Source Matching in the SDSS and RASS: Which Galaxies are Really X-ray Sources?

TL;DR

This study cross-matched SDSS galaxies with RASS X-ray data to identify true X-ray sources, revealing that detection rates depend on galaxy type and providing a statistical validation method for survey cross-matching.

Contribution

It introduces a statistical method to validate cross-matches between surveys with positional uncertainties, applied to SDSS and RASS data to identify genuine X-ray emitting galaxies.

Findings

No star-forming galaxies detected in RASS

Over 0.6% of narrow-line Seyferts are RASS sources

High detection rate in ambiguous optical classifications

Abstract

The current view of galaxy formation holds that all massive galaxies harbor a massive black hole at their center, but that these black holes are not always in an actively accreting phase. X-ray emission is often used to identify accreting sources, but for galaxies that are not harboring quasars (low-luminosity active galaxies), the X-ray flux may be weak, or obscured by dust. To aid in the understanding of weakly accreting black holes in the local universe, a large sample of galaxies with X-ray detections is needed. We cross-match the ROSAT All Sky Survey (RASS) with galaxies from the Sloan Digital Sky Survey Data Release 4 (SDSS DR4) to create such a sample. Because of the high SDSS source density and large RASS positional errors, the cross-matched catalog is highly contaminated by random associations. We investigate the overlap of these surveys and provide a statistical test of the validity of RASS-SDSS galaxy cross-matches. SDSS quasars provide a test of our cross-match validation scheme, as they have a very high fraction of true RASS matches. We find that the number of true matches between the SDSS main galaxy sample and the RASS is highly dependent on the optical spectral classification of the galaxy; essentially no star-forming galaxies are detected, while more than 0.6% of narrow-line Seyferts are detected in the RASS. Also, galaxies with ambiguous optical classification have a surprisingly high RASS detection fraction. This allows us to further constrain the SEDs of low-luminosity active galaxies. Our technique is quite general, and can be applied to any cross-matching between surveys with well-understood positional errors.