ABSTRACT
Thermal infrared (TIR) spectral emissivity is a unique property used to detect, measure, and map the surface composition on Earth and other planetary bodies. Most major rock-forming and clay minerals have diagnostic spectral features in the TIR region. Moreover, the strong absorption coefficients of these minerals produce a linear relationship between their areal abundance and the magnitude of their spectral features. Therefore, linear deconvolution models coupled with spectral libraries can be used to determine both the mineral species and areal percentage for each pixel in a TIR multispectral image. This paper centers on a new surface mineralogy (SM) algorithm developed for global mineralogy and weight percent silica (WPS) mapping. We describe the development and evaluation of the SM algorithm, and its testing on a suite of airborne datasets, to assess detection accuracy and the choice of endmember mineral suites prior to the launch of the next generation of TIR sensors.
