Radio Source Surveys at 408 MHz

Abstract

This thesis describes an investigation centred on deep surveys of celestial radio sources made at 408 MHz with the Mills Cross telescope of the Molonglo Radio Observatory. Source counts were compiled from the surveys, and the cosmological implications were studied. The deep surveys covered small areas at two declination zones, -20 deg and -62 deg, and reached (with very high reliability) sources as weak as 84 mJy, which is about four times fainter than the previous lower limit for reliable source counts with the Molonglo instrument. Particular attention has been given to the development of consistent procedures for cataloguing, to cover such questions as resolution of closely-spaced sources, and the treatment of extended sources. A thorough study has been made of the distribution of random errors to which the observed sources are subject. The principal method used was the insertion of synthetic Monte Carlo sources. Over 1000 such sources were use d at each declination zone, to obtain accurate distributions of the errors. The separate errors due to noise and confusion were also found, and showed that these surveys approached the confusion limit of the telescope. The results of the above analysis were used to calculate the corrections to the source counts due to the effects of random errors. The corrections are not large, but an effort has nevertheless been made to use the best possible methods to obtain them. The effects of confusion errors on the corrections were investigated, and a new method was developed for dealing with them. The completeness and reliability of the catalogues were investigated, using new operational definitions of these quantities. The completeness has been evaluated explicitly from the results of the error analysis. Sources of intermediate flux density were covered by previous surveys at Molonglo. The next step in obtaining source counts over a wide range of flux density wa s therefore to improve the counts for the strongest sources.! Previously only observations from one hemisphere had been used; a thorough re-examination of published data and some further observations with the Molonglo instrument were used to compile a catalogue covering the whole sky on a uniform basis. Because of the greater number of sources this minimises the statistical errors and now supplies the best available counts for the strongest sources. The collected counts from all the above programs are given. Comparisons are made between counts from the deep surveys and those from the Cambridge aperture synthesis surveys and Bologna B2 survey. Although there is some difference in the absolute value of the Molonglo and Cambridge counts, the marked flattening of the Cambridge counts at low flux densities is confirmed. This is of interest because the latter surveys were subject to some criticism. The final phase of the project involved the fitting of cosmological model counts to the observed data in order to find what the data imply about cosmology and the evolution of sources. The investigation is largely based on the homogeneous and isotropic models of general relativity, principally the Einstein-de Sitter model. Various methods of introducing evolution are examined. The parametric method, commonly used in the literature, is able to fit the counts reasonably well after some trial and error. More promising, however, is the free-form method, in which it is not necessary to specify a functional form for the evolution; instead, the entire evolution function is found (as far as possible) from the data. A simple new method of solving for this function is given. The results show that, given the assumptions made, several plausible models can be found which fit the counts well (if evolution is restricted to the stronger sources) and which agree well with available data on redshifts and luminosities. The dependence of the required evolution on the form of the radio luminosity function is exa mined and found to be quite important. Some limitations of t! he models! are discussed, particularly the assumption of homogeneity and isotropy. The analysis concludes with a discussion of a commonly misunderstood point - the steep slope for the counts of unidentified sources. It is shown that this is due to a selection effect, and is not of cosmological significance.