The Inhomogeneous Ionizing Background Following Reionization

TL;DR

This paper investigates the spatial fluctuations of the hydrogen ionizing background after reionization, showing that it is highly inhomogeneous due to source clustering and mean free path effects, but Lya forest observations remain largely insensitive to these fluctuations.

Contribution

The study provides a comprehensive analysis of ionizing background fluctuations using multiple modeling approaches, demonstrating their impact on Lya forest measurements and justifying the uniform background assumption.

Findings

Ionizing background exhibits significant fluctuations during post-reionization epoch.

Power spectrum of ionizing flux is highly sensitive to source properties.

Lya forest spectra are insensitive to UVB fluctuations despite background inhomogeneity.

Abstract

We study the spatial fluctuations in the hydrogen ionizing background in the epoch following reionization (z ~ 5--6). The rapid decrease with redshift in the photon mean free path (m.f.p.), combined with the clustering of increasingly rare ionizing sources, can result in a very inhomogenous ionizing background during this epoch. We systematically investigate the probability density functions (PDFs) and power spectra of ionizing flux, by varying several parameters such as the m.f.p., minimum halo mass capable of hosting stars, and halo duty cycle. In order to be versatile, we make use of analytic, semi-numeric and numeric approaches. Our models show that the ionizing background indeed has sizable fluctuations during this epoch sourced by the clustering of sources, with the PDFs being a factor of few wide at half of the maximum likelihood. The distributions also show marked asymmetries, with a high-value tail set by clustering on small scales, and a shorter low-value tail which is set by the mean free path. The power spectrum of the ionizing background is much more sensitive to source properties than the PDF and can be well-understood analytically with a framework similar to the halo model (usually used to describe dark matter clustering). Nevertheless, we find that Lya forest spectra are extremely insensitive to the details of the UVB, despite marked differences in the PDFs and power spectra of our various ionizing backgrounds. Assuming a uniform ionizing background only underestimates the value of the mean ionization rate inferred from the Lya forest by a few percent. Instead, analysis of the Lya forest is dominated by the uncertainties in the density field. Thus, our results justify the common assumption of a uniform ionizing background in Lya forest analysis even during this epoch.