Automated Detection and Mapping of Volcanic Plumes for ASTER, MODIS, VIIRS, and SBG

Abstract:

The key science objectives of the planned Surface Biology and Geology Mission include (1) measuring the gas and ash emissions of the Earth’s entire inventory of active volcanoes before, during, and after eruptions, and (2) assessing the impact of volcanic materials on atmospheric composition following an eruption. We will map the composition of volcanic gas emissions and eruption plumes with multispectral thermal infrared (TIR) radiance measurements acquired by SBG, based on retrieval procedures developed to analyze the TIR data acquired by ASTER, MODIS, and VIIRS. Currently, the retrieval procedures are components of an interactive toolkit for mapping volcanic plumes with high-performance radiative transfer (RT) modeling.

To process the large volumes of data anticipated from SBG, we are adapting the interactive retrieval procedures for automated operations. The numerical performance of the retrieval procedures improves with limitations on the number of calls to the computationally-expensive RT model. We have achieved significant improvements in performance through the factoring of radiance spectra into surface and atmospheric components, and caching the atmospheric components (transmission, upwelling, and downwelling spectra) for re-use within a scene where similar input parameters to the RT model are needed. In addition, the factoring enables us to estimate surface temperature and gas concentration in separate steps. We utilize the temperature estimates to screen for clouds and detect the presence of volcanic plumes, further reducing the number of calls to the RT model required to process a scene.

We are evaluating additional strategies for improving the performance of the retrieval procedures. One strategy is to calculate only transmission spectra when estimating gas concentrations, re-using the upwelling and downwelling spectra saved during surface temperature estimation. Another strategy is to generate cloud and plume masks without RT modeling, based on brightness temperatures (BT) and BT differences, and then import these masks to the RT-based retrieval procedures.

This research was performed at the Jet Propulsion Laboratory, California Institute of Technology, under contract to the National Aeronautics and Space Administration. Government sponsorship acknowledged.