Modeling LAEs in the epoch of reionization with OBELISK

The connection between Lyman-α spectra and Lyman-continuum escape

¹ Department of Astronomy, University of Geneva, Chemin Pegasi 51, 1290 Versoix, Switzerland
e-mail: emma.giovinazzo@unige.ch
² Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands
³ Cosmic Dawn Center (DAWN), Niels Bohr Institute, University of Copenhagen, Jagtvej 128, København N 2200, Denmark

Received: 28 November 2023
Accepted: 29 April 2024

Abstract

Context. Lyman-α emitters (LAEs) are particularly useful objects in the study of the epoch of reionization. Lyman-α profiles can be used to estimate the number of ionizing photons that are able to escape galaxies, and therefore to understand which objects contributed to reionization. However, Lyman-α is a resonant line and its complex radiative transfer effects make the interpretation of the line challenging and require the use of appropriate radiative transfer methods for anything but the simplest gas distributions, such as uniform gaseous spheres, slabs, or cubes.

Aims. With this work, we aim to study the properties of simulated LAEs, and the robustness of these inferred properties during a change in the dust model. We also explore the Lyman continuum (LyC) escape fraction of these galaxies and compare our results with observationally calibrated methods to infer this quantity from the Lyman-α spectrum.

Methods. We used the radiative transfer code RASCAS to perform synthetic observations of 13 flux-selected galaxies from the OBELISK simulation at a redshift of z = 6, toward the end of the epoch of reionization. Each galaxy was observed in Lyman-α, as well as ionizing and nonionizing continuum from 48 different viewing angles.

Results. We show that the Lyman-α profiles emitted from a galaxy present large variations with a change in viewing angle and that the relation between peak separation and the Lyman-α escape fraction is not as strong as previously found, as we find lines of sight with both a low peak separation and a low escape fraction, due to their dust content. We also show that the properties of the Lyman-α line are reasonably robust during a change in the dust model. Lastly, we compare the LyC escape fractions that we derive from the simulation to three observationally calibrated methods of inferring this quantity. We determine that none of these relations reproduce the scatter that we find in our sample, and that high escape fraction lines of sight have both a low peak separation and a low dust extinction in the ultraviolet (UV).