Campus, Varian 355
Extracting cosmological information from current survey data is increasingly reliant on computationally-intensive n-body simulations of large-scale structure. For weak-lensing surveys, the matter distribution is directly measured from simulations on scales where linear perturbation theory fails. For galaxy-clustering surveys, haloes are identified in simulations and then populated with galaxies to provide mock catalogues. Future surveys have grandiose plans to measure properties of dark energy, modified gravity, neutrinos etc. and, at first glance, the computational resources required to provide constraints on these 'extended' cosmological models will be enormous. In this talk, I will discuss various methods that may overcome some of this computational burden. These methods mainly use the 'halo model' of large-scale structure. I will discuss how the cosmology of an existing simulation can be changed by manipulating the data in post processing, how the power spectrum of matter fluctuations can be modelled semi-analytically, and how to include the non-linear bias of haloes in semi-analytical calculations.