Abstract-In this paper, a statistical approach to evaluate randomly rough surfaces (RRS) in an inverse scattering problem is presented. Whereas in these investigations the roughness criterions possess random variables, the use of deterministic techniques such as the target decomposition (TD) can not be useful by itself as a tool of analysis. In these conditions, a statistical approach is essentially required to evaluate the target parameters. The goal of this study is the estimation of the polarimetric signatures, such as the scattering mechanism α and the entropy H, via a novel approach including the combination of TD and a new statistical model. To validate our work, SAR data sets, provided by the European Space Agency (ESA), are analyzed and compared with the simulation results.