XMM Cluster Survey: X-ray source identification using the Sloan Digital Sky Survey

[ABRIDGED] The XMM Cluster Survey (XCS) is predicted to detect thousands of clusters observed serendipitously in XMM-Newton pointings. We investigate automating optical follow-up of cluster candidates using the SDSS public archive, concentrating on 42 XMM observations that overlap the SDSS First Data Release. Restricting to the inner 11 arcminutes of the XMM field-of-view gives 637 unique X-ray sources across a 3.09 deg^2 region with SDSS coverage. The log N-log S relation indicates survey completeness to a flux limit of around 1x10^{-14} erg s^-1 cm^-2. We have cross-correlated XMM point sources with SDSS quasars, finding 103 unique matches from which we determine a matching radius of 3.8 arcseconds. Using this we make immediate identifications of roughly half of all XMM point sources as quasars (159) or stars (55). We have estimated the typical error on SDSS-determined photometric cluster redshifts to be 5% for relaxed systems, and 11% for disturbed systems, by comparing them to spectroscopic redshifts for eight previously-known clusters. The error for disturbed systems may be problematic for cluster surveys relying on photometric redshifts alone. We use the False Discovery Rate to select 41 XMM sources (25 point-like, 16 extended) statistically associated with SDSS galaxy overdensities. Of the extended sources, 5 are new cluster candidates and 11 are previously-known clusters (0.044 1 clusters rediscovered by our wavelet-based detection software. These results show that the SDSS can provide useful automated follow-up to X-ray cluster surveys, but cannot provide all the optical follow-up necessary for X-ray cluster surveys.