Diagnosing NH ₃ Emission Biases Through Hybrid Inversion and Land‐Use Data: A Case Study Over Eastern China in July 2019

Abstract

Ammonia (NH ₃ ) has become an increasingly important target for reducing fine particulate matter (PM _2.5 ) pollution in China, yet mitigation efforts remain constrained by large uncertainties in bottom‐up emission inventories. Here, we develop a diagnostic framework that integrates satellite‐constrained inversion with land‐use‐based statistical attribution to identify subsector‐specific biases. Our analysis focuses on July 2019 over eastern China, a hotspot region during a peak period for atmospheric NH ₃ concentrations when agricultural activities and NH ₃ volatilization are particularly strong. Optimized NH ₃ emissions (717 Gg) are 98% higher than the base inventory (363 Gg), with the largest underestimation in the agriculture‐intensive North China Plain. Agricultural activities remain the dominant source sector (∼95%) in both the prior and optimized inventories, but the attribution analysis reveals substantial differences among crop and livestock subsectors, reflected in distinct per‐unit emission adjustments based on cropland area and livestock population. Strongest underestimations occur for rice, fruits, and sheep/goats, requiring upward scaling factors of ∼3.0–4.0, whereas cotton, peanuts, maize, cattle, pigs, and poultry show more moderate underestimation (∼2.0–3.0). In contrast, vegetables and soybeans are slightly overestimated. In urban areas, agricultural sources remain dominant, but non‐agricultural emissions are severely underestimated, generally requiring upward corrections by factors of ∼2–6. These results reveal substantial subsector‐specific biases in current NH ₃ inventories during the summer atmospheric NH ₃ peak period over eastern China. More broadly, this study demonstrates the potential of combining satellite inversion and land‐use attribution to diagnose biases in bottom‐up inventories and provide subsector‐specific constraints for improving NH ₃ emission estimates and informing mitigation strategies.