The Properties of Microjansky Radio Sources in the Hubble Deep Field–North, SSA 13, and SSA 22 Fields

Publisher Website
Google Scholar
Abstract
We present multiwavelength observations for a large sample of microjansky radio sources detected in ultradeep 1.4 GHz maps centered on the Hubble Deep Field-North (HDF-N) and the Hawaii Survey Fields SSA 13 and SSA 22. Our spectroscopic redshifts for 169 radio sources reveal a flat median redshift distribution, and these sources are hosted by similarly luminous optical L* galaxies, regardless of redshift. We suggest that the absence of low optical luminosity galaxies at low redshifts, where there are no selection effects, is due to small galaxies not being efficient at retaining the cosmic rays necessary to host microjansky radio sources. From our submillimeter measurements for 278 radio sources, we find error-weighted mean 850 μm fluxes of 1.72 ± 0.09 mJy for the total sample, 2.37 ± 0.13 mJy for the optically faint (I > 23.5) subsample, and 1.04 ± 0.13 mJy for the optically bright (I < 23.5) subsample. We significantly (>3 σ) detect in the submillimeter 50 of the radio sources, 38 with I > 23.5. Spectroscopic redshifts for three of the I < 23.5 submillimeter-detected radio sources are in the range z = 1.0-3.4, and all show active galactic nucleus (AGN) signatures. Using only the submillimeter-mapped regions, we find that 69% ± 9% of the submillimeter-detected radio population are at I > 23.5. We also find that 66% ± 7% of the S850 μm > 5 mJy (>4 σ) sources are radio-identified. We use our spectroscopic sample to determine the evolution with redshift of the radio power, and hence the far-infrared (FIR) luminosity (through the local FIR-radio correlation). We find that millimetric redshift estimates at low redshifts are best made with an FIR template intermediate between a Milky Way-type galaxy and a starburst galaxy, and those at high redshifts with an Arp 220 template.
All Related Versions