The multiplicity of 250-$\mu$m Herschel sources in the COSMOS field

TL;DR

This study examines the multiplicity of Herschel-detected extragalactic sources in the COSMOS field, revealing that many sources are composed of multiple components, which impacts flux attribution and source characterization.

Contribution

It introduces a Bayesian inference method using XID+ to quantify the contributions of multiple components to Herschel 250-$ m{ extmu m}$ sources, highlighting the prevalence of multiplicity.

Findings

Most Herschel sources have multiple components contributing to the flux.

Brightest sources often contain significant flux in secondary components.

Single-source flux estimates may overstate actual flux by up to 150%.

Abstract

We investigate the multiplicity of extragalactic sources detected by the Herschel Space Observatory in the COSMOS field. Using 3.6- and 24-$\mu$m catalogues, in conjunction with 250-$\mu$m data from Herschel, we seek to determine if a significant fraction of Herschel sources are composed of multiple components emitting at 250 $\mu$m. We use the XID+ code, using Bayesian inference methods to produce probability distributions of the possible contributions to the observed 250-$\mu$m flux for each potential component. The fraction of Herschel flux assigned to the brightest component is highest for sources with total 250-$\mu$m fluxes < 45 mJy; however, the flux in the brightest component is still highest in the brightest Herschel sources. The faintest 250-$\mu$m sources (30-45 mJy) have the majority of their flux assigned to a single bright component; the second brightest component is typically significantly weaker, and contains the remainder of the 250-$\mu$m source flux. At the highest 250-$\mu$m fluxes (45-110 mJy), the brightest and second brightest components are assigned roughly equal fluxes, and together are insufficient to reach 100 per cent of the 250-$\mu$m source flux. This indicates that additional components are required, beyond the brightest two components, to reproduce the observed flux. 95 per cent of the sources in our sample have a second component that contains more than 10 per cent of the total source flux. Particularly for the brightest Herschel sources, assigning the total flux to a single source may overestimate the flux contributed by around 150 per cent.