The WSRT wide-field H I survey I. The background galaxy sample

We have used the Westerbork array to carry out an unbiased wide-field survey for H  emission features, achieving an  sensitivity of about 18 mJy/Beam at a velocity resolution of 17 km s −1 over 1800 deg 2 and between −1000 < VHel < +6500 km s −1 . The primary data consists of auto-correlation spectra with an effective angular resolution of 49 � FWHM, although cross-correlation data were also acquired. The survey region is centered approximately on the position of Messier 31 and is Nyquist-sampled over 60 × 30 ◦ in RA × Dec. More than 100 distinct features are detected at high significance in each of the two velocity regimes (negative and positive LGSR velocities). In this paper we present the results for our H  detections of external galaxies at positive LGSR velocity. We detect 155 external galaxies in excess of 8σ in integrated H  flux density. Plausible optical associations are found within a 30 � search radius for all but one of our H  detections in DSS images, although several are not previously cataloged or do not have published red-shift determinations. Our detection without a DSS association is at low galactic latitude. Twenty-three of our objects are detected in H  for the first time. We classify almost half of our detections as "confused", since one or more companions is cataloged within a radius of 30 � and a velocity interval of 400 km s −1 . We identify a handful of instances of significant positional offsets exceeding 10 kpc of unconfused optical galaxies with the associated H  centroid, possibly indicative of severe tidal distortions or uncataloged gas-rich companions. A possible trend is found for an excess of detected H  flux in unconfused galaxies within our large survey beam relative to that detected previously in smaller telescope beams, both as function of increasing distance and increasing gas mass. This may be an indication for a diffuse gaseous component on 100 kpc scales in the environment of massive galaxies or a population of uncataloged low mass companions. We use our galaxy sample to estimate the H  mass function from our survey volume. Good agreement is found with the HIPASS BGC results, but only after explicit correction for galaxy density variations with distance.