Reconstruction of halo power spectrum from redshift-space galaxy distribution: cylinder-grouping method and halo exclusion effect (1611.04165v2)

Abstract: The peculiar velocity field measured by redshift-space distortions (RSD) in galaxy surveys provides a unique probe of the growth of large-scale structure. However, systematic effects arise when including satellite galaxies in the clustering analysis. Since satellite galaxies tend to reside in massive halos with a greater halo bias, the inclusion boosts the clustering power. In addition, virial motions of the satellite galaxies cause a significant suppression of the clustering power due to nonlinear RSD effects. We develop a novel method to recover the redshift-space power spectrum of halos from the observed galaxy distribution by minimizing the contamination of satellite galaxies. The cylinder grouping method (CGM) we study effectively excludes satellite galaxies from a galaxy sample. However, we find that this technique produces apparent anisotropies in the reconstructed halo distribution over all the scales which mimic RSD. On small scales, the apparent anisotropic clustering is caused by exclusion of halos within the anisotropic cylinder used by the CGM. On large scales, the misidentification of different halos in the large-scale structures, aligned along the line-of-sight, into the same CGM group, causes the apparent anisotropic clustering via their cross-correlation with the CGM halos. We construct an empirical model for the CGM halo power spectrum, which includes correction terms derived using the CGM window function at small scales as well as the linear matter power spectrum multiplied by a simple anisotropic function at large scales. We apply this model to a mock galaxy catalog at z=0.5, designed to resemble SDSS-III BOSS CMASS galaxies, and find that our model can predict both the monopole and quadrupole power spectra of the host halos up to k<0.5 h/Mpc to within 5%.