Dust Plume Position and Transport in the NASA GEOS-4 Model Compared With MODIS, TOMS, AERONET, and Ground-Based Measurements

Annually, about 240 Tg of Saharan dust is transported across the Atlantic Ocean. Dust deposited to the surface provides nutrients that support the aquatic ecosystems of the Atlantic Ocean as well as the terrestrial ecosystems of South America, the Caribbean, and North America. Additionally, the long-range transport of dust is a potential source of pollutant particulate matter. Dust has an influence on the Earth radiation budget through scattering and absorption of radiation and modification of cloud properties. These radiative feedbacks potentially influence the cyclogenesis occurring off the west African coast. In order to investigate the impacts of dust on the Earth system we employ a newly developed dust model implemented in the NASA GEOS-4 general circulation model and data assimilation system. Here we present a 27 year simulation (1979-2005) of dust transport. We find that our model consistently transports dust further south in the Caribbean then is seen in space-based imagery from the Moderate Resolution Imaging Spectroradiometer (MODIS) and Total Ozone Mapping Spectrometer (TOMS). This misplacement of the dust plume is also found in our comparisons to in situ surface measurements of dust mass concentration. Our model accurately simulates surface dust mass concentrations in Barbados (South Caribbean), but inaccurately simulates concentrations in Bermuda (Atlantic) and Miami (North Caribbean), particularly in summer months. We suspect that our model is not accurately simulating the lifting processes for dust over Africa, so that the dust is not positioned properly in the vertical to be correctly transported westward within the Saharan Air Layer (SAL) and the African Easterly Jet (AEJ). Here we consider constraints provided by MODIS and TOMS, ground-based sun photometer observations from the Aerosol Robotic Network (AERONET), and in situ surface measurements of dust mass concentration to evaluate the location of the dust plume in our model simulations. Further investigation employs backward-trajectory models to evaluate air parcel history to investigate systematic errors in the large- scale transport dynamics within the model.