Compact structure in southern peaked spectrum radio sources

Data for a sample of forty three Southern radio sources is presented. The majority of the sources have spectra that are steep at high frequencies and that peak, or at least flatten markedly, below a few giga-hertz. The sources have been studied through both single dish observations and VLBI. A dual frequency (2.3 and 8.4 GHz) flux density monitoring program lasting nearly two and a half years and comprising measurements of all sources at intervals of between one and three weeks has shown that the incidence of variability is systematically low in peaked spectrum sources. This observation is consistent with the assertion that the energy source powering the luminous cores in this type object is of a different nature to that driving the compact cores seen in many powerful fl.at spectrum radio sources. In order that the observations could be performed on a regular basis with minimum user intervention, an automated observing and data acquisition system was implemented. To stabilise the receiver gain, the system incorporates a versatile.software implementation of a noise adding radiometer. · The recently established Southern Hemisphere VLBI Experiment (SHEVE) array is described and its capabilities and limitations discussed. The operational procedures and special difficulties inherent in the operation of the array are explained. Images for nine sources made from observations with the SHEVE array are presented. For seven of the sources, images are available at both 2.3 and 8.4 GHz. The images have a mean dynamic range of 65: 1 and provide accurate measu!ements of the detailed structure in these sources, including component sizes, flux densities and spectral indices. All of the peaked spectrum sources imaged are doubles and few show any evidence of significant extended structure. These data constitute the first high resolution observations for six of the sources. Observations with a single baseline interferometer have provided the first survey of the compact structure in thirty eight of the sources in the sample. Simple models representing the basic source structures have been fit to the data for all objects detected in this survey. These models show that about two thirds of the sources surveyed have more than one component at milli-arcsecond resolution. A comparison of the models with the images is used to set limits on the reliability of the model fitting results. A database of all the source properties obtained, both through the observations described, and compiled from the literature is presented. Numerous parameters intrinsic to the sources are calculated using these data and are briefly discussed.