have fairly secure optical counterparts. We find that the matched filter algorithm, when applied to images with galaxy flux sensitivities of m(I) similar to 23, is fairly well-matched to discovering z less than or equal to 1 clusters detected by wavelets in ROSAT PSPC exposures of 8000 60, 000 s. The difference in the spurious fractions between the optical and X-ray (30% and 10%, respectively) cannot account for the difference in source number. In Paper I, we compared the optical and X-ray cluster luminosity functions and we found that the luminosity functions are consistent if the relationship between X-ray and optical luminosities is steep (L-X proportional to L-opt(3-4)). Here, in Paper II, we present the cluster catalogs and a numerical simulation of the ROXS. We also present color-magnitude plots for several of the cluster candidates, and examine the prominence of the red sequence in each. We find that the X-ray clusters in our survey do not all have a prominent red sequence. We conclude that while the red sequence may be a distinct feature in the color-magnitude plots for virialized massive clusters, it may be less distinct in lower mass clusters of galaxies at even moderate redshifts. Multiple, complementary methods of selecting and de ning clusters may be essential, particularly at high redshift where all methods start to run into completeness limits, incomplete understanding of physical evolution, and projection effects.