Bayesian redshift-space distortions correction from galaxy redshift surveys

We present a Bayesian reconstruction method which maps a galaxy distribution from redshift-space to real-space inferring the distances of the individual galaxies. The method is based on sampling density fields assuming a lognormal prior with a likelihood given by the negative binomial distribution function modelling stochastic bias. We assume a deterministic bias given by a power law relating the dark matter density field to the expected halo or galaxy field. Coherent redshift-space distortions are corrected in a Gibbs-sampling procedure by moving the galaxies from redshift-space to real-space according to the peculiar motions derived from the recovered density field using linear theory with the option to include tidal field corrections from second order Lagrangian perturbation theory. The virialised distortions are corrected by sampling candidate real-space positions (being in the neighbourhood of the observations along the line of sight), which are compatible with the bulk flow corrected redshift-space position adding a random dispersion term in high density collapsed regions. The latter are defined according to the eigenvalues of the Hessian. This approach presents an alternative method to estimate the distances to galaxies using the three dimensional spatial information, and assuming isotropy. Hence the number of applications is very broad. In this work we show the potential of this method to constrain the growth rate up to $k$ ~ 0.3 $h$ Mpc$^{-1}$. Furthermore it could be useful to correct for photo-metric redshift errors, and to obtain improved BAO reconstructions.