Though Mars is a cold, dry planet, with respect to the thermal stability of liquid water at low altitudes it is not terribly different from comparably cold places on Earth. In dry air such water would evaporate faster on Mars, at a rate comparable to a 60 C hot spring on Earth, but the heat loss associated with that evaporation would be mitigated by the poor thermal convection in the thin Martian air. Even at higher altitudes where the atmospheric pressure does not reach the triple point of water, liquid water might theoretically exist in a low-vapor pressure form such as wet soil, in a briny solution, or simply under a layer of dust or snow. The theoretical stability of liquid water does not suggest its occurrence, either on Mars or in Antarctica. In fact, global models have suggested that locations capable of providing sufficient heat for melting are, precisely for that reason, too dry for water to be present. However, the temperature of irregular local structures such as trenches or craters can be markedly warmer than those of the uniform surfaces of global models. The work described here suggests a plausible scenario in which seasonal liquid water might be produced locally, in sheltered locations, through a process of condensation, cold-trapping, buffering, and melting. While the amounts produced in the present climate would be small, copious amounts of meltwater may have been produced at other phases of the orbital cycle, as recently as 20,000 years ago.
Lunar and Planetary Science and Exploration
Lunar and Planetary Science XXXIV; LPI-Contrib-1156