Principal component analysis of the reionization history from Planck 2015 data

TL;DR

This paper uses a principal component approach to analyze the reionization history from Planck 2015 data, revealing that reionization was incomplete at high redshifts and impacting cosmological parameter estimates.

Contribution

It introduces a model-independent principal component method to constrain reionization history from Planck data, improving upon the standard instantaneous reionization assumption.

Findings

Reionization was not complete at redshifts z ≥ 8.5 at 95% CL.

Higher matter fluctuation amplitude than standard models, consistent within uncertainties.

Relief of tension between Planck data and astrophysical observations when neutrino masses vary.

Abstract

The simple assumption of an instantaneous reionization of the Universe may bias estimates of cosmological parameters. In this paper a model-independent principal component method for the reionization history is applied to give constraints on the cosmological parameters from recent Planck 2015 data. We find that the Universe are not completely reionized at redshifts $z \ge 8.5$ at 95% CL. Both the reionization optical depth and the matter fluctuation amplitude are higher than but consistent with those obtained in the standard instantaneous reionization scheme. The high estimated value of the matter fluctuation amplitude strengthens the tension between Planck CMB observations and some astrophysical data, such as cluster counts and weak lensing. The tension can significantly be relieved if the neutrino masses are allowed to vary. Thanks to a high scalar spectral index, the low-scale spontaneously broken SUSY inflationary model can fit the data well, which is marginally disfavored at 95% CL in the Planck analysis.