A Spectroscopic Inversion Study Of The Visible And Near-infrared Reflectance Spectra Of Major Endmembers Mixture Of The Lunar Surface

Mineralogy information is the basis for understanding the properties,origin,and evolution of the moon.Visible and Near-infrared spectroscopy is an important method to study the mineral abundance and components of planetary surfaces.At present,there are many methods for the inversion of lunar surface minerals.After presenting the spectral characteristics of typical lunar minerals and common mineral inversion methods,we used the Lunar Soil Characterization Consortium(LSCC)data to compare the look-up tables and the Gaussian model method(MGM).The conclusions are below: 1)The look-up table can be used to retrieve the plagioclase content.The plagioclase accounts for a large proportion of the lunar surface regolith and knowing its content can help to fully understand the material composition and source of the measurement area.The effect of space weathering is considered in the Look-up table,so the spectra in the Look-up table are closer to the ture state of the regolith on the lunar surface.However,compared with the single fixed value of the size and magnesium number(Fo#)which the look-up table set,the Fo# of the olivine can be estimated using the MGM.The MGM is less dependent on the size,and it is more suitable for the sample with different sizes or contains a large amount of iron-mafic minerals.2)The results of LSCC data show that the relative content of ferromagnesian minerals inversion by using these two methods has similar accuracy(the difference between the real value and the inversion result is ~10%),and the accuracy of the look-up table is slightly better than that of MGM.The soft landing of the Chang’e-4 at the Von Karman crater has provided an important opportunity to understand the compositional distribution of the far side of the lunar surface.Considering the characteristics of the Chang’e-4 data,we employed the Look-up table method to infer the major mineral abundances from measurements.We also used multivariate remote sensing data to analyze the source of the landing site.The analysis results show below: 1)The landing area is dominated by plagioclase(>60%),followed by pyroxene,and the abundance of olivine is the least(<10%).Kaguya’s Multiband Imager(MI)data results are consistent with it.The Fo# of olivine calculated by MGM is 72-82.2)A mineral composition analysis using measurements made by the Moon Mineralogy Mapper(M3)and the MI has shown that the surface materials of the landing site may have been ejected from the nearby Finsen crater.This result is also supported by principal component analysis which showed a strong correlation between the compositions of the landing site and the Finsen crater.3)The high plagioclase abundance at the landing site may indicate the associated olivine is inherent of anorthosites,rather than the Mg-suite troctolites erupted from the interior of the Moon.According to the existing magma ocean hypothesis,the lunar mantle material should contain a large amount of olivine with a high magnesium number.Based on the abundance of olivine and the Fo# in the sample,we thought the materials at the landing site are unlikely olivine-and pyroxene-rich materials from the lunar mantle,or the lunar mantle material is covered by the material produced by subsequent impact melting and redifferentiation.Lunar glasses are among the important endmembers of the lunar surface materials,but are often ignored in the spectral analysis due to their complex composition and structure.In recent years,with the improvement of data space and spectral resolution,it is possible to distinguish the glass from the sample’s spectra.In this paper,we introduced several lunar glass materials and chosen the volcanic glass(which have obvious absorption characteristics)as the endmember material for the mineral inversion.With reference to the Look-up table,the Hapke model was used to construct the new reflectance mixing library using the synthetic glass and the other four minerals.The reflectance mixing library was tested by the LSCC and Reflectance Experiment Laboratory(RELAB)data and found that: 1)Adding the glass as the endmember to construct the reflectance mixing library can get more accurate abundances of minerals.Furthermore,the presence of the glass in the mixture can be tested and the abundance of glass can be retrieved qualitatively.2)The best match spectra over full wavelengths from the reflectance mixing library can get the more accurate retrieved abundances of minerals and glass than the spectra(continuum-removed)from the continuum-removed reflectance mixing library.But in some cases,the spectra of the unknown samples cannot find the best match in our reflectance mixing library because of the shallow absorption features.The continuum-removed spectra can also be used in the spectral analysis and the mineral abundance retrieved from the continuum-removed reflectance mixing library with glass are more accurate.We apply the reflectance mixing library to the Chang’e-3 and Chang’e-4 in-situ measurement data and the M3 data.After dividing the stratigraphic units around Chang’e-5,we retrieved the mineral abundance of these units.From the M3 data analysis of the typical areas,the conclusion is below: 1)The retrieved abundance of plagioclase of Chang’e-3/4 data by using the reflectance mixing library are increased compared with the Look-up table results,while the olivine and pyroxene are decreased.Referring to the results of the Chang’e-3 APXS mineralogy analysis,it is believed that the mineral abundances retrieved by using the reflectance mixing library are more accurate.The glass abundance(~5.7% with green glass)of the landing site of Chang’e-3 obtained from the reflectance mixing library is higher than that CE-4landing site(~2.9% with green glass),but similar to that of the LSCC mare sample(~10.8 wt.% orange glass,~7.1 wt.% green glass).That is consistent with the fact that the Chang’e-3 landing site is a typical lunar mare region.2)Analysis the M3 data of the Chang’e-3 and Chang’e-4 landing area,surrounding units and craters show that the glass abundance near the landing site is significantly lower than the Chang’e-3 and Chang’e-4 data results.It is possible that any glass-rich patches in small scales(~cm)that have been detected by in-situ measurements cannot be resolved by remote sensing measurements(~280 m/pixel for M3).Such difference in spatial resolutions may also have caused the measured mineral abundances difference.3)The glass abundance of the Chang’e-5 landing site is higher than that of the Chang’e-3 and Chang’e-4 landing sites and the plagioclase abundance is lower than that of the Chang’e-3 and Chang’e-4 landing areas.Understanding the mineral composition of the Chang’e-5 landing site through the inversion results of the M3 data can provide background support for the study of returned samples and in-situ measurement data.