Analysis of Shuttle Multispecral Infrared Radiometer measurements of the western Saudi Arabian shield.

Abstract

During the November 12–14, 1981, mission of the space shuttle Columbia, the Shuttle Multispectral Infrared Radiometer (SMIRR) recorded radiances in ten channels along a 100 m wide groundtrack across the western Saudi Arabian shield. The ten channels are located in the 0.5 to 2.4 μm region, with five positioned between 2.0 and 2.40 μm for measuring absorption features that are diagnostic of OH‐bearing and CO 3 ‐ bearing "> CO 3 ‐ bearing minerals. This exceptionally well exposed area consists of late Proterozoic metamorphic, intermediate to silicic intrusive, and interlayered clastic sedimentary and intermediate silicic volcanic rocks that have not been studied previously using SMIRR data. Plots or traces of unnormalized SMIRR channel ratios were examined before field studies to locate areas with high spectral contrast, especially in the 2.0 μm to 2.40 μm channels. Reflectance spectra were measured in the laboratory for rock and soil samples collected in these areas, and the mineralogic causes of the main absorption features were determined using X‐ray diffraction. Laboratory SMIRR spectra were produced by convolving the ten SMIRR filters with the laboratory spectra. Then, normalized SMIRR reflectance spectra were generated along the groundtrack using normalization coefficients calculated for a field sample representing a uniform, low‐spectral contrast area. Field evaluation shows that unnormalized SMIRR ratio traces are useful, even without specific mineralogic information, for distinguishing rocks that are characterized by Al‐OH, Mg‐OH, and/or CO 3 "> CO 3 "> CO 3 , Fe 3 + "> Fe 3 + , and Fe2+ absorption features. Analysis of field samples permits suites of minerals causing absorption features to be identified. However, specific mineral identification cannot be achieved consistently using the SMIRR ratio traces or normalized SMIRR spectra, because the Al‐OH and Mg‐OH absorption features can be caused by more than one of the minerals commonly present. The normalized SMIRR spectra are especially useful for identifying subtle Al‐OH and Mg‐OH absorption features that are difficult to identify in the unnormalized ratio traces and for comparing the relative intensities of absorption features. Al‐OH absorption is related to muscovite, smectite, illite, and kaolinite, whereas Mg‐OH absorption is caused by chlorite, amphibole, and biotite. The principal sources of error in using SMIRR spectral measurements for identifying mineral groups along the orbit 27 groundtrack are inaccuracies in field location and lithologic heterogeneity that is not represented adequately by field samples. Calibration errors may account for systematic albedo and absorption intensity differences between calculated laboratory SMIRR spectra and normalized SMIRR spectra. SMIRR instrument noise and atmospheric factors appear to be less important sources of error. However, as higher spectral and spatial resolution systems are developed for mineral identification, radiometric precision and atmospheric factors will become more important.