AGN-driven helium reionization and the incidence of extended HeIII regions at redshift z>3

TL;DR

This study uses hydrodynamic simulations with radiative transfer to investigate helium reionization driven by AGNs at redshifts greater than 3, showing it is extended, inhomogeneous, and consistent with some recent observations, but not conclusively ruling out AGN dominance.

Contribution

The paper presents a detailed simulation-based analysis of AGN-driven helium reionization, incorporating early source populations and improved modeling to reconcile with observational data.

Findings

Reionization is very extended ({}za0a02) and highly inhomogeneous.

Mock spectra show large variability in HeII optical depth evolution.

Model predicts lower HeII optical depth at za0a03.4, reducing tension with observations.

Abstract

We use hydrodynamic simulations post-processed with the radiative-transfer code RADAMESH to assess recent claims that the low HeII opacity observed in z>3 quasar spectra may be incompatible with models of HeII reionization driven by the observed population of active galactic nuclei (AGNs). In particular, building upon our previous work, we consider an early population of sources and start the radiative-transfer calculation at redshifts z>=5. Our model faithfully reproduces the emissivity of optically selected AGNs as inferred from measurements of their luminosity function. We find that HeII reionization is very extended in redshift ({\Delta} z>=2) and highly spatially inhomogeneous. In fact, mock spectra extracted from the simulations show a large variability in the evolution of the HeII effective optical depth within chunks of size {\Delta} z=0.04. Regions with low opacity ({\tau}^{eff}_{HeII}<3) can be found at high redshift, in agreement with the most recent observations of UV-transmitting quasars. At the highest redshift currently probed by observations (z~3.4), our updated model predicts a much lower HeII effective optical depth than previous simulations in the literature relieving most of the tension with the current data, that, however, still persists at about the (Gaussian) 1{\sigma} to 2{\sigma} level. Given the very small number of observed lines of sight, our analysis indicates that current data cannot rule out a purely AGN-driven scenario with high statistical significance.