This is the third paper in a series which combines N -body simulations and semi-analytic modelling to provide a fully spatially resolved simulation of the galaxy formation and clustering processes .
Here we extract mock redshift surveys from our simulations : a Cold Dark Matter model with either \Omega _ { 0 } = 1 ( \tau CDM ) or \Omega _ { 0 } = 0.3 and \Lambda = 0.7 ( \Lambda CDM ) .
We compare the mock catalogues with the northern region ( CfA2N ) of the Center for Astrophysics ( CfA ) Redshift Surveys .
We study the properties of galaxy groups and clusters identified using standard observational techniques and we study the relation of these groups to real virialised systems .
Most features of CfA2N groups are reproduced quite well by both models with no obvious dependence on \Omega _ { 0 } .
Redshift space correlations and pairwise velocities are also similar in the two cosmologies .
The luminosity functions predicted by our galaxy formation models depend sensitively on the treatment of star formation and feedback .
For the particular choices of Paper I they agree poorly with the CfA survey .
To isolate the effect of this discrepancy on our mock redshift surveys , we modify galaxy luminosities in our simulations to reproduce the CfA luminosity function exactly .
This adjustment improves agreement with the observed abundance of groups , which depends primarily on the galaxy luminosity density , but other statistics , connected more closely with the underlying mass distribution , remain unaffected .
Regardless of the luminosity function adopted , modest differences with observation remain .
These can be attributed to the presence of the “ Great Wall ” in the CfA2N .
It is unclear whether the greater coherence of the real structure is a result of cosmic variance , given the relatively small region studied , or reflects a physical deficiency of the models .