Selecting a complete sample of blazars in sub-millimetre catalogues

TL;DR

This paper presents a method to select a complete sample of blazars from sub-millimetre surveys by combining sub-mm and radio data, resulting in a reliable catalog of 39 bright blazars and validating the selection against models.

Contribution

The study develops a novel selection approach that efficiently identifies blazars in sub-mm surveys by leveraging spectral information across multiple wavelengths, ensuring sample completeness.

Findings

Identified 39 bright blazars in H-ATLAS fields.

Demonstrated the selection method's high efficiency and completeness.

Found good agreement between observed counts and theoretical models.

Abstract

The \textit{Herschel} Astrophysical Terahertz Large Area Survey (H-ATLAS), that has covered about 642 sq. deg. in 5 bands from 100 to 500 $\mu\rm m$, allows a blind flux-limited selection of blazars at sub-mm wavelengths. However, blazars constitute a tiny fraction of H-ATLAS sources and therefore identifying them is not a trivial task. Using the data on known blazars detected by the H-ATLAS we have defined a locus for 500\,$\mu$m selected blazars and exploited it to select blazar candidates in the H-ATLAS fields. Candidates and known blazars in the H-ATLAS equatorial and South Galactic Pole fields were followed up with the Australia Telescope Compact Array (ATCA) or with the Karl G. Jansky Very Large Array (VLA), and matched with existing radio- and mm-catalogues to reconstruct the spectral behaviour over at least 6 orders of magnitude in frequency. We identified a selection approach that, combining the information in the sub-mm and radio domains, efficiently singles out genuine blazars. In this way, we identified a sample of 39 blazars brighter than $S_{500\mu\rm m} = 35\,$mJy in the H-ATLAS fields. Tests made cross-matching the H-ATLAS catalogues with large catalogues of blazar candidates indicate that the sample is complete. The derived counts are compared with model predictions finding good consistency with the C2Ex model and with estimates based on ALMA data.