Frequency Based Volcanic Activity Detection through Remotely Sensed Data
Abstract
Satellite remote sensing has proved to offer a useful and relatively inexpensive method for monitoring large areas where field work is logistically unrealistic, and potentially dangerous. Current sensors are able to detect the majority of explosive volcanic activity; those that tend to effect and represent larger scale changes in the volcanic systems, eventually relating to ash producing periods of extended eruptive activity, and effusive activity. As new spaceborne sensors are developed, the ability to detect activity improves so that a system to gauge the frequency of volcanic activity can be used as a useful monitoring tool. Four volcanoes were chosen for development and testing of a method to monitor explosive activity: Stromboli (Italy); Shishaldin and Cleveland (Alaska, USA); and Karymsky (Kamchatka, Russia). Each volcano studied had similar but unique signatures of pre-cursory and eruptive activity. This study has shown that this monitoring tool could be applied to a wide range of volcanoes and still produce useful and robust data. Our method deals specifically with the detection of small scale explosive activity. The method described here could be useful in an operational setting, especially at remote volcanoes that have the potential to impact populations, infrastructure, and the aviation community. A number of important factors will affect the validity of application of this method. They are: (1) the availability of a continuous and continually populated dataset; (2) appropriate and reasonable sensor resolutions; (3) a recorded history of the volcano's previous activity; and, if available, (4) some ground-based monitoring system. We aim to develop the method further to be able to capture and evaluate the frequency of other volcanic processes such as lava flows, phreatomagmatic eruptions and dome growth and collapse. The work shown here has served to illustrate the capability of this method and monitoring tool for use at remote, un-instrumented volcanoes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2015
- Bibcode:
- 2015AGUFM.V23A3085W
- Keywords:
-
- 0555 Neural networks;
- fuzzy logic;
- machine learning;
- COMPUTATIONAL GEOPHYSICS;
- 4315 Monitoring;
- forecasting;
- prediction;
- NATURAL HAZARDS;
- 8414 Eruption mechanisms and flow emplacement;
- VOLCANOLOGY;
- 8419 Volcano monitoring;
- VOLCANOLOGY