Protecting air travel from volcanic ash in the coming decade

Abstract:

Every day, millions of passengers fly over volcanoes that could erupt and compromise safety. Aircraft encounters in the 1980s prompted the international community to establish nine Volcanic Ash Advisory Centers (VAACs) with protocols to warn of ash clouds. Throughout the 2000s, VAACs detected ash clouds by manually scanning satellite images and communicating with volcano observatories, pilots and others. VAAC forecasters then ran models and manually delineated expected cloud locations as polygons on maps. These polygons, along with information on plume height and start time, were issued as Volcanic Ash Advisories (VAAs). During the 2000s, model source parameters were perceived as a limiting factor in forecast accuracy. There was a view that forecasts should be conservative (i.e. polygons should be large) to ensure safety. That view changed during the 2010 Eyjafjallajökull eruption, when flight restrictions caused ~$5B in economic loss and disruption. In Europe, to mitigate the crisis, regulators changed the format of forecasts from polygons to model-based maps of ash concentration, with 0.2, 2 and 4 mg/m³ thresholds that separate areas of increasingly restricted flight. Stakeholders began touting the importance of accuracy over conservatism. During the 2010s, VAACs formalized a process to unify best practices across regions. NASA conducted engine tests, and Rolls Royce analyzed past aircraft encounters, to better constrain the concentration of ash that jet engines can tolerate, lending some support to the 0.2-4 mg/m³ levels set in 2010. Throughout the 2010s, a new global constellation of geostationary, high-resolution satellites has enabled detection of smaller eruption clouds, nearly doubling the number of VAAs issued per year. Algorithms have matured to automatically detect and warn of clouds seen in satellite, and convert infrared images to maps of cloud mass load. Data assimilation and insertion methods now integrate those maps into models to better match observations. These developments proceed at different rates in different regions, but over the next decade, forecasts are likely to take the form of a 3d, time-changing “data cube” of ash concentration that integrates observations and modeling results. Improved monitoring and observation will remain critical.