(Keynote Presentation) Spatial-Temporal Aerosol Composition Trends Over the North Atlantic and the Relationship to African Dust Transport

The Atmosphere/Ocean Chemistry Experiment (AEROCE) was a comprehensive multi-disciplinary and multi-institutional research program that focused on the atmospheric chemistry over the North Atlantic Ocean (NAO). Of particular interest was the impact of aerosols transported from the continents to the oceans. To this end, sampling stations were established on Barbados, Miami, Bermuda, Tenerife, and Mace Head (Ireland), and daily samples were collected from 1987 to 1996. This data set is unique in its temporal and spatial coverage. Various aspects of these data have been presented in past publications. Here we bring these data together to provide a broad overview of dust over the Atlantic. Our main focus is on the elemental composition of dust at Barbados where we have the most complete sample coverage: over the 2327 days of the program, we acquired 2163 samples. The completeness of this record enables us to look for changes in concentration and composition that might be linked to sources in Africa; it also enables us to identify impacts that might be linked to other sources including pollution. Previous studies carried out in proximity to African sources (e.g., in the Mediterranean or along the coast of Africa) have shown changes in composition that were attributed to specific sources. Also it is known that on larger scales, the locus of dust activity changes with season. In late winter and spring, dust sources in the Sahel are most active while in summer and fall, the Saharan sources become more persistently active; thus we might expect the dusts' composition to change seasonally to reflect changes in large-scale source terrains. However even if the emitted dust did change for these reasons, it is not clear that the signatures would be sufficiently unique that they would be retained after the dust had been transported over larger distances because of the mixing that might occur during transit or because of the settling-out of larger particles that might carry the source signature. In our study, samples were collected using high-volume, bulk, filter samplers and analyzed by instrumental neutron activation for a suite of 28 elements: Al, Br, Ca, Cl, Cu, I, Mg, Mn, Na, V, Ba, Co, Cr, Cs, Eu, Fe, Hf, Nd, Ni, Rb, Sb, Sc, Se, Ta, Tb, Th, Yb, Zn. For the most part, the elemental concentrations were highly correlated with total dust concentration. Most elements were present in concentrations that were close to average crustal abundances. A number of elements were distinctly different in that their concentrations were much higher than crustal and the scatter poorly correlated with dust concentration (or with dust proxy elements, e.g., Al, Fe, Sc). These were the usual pollution-linked elements: e.g., Sb, Se, V, and Zn. Furthermore when we stratified the data by season, we saw no substantial change in these relationships. These findings suggest dust-aerosol composition over the Atlantic is, in general, relatively invariant, as will be the dust that is deposited to the ocean. Thus we might expect that it will be difficult to extract source-related information on long-term changes in African dust emissions based on elemental composition of sediments from the remote Atlantic. These statements do not necessarily apply to sediments closer to the coast of West Africa where specific source signatures have indeed been observed.