Scenario based volcanic hazard assessment from ash dispersion

Abstract

Before the eruption of Eyjafjallajökull in 2010, the aviation disasters in such an enormous scale have not realistically been expected. Volcanic eruptions and earthquakes are not only rare but also very sudden events of which prediction and preparedness are often impossible. This study present a scenario based strategy to assess the hazard from the volcanic ash dispersion by combining computational simulation of the ash dispersions and ensemble method to reduce the uncertainty. Due to lack of information in advance, a scenario is designed to consider a range of parameters of the eruption such as particulate volume and eruption strength. The meteorological data are generated over past thousands days by numerical weather forecasting model in this study. With the given conditions in each scenario, the volcanic ash dispersion is simulated using Eulerian based scalar transport model. Once the large amount of the results from ash dispersion simulations are compiled, the meteorological conditions are classified to several groups according to meteorological similarity using K-mean method. Since tens of days are grouped to each similar weather condition, a set of ash dispersion results corresponding to a specified similarity group then is ensemble averaged to generate the representative hazard.