The Constant Average Relationship between Dust-obscured Star Formation and Stellar Mass from $z=0$ to $z=2.5$

TL;DR

This study reveals a consistent relationship between dust-obscured star formation and stellar mass from redshift 0 to 2.5, showing that more massive galaxies are predominantly obscured, with little evolution over cosmic time.

Contribution

It demonstrates a constant average relationship between dust-obscured star formation and stellar mass across a wide redshift range, challenging existing theoretical models.

Findings

Obscured star formation fraction strongly depends on stellar mass.

No significant evolution of obscured fraction with redshift up to 2.5.

High-mass galaxies (>10.5 log M/M_sun) are >90% obscured at all redshifts.

Abstract

The total star formation budget of galaxies consists of the sum of the unobscured star formation, as observed in the rest-frame ultraviolet (UV), together with the obscured component that is absorbed and re-radiated by dust grains in the infrared. We explore how the fraction of obscured star formation depends on stellar mass for mass-complete samples of galaxies at $0<z<2.5$. We combine GALEX and WISE photometry for SDSS-selected galaxies with the 3D-HST treasury program and Spitzer/MIPS 24$\mu$m photometry in the well-studied 5 extragalactic CANDELS fields. We find a strong dependence of the fraction of obscured star formation ($f_{\mathrm{obscured}}$=SFR$_{\mathrm{IR}}$/SFR$_{\mathrm{UV+IR}}$) on stellar mass, with remarkably little evolution in this fraction with redshift out to $z$=2.5. 50\% of star formation is obscured for galaxies with log(M/M$_{\odot}$)=9.4; although unobscured star formation dominates the budget at lower masses, there exists a tail of low mass extremely obscured star-forming galaxies at $z>1$. For log(M/M$_{\odot}$)$>$10.5, $>$90\% of star formation is obscured at all redshifts. We also show that at fixed total SFR, $f_{\mathrm{obscured}}$ is lower at higher redshift. At fixed mass, high-redshift galaxies are observed to have more compact sizes and much higher star formation rates, gas fractions and hence surface densities (implying higher dust obscuration), yet we observe no redshift evolution in $f_{\mathrm{obscured}}$ with stellar mass. This poses a challenge to theoretical models to reproduce, where the observed compact sizes at high redshift seem in tension with lower dust obscuration.