paper

How does radiative feedback from a UV background impact reionization?

arXiv:1301.6781 · doi:10.1093/mnras/stt693

Abstract

An ionizing UV background (UVB) inhibits gas accretion and photo-evaporates gas from the shallow potential wells of small, dwarf galaxies. During cosmological reionization, this effect can result in negative feedback: suppressing star-formation inside HII regions, thus impeding their continued growth. It is difficult to model this process, given the enormous range of scales involved. We tackle this problem using a tiered approach: combining parameterized results from single-halo collapse simulations with large-scale models of reionization. In the resulting reionization models, the ionizing emissivity of galaxies depends on the local values of the reionization redshift and the UVB intensity. We present a physically-motivated analytic expression for the average minimum mass of star-forming galaxies, which can be readily used in modeling galaxy formation. We find that UVB feedback: (i) delays the end stages of reionization by less than 0.5 in redshift; (ii) results in a more uniform distribution of HII regions, peaked on smaller-scales (with large-scale ionization power suppressed by tens of percent); and (iii) suppresses the global photoionization rate per baryon by a factor of < 2 towards the end of reionization. However, the impact is modest, since the hydrodynamic response of the gas to the UVB occurs on a time-scale comparable to reionization. In particular, the popular approach of modeling UVB feedback with an instantaneous transition in the minimum mass of star-forming galaxies, dramatically overestimates its importance. UVB feedback does not significantly affect reionization unless: (i) molecularly-cooled galaxies contribute significantly to reionization; or (ii) internal feedback processes strongly couple with UVB feedback in the early Universe. Since both are considered unlikely, we conclude that there is no significant self-regulation of reionization by UVB feedback.

9 pages, 9 figures

References in corpus (15)

Cited by in corpus (86)