A self-consistent analytical model for both the photoionization rate and reionization history

Abstract

Recent developments at the intersection of cosmology and astrophysics have highlighted the need for improved analytical models of observables that probe the Epoch of Reionization. With few exceptions, fast analytical treatments of reionization suitable for use in Bayesian inference have been limited to modeling the reionization history, $x_i(z)$. Such models cannot take full advantage of observables that constrain $x_i$ indirectly. One such observable is the photoionization rate of neutral hydrogen, $Γ_{\rm HI}(z)$, which can be inferred from the mean transmission of the Lyman-$α$ forest of high-redshift quasars and galaxies. It has been shown by several prior works that the evolution of $Γ_{\rm HI}$ at $5 \lesssim z \lesssim 6$ is highly sensitive to the tail end of reionization, potentially providing a tight astrophysical constraint on the reionization timeline. We present a new analytical formalism, based on the cosmological radiative transfer equation, that self-consistently predicts $x_i$ and $Γ_{\rm HI}$. We test our model against detailed radiative transfer simulations and find it to be percent-level accurate in $x_i$ and $20-30\%$ accurate in $Γ_{\rm HI}$ at $z \lesssim 6$ - better than or comparable to existing observational uncertainties. Finally, we demonstrate that modest shifts in the ionizing photon output of high-redshift galaxies and/or the endpoint of reionization lead to differences in $Γ_{\rm HI}$ much larger that the model's intrinsic uncertainty, highlighting its utility for interpreting existing data. We explore the origin of modeling uncertainty in $Γ_{\rm HI}$ and comment on future pathways for improvement.