Exploring the Intrinsic Scatter of the Star-Forming Galaxy Main Sequence at redshift 0.5 to 3.0

TL;DR

This study measures the true intrinsic scatter of the galaxy main sequence across redshifts 0.5 to 3.0, revealing its dependence on stellar mass and cosmic time, and highlighting the importance of comprehensive spectral energy distribution fitting.

Contribution

It provides the first self-consistent measurement of the intrinsic scatter of the galaxy main sequence accounting for photometric redshift uncertainties and full spectral data.

Findings

Intrinsic MS scatter is 1.4 to 2.6 times larger than measurement errors.

Average intrinsic MS scatter decreases slightly from z=0.5 to 2.0.

MS scatter varies with stellar mass, showing a minimum around 10^10.25 solar masses.

Abstract

Previous studies have shown that the normalization and scatter of the galaxy 'main sequence' (MS), the relation between star formation rate (SFR) and stellar mass ($M_*$), evolves over cosmic time. However, such studies often rely on photometric redshifts and/or only rest-frame UV to near-IR data, which may underestimate the SFR and $M_*$ uncertainties. We use MAGPHYS+photo-z to fit the UV to radio spectral energy distributions of 12,380 galaxies in the COSMOS field at $0.5<z<3.0$ and self-consistently include photometric redshift uncertainties on the derived SFR and $M_*$. We quantify the effect on the observed MS scatter from (1) photometric redshift uncertainties (which are minor) and (2) fitting only rest-frame ultraviolet to near-infrared observations (which are severe). At fixed redshift and $M_*$, we find that the intrinsic MS scatter for our sample of galaxies is 1.4 to 2.6 times larger than the measurement uncertainty. The average intrinsic MS scatter has decreased by 0.1 dex from $z=0.5$ to $\sim2.0$. At low-$z$, the trend between the intrinsic MS scatter and $M_*$ follows a functional form similar to an inverse stellar mass-halo mass relation (SMHM; $M_*$/$M_{\rm halo}$ vs $M_*$), with a minimum in intrinsic MS scatter at log($M_*/M_{\odot})\sim10.25$ and larger scatter at both lower and higher $M_*$; while this distribution becomes flatter for high-$z$. The SMHM is thought to be a consequence of feedback effects and this similarity may suggest a link between galaxy feedback and the intrinsic MS scatter. These results favor a slight evolution in the intrinsic MS scatter with both redshift and mass.