The XMM large scale structure survey: optical vs. X-ray classifications of active galactic nuclei and the unified scheme

TL;DR

This study compares optical and X-ray classifications of AGN in the XMM-LSS survey, revealing a mild correlation and identifying cases where intrinsic properties or host galaxy dilution explain classification discrepancies, supporting the unified scheme.

Contribution

It provides a detailed analysis of the correlation between optical and X-ray AGN classifications, introducing the fourfold point correlation coefficient and assessing the unified scheme’s applicability.

Findings

Up to 30% of sources have differing X-ray and optical classifications.

Most discrepancies can be explained by the standard AGN unified scheme.

Only 12% of sources challenge the unified scheme predictions.

Abstract

Our goal is to characterize AGN populations by comparing their X-ray and optical classifications. We present a sample of 99 spectroscopically identified X-ray point sources in the XMM-LSS survey which are significantly detected in the [2-10] keV band, and with more than 80 counts. We performed an X-ray spectral analysis for all of these 99 X-ray sources. Introducing the fourfold point correlation coefficient, we find only a mild correlation between the X-ray and the optical classifications, as up to 30% of the sources have differing X-ray and optical classifications: on one hand, 10% of the type 1 sources present broad emission lines in their optical spectra and strong absorption in the X-rays. These objects are highly luminous AGN lying at high redshift and thus dilution effects are totally ruled out, their discrepant nature being an intrinsic property. Their X-ray luminosities and redshifts distributions are consistent with those of the unabsorbed X-ray sources with broad emission lines. On the other hand, 25/32 are moderate luminosity AGN, which are both unabsorbed in the X-rays and only present narrow emission lines in their optical spectra. The majority of them have an optical spectrum which is representative of the host galaxy. We finally infer that dilution of the AGN by the host galaxy seems to account for their nature. 5/25 have been defined as Seyfert 2. In conclusion, most of these 32 discrepant cases can be accounted for by the standard AGN unified scheme, as its predictions are not met for only 12% of the 99 X-ray sources. ABRIDGED