The question of whether evaporite minerals exist on Mars is of particular interest because such deposits would mark the locations of potential habitats for past martian life. Currently, the Thermal Emission Spectrometer (TES) experiment on NASA's Mars Global Surveyor spacecraft is mapping the mineralogy of the entire surface of Mars at a spatial resolution of 3km and a spectral resolution of 10cm^-1 in the 6-50 um region. To date, TES has not detected any signatures of carbonates, sulfates, or other evaporite minerals. This year, NASA launched another experiment, the Thermal Emission Imaging System (THEMIS) to map the surface of Mars in 8 broad bands over the 7-14 um region at a spatial resolution of 100m.

We are using airborne visble/near-infrared/thermal-infrared hyperspectral observations of the playas in Death Valley, CA, to set quantitative thresholds for the spatial and spectral resolutions needed to detect evaporite deposits in natural settings. Ground-truthing of our observations is accomplished by laboratory analyses of samples collected in the field (see Baldridge et al., this volume). Our initial results from using spatially degraded airborne thermal IR data have provided tantalizing hints of an explanation for the failure of TES to detect martian evaporites. A spatial resolution of 3 km is very close to the threshold of detectability for carbonates in the Badwater Basin, and it is beyond the threshold of detectability for sulfate evaporites. Further, this coarse spatial resolution is insufficient to associate specific spectral signatures with evaporite-related visible albedo patterns. At 100m resolution, carbonates and sulfates are readily identified in Badwater, appearing in the classic "bathtub ring" pattern of evaporite basins. The halite deposits found interior to these rings are very difficult to detect at any spatial resolution, owing to this mineral's lack of significant thermal or near infrared spectral features.

Ultimately, the experience we gain by simulating current and upcoming Mars experiments with terrestrial data should help us optimize the process of using orbital remote sensing data to select landing sites for future Mars sample collection missions, in order that may go to the locations on Mars most likely to bear signs of past life.