The significant effects of stellar mass estimation on galaxy pair fractions.

ABSTRACT Theoretical models suggest that galaxy mergers may have profoundly shaped galaxy morphologies through cosmic time. However, observational estimates of galaxy pair fractions, and implied galaxy mergers, still do not converge, suggesting different numbers and redshift evolution. This paper explores the connection between stellar mass estimation and the pair fraction of galaxies using $\rm {\small STEEL}$, the Statistical sEmi-Emprical modeL, in which galaxies are mapped on to host dark matter haloes via a stellar mass–halo mass (SMHM) relation. As haloes of different mass are predicted to undergo, on average, different degrees of mergers, we expect that galaxy merger rates at fixed stellar mass will also vary depending on the shape of the input SMHM relation. Using a variety of input SMHM relations, also including one that matches the outputs of the Illustris TNG simulation, we thoroughly explore how systematic variations in the stellar mass functions propagate on to the normalization and redshift dependence of the galaxy pair fraction. We find that (i) stellar mass functions characterized by larger number densities of massive galaxies and thus a steeper SMHM relation, result in a substantial reduction of the pair fractions of massive galaxies; (ii) a time-varying SMHM relation produces galaxy pair fractions that can have varied redshift evolutions. We argue that in a hierarchical, dark matter dominated Universe, stellar mass estimates are a considerable cause of bias that must be accounted for when comparing pair fractions from different data sets or from theoretical models developed independently of the fitted survey data.