The Effects of Lyman-Limit Systems on the Evolution and Observability of the Epoch of Reionization

TL;DR

This study uses large-scale radiative transfer simulations to investigate how Lyman-limit systems influence the timing, morphology, and observability of the epoch of reionization, revealing their significant impact on the growth of ionized regions and 21-cm signals.

Contribution

It introduces the first large-scale, full radiative transfer simulations of reionization that incorporate Lyman-limit systems with variable opacity, providing new insights into their effects.

Findings

LLSs slow down ionization fronts and alter H II region sizes.

Presence of LLSs extends reionization by approximately 0.8 in redshift.

LSSs reduce large-scale 21-cm power by a factor of 2 at certain ionization stages.

Abstract

We present the first large-scale, full radiative transfer simulations of the reionization of the intergalactic medium in the presence of Lyman-limit systems (LLSs). To illustrate the impact of LLS opacity, possibly missed by previous simulations, we add either a uniform or spatially-varying hydrogen bound-free opacity. This opacity, implemented as the mean free path (mfp) of the ionizing photons, extrapolates the observed, post-reionization redshift dependence into the epoch of reionization. In qualitative agreement with previous studies, we find that at late times the presence of LLSs slows down the ionization fronts, and alters the size distribution of H II regions. We quantitatively characterize the size distribution and morphological evolution of H II regions and examine the effects of the LLSs on the redshifted 21-cm signal from the patchy reionization. The presence of LLSs extends the ionization history by $\Delta z \sim 0.8$. The LLS absorbers significantly impede the late-time growth of the H II regions. The position dependent LLS distribution slows reionization further and additionally limits the late growth of the ionized regions. However, there is no "freeze out" of the H II regions and the largest regions grow to the size of the simulation volume. The 21-cm power spectra show that at large scales the power drops by a factor of 2 for 50% and 75% ionization stages (at $k = 0.1$ $\text{h} \, \text{Mpc}^{-1} $) reflecting the limiting effect of the LLSs on the growth of ionized patches. The statistical observables such as the RMS of the brightness temperature fluctuations and the peak amplitudes of the 21-cm power spectra at large-scales ($k = 0.05 - 0.1$ $\text{h} \, \text{Mpc}^{-1} $) are diminished by the presence of LLS.