Tracing Galaxies Through Cosmic Time with Number Density Selection
A central challenge in observational studies of galaxy formation is how to associate progenitor galaxies with their descendants at lower redshifts. One promising approach is to link galaxies at fixed number density, rather than fixed luminosity or mass. This method is effective if stellar mass rank order is broadly conserved through cosmic time. In this paper, we use the Guo et al. (2011) semi-analytical model to analyze under what circumstances this assumption is valid in the context of a cosmological simulation. Specifically, we select progenitor galaxies at a constant number density and compare the stellar mass evolution of their descendants to the evolution at a constant number density. The median stellar mass of the descendants increases by a factor of four (0.6 dex) from z = 3 to z = 0. Constant number density selection reproduces this to within 40% (0.15 dex) over a wide range of number densities. We show that the discrepancy primarily results from scatter in the stellar mass growth rates and merging. After applying simple, observationally-based corrections for these processes, the discrepancy is reduced to 12% (0.05 dex). We conclude that number density selection can be used to predict the median descendant mass of high-redshift progenitor galaxies. The main uncertainty in this study is that semi-analytical models do not reproduce the observed mass evolution of galaxies, which makes the quantitative aggregate effects of star formation, merging, and quenching on the rank order of galaxies somewhat uncertain.