Bayesian analysis of the dynamic Universe

The spectacular progresses of the last decade in observational cosmology have tested the validity of our standard picture of the Universe at unprecedented accuracy. Challenges now arise from studying the inhomogeneities of the cosmic large-scale structure, which may contain a wealth of untamed information on some of main problems facing modern cosmology: the nature of dark energy, the masses of neutrinos, the processes driving cosmic inflation, and the presence of gravitational waves. Ideally, the connection between surveys and cosmological models should be made in a full-scale Bayesian framework, by exploring the multi-million dimensional parameter space. Tackling this idea, several algorithms and software packages have met some success in recent years. This project aims at exploiting these results to further our understanding of cosmology and fundamental physics. The proposed work includes the use of computational techniques to predict and extract physical signals, the application to synthetic and real galaxy and lensing surveys, and the cross-correlation with complementary astronomical data sets. As natural outcomes of the approach, the project will provide unprecedentedly detailed characterisations of the dynamic cosmic web in the local and extreme-scale Universe. Potential applications include inferring the Hubble expansion rate from supernovae, including the derived knowledge of the peculiar velocity distribution; investigating the kinetic Sunyaev-Zel'dovich effect from filaments; exploring correlations between galaxy properties, such as intrinsic alignments, and the matter distribution and tidal fields, to further the understanding of the galaxy formation process.