A New View of the Size-Mass Distribution of Galaxies: Using $r_{20}$ and $r_{80}$ instead of $r_{50}$

Abstract

When investigating the sizes of galaxies it is standard practice to use the half-light radius, $r_{50}$. Here we explore the effects of the size definition on the distribution of galaxies in the size -- stellar mass plane. Specifically, we consider $r_{20}$ and $r_{80}$, the radii that contain 20% and 80% of a galaxy's total luminosity, as determined from a Sersic profile fit, for galaxies in the 3D-HST/CANDELS and COSMOS-DASH surveys. These radii are calculated from size catalogs based on a simple calculation assuming a Sersic profile. We find that the size-mass distributions for $r_{20}$ and $r_{80}$ are markedly different from each other and also from the canonical $r_{50}$ distribution. The most striking difference is in the relative sizes of star forming and quiescent galaxies at fixed stellar mass. Whereas quiescent galaxies are smaller than star forming galaxies in $r_{50}$, this difference nearly vanishes for $r_{80}$. By contrast, the distance between the two populations increases for $r_{20}$. Considering all galaxies in a given stellar mass and redshift bin we detect a significant bimodality in the distribution of $r_{20}$, with one peak corresponding to star forming galaxies and the other to quiescent galaxies. We suggest that different measures of the size are tracing different physical processes within galaxies; $r_{20}$ is closely related to processes controlling the star formation rate of galaxies and $r_{80}$ may be sensitive to accretion processes and the relation of galaxies with their halos.