Anomaly in the Opacity of the Post-reionization Intergalactic Medium in the Lyα and Lyβ Forest

Abstract

We measure the intergalactic medium (IGM) opacity in the Lyα as well as in the Lyβ forest along 19 quasar sightlines between 5.5 z abs 6.1, probing the end stages of the reionization epoch. Owing to its lower oscillator strength the Lyβ transition is sensitive to different gas temperatures and densities than Lyα, providing additional constraints on the ionization and thermal state of the IGM. A comparison of our measurements to different inhomogeneous reionization models, derived from post-processing the Nyx cosmological hydrodynamical simulation to include spatial fluctuations in the ultraviolet background (UVB) or the gas temperature field, as well as to a uniform reionization model with varying thermal states of the IGM, leads to two primary conclusions: First, we find that including the effects of spectral noise is key for a proper data to model comparison. Noise effectively reduces the sensitivity to high opacity regions, and thus even stronger spatial inhomogeneities are required to match the observed scatter in the observations than previously inferred. Second, we find that models which come close to reproducing the distribution of Lyα effective optical depths nevertheless underpredict the Lyβ opacity at the same spatial locations. The origin of this disagreement is not entirely clear but models with an inversion in the temperature-density relation of the IGM just after reionization is completed match our measurements best, although they still do not fully capture the observations at z 5.8.

Observing the tail of reionization: neutral islands in the z = 5.5 Lyman-α forest

“…We rescale the foreground hydro skewers to match the mean fluxes reported by Iršič et al (2017). Fig 10 shows a comparison of τ eff cumulative distributions for Lyα (top row) and Lyβ (bottom row) with the measurements of Eilers et al (2019). 9 All of our models provide reasonable matches to the measurements, with the exception of the 5.7 < z < 5.9 bin, where the models do not accommodate the highest Lyβ opacities observed.…”

“…The Lyβ opacities are a useful window into the fluctuations because they probe different temperatures and densities. Recently, Eilers et al (2019) measured this distribution using a sample of 19 quasar sightlines between 5.5 z 6.1. Particularly at z 5.8, they found that fluctuating UVB and temperature models, akin to those of Davies &Furlanetto (2016) andD'Aloisio et al (2015), under-predict the Lyβ opacities when they are tuned to match the observed distribution of Lyα opacities.…”

“…In this section we follow the convention of Eilers et al (2019) in which we measure τ eff along segments of length 27h −1 Mpc, rather than 50h −1 Mpc. The UVB in our models are iteratively rescaled in the ionized regions to match the Lyα mean fluxes of Eilers et al (2019), excluding nondetections, and for Lyβ we include foreground Lyα absorption.…”

“…In this section we follow the convention of Eilers et al (2019) in which we measure τ eff along segments of length 27h −1 Mpc, rather than 50h −1 Mpc. The UVB in our models are iteratively rescaled in the ionized regions to match the Lyα mean fluxes of Eilers et al (2019), excluding nondetections, and for Lyβ we include foreground Lyα absorption. To each Lyβ skewer we add opacities from a randomly chosen skewer from our hydro simulation at z ≈ (1 + z Lyα )λ Lyβ /λ Lyα − 1, where λ Lyβ and λ Lyα are the restframe Lyα and Lyβ wavelengths, respectively, and z Lyα is the redshift of interest in the Lyα forest.…”

“…We have also examined the correlation between Lyα and 9 Eilers et al (2019) noted the importance of modelling noise when comparing models to observed distributions, particularly for the high-opacity tails of the distributions. However, just as in Fig.…”