Obscured activity and the role of environment on galaxy evolution at high redshift

A significant amount of activity in the Universe is obscured by dust, produced in the final phases of stellar evolution and in the detonation of Type II supernovae. Re-processed radiation from starlight is emitted from this dust at infrared wavelengths, and this must be taken into consideration when performing surveys of star formation (and nuclear activity) in order to form an unbiased picture of galaxy evolution. It is also clear that the star formation histories of galaxies are significantly modified by their local environment, the outcome of which is the characteristic galaxy populations observed in rich clusters and in the field in the local Universe. In this thesis I examine galaxy evolution in the context of environment from z ~ 0.5 to 2 ~ 3, paying attention to obscured activity revealed by observations in the rest-frame infrared. A mid-infrared (24μm) survey of two intermediate redshift clusters reveals a population of luminous infrared galaxies (LIRGs) which are missed in optical surveys (or significantly underestimated in terms of their star formation rates). Despite there being a large difference between the number of LIRGs detected in the two clusters (likely due to varying global cluster properties controlling the survival of starbursts in the cluster environment), these could be a potentially important population of galaxies. Their large star formation rates mean that they could evolve into local passive S0s by the present day. Although the S0s must be assembled after z ~ 0.5, local clusters are also dominated by massive elliptical galaxies which are mostly already in place by z ~ 0.5, and therefore must have assembled their stellar mass at much higher redshifts (z ≥3). At z = 3.1 I examine the nature of extreme activity in a rich, primitive environment - an example of a progenitor of a rich cluster of galaxies, and therefore the likely site of formation of local massive ellipticals. A number of giant (100 kpc-scale) Lyman-α emission-line nebulae (LABs) in the SA 22 protocluster contain bright submillimeter (850μm) galaxies (SMGs). Their extremely luminous rest-frame far-infrared emission suggest very high star formation rates and/or nuclear activity. Given that a large fraction of LABs seem to contain these active galaxies, it is plausible to link LABs' formation with feedback events such as superwind outflows from starburst regions. Indeed, a weak correlation between the SMGs' bolometric luminosity and the LABs' Lya luminosities appears to suggest that SMGs are powering these extended haloes. Although feedback from active galaxies appears to be important at early times, it remains a significant factor in galaxy-environment symbiosis at all epochs. The most profound effect a galaxy can have on its surroundings is to impart energy to the surrounding medium. In clusters, this is important for preventing the cooling of baryons and therefore the truncation of star formation. I investigate the environments of four low-power (L(_1.4GHz) ≤ 10(^25) WHz (^-1)) radio galaxies in the Subaru-XMM-Newton Deep Field at z ~ 0.5. The environments are all found to be moderately rich groups, and at least one shows evidence that it is in a stage of cluster assembly via sub-group merging. The conclusion is that the radio loud active galactic nuclei are triggered by galaxy-galaxy interactions within sub-groups, prior to cluster virialisation. These radio galaxies are destined to become brightest cluster galaxies, providing a low-power, but high-duty cycle feedback on gas in high-density regions at low redshift - necessary to suppress star formation in massive ellipticals at z ~ 0. The hostility of clusters to star formation (or at least the observation that it is suppressed in the highest density regions of the local Universe) is thought to be in part responsible for the rapid decline in the global volume averaged star formation rate (SFRD) since 2 ~ 1. Tracking the evolution of the SFRD beyond z ~ 1 is hard, because optical tracers (e.g. Ha) used in the local Universe become redshifted into the near-infrared, and up until recently the cameras suitable for large surveys have not been available. I have performed the largest ever near-infrared narrowband blank field survey for Hα emission at z = 2.23. Understanding the evolution of the SFRD before its decline to the present day is essential if we are to find the 'epoch' of galaxy formation. I present the Hα luminosity function and measure the SFRD at this epoch, finding little evolution in the time between z = 1.3 and z = 2.23. This is consistent with a flattening of the SFRD, indicating that this is the peak era of star formation in the Universe, before the gradual suppression of activity during the build up of groups and clusters to the present day.EThOS - Electronic Theses Online ServiceGBUnited Kingdo