An ultra-deep multi-band VLA survey of the faint radio sky (COSMOS-XS): New constraints on the cosmic star formation history

TL;DR

This study uses ultra-deep radio observations from the COSMOS field to refine the understanding of the evolution of star-forming galaxies and the cosmic star formation rate density, revealing discrepancies between radio and UV estimates at high redshift.

Contribution

It provides new constraints on the evolution of radio luminosity functions and the cosmic star formation history using the deepest radio data available for the COSMOS field.

Findings

Radio luminosity functions show strong redshift evolution with pure luminosity evolution.

Evidence for significant density evolution when combining deep and wide surveys.

Discrepancy of about 1 dex in star formation rate density estimates at z>3 between radio and UV methods.

Abstract

We make use of ultra-deep 3 GHz Karl G. Jansky Very Large Array observations of the COSMOS field from the multi-band COSMOS-XS survey to infer radio luminosity functions (LFs) of star-forming galaxies (SFGs). Using $\sim$1300 SFGs with redshifts out to $z\sim4.6$, and fixing the faint and bright end shape of the radio LF to the local values, we find a strong redshift trend that can be fitted by pure luminosity evolution with the luminosity parameter given by $\alpha_L \propto (3.40 \pm 0.11) - (0.48 \pm 0.06)z$. We then combine the ultra-deep COSMOS-XS data-set with the shallower VLA-COSMOS $\mathrm{3\,GHz}$ large project data-set over the wider COSMOS field in order to fit for joint density+luminosity evolution, finding evidence for significant density evolution. By comparing the radio LFs to the observed far-infrared (FIR) and ultraviolet (UV) LFs, we find evidence of a significant underestimation of the UV LF by $21.6\%\, \pm \, 14.3 \, \%$ at high redshift ($3.3\,<\,z\,<\,4.6$, integrated down to $0.03\,L^{\star}_{z=3}$). We derive the cosmic star formation rate density (SFRD) by integrating the fitted radio LFs and find that the SFRD rises up to $z\,\sim\,1.8$ and then declines more rapidly than previous radio-based estimates. A direct comparison between the radio SFRD and a recent UV-based SFRD, where we integrate both LFs down to a consistent limit ($0.038\,L^{\star}_{z=3}$), reveals that the discrepancy between the radio and UV LFs translates to a significant ($\sim$1 dex) discrepancy in the derived SFRD at $z>3$, even assuming the latest dust corrections and without accounting for optically dark sources.