Counts of galaxy clusters as cosmological probes: the impact of baryonic physics

TL;DR

This paper demonstrates that baryonic physics significantly affects galaxy cluster counts used for cosmology, emphasizing the need to incorporate baryonic corrections for unbiased parameter estimation in upcoming surveys.

Contribution

It introduces a formalism to account for baryonic mass fraction variability in the cluster mass function, improving cosmological parameter inference from galaxy cluster data.

Findings

Baryonic mass fraction varies with cluster mass and impacts the mass function.

Inclusion of baryonic corrections alters the normalization and shape of the mass function.

Combining cluster counts with baryonic fraction measurements enhances cosmological constraints.

Abstract

The halo mass function from N-body simulations of collisionless matter is generally used to retrieve cosmological parameters from observed counts of galaxy clusters. This neglects the observational fact that the baryonic mass fraction in clusters is a random variable that, on average, increases with the total mass (within an overdensity of 500). Considering a mock catalog that includes tens of thousands of galaxy clusters, as expected from the forthcoming generation of surveys, we show that the effect of a varying baryonic mass fraction will be observable with high statistical significance. The net effect is a change in the overall normalization of the cluster mass function and a milder modification of its shape. Our results indicate the necessity of taking into account baryonic corrections to the mass function if one wants to obtain unbiased estimates of the cosmological parameters from data of this quality. We introduce the formalism necessary to accomplish this goal. Our discussion is based on the conditional probability of finding a given value of the baryonic mass fraction for clusters of fixed total mass. Finally, we show that combining information from the cluster counts with measurements of the baryonic mass fraction in a small subsample of clusters (including only a few tens of objects) will nearly optimally constrain the cosmological parameters.