Identifying African dust sources that contribute to the seasonal cycles of dust transport to the Caribbean Basin and South America.

Abstract:

Decades of aerosol measurements on Barbados have yielded a detailed picture of African mineral dust transport to the Caribbean Basin that shows a strong seasonal cycle with a maximum in boreal summer and a minimum in winter. Recently Prospero et al. (Global Biogeochemical Cycles, 2014) presented 10 years (2002 – 2011) of aerosol measurements made at Cayenne, French Guiana, along with concurrent dust measurements on Barbados. The Cayenne study, coupled with satellite products and other evidence, shows that during spring African dust is carried to a broad region of northeastern South America in quantities comparable to, or greater than, those measured at Barbados in summer. Various lines of evidence suggest that the sources that impact on Cayenne in spring are mainly in the Sahel region, including the Bodélé Depression. In summer transport to Barbados is believed to be most affected by emissions that lie in more northerly regions. Thus the record of measurements at Cayenne and Barbados provide a data set that could be used to test the ability of dust transport models to replicate the seasonal shift of dust sources and the consequent impact on transport to these two sites.

Here we attempt to link the measurements at Cayenne and Barbados to specific source regions using the GFDL global climate model (Donner et al., 2011) which simulates aerosol mass distributions for dust and other aerosol components. Winds are nudged with the NCEP re-analysis as in Li et al. (2008). The model is run repeatedly over the years 1999-2010, activating dust sources in only one North African country in each run (e.g., Mali, Mauritania, Algeria, Niger, etc.). The model accurately depicts the strong seasonal contrast in dust transport to Barbados and Cayenne and shows the changing impact of African sources over the course of the year. In our presentation we will discuss the model results and compare them to the measurements at the receptor sites. It is notable that during the dust seasons at these receptor sites, dust has a dramatic impact on PM10 concentrations so that the average 24-hour PM10 concentrations frequently exceed the World Health Organization air quality guideline, 50 µg m^-3. Our model results will be discussed in this context as well.