Dark-ages reionization & galaxy formation simulation V: morphology and statistical signatures of reionization

TL;DR

This paper uses the DRAGONS simulation framework to study how galaxy formation physics, especially supernova feedback, influences the morphology and statistical signatures of reionization during the Epoch of Reionization, highlighting the importance of realistic galaxy models.

Contribution

It demonstrates that galaxy formation physics significantly affects reionization morphology and statistics, and shows that simplified models can mimic detailed predictions with proper parameter choices.

Findings

Supernova feedback reduces HII region sizes by up to 20%.

Power spectra amplitude varies by up to 17% without feedback.

Reionization morphology depends on halo mass, not just galaxy formation details.

Abstract

We use the Dark-ages, Reionization And Galaxy-formation Observables from Numerical Simulations (DRAGONS) framework to investigate the effect of galaxy-formation physics on the morphology and statistics of ionized hydrogen (HII) regions during the Epoch of Reioinization (EoR). DRAGONS self-consistently couples a semi-analytic galaxy-formation model with the inhomogeneous ionizing UV background, and can therefore be used to study the dependence of morphology and statistics of reionization on feedback phenomena of the ionizing source galaxy population. Changes in galaxy-formation physics modify the sizes of HII regions and the amplitude and shape of 21-cm power spectra. Of the galaxy physics investigated, we find that supernova feedback plays the most important role in reionization, with HII regions up to $\approx 20$ per cent smaller and a fractional difference in the amplitude of power spectra of up to $\approx 17$ per cent at fixed ionized fraction in the absence of this feedback. We compare our galaxy-formation-based reionization models with past calculations that assume constant stellar-to-halo mass ratios and find that with the correct choice of minimum halo mass, such models can mimic the predicted reionization morphology. Reionization morphology at fixed neutral fraction is therefore not uniquely determined by the details of galaxy formation, but is sensitive to the mass of the haloes hosting the bulk of the ionizing sources. Simple EoR parametrizations are therefore accurate predictors of reionization statistics. However, a complete understanding of reionization using future 21-cm observations will require interpretation with realistic galaxy-formation models, in combination with other observations.