Ly α with SPICE: interpreting Ly α emission at z > 5

Abstract

Ly emission is key to understanding the process of cosmic reionization. JWST is finally enabling us to measure Ly emission deep into the epoch of reionization for an increasing number of galaxies. However, discrepancies between measurements of Ly equivalent widths (EW) of Ly emitters (LAEs) have been noted between JWST and ground-based facilities. We employ SPICE, a suite of radiation-hydrodynamical simulations featuring different stellar feedback models, and investigate the impact of radiative transfer effects (e.g. extended emission, Ly-UV spatial offsets) and observational systematics (such as slit placement) on the measured Ly EW. We perform radiative transfer of Ly and UV photons for SPICE galaxies to mimic slit spectroscopy for ground-based slits and JWST-MSA pseudo-slits. We find that spatial Ly-UV offsets exist independently of feedback model, and are common (70$]]> per cent galaxies with \rm 108 M_\odot$]]>) with median values of. The theoretical predictions from SPICE are consistent with the observed spatial offset distribution. In addition, spatial Ly-UV offsets are identified as a major cause of loss of flux for JWST-MSA type observations, with median pseudo-slit losses of per cent, and per cent cases suffering from 95$]]> per cent pseudo-slit losses. Even in the absence of such spatial offsets, the presence of extended emission can cause median pseudo-slit losses of 40 per cent, with 4 per cent cases suffering from 95$]]> per cent pseudo-slit losses. Finally, complex galaxy morphologies or misplaced JWST-MSA pseudo-slit can lead to underestimated UV continuum, resulting in spuriously high estimates of EW from JWST in per cent of galaxies. We compare the predictions from SPICE to a sample of 25 galaxies with Ly emission observations from both the ground and from JWST. The EW and the EW exhibit scatter in line with predictions from SPICE, indicating that both physical and systematic effects are likely at play.