The WIRCam Deep Survey. II. Mass selected clustering.

We present an analysis of the clustering of galaxies from z ≈ 2 to the present day using the WIRCam Deep Survey (WIRDS). WIRDS combines deep optical data from the CFHTLS Deep fields with its own deep near-infrared data, providing a photometric data-set over an effective area of 2.4 deg2, from which accurate photometric redshifts and stellar masses can be estimated. We use the data to calculate the angular correlation function for galaxy samples split by star-formation activity, stellar mass and redshift. Using WIRDS with its large total area and multiple fields gives a low cosmic variance contribution to the error, which we estimate to be less than ~2.8%. Based on power-law fits, we estimate the real-space clustering for each sample, determining clustering lengths and power-law slopes. For galaxies selected by constant mass, we find that the clustering scale shows no evolution up to z ≈ 2. Splitting the galaxy sample by mass, we see a consistent trend for higher mass galaxies to have larger clustering scales at all redshifts considered. We use our results to test the GALFORM semi-analytical model of galaxy formation and evolution. The observed trends are well matched by the model galaxies for both the redshift evolution and the mass dependence of the galaxy clustering. We split the galaxy population into passive and star-forming populations based on rest-frame dust-corrected NUV-r colours. We find that the passive galaxy populations show a significantly larger clustering scale at all redshifts than the star-forming population below masses of M⋆ ~ 1011 h-1 M⊙, showing that even at z ≈ 2 passive galaxies exist in denser environments than the bulk of the star-forming galaxy population. For star-forming galaxies with stellar masses of M⋆ ≳ 1011 h-1 M⊙, we find a clustering strength of ~8 h-1 Mpc across all redshifts, comparable to the measurements for the passive population. Additionally, for star-forming galaxies we see that clustering strength increases for higher stellar mass systems, however little sign of a mass dependence in passive galaxies is observed over the range in stellar mass that is probed. Comparing our results to the model galaxy population produced by GALFORM, we find a qualitative good agreement between the model predictions and the observed clustering. Finally, we investigate the connection between galaxy stellar mass and dark matter halo mass, showing a clear correlation between the two in both the WIRDS data and the GALFORM predictions