Spatial attribution of sulfate and dust aerosol sources in an urban area using receptor modeling coupled with Lagrangian trajectories

PM10 and PM2.5 (particles with diameter less than 10 mu m and 2.5 mu m, respectively) aerosol samples were collected from 2002 to 2004 in Athens, Greece and analyzed for elements ranging from Na to Pb using X-ray fluorescence. Positive matrix factorization (PMF) was applied to identify and quantify the types of PM10 and PM2.5 sources. The agreement between calculated and measured particle mass concentrations was very good for both aerosol fractions. Mineral soil, road dust and sea salt particles were the most significant types of coarse particles (PM10-2.5) while their contributions to PM2.5 fraction were minimal. Secondary sulfate, diesel particles from shipping and other oil combustion activities and primary traffic exhausts were determined as the most important types of PM2.5 sources. The analysis of normalized average air mass residence times showed weak seasonal patterns on the pathways of air masses prior to their arrival in Athens. Using trajectory regression analysis, the four adjacent regions (less than 500 km) accounted for the largest fraction of fine sulfate and diesel particles. On the contrary, shipping emissions and neutralization by sea salt may explain the high contributions of the regions covering the Mediterranean Sea and Dardanelle straights. The four adjacent regions and central Europe accounted for most of road and mineral dust particles. The transport from northern Africa over Mediterranean Sea also appeared to influence the mineral dust particles.