Improved temporal resolution in process-based modelling of agricultural soil ammonia emissions

An emerging environmental issue in Canada is how to quantify the contribution of agricultural soil emissions of ammonia (NH3) to environmental pollution. Emission inventories are essential to predict these emissions and their subsequent atmospheric transportation, transformation, and deposition. Due to the high spatial and temporal variability associated with NH3 emissions, emission inventories based on measurements become expensive and emission factors lose accuracy. Process-based models are capable of accounting for the complex soil interactions, but current models lack temporal refinement and few models consider NH3 emissions. This paper presents the development of a one-dimensional (vertical), time-dependent model capable of predicting NH3 emissions from a slurry applied to a bare soil. The model is based on chemical, physical and biological relationships that govern soil heat, moisture, and nitrogen movement. Processes considered include convection, diffusion, decomposition, nitrification, denitrification, and surface to atmosphere transport. The model is tested with experimental data from Agriculture and Agri-Food Canada which conducted NH3 measurements following application of dairy cattle slurry to a bare field. An investigation into the sensitivity of emissions to pH and slurry infiltration rate is conducted and model predictions are best fit to measurements based on this investigation. Testing demonstrated the model's ability to predict the large NH3 emissions immediately following application and subsequent emission trends associated with diurnal patterns that emission factors cannot capture. Results showed that model performance could benefit from a more in depth measurement program and empirical or process models of surface pH. Potential exists for the model to become a useful tool in predicting emissions on local, regional, or national scales.