2000MNRAS.319..813D

Many of the faint submm sources uncovered by deep SCUBA surveys still remain unidentified at most, if not all, other wavelengths. The most successful method so far of obtaining accurate positions and hence allowing more secure identifications has been to use centimetre-wavelength imaging with the VLA. This has led to a tempting approach for obtaining redshift estimates for this population by Carilli & Yun, which relies on the tight FIR-radio relationship and takes advantage of the steep spectral slope in the submm. In this paper we use the submm data from the SCUBA Local Universe Galaxy Survey (SLUGS) to estimate the usefulness of, and the uncertainties in, the radio-submm redshift estimator. If the submm-radio spectral index were correlated with either dust temperature or 850-µm luminosity, this method could produce biased redshift estimates for 850-µm selected galaxies. We find, however, that within SLUGS, these correlations are not significant. The ratio of 850-µm to 1.4-GHz flux is found to decrease with increasing radio and FIR luminosity, and we propose that this is due to a component of the dust not associated with recent star formation, but which is instead heated to 15-20K by the general interstellar radiation field.