We study the observational consequences of several unknown properties of Population III (Pop III) stars using large-scale cosmological simulations that include a subgrid model to track the unresolved mixing of pollutants. Varying the value of the critical metallicity that marks the boundary between Pop III and Population II (Pop II) star formation across 2 dex has a negligible effect on the fraction of Pop III stars formed and the subsequent fraction of Pop III flux from high-redshift galaxies. However, adopting a log normal initial mass function (IMF) for Pop III stars, in place of a baseline Salpeter IMF, results in a Pop III star formation rate density (SFRD) that is 1/4 of the baseline rate. The flux from high-redshift galaxies modeled with this IMF is highly bimodal, resulting in a tiny fraction of $z leq 8$ galaxies with more than 75% of their flux coming from Pop III stars. However, at $z=9$, right before reionization in our simulations, $approx$ 20% of galaxies are Pop III-bright with $m_{rm UV} le 31.4$ mag and at least 75% of their flux generated by Pop III stars . Additionally, the log normal Pop III IMF results in a population of carbon enhanced, metal poor stars in reasonable agreement with MW halo observations. Our analysis supports the conclusion that the Pop III IMF was dominated by stars in the 20-120$M_{odot}$ range that generate SN with carbon-enhanced ejecta.
Download