A Lagrangian model for the emission, transport, transformation, and deposition of NH3 and species derived from NH3 emissions is described. This model also involves SO2 and NOx, since the chemistries of NH3, NO3, NOx, and SO2 are interlinked. The model evaluated by comparing it with measurements of sulfur and nitrogen compounds in air and precipitation during an intensive 3-month period in August--October 1985 at six Scandinavian European Monitoring and Evaluation Programme sites. The concentrations of NO2 are underestimated at all sites except one, and they almost follow the curve for a 1:2 ratio when they are compared to the measured values. This is a consistent feature which is not strongly linked to meteorological variability or emission changes but rather is caused by the chemical formulation of the model and its physical assumptions. For nitrate aerosol there is a systematic overestimation in the model, while for the sum of ammonia and ammonium aerosol the observed average concentration is twice the calculated one, with particularly large discrepancies where there are local ammonia sources.

The average of all observations of SO2 is matched by the calculations, but the regression line has a wrong slope and a nonzero intercept. For sulfate aerosol the calculations give concentrations which are about 50% of the measured values, while for nitrate, ammonium, and sulfate in precipitation, the correlations between observations and calculated values are lower than those for the airborne species. This difference arises in part from the difference between precipitation in the model and as measured at the sites. ¿ American Geophysical Union 1994