SIMBAD references

Despite extensive observational efforts, a convincing optical/infrared identification of the brightest submm source in the Hubble Deep Field, HDF 850.1, has remained elusive after almost four years. This failure is all the more notable given the availability of supporting multifrequency data of unparalleled depth, and subarcsec positional accuracy for the submm/mm source. Consequently, HDF 850.1 has become a test case for the possibility that the most violently star-forming objects in the Universe are too red and/or distant to be seen in the deepest optical images.

Here we report the discovery of the galaxy counterpart of HDF 850.1. This object has been revealed by careful analysis of a new, deep K' image of the HDF obtained with the Subaru 8.2-m telescope. Its reality is confirmed by a similar analysis of the HST NICMOS F160W image of the same region. This object is extremely faint (K≃ 23.5), clumpy (on subarcsec scales) and very red (I-K > 5.2; H-K= 1.4±0.35). The likelihood that it is the correct galaxy counterpart is strongly reinforced by a reanalysis of the combined MERLIN+VLA 1.4-GHz map of the field, which provides a new radio detection of HDF 850.1 only 0.1 arcsec from the new near-infrared counterpart, and with sufficient positional accuracy to exclude all previously considered alternative optical candidates.

We have calculated new confidence limits on the estimated redshift of HDF 850.1 in the light of the new radio detection, and find z= 4.1±0.5. We have also determined the scalelength, and hence estimated the mass of the apparently nearby (0.5 arcsec distant) z≃ 1 elliptical galaxy 3-586.0. From this we calculate that the flux density of HDF 850.1 has been boosted by a factor of ≃3 through gravitational lensing by this intervening elliptical, consistent with predictions that a small but significant fraction of blank-field submm sources are lensed by foreground galaxies. We discuss the wider implications of these results for the submm population and cosmic star formation history.