Distribution of red and blue galaxies in groups: an empirical test of the halo model

Abstract

The popular halo model predicts that the power spectrum of the galaxy fluctuations is simply the sum of the large-scale linear halo-halo power spectrum and the weighted power spectrum of the halo profile. Previous studies have derived halo parameters from the observed galaxy correlation function. Here we test the halo model directly for self-consistency with a minimal set of theoretical assumptions by utilizing the 2dF galaxy redshift survey (2dFGRS). We derive empirically the halo occupation and galaxy radial distributions in the haloes of the 2dFGRS percolation-inferred galaxy group (2PIGG) catalogue. The mean halo occupation number is found to be well-fitted by a power law, 〈N|M〉∝Mβ, at high masses, with β= 1.05, 0.88 and 0.99 for red, blue and all galaxies, respectively (with 1σ errors of 151–19 per cent). We find that the truncated Navarro, Frenk & White profile provides a good fit to the galaxy radial distributions, with concentration parameters c= 3.9, 1.3 and 2.4 for red, blue and all galaxies, respectively (with 1σ errors of 81–15 per cent). Adding the observed linear power spectrum to these results, we compare these empirical predictions of the halo model with the observed correlation functions for these same 2dF galaxy populations. We conclude that subject to some fine-tuning it is an acceptable model for the two-point correlations. Our analysis also explains why the correlation function slope of the red galaxies is steeper than that of the blue galaxies. It is mainly due to the number of red and blue galaxies per halo, rather than the radial distribution within the haloes of the two galaxy species.