Gaussian Process Reconstruction of Reionization History

Abstract We reconstruct the history of reionization using Gaussian process regression. Using UV luminosity data compilation from Hubble Frontiers Fields we redshift evolution density and thereby source term in ionization equation. This model-independent reconstruction rules out single power-law but supports logarithmic double parameterization. obtain by integrating equations with reconstructed term. optical depth constraint Planck cosmic microwave background observation, measurement function integrated until truncation magnitude −17 −15, derived fraction high quasar, galaxies, gamma-ray burst observations, constrain reionization. In conservative case find on as τ = 0.052 ± 0.001 0.002 at 68% 95% confidence intervals. duration between 10% 90% to be 2.05 − 0.21 0.30 + 0.11 0.37 . Longer is supported if −15 used joint analysis. Our results point that even a reconstruction, combination cosmological astrophysical observations can provide stringent constraints epoch