Evidence for exposed water ice in the Moon's south polar regions from Lunar Reconnaissance Orbiter ultraviolet albedo and temperature measurements

We utilize surface temperature measurements and ultraviolet albedo spectra from the Lunar Reconnaissance Orbiter to test the hypothesis that exposed water frost exists within the Moon's shadowed polar craters, and that temperature controls its concentration and spatial distribution. For locations with annual maximum temperatures Tmax greater than the H2O sublimation temperature of ∼110 K, we find no evidence for exposed water frost, based on the LAMP UV spectra. However, we observe a strong change in spectral behavior at locations perennially below ∼110 K, consistent with cold-trapped ice on the surface. In addition to the temperature association, spectral evidence for water frost comes from the following spectral features: (a) decreasing Lyman-α albedo, (b) decreasing “on-band” (129.57-155.57 nm) albedo, and (c) increasing “off-band” (155.57-189.57 nm) albedo. All of these features are consistent with the UV spectrum of water ice, and are expected for water ice layers >∼100 nm in thickness. High regolith porosity, which would darken the surface at all wavelengths, cannot alone explain the observed spectral changes at low temperatures. Given the observed LAMP off-band/on-band albedo ratios at a spatial scale of 250 m, the range of water ice concentrations within the cold traps with Tmax < 110 K is ∼0.1-2.0% by mass, if the ice is intimately mixed with dry regolith. If pure water ice is exposed instead, then up to ∼10% of the surface area on the 250-m scale of the measurements may be ice-covered. The observed distribution of exposed water ice is highly heterogeneous, with some cold traps <110 K having little to no apparent water frost, and others with a significant amount of water frost. As noted by Gladstone et al. (Gladstone, G.R. et al. [2012]. J. Geophys. Res.: Planets 117(E12)), this heterogeneity may be a consequence of the fact that the net supply rate of H2O molecules to the lunar poles is very similar to the net destruction rate within the cold traps. However, an observed increase in apparent H2O abundance with decreasing temperature from ∼110 K to 65 K suggests that destruction of surface frosts by impact gardening and space weathering is spatially heterogeneous. We find a loosely bimodal distribution of apparent ice concentrations with temperature, possibly due to competition between vertical mixing by impact gardening and resupply of H2O by vapor diffusion at sites ∼110 K. Finally, we cannot rule out the possibility that the colder population of ice deposits is in fact primarily carbon dioxide ice, although peak temperatures of ∼65 K are slightly higher than the usual CO2 sublimation temperature of ∼60 K.