Ice Accumulation at the Lunar Poles during a Transient Atmospheric Period

Abstract:

Needham and Kring (2017) proposed that the Moon could have had a volcanically induced, collisional atmosphere at ~3.5 Ga composed mainly of CO and S, with roughly 1% H₂O, that would have lasted for ~70 million years. Polar cold trap temperatures were likely low enough for this atmospheric water vapor to condense onto the surface, implying that a significant amount of lunar ice could be atmospheric in origin. We investigate where this water would have condensed onto the surface, the condensation timescales, and its subsequent transport during this atmospheric period.

We created a simple model that couples the lunar atmosphere to an arbitrary number of surface regions with given temperatures that could serve as ice reservoirs over time. For our initial investigation, we used annual maximum temperature maps from the Diviner Lunar Radiometer Experiment onboard the Lunar Reconnaissance Orbiter. The model assumes a collisional atmosphere directly coupled to the surface, with rapid mixing of water into a single atmospheric reservoir. Rapid mixing should occur since the overturn timescale for a global-scale lunar Hadley cell is <<1 yr assuming Mars-like wind speeds. We ran our model for a period of ~10 million years assuming an initial atmospheric mass of ~10¹⁶ kg and a water fraction of 1%. We used this model to produce maps of the distribution and concentration of ice mass throughout the polar regions from 60-90° latitude.

Colder regions accumulated more ice, meaning areally averaged ice mass increased toward the poles. Virtually all of the water condensed onto the surface in 100s of years. However, once the atmospheric water vapor pressure became comparable to the saturation vapor pressures above the surface ice, water began to exchange between ice reservoirs via the atmosphere. This resulted in the slow migration of ice from warmer to colder reservoirs over the rest of the model run. The transport of ice from warmer to colder reservoirs suggests that any ice that was accumulated during this atmospheric period would have been concentrated in the coldest cold traps, even if warmer cold traps were initially able to trap ice. This model may be a useful tool to evaluate the potential atmospheric origin of lunar ices, particularly in cases where lunar ices display regional or pole-to-pole asymmetries (e.g. Rubanenko et al., 2019; Siegler et al., 2016).