Identifying Spatial Heterogeneity in Ammonia Emissions from Agricultural Fertilization

摘要

Improved predictions for regional-scale modeling of atmospheric particulate matter (PM) formation and reactive nitrogen deposition to different ecosystems can be provided by developing improved NH3 emission inventories (EI), especially from agricultural fertilization (AF) and soils at spatial and temporal resolutions required by chemical transport models (CTMs). The Sparse Matrix Operator Kernel Emissions (SMOKE) model produces gridded emission inputs by disaggregating the National Emission Inventory (NEI) data using spatial surrogates and temporal factors to scales required by CTMs. For AF, NEI provides county-level NH_3 estimates produced by annual fertilizer sales data and approximate emission factors. Thus, SMOKE can only produce spatially averaged NH3 emissions that do not capture localized hotspots from crop-specific AF practices and soil-atmosphere interactions. According to NEI, AF contributed 55% of the anthropogenic NH_3 emissions to the atmosphere in Illinois (IL) in 2008. We have developed a high-spatial resolution; 4 km × 4 km EI for AF based NH_3 emissions for IL by improving existing spatial surrogate ratios in SMOKE by combining crop-specific fertilization rates with cropland distribution. While our total emissions estimates are within 2% of NEI, since both methods are constrained by fertilizer sales data; there are two major improvements. Firstly, it captures localized NH_3 emission hotspots at sub-county resolutions absent in NEI. Secondly, it also apportions NH_3 emissions to specific crops with corn fertilization contributing to 48% of AF emissions. The identified hotspots are being evaluated by spatial autocorrelation techniques to assess the suitability of this newly developed approach. Local crop schedules are being used to estimate temporal factors to further refine inputs to SMOKE. The newly developed EI will provide improved NH_3 emission estimates as direct inputs to SMOKE for high-spatial and temporal resolution regional-scale CTM modeling studies in PM formation and reactive nitrogen deposition to different ecosystems.