Entropy-Based Classification of Subsurface Scatterers: A Valuable Tool for the Analysis of Data Obtained by the Fully Polarimetric WISDOM Radar

The "Water Ice Subsurface Deposition on Mars" Experiment (WISDOM) is a Ground Penetrating Radar (GPR) and part of the 2020 ExoMars Rover payload. It will be the first GPR operating on a planetary rover and the first fully polarimetric radar tasked at probing the subsurface of Mars. WISDOM operates at frequencies between 500 MHz and 3 GHz yielding a centimetric resolution and a penetration depth of about 3 meters in Martian soil. Its prime scientific objective is the detailed characterization of the material distribution within the first few meters of the Martian subsurface as a contribution to the search for evidence of past life. For the first time, WISDOM will give access to the geological structure, electromagnetic nature, and hydrological state of the shallow subsurface by retrieving the layering and properties of the buried reflectors at an unprecedented resolution and, due to the fully polarimetric measurements, amount of information. Furthermore, a "real time" subsurface analysis will support the drill operations by identifying locations of high scientific interest and low risk. Key element in the WISDOM data analysis is the fast and reliable classification and correct localization of subsurface scatterers and layers. The fully polarimetric nature of the WISDOM measurements allows the use of the entropy-alpha decomposition (H-alpha). This method enables the classification of reconstructed images of the subsurface (obtained by inverse imaging algorithms, e.g. f-k migration) with regard to the main scattering mechanisms of geological features present in the image of the subsurface. It is, for example, possible to differentiate smooth surfaces, rough surfaces, isolated spherical scatterers, double- and bounce scattering, anisotropic scatterers, clouds of small scatterers of similar shape as well as layers of oblate spheroids. Preliminary tests under laboratory conditions suggest the feasibility and value of the approach for the classification of geological features in the Martian subsurface in the context of WISDOM data processing and operations. It is a fast and reliable tool leveraging the whole amount of information provided by the fully polarimetric WISDOM Radar.