Surface-sediment dynamics in a dust source from spaceborne multispectral thermal infrared data

Abstract

This study utilized spaceborne multispectral thermal infrared (TIR) data to document spatial relationships of surface sediments over time in a modern depositional environment associated with dust emissions, Soda Lake playa, Mojave Desert, United States. The approach employed here involved time-series TIR data acquired from Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and a linear spectral mixture analysis. An automated algorithm was applied to derive emissivity image endmembers. Evaluation of the chosen endmembers revealed that they can be categorized into five major spectra classes based on diagnostic absorption features. Each spectrum has been identified in relation to mineral abundance and soil arrangement that are common in playa settings: A, “clayey silt-rich crust”; B, “intermediate-salt crust”; C, “quartz-rich deposit”; D, “salt-rich rough crust”; E, “sulfate-rich crust”. Spectral classes A–B–C–D yielded the lowest RMS errors (0–0.025) over time in the iterative deconvolution algorithm between the measured and modeled spectra. The produced fractional abundance images show high areal concentrations for clayey silt-rich crust, salt-rich rough crust, and quartz-rich deposit, as the first surficial mapping of Soda Lake. Significant changes in the spatial relationships of the major surface sediments in Soda Lake were observed after a flooding event (2005) and a relatively dry period (2006). The approach utilized in this study can be advantageous for continuous monitoring of environments characterized by a small area and a complex surface, which may enable a better understanding of their responses to climate changes and potential for dust emissions.