Simulating ammonia loss from surface applied manure

Abstract

The land spreading of manure can result in significant nitrogen (N) losses to the atmosphere through ammonia (NH3) volatilization. It is estimated that agricultural activities, including manure spreading contribute approximately 50% of the total global NH3 emissions. Computer simulation models have been developed in an attempt to predict NH3 losses resulting from manure spreading. Few models have been validated with success and no validation has been completed for conditions found in eastern Canada. The mechanistic model Volt’Air was implemented and tested to estimate NH3 emissions using field-based wind tunnel data from three field locations in eastern Canada. The model was validated using data from 27 wind tunnel trials, which measured NH3 loss over 5-19 d. Sensitivity analysis revealed that Volt’Air was most sensitive to pH. Good overall agreement was observed between measured results and model simulations over 5-19 d (R2 = 0.91, RMSE = 2.9 kg ha-1). Short-term flux simulations (kg ha-1 h-1) within the first 24 h after spreading were underestimated by 29.5% and fluxes on subsequent days were slightly underestimated by 3.7%. Error with Volt’Air appears to be largely confined to the first day following manure application. Volt’Air offers potential for predicting NH3 losses (i.e., 5-19 d) from manure application within eastern Canada. Improving the simulation of short-term pH changes, however, may result in enhanced overall model performance.Key words: Ammonia volatilization, flux, mechanistic model, swine manure, wind tunnels