Scientific Motivation

Present day large-scale VLBI surveys (e.g. Polatidis et al. 1995, Henstock et al. 1995, Kellermann et al. 1997) have previously been limited to observations of relatively bright radio sources with fluxes > 0.35Jy. Further restrictions have often been imposed e.g. pre-selecting sources which have a flat radio spectrum at cm wavelengths. By employing these selection criteria VLBI surveys have been remarkably succssful but a price has been paid: the source samples are clearly biased. Our current knowledge and understanding of compact radio sources is thus incomplete and quite probably distorted, since we have restricted the majority of investigations to bright, (flat-sepctrum) sources which may not be representative of the general population of compact, extra-galactic radio sources.

The technique of phase-referencing allows sources as weak as a 1~mJy/beam to be reliably detected and imaged (Beasley \& Conway 1993). Hence in principle, it is now possible to investigate samples of radio sources selected at much fainter flux levels than was hitherto possible. For the first time, sources which are 1 to 2 orders of magnitude fainter than those targeted in previous surveys can be systematically imaged and reliably classified by VLBI. Hopefully there will be a few surprises along the road, but there are already some guarantees:

  • the first systematic classification of the milliarcsecond scale structure of a relatively faint sample of compact radio sources,

  • investigate whether the properties of weak sources are similar to those of their brighter cousins e.g. is the statistical distribution of source class similar between faint and bright source samples; are faint sources more compact than bright sources ?

  • conduct simple tests of unified schemes. Evolutionary models constructed by Wall & Jackson 1997 predict that the fraction of flat spectrum objects should decrease to 10-20% at these flux levels. Why is our detection rate with MERLIN at 6cm so high ?

  • CSO objects constitute about 5-10% of objects in flux limited VLBI surveys (Taylor et al. 1996). We can test if the apparent tendency for a lower CSO fraction in lower luminosity sources holds true at these faint flux levels. We will also be able to test evolutionary models (Readhead et al 1996) - any CSO's in our faint sample should be, on average, larger than the ones seen in the CJ surveys

  • By going deeper we hope to uncover a few surprises too.

    Last modified: May 20, 1997

    JIVE Home Page

    Questions ? Comments ? e-mail Mike Garrett