UC Santa Cruz

We assess a model of late cosmic reionization in which the ionizing background radiation arises entirely from highredshift quasars and other active galactic nuclei (AGNs). The low optical depth to Thomson scattering reported by the Planck Collaboration pushes the redshift of instantaneous reionization down to z = 8.8-1.4+1.7and greatly reduces the need for significant Lyman-continuum emission at very early times. We show that if recent claims of a numerous population of faint AGNs at z = 4-6 are upheld and the high inferred AGN comoving emissivity at these epochs persists to higher, z ≳ 10, redshifts, then active galaxies may drive the reionization of hydrogen and helium with little contribution from normal star-forming galaxies. We discuss an AGN-dominated scenario that satisfies a number of observational constraints: the H I photoionization rate is relatively flat over the range 2 < z < 5, hydrogen gets fully reionized by z ; 5.7, and the integrated Thomson scattering optical depth is τ ; 0.056, in agreement with measurements based on the Lyα opacity of the intergalactic medium (IGM) and cosmic microwave background polarization. It is a prediction of the model that helium gets doubly reionized before redshift 4, the heat input from helium reionization dominates the thermal balance of the IGM after hydrogen reionization, and z > 5 AGNs provide a significant fraction of the unresolved X-ray background at 2 keV. Singly and doubly ionized helium contribute about 13% to τ, and the He III volume fraction is already 50% when hydrogen becomes fully reionized.