Abstract

Redshifts are fundamental for our understanding of extragalactic X-ray sources. Ambiguous counterpart associations, expensive optical spectroscopy and/or multi-mission multiwavelength coverage to resolve degeneracies make estimation often difficult in practice. Here we present a new method to constrain redshifts from the full X-ray spectrum directly, by Bayesian model fitting of low-resolution spectra. We demonstrate our approach on real Chandra Deep Field South data, determining 29 redshifts which have optical spectroscopic redshifts. Our accuracy is up to ∆z=0.5, with only three outliers which can be explained by incorrect optical associations. Through extensive simulations we verify the method. To work optimally, AGN need be obscured and have at least ∼200 (XMM-Newton) or ∼150 (Chandra/ACIS) counts. Not meeting these conditions does not lead to wrong redshifts, only large uncertainties. We conclude with predictions for the upcoming eROSITA and ATHENA missions.