Quantifying Water Ice in Regolith Mixtures via Infrared (1.2–2.5 μm) Reflectance Model Parametrization

Abstract

Water ice deposits occur on inner solar system bodies in sunlit and shadowed terrains. We develop approaches to remotely determine the mixing mode with regolith (areal vs. intimate), ice grain size, and abundance that reveal processes of ice formation, emplacement, and evolution. Given the sensitivity of existing inversion techniques to the accuracy of absolute wavelength‐dependent reflectance, we evaluated and parameterized robust observables between 1.2 and 2.5 μm—band depth and band area at 1.5 and 2.0 μm, and near‐infrared spectral slope—based on Hapke equations, using a data‐derived regolith endmember mixed with modeled ice of varying properties. As a case study, we applied the approach to three lunar polar craters with permanent shadows—de Gerlache, Faustini, and Rozhdestvenskiy—and, in a separate companion publication, to ice on Ceres (Kachmar & Ehlmann, 2026, https://doi.org/10.1029/2025JE009421 ). Our inferred abundances for lunar ice (10–30 wt%, 10–100 μm) are in agreement with the ShadowCam results. The derivation of the mixing mode, grain size, and abundance is robust (<20 wt% difference) to choices or uncertainties about ice crystallinity, temperature, and regolith endmember, and is only moderately sensitive to porosity. In contrast, the model choices for surface illumination (Sun vs. shadow), ice optical constants, and phase behavior (isotropic vs. anisotropic) substantially influence the results. The approach can be applied to any planetary infrared spectroscopic data set, and its promise will be fully realized with refined ice optical constants and laboratory tests with ice‐regolith mixtures at various porosities and observational geometries.