Bursty star formation during the Cosmic Dawn driven by delayed stellar feedback

TL;DR

This paper models how delayed stellar feedback causes bursty star formation in small galaxies during the Cosmic Dawn, significantly affecting early galaxy evolution and reionization.

Contribution

It introduces a delay in feedback mechanisms in analytic models, revealing bursty star formation in small high-redshift galaxies, a novel insight into early galaxy behavior.

Findings

Delayed feedback induces star formation bursts in small galaxies.

Burstiness increases star formation efficiency by up to tenfold.

Large galaxies show damped burst cycles, maintaining continuous star formation.

Abstract

In recent years, several analytic models have demonstrated that simple assumptions about halo growth and feedback-regulated star formation can match the (limited) existing observational data on galaxies at z>6. By extending such models, we demonstrate that imposing a time delay on stellar feedback (as inevitably occurs in the case of supernova explosions) induces burstiness in small galaxies. Although supernova progenitors have short lifetimes (~5-30 Myr), the delay exceeds the dynamical time of galaxies at such high redshifts. As a result, star formation proceeds unimpeded by feedback for several cycles and "overshoots" the expectations of feedback-regulated star formation models. We show that such overshoot is expected even in atomic cooling halos, with masses up to ~10^10.5 Msun at z>6. However, these burst cycles damp out quickly in massive galaxies, because large haloes are more resistant to feedback so retain a continuous gas supply. Bursts in small galaxies - largely beyond the reach of existing observations - induce a scatter in the luminosity of these haloes (of ~1 mag) and increase the time-averaged star formation efficiency by up to an order of magnitude. This kind of burstiness can have substantial effects on the earliest phases of star formation and reionization.