A Bayesian approach to strong lensing modelling of galaxy clusters

Abstract. In this paper, we explore how well, one can recover the mass distribution in strong lensing cluster cores where different set of multiple images with different redshifts have been identified. To be able to quantify the uncertainty in the mass reconstruction, we have used a Bayesian Monte Carlo Markov Chain (MCMC) sampler (“Bayesys”). In particular, such optimization method allows to avoid local minima in the likelihood distributions which can be frequent in large parameter spaces modelling. To illustrate the power of the MCMC technique, we have simulated three clusters of galaxies with a set of underlying galaxy-scale subhalos and a clusterscale halo modelled with a Pseudo-Isothermal Elliptical Mass Distribution, a pseudo-elliptical Navarro, Frenk White and a pseudo-elliptical Sérsic potential. For each of them, we study the degeneracies between the various model parameters. Indeed, the Bayesian sampler easily provides us with an accurate and straightforward estimation of the errors. In particular, we find that the mass of the galaxies can be strongly degenerated with the cluster mass for certain image