VIRMOS RADIO survey

People involved: Ciliegi, de Ruiter, Zamorani.

The radio survey has been obtained with the VLA at 1.415 GHz in one of the VIRMOS region where deep $BVRI$ band photometry has been already obtained with the CFHT to a limiting magnitude (5 $\sigma$) of about 25.5 and, for a fraction of the area, in the $U$ and $K$ bands with the ESO telescopes to a limiting magnitude of $\sim 25$ and $\sim 21.5$ respectively. The VIRMOS VLA radio survey has mapped an area of 1 square degree with a uniform noise of $\sim$85 $\mu$Jy (5 $\sigma$). A catalogue of radio sources brighter than the local 5$\sigma$ threshold has been extracted from the 1 degree radio map. It contains 1054 radio sources, 19 of which are considered as multiple, i.e. fitted with at least two separate components. The sources counts obtained with this radio catalogue are in very good agreement with those obtained with other surveys. In particular, our point at the faintest flux level ($\sim$ 0.1 mJy) is fully consistent (with a more robust statistics thanks to the higher number of sources and the larger area covered) with the points obtained with a very deep radio observation in the Hubble Deep Field (HDF). The preliminary results of the VIRMOS VLA survey have been presented at the meeting Where's the Matter? Tracing dark and bright matter with the new generation of large scale surveys (Marseille, June 2001).

Subsequently, using the already available optical data, we performed a preliminary cross-correlation between the radio and optical catalogues considering as possible counterparts all the objects for which the offset between radio and optical positions is smaller than 1.5 arcsec. The final optical identification of all the radio sources will be performed using statistical methods like, for example, the Likelihood Ratio Analysis.

We found 770 optical counterparts in the $I$ band catalogue with an offset smaller than 1.5 arcsec from the radio position. The expected number of spurious identifications within this radius is $\sim$80 leading to an identification percentage of $\sim$65%. This percentage has to be considered as a lower limit, mainly because the automatic procedure adopted so far in cross-correlating radio and optical catalogues can miss the optical counterparts for the more extended and/or complex radio sources. We have verified visually that this is indeed happening in a number of cases. On this basis and taking into account the fact that additional catalogues in other bands will soon be available, we expect that the final percentage of identifications will be at least 80%, which would be one of the highest obtained up to now for a large sub-mJy survey and comparable to that of the HDF radio surveys.