Early Growth of the Star Formation Rate Function in the Epoch of Reionization: an Approach with Rest-frame Optical Emissions

TL;DR

This study derives the star formation rate function at redshift 6 using spectral energy distribution fitting, revealing a higher number of dust-obscured star-forming galaxies than previously estimated, and suggests rapid growth in high-SFR galaxies during reionization.

Contribution

It introduces a new approach using rest-frame optical data to estimate the SFRF at z~6, highlighting the underestimation of dust-obscured galaxies in prior UV-based methods.

Findings

The SFRF at z~6 shows an excess compared to UV-based estimates.

The characteristic SFR at z~6 is similar to that at z~2.

The high-SFR end of the SFRF grows rapidly during reionization.

Abstract

We present a star formation rate function (SFRF) at $z\sim6$ based on star formation rates (SFRs) derived by spectral energy distribution (SED) fitting on data from rest-frame UV to optical wavelength of galaxies in the CANDELS GOODS-South and North fields. The resulting SFRF shows an excess compared to the previous estimations by using rest-frame UV luminosity functions (LFs) corrected for the dust attenuation, and is comparable to that estimated from a far-infrared LF. This suggests that the number density of dust-obscured intensively star-forming galaxies at $z\sim6$ has been underestimated in the previous approach based only on rest-frame UV observations. We parameterize the SFRF with using the Schechter function and obtain the best-fit parameter of the characteristic SFR (${\rm SFR}^*$) when the faint-end slope and characteristic number density are fixed. The best-fit ${\rm SFR}^*$ at $z\sim6$ is comparable to that at $z\sim2$, when the cosmic star formation activity reaches its peak. Together with SFRF estimations with similar approach using rest-frame UV to optical data, the ${\rm SFR}^*$ is roughly constant from $z\sim2$ to $z\sim6$ and may decrease above $z\sim6$. Since the ${\rm SFR}^*$ is sensitive to the high-SFR end of the SFRF, this evolution of ${\rm SFR}^*$ suggests that the high-SFR end of the SFRF grows rapidly during the epoch of reionization and reaches a similar level observed at $z\sim2$.