3.4. Clustering properties

The deep submm-wave surveys made to date provide relatively little information about the spatial distribution of the detected galaxies. There is some indication from the UK 8-mJy SCUBA survey (Almaini et al., 2002; Fox et al., 2002; Ivison et al., 2002; Scott et al., 2002) and from the widest-field MAMBO surveys (Carilli et al., 2001) that the clustering strength of SCUBA-selected galaxies is greater than that of faint optically-selected galaxies, yet less than that of K 20 ERO samples (Daddi et al., 2000). Webb et al. (2002b) point out that the angular clustering signal expected in submm surveys is likely to be suppressed by smearing in redshift, as the submm galaxies should have a wide range of redshifts, and so the spatial correlation function might in fact be stronger than that of the EROs. As most K 20 EROs are evolved elliptical galaxies at z 1, which are likely to be amongst the first galaxies to form in the most overdense regions of the Universe, their strong clustering is easily explained. However, a definitive result on the clustering of submm galaxies awaits a much larger sample of galaxies than considered by Scott et al. (2002) and Webb et al. (2002b). Characteristic brightness fluctuations on the angular scales expected from faint unresolved dusty galaxies have been found in both the 850-µm SCUBA image of the HDF-N (Peacock et al., 2000) and in deep, confused 175-µm ISO images (Lagache and Puget, 2000; Kiss et al., 2001).

Haiman and Knox (2000) have discussed the details of measurements of the correlation function of unresolved submm galaxies on arcminute angular scales in the context of CMB experiments, finding that the correlated signal can carry important information about the nature and evolution of the submm galaxy population. A simpler investigation by Scott and White (1999) drew similar conclusions, while there is further discussion by Magliocchetti et al. (2001).