The clustering properties of the reddest galaxies
with UKIDSS-DXS

Abstract

Deep, photometric near-infrared surveys are the most effective ways to select massive galaxies at redshift , where their characteristic spectral features are observed at wavelengths longer than optical surveys are designed to measure.
In this Thesis, we present and study the clustering properties of such galaxies using the UKIDSS-DXS near-infrared survey.

Firstly we produce deep, uniform mosaics for the of and band imaging, which reach depths of ( for point sources), and produce galaxy catalogues which we match with deep optical surveys.
We demonstrate that these catalogues have reliable photometry (by way of comparison with 2MASS), we are able to efficiently remove foreground galactic stars, and that we produce the expected number counts galaxies.
These catalogues will be made available at some point in the near future.

From these catalogues we select Extremely Red Objects (EROs), Distant Red Galaxies (DRGs) and s, all of which aim to exploit the strong break which is present in the oldest stellar populations.
The combination of area and depth provided by DXS allows us to probe the large scale structure of the universe to tens of megaparsecs at .
We demonstrate using the small area of -band DXS coverage that selection of EROs is still effective with similar criteria in and , although do not suggest any new criteria different to those already existing in literature.

We measure the angular two point correlation of the EROs and passive selection of galaxies.
We show that these selections of galaxies exhibit significantly stronger clustering on the smallest scales than what is often assumed by the clustering studies of main sequence galaxies, indicating that these galaxies are residing in the most massive dark matter haloes.
Further, we demonstrate that there is a strong break in the correlation function in these selections of galaxies.

The correlation length we measure when considering the small scale (intra-halo) clustering alone is roughly constant for across the range of magnitudes DXS detects EROs, to .
Given that the halo mass of will be determined by the largest mass galaxy, we interpret this (in a qualitative sense) to mean that the dark matter host halo mass of the most massive EROs is roughly constant over the redshift range DXS measures EROs ().
A comparison of the angular correlation function for passive and star-forming galaxies (for three magnitude cuts where we have sufficient number counts) indicates that the passive galaxies are indeed much more strongly clustered than the star-forming: evidence of environmental quenching.

We present -band luminosity functions of our galaxy catalogues, and the ERO selection for .
We observe that the faint end slope of the ERO luminosity function is much steeper than those of the full galaxy catalogues, confirming that the ERO criteria selects galaxies that are amongst the most luminous in this redshift range.
Finally, we measure the galaxy stellar mass function of the full sample of galaxies, and of restframe selected passive galaxies over the range .
We compute the integrated stellar mass density of the full galaxy sample, and our values match well with those in the literature.
By computing the galaxy stellar mass density of the passive galaxy sample, and the contribution of the very massive end, we observe that the stellar mass density in the most massive passive galaxies is increasing faster than the stellar mass density of the passive galaxies as a whole, an observation which agrees with previous studies.
Given that the galaxy stellar mass function of passive galaxies is well-described by a single Schechter function, which implies a single mechanism of galaxy quenching, we rule out a significant amount of dry mergers at the massive end, and instead postulate that this tentatively suggests that the mass quenching efficiency of the most massive galaxies is lower at later times.

UKIDSS-DXS represents only a fraction of the vast datasets that the upcoming generation of instruments and surveys such as -LSST, and will produce, and this work is a precursor to investigations which use these surveys.