Clustering of Extremely Red Objects in Elais-N1 from the UKIDSS DXS with optical photometry from Pan-STARRS1 and Subaru
We measure the angular clustering of 33 415 extremely red objects (EROs) in the Elais-N1 field covering 5.33 deg$^{2}$, which cover the redshift range $z=0.8$ to $2$. This sample was made by merging the UKIDSS Deep eXtragalactic Survey (DXS) with the optical Subaru and Pan-STARRS PS1 datasets. We confirm the existence of a clear break in the angular correlation function at $\sim 0.02^{\circ}$ corresponding to $1 h^{-1}$ Mpc at $z\sim1$. We find that redder or brighter EROs are more clustered than bluer or fainter ones. Halo Occupation Distribution (HOD) model fits imply that the average mass of dark matter haloes which host EROs is over $10^{13} h^{-1} M_{\odot}$ and that EROs have a bias ranging from 2.7 to 3.5. Compared to EROs at $z\sim1.1$, at $z\sim1.5$ EROs have a higher bias and fewer are expected to be satellite galaxies. Furthermore, EROs reside in similar dark matter haloes to those that host $10^{11.0} M_{\odot}<M_{*}<10^{11.5} M_{\odot}$ galaxies. We compare our new measurement and HOD fits with the predictions of the GALFORM semi-analytical galaxy formation model. Overall, the clustering predicted by GALFORM gives an encouraging match to our results. However, compared to our deductions from the measurements, GALFORM puts EROs into lower mass haloes and predicts that a larger fraction of EROs are satellite galaxies. This suggests that the treatment of gas cooling may need to be revised in the model. Our analysis illustrates the potential of clustering analyses to provide observational constraints on theoretical models of galaxy formation.