V028-0016
A Comparison of Meteorological Deterministic and Ensemble Inputs to HYSPLIT For Volcanic Ash Transport in Small to Moderate Sized Eruptions

Friday, 11 December 2020
Poster
Eric Roy1, Alice Crawford2, Barbara Stunder2 and Binyu Wang3, (1)University of Massachusetts Lowell, Lowell, MA, United States, (2)NOAA Air Resources Laboratory, College Park, MD, United States, (3)IMSG, Rockville, MD, United States
Abstract:
Volcanic ash (VA) can cause a number of malfunctions in modern airplanes including engine failure, and as a result is considered a significant hazard to the aviation industry. In order to mitigate the loss of life and property during VA events, Volcanic Ash Advisory Centers (VAAC) issue Volcanic Ash Advisories (VAA). The HYSPLIT dispersion model is developed and maintained by the NOAA Air Resources Laboratory (ARL) and is commonly used by VAACs for VA modelling. In this study, a comparison of HYSPLIT VA transport model solutions using deterministic and ensemble meteorological inputs for small to moderate sized eruptions was conducted. The focus of this study was on eruptions for several volcanoes in the Washington VAAC region in June and July of 2020.

Forecasting ash transport for small to moderate sized eruptions poses its own set of challenges. Smaller eruptions often have lengthy intermittent and inconsistent emission patterns that are difficult to observe or predict. Relatively small errors in the meteorological wind field and estimated source parameters produce forecasts which often significantly overestimate the extent of the ash cloud and/or show displacement of the plume. These eruptions take up a large portion of the VAAC’s resources. An average of over eight Washington VAAC VAAs per day were issued during the period of this study. By using Global Ensemble Forecast System (GEFS) data as a HYSPLIT input and comparing this to results from deterministic models such as the Global Forecast System (GFS), the effects of uncertainties in the meteorological model and the feasibility of using HYSPLIT for probabilistic and quantitative ash forecasts for small to moderate sized eruptions is explored.