2 The reduction of the IRAS data

The original SCGG lists (Shakhbazian 1973; Shakhbazian & Petrosian 1974; Baier et al. 1974; Petrosian 1974, 1978; Baier & Tiersch 1975, 1976a,b, 1978, 1979) contain 377 groups. In our study the groups Shkh 12, 78, 146, 180 and 275 were omitted because they have been found to contain mainly foreground stars. We were not able to reidentify four other groups (Shkh 206, 241, 301 and 353) since their positions in the original lists were incorrect. There was also one duplication (Shkh 214 = Shkh 252). Thus we searched for FIR emission in the IRAS archives at the positions of 367 SCGGs with accurate coordinates measured from the Digitized Sky Survey or other data (Stoll et al. 1993a,b, 1994a,b, 1996a,b,c, 1997a,b,c). The IRAS Point Source Catalog (PSC), the Faint Source Catalog (FSC) and the Faint Source Catalog Rejects (FSCR) were first searched for positional matches within SCGG boundaries. The fluxes of suspected weak sources with no cirrus contamination flags in the catalogs were measured using the Scan Processing and Integration (SCANPI) method of coadding the 1-dimensional IRAS scans (Helou et al. 1988). A detection was considered reliable if the FIR source was found within about 1 arcmin from the brightest galaxy in the group (consistent with the positional uncertainty ellipse from the FSC or FSCR), if one or more galaxies was found within the 60 $\mu$m IRAS beam, and if the signal to noise ratio (SNR) at 60 $\mu$m was larger than 4. The 60 $\mu$m IRAS band was by far the most sensitive for the detection of SCGGs, as for extragalactic objects in general. Full Resolution Coadded (FRESCO) images of 1 square degree fields were also examined for groups with candidate detections from the FSC, FSCR, or SCANPI. Since FRESCO images do not have the large-scale background removed (they are not point-source filtered), they provide additional information about the group environments, such as possible confusion due to nearby stars or Galactic cirrus.

In the SCANPI processing, the median of all acceptable scans was used to estimate the flux densities from each source. Most sources are weak point sources, and the flux density based on the template amplitude fit was used. In a few cases there were signs of slightly extended emission, and the total flux within the beam area was used as the best estimator. In cases where a source is listed in the FSC or FSCR, SCANPI often provides a reliable 100 $\mu$m flux density estimate that is not in the catalogs.