A SIMULATED GEOCHEMICAL ROVER MISSION TO THE TAURUS-LITTROW VALLEY OFTHE MOON

We test the effectiveness of using an alpha backscatter, alpha-proton, X ray spectrometer on a remotely operated rover to analyze soils and provide geologically useful information about the Moon during a simula ted mission to a hypothetical site resembling the Apollo 17 landing si te. On the mission, 100 soil samples are ''analyzed'' for major elemen ts at moderate analytical precision (e.g., typical relative sample sta ndard deviation from counting statistics: Si [11%], Al [18%], Fe [6%], Mg [20%], Ca [5%]). Simulated compositions of soils are generated by combining compositions of components representing the major lithologie s occurring at the site in known proportions. Simulated analyses are g enerated by degrading the simulated compositions according to the expe cted analytical precision of the analyzer. Compositions obtained from the simulated analyses are modeled by least squares mass balance as mi xtures of the components, and the relative proportions of those compon ents as predicted by the model are compared with the actual proportion s used to generate the simulated composition. Boundary conditions of t he modeling exercise are that all important lithologic components of t he regolith are known and are represented by model components, and tha t the compositions of these components are well known. The effect of h aving the capability of determining one incompatible element at modera te precision (25%) is compared with the effect of the lack of this cap ability. We discuss likely limitations and ambiguities that would be e ncountered, but conclude that much of our knowledge about the Apollo 1 7 site (based on the return samples) regarding the distribution and re lative abundances of lithologies in the regolith could be obtained. Th is success requires, however, that at least one incompatible element b e determined.