Solar Wind Proton Implantation and Hydrogen Diffusion: Model Comparisons to M3 Observations

Abstract

The Moon Mineralogy Mapper (M3) observed a widespread 3-micron absorption feature on the lunar surface, commonly attributed to surficial OH and/or H2O. The feature varies with latitude, local time, and lunar phase, suggesting that at least part of the optically active reservoir is dynamic. We compare phase- and local-time-resolved M3 abundance estimates with a global solar wind proton implantation, hydrogen diffusion, and H2 exosphere model. The model tracks implanted H in the upper grain rim, thermally activated retention and loss, recombinative H2 release, and phase-driven source variations associated with the solar wind, magnetosheath, and magnetotail. We compute the surface abundance using a weighted activation-energy-bin representation, in which the activation-energy distribution is discretized into fixed probability-weighted bins rather than sampled stochastically by Monte Carlo particles. This implementation reduces particle sampling noise in low-flux regions. A refined grid of Gaussian effective activation-energy distributions shows that the low- and mid-latitude M3 trends are best matched by distributions centered near Ec = 0.520 eV with width Ew = 0.090 eV. The preferred case gives a combined low- and mid-latitude RMSE of approximately 27 ppm and a small mean bias of only a few ppm. These fitted parameters should be interpreted as an effective response of the M3-sensitive retained H/OH reservoir, not as a unique mineralogical activation energy. The model reproduces the dominant local-time and phase-dependent abundance structure, including the reduced retained abundance near local noon, while model-data differences near plasma-transition intervals and high-latitude behavior may reflect additional sampling, saturation, or surface-plasma processes.