Recent observations of the surface of Mars have shown several fresh mid-latitude craters. Some of these craters show exposed ice (Byrne, S. et al. [2009]. Science 325, 1674–1676.). In some craters, albedo of ice slowly decreases, while in others, it remains nearly constant. We attempt to determine influence of the regolith structure on the rate of sublimation of ice. For this purpose we performed numerical simulations describing evolution of the exposed ice in model craters located at middle latitudes.

We consider a new model for the structure and evolution of the material at- and beneath the crater floors. In contrast to the previous study by Dundas and Byrne (Dundas, C.M., Byrne, S. [2010]. Icarus 206, 716–728.) we do not investigate sublimation of dirty ice, and the related formation of a sublimation lag. Instead, we consider sublimation of a pure ice layer on top of layered regolith. In our model the observed reflectivity decreases due to the sublimation-driven changes of the optical properties of thinning clean ice. This offers an alternative to the deposition of the dust embedded in ice (sublimation lag).

We have shown that in our model among many parameters affecting ice sublimation rate, volumetric fraction of water ice in the subsurface beneath the crater has the strongest influence. Hence observed darkening of the ice patch on the crater floor might be sufficient to determine the content of water ice in the subsurface. Our calculations show that an albedo decrease of fresh ice patches in mid-latitude craters can be explained by either strong dust sedimentation or, if this is excluded, by sublimation of a thin layer of water ice from the regolith with large thermal inertia. This is consistent with a large volumetric fraction of water ice beneath the crater floor and contributes to evidence for an extended subsurface water reservoir on Mars.

The overall conclusion of our work is that a thin post-impact surface ice coating over ice-rich ground beneath the crater floors is consistent with the observations.