Spectral Unmixing of Aluminum-Rich Clay Minerals with Radiative Transfer Models and Its Implications for Martian Remote Sensing

The relative abundances of aluminum-rich clay minerals such as kaolinite and montmorillonite in Martian weathering profiles reflect variations in chemical weathering intensity and paleoclimatic conditions. Accurate quantification of these minerals is essential for reconstructing early Martian environments and evaluating their potential habitability. Spectral unmixing provides a powerful means to retrieve mineral abundances, yet its accuracy remains uncertain due to limited laboratory validation, particularly for hydrated mineral mixtures with overlapping absorption features. In this study, two sets of laboratory-prepared mixtures (kaolinite–montmorillonite and kaolinite–opal, comprising 12 spectra) were analyzed using two widely applied radiative transfer models for planetary spectroscopy, the Hapke model and the Shkuratov model. The results show that the Shkuratov model achieved abundance errors below 10% in seven samples and between 10–20% in two samples, while the Hapke model produced only three samples below 10% and seven between 10–20%. Overall, the Shkuratov model demonstrated higher inversion accuracy. This work provides an experimental validation framework for quantitative spectral unmixing of Martian aluminum clays and offers methodological guidance for analyzing more complex hydrated mineral mixtures in future study.