The role of obscured activity in galaxy formation

In this thesis I investigate the formation and evolution of the galaxies that eventually form the colour-magnitude relation (CMR) in local galaxy clusters. I survey galaxies that lie on the CMR in nine massive clusters at z~0.2, environments in which the build-up of the faint end of the CMR is still underway. I show that there are relatively few dwarf galaxies on the CMR in the outer, low-density regions of clusters, but that their fraction increases towards higher-density regions as the cluster environment transforms infalling, blue, star-forming galaxies into red, passive, CMR galaxies. However, in the highest density regions, at the very centres of clusters the relative fraction of dwarf galaxies on the CMR is suppressed, evidence that, dwarf galaxies in the highest density regions at z~0.2 are dynamically disrupted. I then use 1.1-mm observations of a massive cluster at z~0.54 to search for active, star-forming cluster galaxies, which would transform into CMR galaxies at lower redshifts as the star-formation terminates. I detect 36 sources in observations of 0.1 deg² of the cluster centre and identify counterparts to ~50% of these submillimetre galaxies (SMGs) using radio, 24-μm and IRAC data. Photometric redshifts suggest that at most two of the SMGs are potential cluster members. If this is the case they each have far-infrared luminosities of ~5x10¹¹ solar luminosities and star-formation rates (SFRs) of ~50 solar masses per year - a significant fraction of the combined SFR of the cluster. I next consider 126 SMGs detected in an 870-μm survey of the Extended Chandra Deep Field South (ECDFS). I derive a photometric redshift distribution of 74 robust radio, 24-μm and IRAC-identified counterparts that peaks at z=2.2. An analysis of sources within the positional error circles of unidentified SMGs identifies a population of likely counterparts with a redshift distribution that peaks at z=2.5±0.3 and likely comprises ~60% of the unidentified SMGs. The remainder are not detected in our IRAC imaging and likely lie at z≳3. In total, I find that ~30% of all SMGs are at z≳3, and the median redshift of all S₈₇₀μm>4 mJy SMGs is z=2.5±0.6. The contribution of SMGs to the global SFRD also peaks at z~2 and SMGs with S₈₇₀μm≳4 mJy and S₈₇₀μm≳1 mJy provide ~5% and ~50% of the global total at z~2, respectively. Analysis of the projected real-space cross-correlation function of SMGs at z=1-3 with IRAC-selected galaxies shows that SMGs are strongly clustered and reside in dark-matter halos of mass (6⁺¹²₋₅)x10¹² solar masses. This halo mass is comparable to that of quasars and the mass at which major mergers are most efficient at triggering starburst activity. I conclude that SMGs at z~2 have star-formation rates, stellar masses and clustering properties that suggests that they are the likely progenitors of the massive CMR galaxies that dominate local clusters.EThOS - Electronic Theses Online ServiceGBUnited Kingdo