Star formation rates in Lyman break galaxies: radio stacking of LBGs in the COSMOS field and the sub-$\mu$Jy radio source population

TL;DR

This study uses radio stacking to analyze star formation in high-redshift Lyman Break Galaxies, revealing lower dust attenuation factors and potential suppression of radio luminosity at z~3, with implications for galaxy evolution models.

Contribution

It provides the first statistical radio detection of z~3 LBGs via stacking, and discusses the implications of lower-than-expected dust attenuation and radio luminosity suppression at high redshift.

Findings

Median stacked radio flux density of 0.90 μJy for z~3 LBGs.

Derived star formation rate of 31±7 M⊙/yr from radio data.

Dust attenuation factor at z~3 is approximately 1.8, lower than standard assumptions.

Abstract

We present an analysis of the radio properties of large samples of Lyman Break Galaxies (LBGs) at $z \sim 3$, 4, and 5 from the COSMOS field. The median stacking analysis yields a statistical detection of the $z \sim 3$ LBGs (U-band drop-outs), with a 1.4 GHz flux density of $0.90 \pm 0.21 \mu$Jy. The stacked emission is unresolved, with a size $< 1"$, or a physical size $< 8$kpc. The total star formation rate implied by this radio luminosity is $31\pm 7$ $M_\odot$ year$^{-1}$, based on the radio-FIR correlation in low redshift star forming galaxies. The star formation rate derived from a similar analysis of the UV luminosities is 17 $M_\odot$ year$^{-1}$, without any correction for UV dust attenuation. The simplest conclusion is that the dust attenuation factor is 1.8 at UV wavelengths. However, this factor is considerably smaller than the standard attenuation factor $\sim 5$, normally assumed for LBGs. We discuss potential reasons for this discrepancy, including the possibility that the dust attenuation factor at $z \ge 3$ is smaller than at lower redshifts. Conversely, the radio luminosity for a given star formation rate may be systematically lower at very high redshift. Two possible causes for a suppressed radio luminosity are: (i) increased inverse Compton cooling of the relativistic electron population due to scattering off the increasing CMB at high redshift, or (ii) cosmic ray diffusion from systematically smaller galaxies. The radio detections of individual sources are consistent with a radio-loud AGN fraction of 0.3%. One source is identified as a very dusty, extreme starburst galaxy (a 'submm galaxy').