Wednesday

Upcoming surveys (such as LSST, Euclid, eROSITA, CMB-S4 etc.) will require robust and accurate theoretical predictions for the distribution of matter on large scales, so that these surveys can constrain the standard model of cosmology and to test its possible extensions. Cosmological hydrodynamical simulations can, at least in principle, provide such predictions, but the task is too computationally demanding at present to simulate the full cosmological landscape.
The halo model formalism is an attractive alternative, given the speed at which the predictions can be generated. In fact, the halo model formalism makes use of simple mass-redshift dependent functions (that strongly depends on the chosen cosmology) to trace several LSS observables without running various simulations. However, it suffers from well-known issues in the non-linear regime and the handling of the complex physics of galaxy formation.
Here we present some interesting results from a project that aims to characterise and improve the performance of the halo model by making detailed comparisons to the BAHAMAS suite of cosmological hydrodynamical simulations and real observations (Planck, SPT, ACT, DES etc). We explore how well the halo model performs at recovering several cosmological probes like the true Sunyaev-Zel'dovich, weak lensing, and X-ray auto- and cross-correlation functions and make physical adjustments to the halo model to better reproduce the simulations.