Surface slope characteristics from Thermal Emission Spectrometer emission phase function observations

It is possible to obtain surface roughness characteristics, by measuring a single surface from multiple emission angles and azimuths in the thermal infrared. Surfaces will have different temperatures depending on their orientation relative to the sun. A different proportion of sunlit versus shaded surfaces will be in the field of view based on the viewing orientation, resulting in apparent temperature differences. This difference in temperature can be utilized to calculate the slope characteristics for the observed area. This technique can be useful for determining surface slope characteristics not resolvable by orbital imagery. There are two main components to this model, a surface DEM, in this case a synthetic, two dimensional sine wave surface, and a thermal model (provided by H. Kieffer). Using albedo, solar longitude, slope, azimuth, along with several other parameters, the temperature for each cell of the DEM is calculated using the thermal model. A temperature is then predicted using the same observation geometries as the Thermal Emission Spectrometer (TES) observations. A temperature difference is calculated for the two complementary viewing azimuths and emission angles from the DEM. These values are then compared to the observed temperature difference to determine the surface slope. This method has been applied to TES Emission Phase Function (EPF) observations for both the spectrometer and bolometer data, with a footprint size of 10s of kilometers. These specialized types of TES observations measure nearly the same surface from several angles. Accurate surface kinetic temperatures are obtained after the application of an atmospheric correction for the TES bolometer and/or spectrometer. Initial results include an application to the northern circumpolar dunes. An average maximum slope of ~33 degrees has been obtained, which makes physical sense since this is near the angle of repose for sand sized particles. There is some scatter in the data from separate observations, which may be due to the large footprint size. This technique can be better understood and characterized by correlation with high resolution imagery. Several different surface maps will also be tested in addition to the two dimensional sine wave surface. Finally, by modeling the thermal effects on different particle sizes and land forms, we can further interpret the scale of these slopes.