Joint reconstruction of galaxy clusters from gravitational lensing and thermal gas. II. Inversion of the thermal Sunyaev-Zeldovich effect

This paper continues a series in which we intend to show how all observables of galaxy clusters can be combined to recover the two-dimensional, projected gravitational potential of individual clusters. Our goal is to develop a non-parametric algorithm for joint cluster reconstruction taking all cluster observables into account. In this paper, we begin with the relation between the Compton-y parameter and the Newtonian gravitational potential, assuming hydrostatic equilibrium and a polytropic stratification of the intracluster gas. We show how Richardson-Lucy deconvolution can be used to convert the intensity change of the CMB due to the thermal Sunyaev-Zeldovich effect into an estimate for the two-dimensional gravitational potential. Synthetic data simulated with characteristics of the ALMA telescope show that the two-dimensional potential of a cluster with mass 5*10^14 M_sun/h at redshift 0.2 is possible with an error of < 5% between the cluster centre and a radius r < 0.9 Mpc/h.

Joint reconstruction of galaxy clusters from gravitational lensing and thermal gas I. Outline of a non-parametric method

We present a method to estimate the lensing potential from massive galaxy clusters for given observational X-ray data. The concepts developed and applied in this work can easily be combined with other techniques to infer the lensing potential, e.g. weak gravitational lensing or galaxy kinematics, to obtain an overall best fit model for the lensing potential. After elaborating on the physical details and assumptions the method is based on, we explain how the numerical algorithm itself is implemented with a Richardson-Lucy algorithm as a central part. Our reconstruction method is tested on simulated galaxy clusters with a spherically symmetric NFW density profile filled with gas in hydrostatic equilibrium. We describe in detail how these simulated observational data sets are created and how they need to be fed into our algorithm. We test the robustness of the algorithm against small parameter changes and estimate the quality of the reconstructed lensing potentials. As it turns out we achieve a very high degree of accuracy in reconstructing the lensing potential. The statistical errors remain below 2.0% whereas the systematical error does not exceed 1.0%.