Key points are not available for this paper at this time.
The VLA-COSMOS Large Project has imaged the 2 deg2 COSMOS field with a resolution of 1.5" and a sensitivity of about 11 muJy (1 sigma), yielding a catalog of ~3600 radio sources. In this paper we present a further analysis of the VLA-COSMOS Large Project catalog of radio sources aimed to (1) quantify and correct for the effect of bandwidth smearing in the catalog, (2) determine the incompleteness produced by the noise bias and the resolution bias in the new catalog, and (3) derive the radio source counts at 1.4 GHz. The effect of bandwidth smearing on the radio sources in the catalog was quantified comparing the peak and total flux densities in the final mosaic and in each of the individual pointings where the source was closest to the center of the field. We find that the peak flux densities in the original VLA-COSMOS Large Project catalog have to be divided by a factor of about 0.8 or 0.9, depending on the distance from the mosaic center. The completeness of the radio catalog has been tested using samples of simulated radio sources with different angular size distributions. These simulated sources have been added to the radio image and recovered using the same techniques used to produce the radio catalog. The fraction of missed sources as a function of the total flux density is a direct measure of the incompleteness. Finally, we derived the radio source counts down to 60 muJy with unprecedented good statistics. Comparison to the findings of other surveys shows good agreement in the flux density range 0.06-1 mJy confirming the upturn at ~0.5 mJy and a possible decline of the source counts below ~0.1 mJy.
Bondi et al. (Thu,) studied this question.