Oral Presentation

SCUBA-2 Ultra Deep Imaging EAO Survey (STUDIES): Multi-wavelength properties and luminosity function of 450-μm selected galaxies

Author(s): Chen-Fatt Lim(NTU/ASIAA), Wei-Hao Wang(ASIAA), Wei-Ching Lin(NTU), Chian-Chou Chen(ESO/ASIAA), Yu-Yen Chang(ASIAA), James M. Simpson(ASIAA), Yoshiki Toba (Kyoto University/ASIAA) et al.

Presenter: Chen-Fatt Lim (ASIAA/NTU)

The majority of star formation in the Universe is obscured by dust and produces strong infrared and sub-millimeter radiation. Far-infrared or sub-millimeter observations are therefore required to fully understand the formation history of galaxies. Sub-millimeter surveys have revealed a population of dusty galaxies (sub-millimeter galaxies, SMGs) at z ~ 1–3.5, occupying the same epoch as the peak of quasar activities. SMGs dominate the massive end of star-formation with star formation rates of > 100 M⦿yr^-1 and reside in massive dark matter halos (10^12 to 10^13 h^-1 M⦿), suggesting that SMGs may be the progenitors of elliptical galaxies in the local Universe. In this talk, I will present the results of the SCUBA-2 Ultra Deep Imaging EAO Survey (STUDIES) project. STUDIES is a JCMT Large Project aiming at obtaining confusion limited 450-μm images in the COSMOS and SXDS regions, to detect the majority of the dusty galaxy population. By the end of 2019, the STUDIES project leads to a very deep 450-μm image with an area of ~300 arcmin^2 and a noise level of ~0.56 mJy in the COSMOS field. I will also present some physical properties of our sources with multi-wavelength counterparts, including infrared luminosities, star formation rates, dust temperatures, stellar masses, and dust attenuations. Because of the unprecedented depth of the STUDIES image, these analyses reached ~ 30 M⦿yr^-1 galaxies at z ~ 2. Our data provide new measurements of obscured SFR densities at z = 0–4, without any assumptions in the faint-end slopes of the infrared luminosity function. We are also building a 450-μm-based machine-learning algorithm to identify the SMG candidates in the entire COSMOS region. There will be roughly 6000 such machine-learning candidates and they will be used for clustering analyses and halo mass estimations for faint SMGs. All the above will advance our understanding of the typical members in the high-redshift dusty galaxy population.