Abstract
The results of an extensive optical and infrared investigation of a complete sub-sample of the Leiden-Berkeley Deep Survey of radio sources at 1.4 GHz are presented. Optical counterparts have been identified for 69 of the 73 sources in the two Hercules fields, and redshifts obtained for 49 of them. Photometric redshifts are computed from the griK data for the remaining 20 sources.
Redshifts have been measured for fifteen sources in the brighter Parkes Selected Regions radio survey, in order to test the accuracy of redshift estimates based on the K -z relation for brighter radio sources. The results enhance the evidence for a high-redshift cut-off in the2.7 GHz radio luminosity function.
The data from the LBDS Hercules sample is compared with the radio luminosity functions (RLFs) of Dunlop and Peacock (1990). Two of the RLF models successfully trace the evolution of the radio sources with redshift, but there is some disagreement between the luminosity-dependence of the models and the data. The observed RLF for the lower luminosity population (log10 P2.7 < 26) shows evidence for a cut-off at lower redshifts (z ~ 0.5-1.5) than for the more powerful objects in the Parkes Selected Regions survey.
The spectral evolution of these radio sources is investigated, with particular emphasis on modelling the red envelope of galaxies. Together with a sample of passively-evolving galaxies believed to be at z ~ 2.4, these data are used to show that the oldest galaxies at high- redshift are incompatable with a critical-density (Ho = 1) universe, unless there is a significant cosmological constant.
