There is anecdotal evidence that astronomers have been talking about a VLA Sky Survey since the VLA's inception. In the spring of 1990, having just completed the analysis of a VLA Galactic Plane survey, Rick White, David Helfand, and Bob Becker in a mood of elation were contemplating their next move and, as may have happened many times before to other astronomers, began fantasizing about a VLA Sky Survey. But this time, instead of just going out for a drink, we convinced ourselves that such a project was truly feasible. Three months later, we submitted what was, to the best of our knowledge, the first serious proposal to survey the sky at 20-cm wavelength with the VLA. The proposal called for observing fields for 90 s each in a hybrid BC configuration with the objective of surveying all the sky north of declination. The proposal generated a great deal of discussion among the staff at NRAO, but ultimately was rejected without formal review. But the stage had been set.
Nine months later, NRAO received two all-sky proposals; a revised version of our original proposal, this time in C configuration, and a D-configuration proposal from Jim Condon. In fact, this was a healthy development. In writing a C configuration proposal, we were well aware of the scientific compromises involved. The primary difference among VLA configurations is angular resolution: the FWHM ranges from 54 arcsec in D-configuration to 2 arcsec in A-configuration. A secondary consideration is available bandwidth (high-resolution images must be made in spectral line mode). High resolution images are desirable because they facilitate the identification of optical counterparts. Low resolution images are desirable because they are more sensitive to extended, low surface brightness sources. The submission of both high and low resolution proposals insured a serious discussion of the best way to carry out a VLA survey.
NRAO sent both proposals out for review and there appeared to be a consensus that the time was right for a VLA Sky survey, but a great deal of uncertainty as to which survey to do.
To facilitate an objective, optimal decision, trial observations were scheduled in the spring and summer of 1992 to compare the merits of survey results from C and D-configuration observations. Two scientific panels were convened in October 1992 and March 1993 to review the situation and make recommendations to NRAO. The outcome of these two meetings was a surprise to everyone involved. Rather than sacrifice the advantages derived from either a high or a low resolution survey, it was decided that two surveys would be made, one in D configuration and one in B configuration. The D-configuration survey was chartered to cover 85%of the sky to a 2 mJy flux density limit. The B-configuration survey was chartered to cover the 25%of the sky centered on the North Galactic Pole (NGP; chosen to coincide with the Sloan Digital Sky Survey [Gunn &Knapp 1993]) to a 1 mJy flux density limit. Observations for both surveys would start in 1993. The rest of this paper will be restricted to discussion of the high-resolution, B-configuration survey.