Global and Regional Dust Forecasting

Forecasting of dust storms is important in order to anticipate, and possibly mitigate their impacts on human health, airplanes, or road traffic. The city of Phoenix, for example, experienced huge dust storms, which caused black-out and deadly accidents on highways. Dust storms are challenging to model due to their short duration and local spatial extent. For this purpose, models need to run at a very fine resolution to be able to resolve the convection explicitly. With an improved version of the NCEP Non-hydrostatic Mesoscale Model coupled online with the BSC-Dust model, developed at the Barcelona Supercomputing Center - NOAA/NCEP - NASA Goddard Institute for Space Studies, this goal could be achieved.

The improvements of NCEP dust prediction capabilities at regional and global levels are made by the implementation of a new global-scale high-resolution (0.1°) mapping of sources and parameters of dust emission based on MODIS satellite retrievals and a new dust particles size distribution at emission. The model calibration strategy presented here is based on of comparisons with several dust observational datasets: Modeled Aerosol Optical Depth and Particulate Matter are compared with ground based measurements, and with MODIS-AOD satellite data. The model performance in representing the dust cycle are evaluated in emission areas in Africa and in North America as well as in remote places, based on climatologic, annual, monthly and daily data. Lastly, the ability of the model to forecast dust storms is discussed.