Characterization of SCUBA-2 450 μm and 850 μm selected galaxies in the COSMOS field.

We present deep 450 μm and 850 μm observations of a large, uniformly covered 394 arcmin2 area in the Cosmic Evolution Survey (COSMOS) field obtained with the SCUBA-2 instrument on the James Clerk Maxwell Telescope (JCMT). We achieve root-mean-square noise values of σ450 = 4.13 mJy and σ850 = 0.80 mJy. The differential and cumulative number counts are presented and compared to similar previous works. Individual point sources are identified at >3.6σ significance, a threshold corresponding to a 3–5 per cent sample contamination rate. We identify 78 sources at 450 μm and 99 at 850 μm, with flux densities S450 = 13–37 mJy and S850 = 2–16 mJy. Only 62–76 per cent of 450 μm sources are 850 μm detected and 61–81 per cent of 850 μm sources are 450 μm detected. The positional uncertainties at 450 μm are small (1–2.5 arcsec) and therefore allow a precise identification of multiwavelength counterparts without reliance on detection at 24 μm or radio wavelengths; we find that only 44 per cent of 450 μm sources and 60 per cent of 850 μm sources have 24 μm or radio counterparts. 450 μm selected galaxies peak at 〈z〉 = 1.95 ± 0.19 and 850 μm selected galaxies peak at 〈z〉 = 2.16 ± 0.11. The two samples occupy similar parameter space in redshift and luminosity, while their median SED peak wavelengths differ by ∼20–50 μm (translating to ΔTdust = 8–12 K, where 450 μm selected galaxies are warmer). The similarities of the 450 μm and 850 μm populations, yet lack of direct overlap between them, suggests that submillimetre surveys conducted at any single far-infrared wavelength will be significantly incomplete (≳30 per cent) at censusing infrared-luminous star formation at high z