The second Herschel-ATLAS Data Release - III: optical and near-infrared counterparts in the North Galactic Plane field

TL;DR

This paper reports on the identification of optical and near-infrared counterparts to Herschel submillimetre sources in the North Galactic Plane, achieving high reliability and completeness using the likelihood ratio method across multiple surveys.

Contribution

It presents a detailed methodology for cross-matching Herschel sources with optical and near-infrared data, improving counterpart identification accuracy in large-area surveys.

Findings

Reliable optical counterparts identified for 37.8% of sources.

Near-infrared counterparts found for 61.8% of sources within a smaller area.

High correctness (80-90%) of the identified counterparts confirmed by radio data.

Abstract

This paper forms part of the second major public data release of the Herschel Astrophysical Terahertz Large Area Survey (H-ATLAS). In this work, we describe the identification of optical and near-infrared counterparts to the submillimetre detected sources in the $177$ deg$^2$ North Galactic Plane (NGP) field. We used the likelihood ratio method to identify counterparts in the Sloan Digital Sky Survey and in the UKIRT Imaging Deep Sky Survey within a search radius of $10$ arcsec of the H-ATLAS sources with a $4\sigma$ detection at $250 \, \mu$m. We obtained reliable ($R \ge 0.8 $) optical counterparts with $r< 22.4$ for 42429 H-ATLAS sources ($37.8$ per cent), with an estimated completeness of $71.7$ per cent and a false identification rate of $4.7$ per cent. We also identified counterparts in the near-infrared using deeper $K$-band data which covers a smaller $\sim25$ deg$^2$. We found reliable near-infrared counterparts to $61.8$ per cent of the $250$-$\mu$m-selected sources within that area. We assessed the performance of the likelihood ratio method to identify optical and near-infrared counterparts taking into account the depth and area of both input catalogues. Using catalogues with the same surface density of objects in the overlapping $\sim25$ deg$^2$ area, we obtained that the reliable fraction in the near-infrared ($54.8$ per cent) is significantly higher than in the optical ($36.4$ per cent). Finally, using deep radio data which covers a small region of the NGP field, we found that $80 - 90$ per cent of our reliable identifications are correct.