The Clustering of Galaxies on the Smallest Scales Across Cosmic Time

Abstract

This work aims to quantify the changing clustering of satellite galaxies in their host dark matter halo. I use three samples of Sloan Digital Sky Survey (SDSS) galaxies; the first sample is of galaxies in the local universe, redshifts from z = 0.02 to z = 0.165. I then compare the clustering of two higher redshift samples, the Luminous Red Galaxy (LRG) sample, z = 0.16 to z = 0.36, and the SDSS-III Constant Mass Galaxy (CMASS) sample, z = 0.43 to z = 0.7. First, I measure the angular correlation function of local SDSS galaxies on very small scales 7†< θ < 320†in a range of luminosity threshold samples (Mr <-18 up to Mr <-21) that are constructed from the subset of SDSS that has been spectroscopically observed more than once (the plate overlap region). I measure clustering in this reduced survey footprint in order to minimize the effects of fiber collision incompleteness, which are otherwise substantial on these small scales. I model our clustering measurements using a fully numerical halo model that populates dark matter halos in N-body simulations to create realistic mock galaxy catalogs. The model has free parameters that specify both the number and spatial distribution of galaxies within their host halos. I adopt a flexible density profile for the spatial distribution of satellite galaxies that is similar to the dark matter Navarro-Frenk-White (NFW) profile,  except that the inner slope is allowed to vary. I find that the angular clustering of our most luminous samples suggests that luminous satellite galaxies have substantially steeper inner density profiles than NFW. Lower luminosity samples are less constraining, however, and are consistent with satellite galaxies having shallow density profiles. Our results confirm the findings of Watson et al. (2012) while using different methodology. Moving onto the higher redshift regime, I then measure the very small scale clustering of LRGs and their progenitor CMASS galaxies. I use a cross correlation technique to overcome the limitations due to fiber collisions. I see significant evolution in the small scale clustering between z = 0.7 and z = 0.43, and even more striking, a significant difference between the clustering of LRGs and the CMASS galaxies. The clustering of LRGs is substantially steeper, indicative of the satellite galaxy distribution becoming more compact over the course of cosmic time.