Lyman-alpha emitters gone missing: the different evolution of the bright and faint populations

We model the transmission of the Lyman-alpha line through the circum- and intergalactic media around dark matter haloes expected to host Lyman-alpha emitters (LAEs) at z > 5.7, using the high-dynamic-range Sherwood simulations. We find very different CGM environments around more massive haloes (~10^11 M_sun) compared to less massive haloes (~10^9 M_sun) at these redshifts, which can contribute to a different evolution of the Lyman-alpha transmission from LAEs within these haloes. Additionally we confirm that part of the differential evolution could result from bright LAEs being more likely to reside in larger ionized regions. We conclude that a combination of the CGM environment and the IGM ionization structure is likely to be responsible for the differential evolution of the bright and faint ends of the LAE luminosity function at z > 6. More generally, we confirm the suggestion that the self-shielded neutral gas in the outskirts of the host halo can strongly attenuate the Lyman-alpha emission from high redshift galaxies. We find that this has a stronger effect on the more massive haloes hosting brighter LAEs. The faint-end of the LAE luminosity function is thus a more reliable probe of the average ionization state of the IGM. Comparing our model for LAEs with a range of observational data we find that the favoured reionization histories are our previously advocated Late' and Very Late’ reionization histories, in which reionization finishes rather rapidly at around z ~ 6.