In-Situ Segregation of Ground Ice on Mars

Several lines of evidence indicate the presence of nearly pure, segregated ground ice in the martian high latitudes. In particular, shallow ice containing only 1-2% soil was excavated by Phoenix. One hypothesis for the excess ice is that it developed in situ, via a mechanism analogous to terrestrial ice lenses. Problematically, terrestrial soil-ice segregation is driven by freeze/thaw cycles, which have not occurred recently on Mars. Here we investigate ice lens formation at T < 273 K, with attention to the possibility on interannual accumulation of segregated ice, and the effects of salts. We developed a numerical model that applies premelting physics to track phase partitioning and lens growth on Mars. The model balances forces arising from intermolecular interactions against gravity and overburden pressure. Thin films of premelted ice minimize the interfacial free energy between ice and soil particles, leading to strong repulsive forces that are ultimately responsible for frost heave. In a freezing soil, gravity and the repulsive intermolecular forces are balanced by the force transmitted vertically between soil grains. Integrating the force balance equation downward from the surface, we identify layers in which interparticle pressures become negative. At those depths the interparticle forces unload, initiating lens formation. Then, given circumstances in which lens initiation is indicated, we ask how quickly lenses grow, how long growth accumulates, and how rapidly lenses are destroyed. We have modeled the last 10⁶ years, assuming two different soils (silt and clay), ice depth determined by vapor-phase equilibration, and, (initially), salt-free liquid phases. Although intermolecular forces are frequently capable of unloading soil grains, rates of vertical H₂O transport typically limit lens growth to << 1 μm/year, while ice table migration due to vapor phase transport might average a few μm/year. Thus, with the possible exception of a single episode at ~ 630 ka bp, vapor phase exchange with the atmosphere would be expected to outstrip and prevent in situ segregated ice lens formation in a salt-free soil. (Earlier in Mars' history however, we find that warmer temperatures frequently lead to macroscopic lens development.) The soils measured by Phoenix of course were not salt-free; in particular the presence of per-chlorate argues that our model must be expanded. The inclusion of even a single salt has multiple complicating effects. For saturated Mg(ClO₄)₂ solutions, the eutectic temperature is as low as 206 K. The resulting decrease in pore ice at low T leads to higher hydraulic permeability which would enhance growth rates. Liquid phase density increases, which increases the buoyancy forces on soil grains. Conversely, dynamic viscosity also increases, inhibiting lens growth. Both the heat capacity and thermal conductivity of the melt are also affected. At high concentrations, osmotic potentials begin to play a role in determining the movement of melt. Finally, all of these properties evolve continually with temperature, as the composition of the liquid phase changes. Understanding the ways in which these processes might affect in situ segregation of martian ground ice is a challenging and exciting undertaking.