Constraints on photoionization feedback from number counts of ultra-faint high-redshift galaxies in the Frontier Fields

TL;DR

This study uses ultra-faint high-redshift galaxy counts from the Frontier Fields to constrain the impact of photoionization feedback on star formation, finding a cutoff halo mass below which star formation is suppressed.

Contribution

First observational constraint on the halo mass threshold for star formation suppression due to photoionization feedback at high redshift.

Findings

Photoionization feedback suppresses star formation in halos below ~5.6×10^9 M_sun at z=5.

The cutoff circular velocity for star formation suppression is less than 50 km/s.

Uncertainties in lens modeling limit the precision of the constraints.

Abstract

We exploit a sample of ultra-faint high-redshift galaxies (demagnified HST $H_{160}$ magnitude $>30$) in the Frontier Fields clusters A2744 and M0416 to constrain a theoretical model for the UV luminosity function (LF) in the presence of photoionization feedback. The objects have been selected on the basis of accurate photometric redshifts computed from multi-band photometry including 7 HST bands and deep $K_s$ and IRAC observations. Magnification is computed on an object-by-object basis from all available lensing models of the two clusters. We take into account source detection completeness as a function of luminosity and size, magnification effects and systematics in the lens modeling of the clusters under investigation. We find that our sample of high-$z$ galaxies constrain the cut-off halo circular velocity below which star-formation is suppressed by photo-ionization feedback to $v_c^{\rm cut} < 50$ km s$^{-1}$. This circular velocity corresponds to a halo mass of $\approx5.6\times10^9~M_\odot$ and $\approx2.3\times10^9~M_\odot$ at $z=5$ and 10 respectively: higher mass halos can thus sustain continuous star formation activity without being quenched by external ionizing flux. More stringent constraints are prevented by the uncertainty in the modeling of the cluster lens, as embodied by systematic differences among the lens models available.