Using a genetic algorithm to estimate source term parameters of volcanic ash clouds

Ash clouds generated by volcanic eruptions pose a major hazard to aircraft. Consequently, they must be closely monitored on a regular basis. The Federal Aviation Administration (FAA) is responsible for modeling volcanic ash clouds to reroute aircraft effectively and efficiently if needed. Atmospheric transport and dispersion models, used for prediction of cloud movement, rely on accurate knowledge of the source parameters in order to make a prediction about the future state of the cloud. In this study we focus specifically on back-calculating the emission rate which is a measurement of aerosols being pumped into the atmosphere.

A method is presented that applies a genetic algorithm (GA) to observation data in order to back-calculate the source parameters governing the eruption. The GA is an optimization technique where trial source terms, in this case emission rates, are created and evolved through mating and mutation operators. The GA continually evolves the population of trial solutions until it converges on the most likely source term. Observed concentration data are derived from satellite data to determine the observed ash concentrations. A case study is made of the March 2009 eruption of Mt. Redoubt in Alaska.