| Summary: | The number density and clustering evolution of massive galaxies, particularly at redshift <I>z </I>> 1, can be used to place strong constraints on galaxy evolution models. On the one hand passive evolution models predict a constant comoving density of massive galaxies to high redshifts while on the other hand hierarchical merging models predict a strong decline in number density beyond <I>z</I> ~ 1. Comparisons between observations and models are best done in the near-infrared since, unlike optical observations which are affected by star-formation and dust, and near-infrared flux of a galaxy is more closely related to its total stellar mass. However, obtaining statistically significant of near-infrared selected galaxies has only recently become possible with the advent of wide-field near-infrared detectors. This thesis presents the Las Campanas Infra-Red (LCIR) survey - one of the widest-field deep near-infrared surveys to date. The survey covers one square degree to near-infrared magnitude <I>H</I> ~ 20.5, complemented with deep follow-up in 5-6 optical filters. The observing strategy, data reduction and catalogue generation are described. Photometric redshifts are determined and the galaxy redshift distributions are compared with the predictions of passive evolution and semi-analytic hierarchical merging models. Special attention is given to the population of Extremely Red Objects, many of which are believed to be massive evolved galaxies at <I>z</I> < 1. Angular and spatial clustering statistics are also calculated. The evolution of the red galaxy number counts and clustering from <I>z</I> = 1.2 (corresponding to a time when the Universe was less than half its present age, in currently favoured cosmologies) to <I>z</I> = 0 is assessed by combining the LCIR survey results with previous and new measurements at lower redshifts. The results are compared with models and the consequences for galaxy evolution are discussed. A separate chapter is devoted to a detailed review of photometric redshift techniques. Their reliability is reassessed as a function of template set, filter set and the target galaxy redshift and spectral type. Additionally, new techniques - including the use of semi-analytic models, principal components analysis and artificial neural networks - are investigated.
|
|---|
