| The physical characterization of the upper few centimeters to meters of the Martian surface has greatly benefited from remote temperature measurements. Typical grain sizes, rock abundances, subsurface layering, soil cementation, bedrock exposures, and ice compositions have been derived and mapped using temperature data in conjunction with subsurface models of heat conduction. Yet, these models of heat conduction are simplistic, precluding significant advances in the characterization of the physical nature of the Martian surface.;Preliminary results confirm the influence of the temperature on the bulk conductivity, as well as the effect of changing the gas composition. Cemented samples are shown to conduct heat more efficiently than their uncemented counterparts.;A new model of heat conduction for homogeneous particulated media accounting for the grain size, porosity, gas pressure and composition, temperature, and the effect of any cementing phase is presented. The incorporation of the temperature effect on the bulk conductivity results in a distortion of the predicted diurnal and seasonal temperatures when compared to temperatures predicted with a temperature-independent conductivity model. Such distortions have been observed and interpreted to result from subsurface heterogeneities, but they may simply be explained by a temperature-dependency of the thermal inertia, with additional implications on the derived grain sizes. Cements are shown to significantly increase the bulk conductivity of a particulated medium and bond fractions <5% per volume are consistent with Martian thermal inertia data previously hypothesized to correspond to a global duricrust. A laboratory setup has been designed, built, calibrated and used to measure the thermal conductivity of particulated samples in order to test and refine the models mentioned above. |
