The OVI mystery: mismatch between X-ray and UV column densities

TL;DR

This paper resolves the longstanding discrepancy between UV and X-ray measurements of OVI column densities at z=0 by identifying blending with OII lines in X-ray spectra, and discusses implications for future high-resolution observations.

Contribution

It demonstrates that the OII K-beta line blends with the OVI K-alpha line in X-ray spectra, explaining the discrepancy for z=0 systems and clarifying the origin of the observed mismatch.

Findings

OII line accounts for the entire blended line in most cases.

The small OVI equivalent width is consistent with UV measurements.

The solution does not apply to high column density systems like AGN outflows.

Abstract

The UV spectra of Galactic and extragalactic sightlines often show OVI absorption lines at a range of redshifts, and from a variety of sources from the Galactic circumgalactic medium to AGN outflows. Inner shell OVI absorption is also observed in X-ray spectra (at lambda=22.03 AA), but the column density inferred from the X-ray line was consistently larger than that from the UV line. Here we present a solution to this discrepancy for the z=0 systems. The OII K-beta line ^4S^0 --> (^3D)3p ^4P at 562.40 eV (==22.04 AA) is blended with the OVI K-alpha line in X-ray spectra. We estimate the strength of this OII line in two different ways and show that in most cases the OII line accounts for the entire blended line. The small amount of OVI equivalent width present in some cases has column density entirely consistent with the UV value. This solution to the OVI discrepancy, however, does not apply to the high column density systems like AGN outflows. We discuss other possible causes to explain their UV/X-ray mismatch. The OVI and OII lines will be resolved by gratings on-board the proposed mission Arcus and the concept mission Lynx and would allow detection of weak OVI lines not just at z=0 but also at higher redshift.