A Burst of Reionization in the Dark Ages

Abstract

Primordial overdensities at small scales can be significantly larger than those observed at large scales. As a result, they can collapse much earlier, at redshifts $z \sim 100$. In these collapsed halos, stars and black holes can form, where they can then contribute to an early reionization of the universe. If this epoch is temporary, the free electrons will recombine yielding a low level of ionization leading into cosmic dawn. This contributes to the optical depth of the universe, and adds more weakly to low-$\ell$ polarization signals of ionization compared to later sources. We study the effects of a redshift-localized source of reionization $Δτ_{burst}$ in the context of a broad suite of cosmological data to constrain this scenario, finding a preference of approximately $\sim 2-2.5σ$ for a new contribution. Regardless of Planck reprocessing used (PR3 - SRoll2+plik or PR4 - LoLLiPoP+CamSpec/HiLLiPoP), the results are consistent, favoring a new contribution $Δτ_{burst} \sim 0.012$ ($0.001 < Δτ_{burst} < 0.022$ at 95\%) and a similar total optical depth $τ\sim 0.073$. The precise era of reionization is somewhat discrepant depending on the higher-$\ell$ reprocessing used, with HiLLiPoP favoring $z\sim 45$, CamSpec $z\sim 100$ and plik $z\sim 165$. If such a source is present at $40\lesssim z\lesssim150$, LiteBIRD should be able to detect it, while Simons Observatory should also constrain it.