Heterogeneous drivers and transition typologies of agricultural reactive nitrogen emissions: Evidence from 340 Chinese cities (2000–2020)

Agricultural reactive nitrogen (Nr) emissions affect climate, ecosystems, and human health, yet coarse-scale assessments often mask prefecture-level heterogeneity and hinder targeted mitigation. This study developed a long-term prefecture-level Nr emission inventory covering NH 3 , N 2 O, and NO x from crop, livestock, and aquaculture sectors across 340 Chinese cities for the period 2000-2020. Using LMDI decomposition and K-means clustering, we reveal pronounced inter-city disparities hidden by provincial averages and identify five distinct emission-driving typologies, input-intensity-dominated, high-input-high-population driven, efficiency-input balanced, agricultural reverse-transition, and population-sensitive, each characterized by distinct combinations of emission intensity, intermediate-input productivity, per-capita intermediate input, and population scale effects. Although national emissions peaked in 2014, approximately 15% of cities peaked after 2017, revealing asynchronous city-level nitrogen transitions that are invisible in national statistics. These heterogeneous transition pathways highlight the limitations of one-size-fits-all mitigation policies and underscore the need for city-centered differentiated governance. This study provides a transferable framework for diagnosing emission transitions and designing differentiated nitrogen mitigation strategies in rapidly transforming agricultural systems.