Reconstruction Analysis of Galaxy Redshift Surveys: A Hybrid Reconstruction Method

Abstract

In reconstruction analysis of galaxy redshift surveys, one works backwards from the observed galaxy distribution to the primordial density field in the same region, then evolves the primordial fluctuations forward in time with an N-body code. This incorporates assumptions about the cosmological parameters, the properties of primordial fluctuations, and the biasing relation between galaxies and mass. These can be tested by comparing the reconstruction to the observed galaxy distribution, and to peculiar velocity data. This paper presents a hybrid reconstruction method that combines the `Gaussianization'' technique of Weinberg(1992) with the dynamical schemes of Nusser & Dekel(1992) and Gramann(1993). We test the method on N-body simulations and on N-body mock catalogs that mimic the depth and geometry of the Point Source Catalog Redshift Survey and the Optical Redshift Survey. This method is more accurate than Gaussianization or dynamical reconstruction alone. Matching the observed morphology of clustering can limit the bias factor b, independent of Omega. Matching the cluster velocity dispersions and z-space distortions of the correlation function xi(s,mu) constrains the parameter beta=Omega^{0.6}/b. Relative to linear or quasi-linear approximations, a fully non-linear reconstruction makes more accurate predictions of xi(s,mu) for a given beta, thus reducing the systematic biases of beta measurements and offering further scope for breaking the degeneracy between Omega and b. It also circumvents the cosmic variance noise that limits conventional analyses of xi(s,mu). It can also improve the determination of Omega and b from joint analyses of redshift & peculiar velocity surveys as it predicts the fully non-linear peculiar velocity distribution at each point in z-space.