Earth-based radar observations and results from the MESSENGER mission have provided strong evidence that permanently shadowed regions near Mercury's poles host deposits of water ice. MESSENGER's complete orbital image and topographic datasets enable Mercury's surface to be observed and modeled under an extensive range of illumination conditions. The shadowed regions of Mercury's north polar region from 65°N to 90°N were mapped by analyzing Mercury Dual Imaging System (MDIS) images and by modeling illumination with Mercury Laser Altimeter (MLA) topographic data. The two independent methods produced strong agreement in identifying shadowed areas. All large radar-bright deposits, those hosted within impact craters ≥6 km in diameter, collocate with regions of shadow identified by both methods. However, only ∼46% of the persistently shadowed areas determined from images and ∼43% of the permanently shadowed areas derived from altimetry host radar-bright materials. Some sizable regions of shadow that do not host radar-bright deposits experience thermal conditions similar to those that do. The shadowed craters that lack radar-bright materials show a relation with longitude that is not related to the thermal environment, suggesting that the Earth-based radar observations of these locations may have been limited by viewing geometry, but it is also possible that water ice in these locations is insulated by anomalously thick lag deposits or that these shadowed regions do not host water ice.

Images acquired by NASA's MESSENGER spacecraft have revealed the morphology of frozen volatiles in Mercury's permanently shadowed polar craters and provide insight into the mode of emplacement and evolution of the polar deposits. The images show extensive, spatially continuous regions with distinctive reflectance properties. A site within Prokofiev crater identified as containing widespread surface water ice exhibits a cratered texture that resembles the neighboring sunlit surface except for its uniformly higher reflectance, indicating that the surficial ice was emplaced after formation of the underlying craters. In areas where water ice is inferred to be present but covered by a thin layer of dark, organic-rich volatile material, regions with uniformly lower reflectance extend to the edges of the shadowed areas and terminate with sharp boundaries. The sharp boundaries indicate that the volatile deposits at Mercury's poles are geologically young, relative to the time scale for lateral mixing by impacts, and either are restored at the surface through an ongoing process or were delivered to the planet recently.

Water ice has been delivered to the lunar poles from different sources over billions of years, but this accumulation was punctuated by large impacts that excavated dry regolith from depth and emplaced it in layers over the poles. Here, we model the resulting stratigraphies of ice and ejecta deposits in the lunar polar regions. Large polar craters were age dated, and their ejecta distributions calculated with standard scaling relations. We then created a Monte Carlo model for ice deposition and ejecta emplacement. Typical model runs showed that deposits in older cold traps (>4 Ga) are divided into two zones: buried ice-rich gigaton deposits and younger more gardened mantles. The latter are consistent with small crater morphometry measurements, but the existence of substantial ice buried at great depths is more difficult to confirm. Rare outlier model runs included Mercury-like cases with significant deposition events in recent history (<200 Ma). Plain Language Summary The polar regions of Earth's Moon have topographic depressions that are never directly exposed to the Sun, so they are cold enough for deposits of ice to exist. Water can get into these regions by water-bearing asteroids colliding with the Moon, or from lunar volcanoes erupting gases that travel to the poles. At the same time, large impact craters that form at the poles eject an enormous amount of soil and rock that could bury existing ice. It is not well understood how these two processes work together to build up deposits that may have alternating layers of ice-rich and ice-poor soil. In this study, we used computer simulations to predict what these layered deposits may look like. We found it is likely that large amounts of relatively pure ice are buried at depth in the oldest deposits, covered with thinner layers hosting less ice. Impact cratering has been the dominant process affecting the lunar poles, but the effects of large polar craters on nearby ice deposits have not been previously addressed. Impact effects have been considered for micrometeoroids (e.g.,