The Herschel Bright Sources (HerBS): sample definition and SCUBA-2 observations

Abstract

We present the Herschel Bright Sources (HerBS) sample, a sample of bright, high-redshift Herschel sources detected in the 616.4 square degree H-ATLAS survey. The HerBS sample contains 209 galaxies, selected with a 500 µm flux density greater than 80 mJy and an estimated redshift greater than 2. The sample consists of a combination of HyLIRGs and lensed ULIRGs during the epoch of peak cosmic star formation. In this paper, we present SCUBA-2 observations at 850 µm of 189 galaxies of the HerBS sample, 152 of these sources were detected. We fit a spectral template to the Herschel-SPIRE and 850 µm SCUBA-2 flux densities of 22 sources with spectroscopically determined redshifts, using a two-component modified blackbody spectrum as a template. We find a cold-and hot-dust temperature of 21.29 +1.35 −1.66 K and 45.80 +2.88 −3.48 K, a cold-to-hot dust mass ratio of 26.62 +5.61 −6.74 and a β of 1.83 +0.14 −0.28 . The poor quality of the fit suggests that the sample of galaxies is too diverse to be explained by our simple model. Comparison of our sample to a galaxy evolution model indicates that the fraction of lenses is high. Out of the 152 SCUBA-2 detected galaxies, the model predicts 128.4 ± 2.1 of those galaxies to be lensed (84.5%). The SPIRE 500 µm flux suggests that out of all 209 HerBS sources, we expect 158.1 ± 1.7 lensed sources, giving a total lensing fraction of 76 per cent.

A search for the lenses in the Herschel Bright Sources (HerBS) sample

“…Even with 10 missed lenses, the total number of lenses found on the SDSS is well below the number expected from the model (∼90 sources across the 121 sources with SDSS coverage; Bakx et al 2018). One obvious possibility, which we test in the next section, is that many of the lenses are too faint to be detected on the SDSS.…”

“…We remove sources within 10 arcseconds of an NVSS source, which removes all bright blazar objects, and potentially some of the brightest non-blazar sub-mm sources. This survey is also able to find all blazar-objects in the early HerBS catalogue (Bakx et al 2018), and identified only three non-blazar sources.…”

“…We start with the H-ATLAS catalogue, neglecting the NGP field, which does not have coverage in the VIKING fields. We remove sources with photometric redshifts smaller than 2 for the brightest Herschel sources -estimated by fitting the template from Bakx et al 2018) -and for sources fainter than 70 mJy at 500µm, we remove sources with photometric redshifts smaller than 3 to account for an increase in false-positives due to the increase in sample size. The template from Bakx et al (2018) assumes a two-temperature modified blackbody SED.…”

“…This sample consists of the 209 brightest (S 500µm > 80 mJy) sources from H-ATLAS with estimated redshifts > 2. Models predict that roughly three quarters of these sources are strongly lensed (Negrello 2007;Cai et al 2013;Bakx et al 2018). Since we expect a large fraction of the optical/near-IR counterparts to be foreground lenses, we have adapted the standard method used to find the optical/near-IR counterparts to H-ATLAS sources (Bourne et al 2016) to make the method more sensitive for finding the lenses.…”

“…This table contains the results of our VIKING cross-analysis. The HerBS number and sub-mm redshift are from Bakx et al (2018). R is the reliability, K mag is the SExtractorderived AUTO-magnitude and error.…”