HR: 0800h
AN: V31A-0294    [Abstracts]
TI: Testing and Adapting a Daytime Four Band Satellite Ash Detection Algorithm for Eruptions in Alaska and the Kamchatka Peninsula, Russia
AU: * Andrup-Henriksen, G
EM: gry@gi.alaska.edu
AF: Alaska Volcano Observatory, Geophysical Institute, University of Alaska Fairbanks, 903 Koyukuk, Fairbanks, AK 99775, United States
AU: Skoog, R A
EM: skoog@gi.alaska.edu
AF: Alaska Volcano Observatory, Geophysical Institute, University of Alaska Fairbanks, 903 Koyukuk, Fairbanks, AK 99775, United States
AB: Volcanic ash is detectable from satellite remote sensing due to the differences in spectral signatures compared to meteorological clouds. Recently a new global daytime ash detection algorithm was developed at University of Madison, Wisconsin. The algorithm is based on four spectral bands with the central wavelengths 0.65, 3.75, 11 and 12 micrometers that are common on weather satellite sensors including MODIS, AVHRR, GOES and MTSAT. The initial development of the algorithm was primarily based on MODIS data with global coverage. We have tested it using three years of AVHRR data in Alaska and the Kamchatka Peninsula, Russia. All the AVHRR data have been manually analyzed and recorded into an observational database during the daily monitoring performed by the remote sensing group at the Alaska Volcano Observatory (AVO). By taking the manual observations as accurate we were able to examine the accuracy of the four-channel algorithm for daytime data. The results were also compared to the current automated ash alarm used by AVO, based on the reverse absorption technique, also known as the split window method, with a threshold of -1.7K. This comparison indicates that the four- banded technique has a higher sensitivity to volcanic ash, but a greater number of false alarms. The algorithm was modified to achieve a false alarm rate comparable to current ash alarm while still maintaining increased sensitivity.
DE: 3360 Remote sensing
DE: 8419 Volcano monitoring (7280)
DE: 8485 Remote sensing of volcanoes
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2007 Fall Meeting