One-parameter family of baryonic physics models in the large-scale structure

Abstract

Baryonic physics significantly influences the large-scale distribution of matter, posing a major challenge to weak gravitational lensing analyses. Numerous models have been proposed to account for these baryonic effects, often by comparing the power spectrum with and without baryonic contributions. While many of these models can be reduced to cases with only one free parameter, there is a pressing need to unify these approaches for future cost-effective simulations and surveys. In this paper, we examine five prevalent models, both analytically and numerically: an empirical model by van Daalen, an effective field theory (EFT) model by Lewandowski, two halo models (by Mead and Fedeli), and a phenomenological model by Amon. Our analysis reveals that only the empirical and EFT models exhibit a single-parameter dependence on k, while the others do not. Additionally, the EFT model shows a significant reduction in the final reduced ratio function due to the inclusion of the non-negligible $P_{1−loop/P11}$ term. This effect is also observed in Amons model. Notably, the halo models demonstrate a unique parameter dependency; for instance, Meads model shows that the reduced ratio function depends solely on the mass ratio.