Forecasting large explosions at Bezymianny Volcano using thermal satellite data

van Manen, S. M.; Blake, S.; Dehn, J. and Valcic, L. (2013). Forecasting large explosions at Bezymianny Volcano using thermal satellite data. Geological Society, London, Special Publications, 380 pp. 187–201.

DOI: https://doi.org/10.1144/SP380.3

Abstract

Large volcanic explosions pose a severe risk to life and cargo by injecting ash into local and international air traffic routes. Prior to exploding, Bezymianny (Kamchatka) commonly shows an increase in lava extrusion rate, which can be detected by satellites as an increase in thermal radiance. Here we present the first method of forecasting explosive eruptions based solely on satellite data. A pattern recognition algorithm using Advanced Very High Resolution Radiometer (AVHRR) data has been developed based on known precursory trends of increasing radiance prior to 19 explosions at Bezymianny Volcano in 1993–2008. The algorithm retrospectively forecasts 89% of the explosions (100% of the explosions that show precursory increases in thermal radiance), with 71% of alerts issued in the 30 days beforehand. The method also provides the probability of an explosion occurring within a given number of days after an alert is triggered by the algorithm. When applied to independent data, the algorithm correctly provided alerts before the 16 December 2009, 31 May 2010 and 13 April 2011 explosions.