ETHOS -- an effective theory of structure formation: predictions for the high-redshift Universe -- abundance of galaxies and reionization

TL;DR

This study compares predictions of galaxy formation and reionization between ETHOS, an alternative dark matter model, and CDM, finding that current observations cannot distinguish them but future JWST surveys might.

Contribution

First hydrodynamical simulations of ETHOS model at high resolution, showing differences in galaxy luminosities and reionization impact compared to CDM.

Findings

ETHOS predicts brighter high-redshift galaxies than CDM.

Faster buildup of faint galaxies in ETHOS affects UV luminosity function.

Reionization optical depth differs by only 10% between ETHOS and CDM.

Abstract

We contrast predictions for the high-redshift galaxy population and reionization history between cold dark matter (CDM) and an alternative self-interacting dark matter model based on the recently developed ETHOS framework that alleviates the small-scale CDM challenges within the Local Group. We perform the highest resolution hydrodynamical cosmological simulations (a 36~Mpc$^3$ volume with gas cell mass of $\sim10^5\mathrm{M}_{\odot}$ and minimum gas softening of $\sim180$~pc) within ETHOS to date -- plus a CDM counterpart -- to quantify the abundance of galaxies at high redshift and their impact on reionization. We find that ETHOS predicts galaxies with higher ultraviolet (UV) luminosities than their CDM counterparts and a faster build-up of the faint end of the UV luminosity function. These effects, however, make the optical depth to reionization less sensitive to the power spectrum cut-off: the ETHOS model differs from the CDM $\tau$ value by only 10 per cent and is consistent with Planck limits if the effective escape fraction of UV photons is 0.1-0.5. We conclude that current observations of high-redshift luminosity functions cannot differentiate between ETHOS and CDM models, but deep JWST surveys of strongly-lensed, inherently faint galaxies have the potential to test non-CDM models that offer attractive solutions to CDM's Local Group problems.