Multi-sensor/Multi-wavelength Data Fusion Over Steep Volcanic Terrain: Analysis Challenges in the Next Era of Remote Sensing

Rick L. Wessels 1 and Michael S. Ramsey 2

1 U.S. Geological Survey, 2255 N. Gemini Dr., Flagstaff, AZ 86001, United States; rwessels@flagmail.wr.usgs.gov

2 Department of Geology and Planetary Science, University of Pittsburgh, Pittsburgh, PA 15260, United States; ramsey@ivis.eps.pitt.edu

Now that a virtual flood of remote sensing data over many wavelengths and resolutions is becoming globally available, research needs to capitalize on the strengths of several instruments to provide new ways of monitor volcanic activity. Our work focuses on merging visible/near-infrared (VNIR), thermal infrared (TIR) from the US-Japanese Advanced Spaceborne Thermal Emission and Reflectance Radiometer (ASTER) instrument, and synthetic aperture radar (SAR) data sets from AIRSAR, SIR-C/XSAR, and ERS. This work is intended to expand upon an existing thermal infrared (TIR) texture mapping technique, developed by the co-author, to detect and track textural and morphological changes on the surface of silicic domes and flows.

Changes in the surface texture and topography of volcanic domes, such as those observed on Mount Saint Helens and the older domes of Cascade volcanoes, may indicate variations in volatile content or lava extrusion rates. Tracking these textural changes on active domes would provide insight into the eruption dynamics and the probability of future explosive eruptions. However, these textures have been difficult to directly observe due to the dangers of continued activity and their complex interaction with most electromagnetic radiation. Mapping the dome textures with remote sensing data not only provides a measure of safety for field investigators, but is also a potential tool for hazard mitigation.

We have developed techniques for merging multiple remote sensing data sets over the extremely rough terrain of silicic volcanoes. A major focus has been on overcoming coregistration errors from local topography induced geometric distortion. Accurate location of sub-pixel, vesicularity variation and thermal features on a dome will be critical to future monitoring with ASTER (especially with its 90m TIR resolution). We compensate for the geometric distortion for by first creating an accurate base image using a combination of global positioning system (GPS) and Laser Range Finder ground control, high resolution digital elevation models (DEM), and orthorectified aerial photographs. The DEMs are currently generated from air photos (1 to 5 m spatial resolution). Future DEMs will be created using ASTER VNIR bands 3n and 3b (15 m spatial resolution). The individual sensor data are then rectified to the new reference base using triangulation geocoding.

We have field-measured the dome textures at the centimeter to meter scales and have located the pixels on the final multi-layered, geo-corrected images using real-time differential GPS locations. We are using these measurements to evaluate the response of each instrument to a variety of dome textures. New results suggest that the shortest wavelength SAR (X and C bands) responds to the vesicularity textures whereas signal at the longer wavelengths is dominated by the blocky surface. These observations offer the potential to increase the accuracy of the TIR texture mapping using weightings from the SAR observations.

------------------
Submitted to: American Geophysical Union Fall Meeting, (V15 Volcanology 2010: How Will the Science and Practice of Volcanology Change in the Coming Decade?
Date:2000