Topographic Correction Based Retrieval Of Lunar Mineral Abundance

After entering the second period of lunar exploration, Lunar highspectrumexploration becomes a focus of many contries. Their aim is to study their distributionand abundance of the primary minerals on the lunar surface, such as the clinopyroxene,orthopyroxene, plagioclase, olivine, ilmenite, and to provide fundament for theresearch of strata formation and history evolution.The data used for perivious mineral abundance retreval is Clementine, but it hasthe disadvantage of too little bands and low resolution. What’s more, topographiceffect is not considered. Based on these shortages, the M3data obtained byChandrayaan-1satallite is chosed, and before the mineral abundance retrieval, thehighspetrum is corrected to eliminate the topographic effect.After comparing the several topographic correction methods, the Sandmeiermodel is choesd, because it has specific physical significance. The irradiance receivedby the moon target, including solar direct irradiance and adjacent terrain-reflectedirradiance. According to the two parts, two kinds of reflectance are introduced, whichare directional-directional reflectance and the hemispheric-directional reflectance on aslope surface. Their relationship with the horizontal surface directional-directionalreflectance is given. The at-satellite radiance is obtanined by summating the reflectedradiance from lunar surface. Based on this, the flat surface directional-directionalreflectance is deduced. The slope and the aspect are two basic topographic factors.According to them and the solar spectrum, the calculating method about theparameters on the horizontal surface reflectance formula is analysed, including theincident angle and emergence angle on a slope surface, the terrain configuration factor,the solar direct irradiance and the the total irradiance on a horizontal surface, thebinary coefficient indicating whether the pixel is sunlit. Applying the correctionmethod to part region of Sinus Iridium, the result indicates that the high spectralreflectance facing the sun is decreased and low spectral reflectance with back to thesun is compensated. The variation of reflectance has been reduced to some degree.The method for lunar minerals inversion includes mixture spectrum analysis andregression model. When mixture spectrum analysing, the different reflectancerespectively used in the satallite and lab measurement is introduced. Withoutconsidering the backscatter, the Hapke model is used to do the change betweenreflectance and albedo. The linear equation about mineral albedo and abundance isestablished and it is calculated using the least square method. Base on LSCC spectrum and measured minerl abundance, two regression modelsare established, respectively linear and band ratio. The correlation coefficient, standarddeviation and scatterplot chart between the measured values by LSCC and thepredicted values using the models are compared.Both of the mixture spectrum analysis and the chosed regression model arerespectively applied to the Apollo16, and the calculated results are compeared with themeasured abundance. Accroding to the result, the scheme is decided. The method forboth of ilmenite and clinopyroxene is mixture spectrum analysis, and for olivine,orthopyroxene and agglutinate are regression model. The method for plagioclase is thedifference of the plagioclase result by mixture spectrum analysis and the agglutinatecalculated by regression model.According to the methods analysed above, the mineral abundance of SinusIridium is calculated. In order to analyse the topographic effect,the results arecompeared with the mineral abundance before spetrum correction. The conclusion isobtained that topography will affect on the mineral abundance retrieval, but not for themineral distribution. So before mineral abundance being retrived, topography shouldbe considered.