Simulating atmospheric transport of the 2011 Grímsvötn ash cloud using a data insertion update scheme

Effective modelling of atmospheric volcanic ash dispersion is important to ensure aircraft safety, and has been the subject of much study since the Eyjafjallajökull ash crisis in Europe in 2010. In this paper, a novel modelling method is presented, where the atmospheric transport of the 2011 Grímsvötn ash cloud is simulated using a data insertion update scheme. Output from the volcanic ash transport and dispersion model, NAME, is updated using satellite retrievals and the results of a probabilistic ash, cloud and clear sky classification algorithm. A range of configurations of the scheme are compared with each other, in addition to a simple data insertion method presented in a previous study. Results show that simulations in which ash layer heights and depths are updated using the model output generally perform worse in relation to satellite derived ash coverage and ash column loading than simulations that use satellite-retrieved heights and an assumed layer depth of 1.0 km. Simulated ash column loading and concentration tends to be under-predicted using this update scheme, but the timing of the arrival of the ash cloud at Stockholm is well captured, as shown by comparison with lidar-derived mass concentration profiles. Most of the updated simulations in this comparison make small gains in skill on the simple data insertion scheme.