LAUSR

Abstract Spatially-explicit data describing commercial nitrogen (N) fertilizer and animal manure inputs are needed to inform modeling and life cycle analysis of agricultural impacts associated with corn production. The currently available N datasets based on farm surveys and sales for the Conterminous U.S. are inappropriate for corn-based modeling at fine resolutions because they are either too coarse in scale or are not corn-specific. This work developed county-level N input data for corn production by harmonizing multiple U.S. datasets using two data fusion approaches. A top-down area-based approach allocates N fertilizer inputs into corn producing areas by combining state-level crop-specific N fertilizer application rates and percentage of area receiving N fertilizer with the county-level proportion of crop-specific planted area. Similarly, county-level manure N rates are calculated based on county-level corn planted area and livestock populations coupled with state-level application data. An alternative approach derives N needs from corn yields, crop rotations, and soil characteristics before N surplus is estimated by subtracting N needs from N rates. Nationally, the weighted averages of corn N inputs (188 kg N ha-1) based on corn planted area exceeded N needs (128 kg N ha-1) by 60 kg N ha-1 with N surplus found in 80% of all U.S. corn producing counties. Results distinguished regions of high (Midwest), moderate (Northern Plains), and low (Southeast and Northwest) N application rates and surpluses. Estimates for Western states had the greatest variability and uncertainty associated with the frequency of N rate outliers where corn production is low. The estimated N inputs for major corn producing areas generally aligned with source datasets, while further evaluation is needed for manure application rates using independent sources. This work shared the first spatially-explicit datasets for U.S. corn fertilizer and manure inputs and N needs together with methods for evaluation. Steps needed to expand access and coverage of detailed N data were identified to improve assessments of agricultural and environmental impacts.