Identifying eruptive sources of drifting volcanic ash clouds using back-trajectory modeling of spaceborne thermal infrared data

Abstract

A new approach is presented for determining the source region of disconnected volcanic ash clouds using a combination of orbital thermal infrared (TIR) image data, HYSPLIT-generated backward trajectories, and spatial geostatistics. Interpolated surfaces derived from the TIR data are created to find the most likely ash cloud travel path and the potential source volcanoes identified from that path. The ability to use backward trajectories to determine the ash cloud source region will become an important triggering tool to target high spatial resolution orbital sensors, which normally rely on thermal anomalies for new targeting. During cases where thermal anomalies are not present or masked by meteorological cloud, ash cloud detection and predicted source location become more important. Image data from several well-documented past eruptions are presented to validate and determine the overall accuracy of this technique. Because this is seen as a limited range source region validation tool, the ash clouds examined were < 1000 km from their source vent. The approach and analysis are deemed successful if 80% of the model results produce one or more trajectories that pass within 60 km of the source volcano. This methodology could be improved further with the ability to determine the cloud location more accurately using higher image data frequency, and most importantly, greater accuracy in determining the cloud height in the atmosphere.