Multi-sensor data fusion for remote sensing of post-eruptive deformation and depositional features at Redoubt Volcano

Monitoring volcanic activity by remote sensing is an essential component of volcanology. Remote sensing includes a variety of different sensing methods and instruments that collect data across a wide range of the electromagnetic spectrum. This study presents an overview of the improvements that are available to remote sensing imaging with multi-sensor and multi-temporal data fusion of optical and radar data, using Redoubt Volcano and related 2009 Drift River lahar deposits as a target area. From ALOS-PRISM data, high resolution DEMs were produced and used to generate elevation change maps of Redoubt Volcano and the Drift River, and to estimate the volcano's dome volume; these DEMs were then fused with ALOS-PALSAR radar images to produce differential interferograms demonstrating the effect of high-resolution DEMs on surface deformation measurements from interferometric radar data; and finally, multi-temporal InSAR coherence data were used to plot the boundaries of lahar flows at the distal end of the Drift River with high accuracy. These techniques demonstrate: (1) how the fusion of data from multiple sensors acquired at multiple temporal intervals can substantially increase the accuracy and precision of remote sensing measurements compared to those from one sensor alone; (2) how data fusion techniques can improve remote sensing change detection in areas otherwise ill-suited for single sensor observations; and (3) how data subject to temporal decorrelation may be used for boundary mapping with high accuracy. In addition to volcanic deformation, these methods can be applied to a number of disciplines, and will become more essential as the number of earth observing satellites increase. (c) 2012 Elsevier B.V. All rights reserved.