2010A&A...523A..66B

We use a combination of CFHTLS deep optical data, WIRcam Deep Survey (WIRDS) near-infrared data and XMM-Newton survey data to identify z>1.1 clusters in the CFHTLS D1 and D4 fields. Counterparts to such clusters can not be identified without deep near-infrared data and as such the total of ≃1°² of J, H and K_s band imaging provided by WIRDS is an indispensable tool in such work. Using public XMM X-ray data, we identify extended X-ray sources in the two fields. The resulting catalogue of extended X-ray sources was then analyzed for optical/near-infrared counterparts, using a red-sequence algorithm applied to the deep optical and near-infrared data. Redshifts of candidate groups and clusters were estimated using the median photometric redshifts of detected counterparts and where available these were combined with spectroscopic data (from VVDS deep and ultra-deep and using AAT AAOmega data). Additionally, we surveyed X-ray point sources for potential group systems at the limit of our detection range in the X-ray data. A catalogue of z>1.1 cluster candidates in the two fields has been compiled and cluster masses, radii and temperatures have been estimated using the scaling relations. The catalogue of group and cluster candidates consists of 15 z>1.1 objects. We find several massive clusters (M>10¹⁴M_☉) and a number of lower mass clusters/groups. Three of the detections are previously published extended X-ray sources. Of note is JKSC 041 (previously detected via Chandra X-ray data and reported as a z=1.9 cluster based on UKIDSS infrared imaging) for which we identify a number of structures at redshifts of z=0.8, z=0.96, z=1.13 and z=1.49 (but see no evidence of a structure at z=1.9). We also make an independent detection of the massive cluster, XMMXCS J2215.9-1738, for which we estimate a redshift of z=1.37 (compared to the spectroscopically confirmed redshift of z=1.45). We use the z> 1.1 catalogue to compare the cluster number counts in these fields with models based on WMAP 7-year cosmology and find that the models slightly over-predict the observations, whilst at z>1.5 we do not detect any clusters. We note that cluster number counts at z>1.1 are highly sensitive to the cosmological model, however a significant reduction in present statistical (due to available survey area) and systematic (due to cluster scaling relations) uncertainties is required in order to confidently constrain cosmological parameters using cluster number counts at high redshift.