Effects of urban dust emissions on fine and coarse PM levels and composition

Abstract

Simulation of the concentration of urban dust and its interactions with other inorganic components of particulate matter (nitrate, ammonium, sulfate) remains a major modeling challenge. In this study the chemical transport model PMCAMx is applied over Europe during the EUCAARI 2008 summer campaign. The default emission inventory leads to significant underprediction of PM10 levels in both urban and rural areas. We test the hypothesis that this is due to a large extent to an underprediction of urban dust by increasing the corresponding emissions by a factor of ten. This leads to improved PM10 predictions in all sites in Europe and especially in urban areas. The PM10 fractional bias of the model decreased by 23% and the fractional error by 13%. Average predicted PM1 nitrate decreases on average by 20% over the modeling domain due to the increased urban dust. Fine nitrate reductions of 0.5–1 μg m−3 are predicted for areas of the Netherlands, Belgium, France and United Kingdom. Fine ammonium and sulfate also decrease by 9% and 7.5% respectively, when the higher urban dust emissions are used. At the same time predicted coarse nitrate concentrations increase on average by 10% with absolute increases of 1.2 μg m−3 in the Netherlands and Belgium and 1 μg m−3 in Paris. Coarse ammonium and sulfate increase by 16% and 6% respectively due to the higher levels of urban dust.