Formation of nitrate and sulfate in the plume of Berlin

Secondary inorganic aerosol (SIA), i.e. particulate sulphate (S(VI)), ammonium and nitrate (N(V)), is formed from gaseous precursors, i.e. sulfur dioxide (S(IV)), ammonia and nitrogen oxides, in polluted air on the time-scale of hours to days. Besides particulate ammonium and nitrate, the respective gaseous species ammonia and nitric acid can be formed as well. SIA contributes significantly to elevated levels of respirable particulate matter in urban areas and in strongly anthropogenically influenced air in general.

The near-ground aerosol chemical composition was studied at two stationary sites in the vicinity of Berlin during a field campaign in the summer of 1998. By means of analysis of the wind field, two episodes were identified which allow one to study changes within individual air masses during transport, i.e. a Lagrangian type of experiment, with one station being upwind and the other downwind of the city. By reference to a passive tracer (Na+) and estimates on dry depositional losses, the influences of dispersion and mixing on concentration changes can be eliminated from the data analysis.

Chemical changes were observed in N(-III), N(V) and S(VI) species. SIA, i.e. N(V) and S(VI), was formed from emissions in the city within a few hours. Furthermore, the significance of emissions in the city was confirmed by the lacking SIA formation in the case of transport around the city. For the two episodes, SIA formation rates could be derived, albeit not more precisely than by an order of magnitude. N(V) formation rates were between 1.4 and 20 and between 1.9 and 59% h(-1) on the two days, respectively, and S(VI) formation rates were >17 and >10% h(-1). The area south of the city was identified as a source of ammonia.

The probability of occurrence of situations during which the downwind site (50 km downwind of Berlin) would be hit by an urban plume is > 7.4%. Furthermore, for the general case of rural areas in Germany, it is estimated that there is a significant probability to be hit by an urban plume (>8%, corresponding to >1 month per year) for more than half of these. The S(VI) formation rates are higher than explainable by homogeneous, gas-phase chemistry and suggest the involvement of heterogeneous reactions of aerosol particles.

The possible contribution of heterogeneous processes to S(VI) formation should be addressed in laboratory studies. Measurements at more than two sites could improve the potential of Lagrangian field experiments for the quantification of atmospheric chemical transformations, if a second downwind site is chosen in such a way that, at least under particularly stable conditions, measurements there are representative for the source area.