Discrimination between maturity and composition of lunar soils from integrated Clementine UV-visible/near-infrared data:: Application to the Aristarchus Plateau

The reflectance spectrum of a lunar soil is mainly dominated by the composition and the degree of exposure to space weathering processes such as micrometeorite bombardment and solar wind implantation. The spectral alteration effects of space weathering should be removed for accurately investigating the composition of the lunar surface using remote sensing data. In this paper we show that the integration of the Clementine UV-visible (UVVIS) and near-infrared (NIR) channels provides an improved evaluation of the spectral alteration. The depth of the mafic absorption feature at 0.95 mu m is also better defined by combining the UVVIS and NIR data. Laboratory spectra of lunar soil samples indicate that the continuum slope derived from the 1500/750 nm ratio is closely related to the concentration of fine-grained submicroscopic iron (Is). The continuum slope therefore provides an evaluation of the spectral alteration of the surface, which can be subtracted from the 1 or 2 mu m absorption band depths to retrieve compositional information. This method has been applied to the Aristarchus plateau, which exhibits a broad range of mineralogical composition and maturity. A nine-channel multispectral mosaic of 680 Clementine images of the Aristarchus plateau has been processed. Eight telescopic spectra have been used to check the validity of the reduction process for the near infrared bands. The 1 mu m absorption band, once corrected for spectral alteration, provides an evaluation of the initial FeO content in mafic silicates (mafic iron). Lunar soil samples show that it is possible to quantitatively map mafic iron with this technique. Our results are in good agreement with those obtained using the algorithm of Lucey et al. [1995, 1998a], which is based on UVVIS bands alone. The mafic iron content and total iron content which can be derived from the combined UVVIS and NIR data sets are less sensitive to local slopes than that derived from Lucey et al.'s method. This new method could therefore be useful for investigating areas at middle to high latitudes. Removing spectral alteration from the 2000/1500 nm ratio also makes possible a better discrimination between olivine and pyroxene within identified mare basalts on the Aristarchus plateau.