Speaker
Description
The current and forthcoming Stage-IV galaxy surveys are going to map the Unvierse in a large sky area and redshift range. The statistical error will be dramatically reduced, hence, controlling the modelling systematics below the statistical error becomes crucial in the era of precision cosmology. Usually, N-body simulations are adopted to study theoretical systematics of models. In order to mitigate the sample variance of simulations, it is typical to conduct a significant number of simulations. However, it is too expensive in terms of computational time and disk usage to run simulations with a large volume. In our study, we apply the CARPool method to suppress the sample variance of AbacusSummit galaxy clustering with the assistance of FastPM simulations. For the paired FastPM with the same initial conditions from AbacusSummit, we apply the halo occupation distribution model to obtain the galaxy number density and two-point clustering well matched to those of AbacusSummit. For both luminous red galaxies and emission line galaxies of AbacusSummit, we obtain a significant reduction factor on the sample variance. Therefore, our study can be substantial for a tighter constraint on the systematics of BAO and RSD models for the DESI survey.
