Ash from a sudden volcanic eruption can grind an entire continent’s air traffic to a halt. Europe learned this lesson the hard way when Iceland’s Eyjafjallajökull volcano repeatedly blew its top in the spring of 2010.
Delen
TU Delft PhD graduate Guangliang Fu from the Faculty of Electrical Engineering, Mathematics and Computer Science studied the eruption while conducting research for his doctoral thesis. The incident, which happened with little warning, shut down airspace across Europe and the North Atlantic.
“Many flights were cancelled and it cost a lot of money,” Fu said. “But it created a fantastic research question regarding ash forecasts.”
The aviation community was largely unprepared for the eruption and their inability to predict where the ash would spread eventually resulted in a global economic loss of nearly $5 billion US. In the years since the incident, researchers like Fu have made efforts to improve volcanic ash forecasts and create computer modelling systems that can quickly estimate how a volcano will impact air traffic in the hours and days following an eruption.
Fu’s research suggests that a method called ensemble-based data assimilation could do the trick. It involves the sequential use of direct measurements to create accurate conditions for the development of predictive models. It’s currently one of the most commonly featured approaches utilised in real-time forecasting applications.
The further a volcanic ash plume travels away from an eruption site, the more difficult it becomes to effectively track and monitor its behaviour. Fu thinks that error rates could be reduced by the incorporation of a certain model of aircraft designed to monitor them. Sending these airplanes out once a day after an eruption to take measurements and directly monitor these plumes could greatly improve forecasts.
Currently, the aviation community often relies on measurements from satellites but they can be lacking. Their data is often presented as two-dimensional and typically isn’t compatible with three-dimensional ash models. Ideally, combining the two could create a detailed forecast that would be valid and accurate for up to half a day, if not longer. One method has been created but thus far its accuracy remains in question. Getting the two to work together well could require at least a few more years of research.
Fu, G., Improving Volcanic Ash Forecasts With Ensemble-Based Data Assimilation, Promoters: Heemink, A.W. and Lin, H.X., Defence: January 10, 2016
Comments are closed.