Gullies and debris flows on Mars: Liquid water or dry ice phenomena???

Dark spots attributed to dry ice along gullies in Lyell crater.

In the January 2016 issue of Nature Geoscience, there is one paper and one News and Views commentary that take on the currently popular idea that gullies and debris flows on Mars are caused by liquid water. These two papers advocate a major role for dry ice, CO2. (Unfortunately, neither article references an important paper by Troy Shinbrot and colleagues in 2004 in which the title says a lot: "Dry granular flows can generate surface features resembling those seen in Martian gullies," PNAS, 101(23), 5 pages. This model appears relevant to the mechanism in part (b) of the cartoon below.)
      The commentary by Colin Dundas (Nature Geoscience 9, pp. 10-11, 2016) sets the stage by summarizing the context. Although the gullies look like terrestrial landforms caused by running water, there is a real problem finding the source of the putative water. Groundwater discharge is frequently cited, but this is inconsistent with the occurrence of gullies on sand dunes and isolated peaks. "Whatever the water source, wet models imply the repeated occurrence of thousands of cubic meters of liquid water at each gully, which would have profound implications for both climate and possible biology on Mars." Over the past two decades photographs of the surface have documented channel erosion and the deposition of debris flows in locations where the present climate is too cold for substantial liquid water. However, many (most?) of these locations occur where seasonal CO2 frost occurs, and gully activity occurs mainly in the winter and spring when CO2 frost is observed on the slopes and is available for participating in gully formation.
    

      In the major article, Pilorget and Forget (Nature Geoscience, 9, 65-69, 2016) develop the model with numerical simulations. Referring to the second figure here, the model is as follows: (a) At the end of winter, the H2O-ice cemented soil (dark blue) is overlain by a regolith and a mantle of CO2 ice (light blue) that is ~1 m thick at the poles and thinner toward mid-latitudes. As the solar intensity increases solar rays penetrate through the CO2 ice and are absorbed by the regolith. This leads to a pressure and temperature increase in the regolith, and the formation of CO2 ice in the pores (light blue circles) of the regolith. (b) When the pressure in the regolith reaches and slightly exceeds the cryostatic pressure of the CO2 ice layer, it rises and cracks, releasing the pressurized CO2. Any remaining CO2 ice in the pores quickly sublimates, and the now-mostly-dry regolith destabilizes and flows out forming viscous granular debris flows because the cryostatic pressure is sufficient to mobilize the grains. (c) Flow of the material thins the regolith, destabilizing the H2O ice underneath and it loses the H2O through sublimation. (d) The process results in the local ice table moving deeper into the crust and the formation of incisions that become, or enhance the formation of, gullies.
     The model used is one-dimensional and examines the evolution of a column that consists of a regolith underlying a CO2 ice layer and an atmosphere. The atmosphere is in radiative-convective equilibrium and the incident radiation on slopes of varying angles is computed. In the CO2 ice layer and the regolith, heat conduction and radiative transfer through the ice are calculated, as well as diffusion, condensation and sublimation of CO2 and the latent heat exchanges (all described in the Methods section). CO2 is predicted to condense above 50 degrees latitude on flat surfaces and down to ~30 degrees latitude on pole-facing slopes.  In such locations, subsurface H2O-ice in equilibrium with the atmospheric water vapor is expected to be present below a layer of dry regolith ranging up to several centimeters thickness. This justified treating the model regolith as a dry porous layer lying above an impermeable, ice-cemented soil.