Impact of interannual weather variation on ammonia emissions and concentrations in Germany

Ammonia is one of the most impactful pollutants emitted from agricultural activities, harming human health and contributing to biodiversity loss. In ammonia emission inventories, the spatial distribution of annual emissions is mostly approximated by constant empirical emission fractions, which do not account for spatial variability, nor for temporal variability within a year or between years caused by weather variations. Besides, factors like manure properties, soil properties, and manure application techniques also lead to differences in the amount of ammonia emitted into the atmosphere. By not or only partly accounting for these factors, significant uncertainties are introduced into ammonia emission estimates at regional and national scales. In this study, we applied the empirical ALFAM2 model to derive spatially explicit slurry application emission fractions from cropland for use in the large-scale INTEGRATOR model, using the information on slurry properties (dry matter content and pH), manure application rate, application technique, incorporation time, air temperature, wind speed, and rainfall rate. In addition, the impact of weather on the ammonia emissions from animal housing and manure storage systems was included through a temperature-dependent scaling. We applied the method to investigate the year-to-year spatio-temporal variabilities of ammonia emissions and modeled concentrations across Germany from 2015 to 2018. Through the comparison with in situ measurements and satellite-derived observations, we studied how surface concentrations and total columns relate to local meteorology. We found that the spatio-temporal variability in emission fractions improves the ability to reproduce the interannual variability observed in ammonia concentration and total column measurements. This study shows that the developed approach to derive spatially explicit emission fractions can significantly improve ammonia emission modeling and is of great importance for studying the temporal variability between years.