First Airborne Thermal Infrared Hyperspectral Imaging of a Dry Lake: Site Geology

Researchers seek to identify dry lake bed deposits on Mars, if present. Primary tools in this search include the 1996 Mars Global Surveyor Thermal Emission Spectrometer (TES) and the 2001 Mars Odyssey multi-channel radiometer THEMIS. Interpretations are based on a comparison of the recorded signatures to laboratory spectra. Texture affects both the shape and contrast of the spectral signatures [1]. Materials in the field have a diverse range of textures over multiple scales. In contrast, TES/THEMIS interpretations use laboratory spectra measured of a very narrow range of textures. This leaves a critical gap in the search for evaporites on Mars: the TES laboratory data have never been compared to hyperspectral (spectrometer) airborne spectra of dry lake deposits. We present the first such study, and here focus on the impact of texture. Background: Kirkland et al. published the first geologic study that used an airborne thermal infrared hyperspectral imager [1]. It uncovered key variations in spectral signatures of real-world carbonate deposits that alter TES/THEMIS interpretations. Here we extend this approach to cover some dry lake deposits, using spectra described in a companion abstract [2]. Site: Bristol Lake is in the California Mojave desert, near Amboy. Two interesting materials present are sulfates and halite. Fig. 1 shows a geologic map. The labeled geologic units are [3]: Qpsh: Silt, clay, and halite salt (Holocene). Qpsg: Silt, clay, gypsum, and celestite (Holocene) Gypsum caps some wind eroded pedestals. Qps: Silt and clay (Holocene). Qya: Alluvium (Holocene and Pleistocene?). Sand and poorly sorted sandy gravel of angular and subangular clasts derived from adjacent mountains. Spectral data: TES hyperspectral (spectrometer) data covers ~6.5–50 μm in 143 channels. However, TES terrestrial airborne foundation studies use low spectral resolution radiometers (multi-spectral), mainly TIMS (6 bands) or MASTER (10 bands). THEMIS is also a multi-channel radiometer (9 bands). In contrast, we analyze unique hyperspectral (spectrometer) data from SEBASS (7.6–13.5 μm, 128 channels) [2]. Our studies bridge the looming gap between the very restricted laboratory spectra used to interpret TES, and spectra measured in the real-world. Geology: Uncovering textural spectral effects present field materials requires a correlation of airborne or satellite spectrometer data with the surface texture. Bristol Lake