Quantifying Uncertainty in Daily Temporal Variations of Atmospheric NH3 Emissions Following Application of Chemical Fertilizers

Abstract:

Improving modeling predictions of atmospheric particulate matter and deposition of reactive nitrogen requires representative emission inventories of precursor species, such as ammonia (NH₃). Anthropogenic NH₃ is primarily emitted to the atmosphere from agricultural sources (80-90%) with dominant contributions (56%) from chemical fertilizer usage (CFU) in regions like Midwest USA. Local crop management practices vary spatially and temporally, which influence regional air quality. To model the impact of CFU, NH₃ emission inputs to chemical transport models are obtained from the National Emission Inventory (NEI). NH₃ emissions from CFU are typically estimated by combining annual fertilizer sales data with emission factors. The Sparse Matrix Operator Kernel Emissions (SMOKE) model is used to disaggregate annual emissions to hourly scale using temporal factors. These factors are estimated by apportioning emissions within each crop season in proportion to the nitrogen applied and time-averaged to the hourly scale. Such approach does not reflect influence of CFU for different crops and local weather and soil conditions.

This study provides an alternate approach for estimating temporal factors for NH₃ emissions. The DeNitrification DeComposition (DNDC) model was used to estimate daily variations in NH₃ emissions from CFU at 14 Central Illinois locations for 2002-2011. Weather, crop and soil data were provided as inputs. A method was developed to estimate site level CFU by combining planting and harvesting dates, nitrogen management and fertilizer sales data. DNDC results indicated that annual NH₃ emissions were within ±15% of SMOKE estimates. Daily modeled emissions across 10 years followed similar distributions but varied in magnitudes within ±20%. Individual emission peaks on days after CFU were 2.5-8 times greater as compared to existing estimates from SMOKE. By identifying the episodic nature of NH₃ emissions from CFU, this study is expected to provide improvements in predicting atmospheric particulate matter concentrations and deposition of reactive nitrogen.