Experimental determination of the surface photometric contribution in the spectral reflectance deconvolution processes for a simulated martian crater-like regolithic target

We produce bidirectional reflectance experimental measurements in the visible and near-infrared range of a macroscopic target simulating the case of a martian crater. Using Hapke's equation of radiative transfer, we compare the performance, in terms of mineralogical abundance determination, of different deconvolution processes on a multispectral image of the experimental target. In particular, we study the effects of the topography and the physical properties of natural rocky surfaces (e.g., local variations of incidence and emergence angles, grain size variations, mixtures of materials) on the data interpretation. For this purpose, we increase progressively the amount of quantitative knowledge available in terms of Hapke parameters description, textural properties and topography for the target. We estimate the accuracy of results in comparison with the known ground truth as a function of the level of knowledge we have of the target and carry out a critical assessment on the relative applicability of the different processes. This study shows that the more important parameters to take into account are (in decreasing order): (1) the textural roughness which is shown essential for the accurate determination of mineralogical abundances; (2) the disparity of Hapke parameters across the target (3) the topography (DEM) that has a limited influence on the results. These findings have obvious implications for interpreting planetary regolith reflectance properties in terms of photometry, spectroscopy and mineralogy, measured either from spaceborne (e.g., Io observations from Galileo, Mars from Mars-Express/HRSC and OMEGA) or in situ (Mars Pathfinder, MER) instruments.