Push it to the limit: Local Group constraints on high-redshift stellar mass functions for Mstar > 10^5 Msun

TL;DR

This study extends the understanding of galaxy stellar mass functions at high redshift down to low masses by combining star formation histories with simulations, revealing a weak evolution of the faint end slope from z=5 to 2.

Contribution

It introduces a novel method combining deep HST observations and N-body simulations to constrain low-mass galaxy populations at high redshift, surpassing current direct imaging limits.

Findings

Faint end slope at z=2: -1.42

Faint end slope at z=5: -1.57

Weak evolution of slope from z=5 to 2

Abstract

We constrain the evolution of the galaxy stellar mass function from 2 < z < 5 for galaxies with stellar masses as low as 10^5 Msun by combining star formation histories of Milky Way satellite galaxies derived from deep Hubble Space Telescope observations with merger trees from the ELVIS suite of N-body simulations. This approach extends our understanding more than two orders of magnitude lower in stellar mass than is currently possible by direct imaging. We find the faint end slopes of the mass functions to be alpha= -1.42(+0.07/-0.05) at z = 2 and alpha = -1.57^(+0.06/-0.06) at z = 5, and show the slope only weakly evolves from z = 5 to z = 0. Our findings are in stark contrast to a number of direct detection studies that suggest slopes as steep as alpha = -1.9 at these epochs. Such a steep slope would result in an order of magnitude too many luminous Milky Way satellites in a mass regime that is observationally complete (Mstar > 2*10^5 Msun at z = 0). The most recent studies from ZFOURGE and CANDELS also suggest flatter faint end slopes that are consistent with our results, but with a lower degree of precision. This work illustrates the strong connections between low and high-z observations when viewed through the lens of LCDM numerical simulations.