Statistical Analysis of an AGN sample with Simultaneous UV and X-ray Observations with Swift

TL;DR

This study statistically analyzes around 100 AGN with simultaneous UV and X-ray data from Swift, revealing correlations with the Eddington ratio and suggesting a new classification scheme based on accretion properties.

Contribution

It introduces multivariate statistical tools like PCA and cluster analysis to identify the Eddington ratio as the main driver of AGN properties and proposes a new classification scheme based on this parameter.

Findings

Correlations between spectral slopes and Eddington ratio.

Bolometric corrections for specific luminosities.

Eddington ratio as the main driver of AGN properties.

Abstract

I report on the statistical analysis of a sample of about 100 AGN with simultaneous UV and X-ray observations with Swift. I found clear correlations between the X-ray spectral slope alpha-X, the UV slope alpha-UV, and the optical-to-x-ray spectral slope alpha-ox with the Eddington ratio L/Ledd. I also report on the bolometric corrections for L(0.2-2.0 keV) and L(5100A). A major aspect of the statistical analysis is multi-variate analysis statistical tools such as the Principal Component Analysis (PCA) and cluster analysis. This analysis shows that the main driver of the AGN properties in this sample is the Eddington ratio L/Ledd. Although separating Seyfert 1s into NLS1s and BLS1s is a good classification, with the 2000 km/s cutoff line it is still arbitrary. A better classification scheme may be to separate AGN into low and highL/Ledd AGN as suggested from the cluster analysis.