Long-term simulation of global dust distribution with the GOCART model: correlation with North Atlantic Oscillation

Global distribution of aeolian dust is simulated from 1981 to 1996 with the Global Ozone Chemistry Aerosol Radiation and Transport (GOCART) model. The results are compared with in situ measurements and satellite data. An index is calculated from the model results and the satellite viewing angles to allow quantitative comparison with the Total ozone mapping spectrometer (TOMS) absorbing aerosol index. The annual budget over the different continents and oceans are analyzed. The simulated annual emission varies from a minimum of 1950 Tg in 1996 to a maximum of 2400 Tg in 1988. Of these emissions, 65% is from North Africa and 25% from Asia. It is found that North America received twice as much dust from other continents than it emits per year. There is no significant trend over the 16-year simulation. The inter-annual variability of dust distribution is analyzed over the North Atlantic and Africa. It is found that in winter a large fraction of the North Atlantic and Africa dust loading is correlated with the North Atlantic Oscillation (NAO) index. It is shown that a controlling factor of such correlation can be attributed to dust emission from the Sahel. The Bodele depression is the major dust source in winter and its inter-annual variability is highly correlated with the NAO. However, the long record of dust concentration measured at Barbados indicates that there is no correlation with the NAO index and surface concentration in winter. Longer simulation should provide the information needed to understand if the effects of the NAO on dust distribution is rather limited or Barbados is at the edge of the affected region. (C) 2003 Elsevier Ltd. All rights reserved.