Semi-analytic galaxy formation during the epoch of the reionisation

Semi-analytic models play an important role in modelling the epoch of reionisation. This thesis presents three studies that are related to this topic. First, we measure clustering segregation with both UV-luminosity and stellar mass at z > 4, which is then compared with predictions from the Meraxes semi-analytic model. Our results suggest that the dependence of clustering strength on UV-luminosity is stronger than stellar mass, indicating that compared with stellar mass, UV-luminosity is more tightly correlated with halo mass. Secondly, we investigate dust extinction in the early Universe. Our method utilises the Meraxes semi-analytic model to produce intrinsic galaxy luminosity and adopts parametric relations to estimate dust extinction. A novelty of our approach is that intrinsic luminosity and dust extinction are determined simultaneously by calibrating both galaxy formation and dust models against only UV observations. Our results suggest that there is a factor of two systematic error in the estimations of the cosmic star formation rate density based on the dust law in the local Universe. Finally, we present a method to augment N-body simulations using a Monte Carlo algorithm, which increases the mass resolution of the simulations. The results can be used by semi-analytic models of reionisation to overcome the challenge that convergent predictions of the reionisation history require both high mass resolution and large simulation volume. The effectiveness of our method is tested using a high resolution small volume N-body simulation.