Predicting and validating the tracking of a volcanic ash cloud during the 2006 eruption of Mt. Augustine volcano

Abstract

The 2006 eruption of the Augustine Volcano provided a unique opportunity to validate the ash cloud dispersion models being used operationally to provide guidance to aircraft operations. During 28 January-2 February, an ash cloud from Mt. Augustine Volcano traveled over several lidar facilities and aerosol samplers in Alaska. Ash and sulfur dioxide clouds were also detected by satellite-borne sensing platforms. The AVO Remote Sensing group at UAF-GI provided Puff model predictions of the ash cloud trajectory that were validated by these remote-sensing and sampling measurements. A summary of the observations and associated predictions is presented in Table 2. We note that the lidar measurements at two sites (Chatanika and Fairbanks) were collected in response to the 29 January forecast by the Puff model. Samplers detected the presence of aerosols at the surface, which had characteristics of volcanic aerosols, at sites where Puff had predicted the ash clouds would be. Satellite data showed the evolution and movement of the ash cloud in the neighborhood of Mt. Augustine Volcano. In addition, satellite data showed the dispersion of an SO2 cloud further from the volcano, where the ash cloud was too thin to be observed directly using the split-window method for detecting ash from remote-sensing data. The AIRS data showed SO2 where ash was very difficult to observe and, in that the eruption occurred during a high-latitude winter, the OMI and other UV sensors were unable to detect the signal. Lidars verified the presence of volcanic aerosol with consistent spectral/polarization characteristics over Alaska. The lidar signals were capable of detecting the aerosol, even in the presence of water vapor clouds, where the ash cloud is too thin or dispersed to be detected by remote-sensing satellite data. The lidar measurements revealed the different trajectories of lower- and upper-level ash consistent with the Puff predictions. Using this suite of observational support, the utility of the Puff model to accurately forecast the path of a volcano ash cloud across Alaska from Augustine Volcano to Barrow (∼1300 km) has been verified. Dispersion models provide a forecast of volcanic ash movement in both proximal and distal areas that might be undetectable but still represent a potential hazard to aircraft. Validation of the predictions from Puff is key to assessing the accuracy of any future predictions. The study highlights the use of multiple and complementary observations in detecting the trajectory of an ash cloud, both at the surface and aloft within the atmosphere. © 2008 American Meteorological Society.