Towards sustainable low-carbon agriculture: Geospatial pattern decoding and differentiated emission mitigation potential across China's intensive cropping regions

Accurate quantification of greenhouse gas (GHG) emissions from crop production, with explicit spatial delineation and traceable sources, together with the identification of mitigation potential, forms the basis for targeted agricultural emission reduction. However, systematic studies on multiple crop systems in the major plains of China remain limited. This study systematically quantified city-level GHG emissions from major crops across three major plains of China, and analyzed their emission characteristics and spatial patterns. Based on this, a Random Forest (RF) model and scenario simulations were employed to identify key components and quantify regional mitigation potential. The results revealed substantial variation in GHG emission intensity among crop types, with vegetables reaching 13.21 t CO₂e ha⁻¹ , whereas soybeans exhibited much lower emissions of 0.74 t CO₂e ha⁻¹ . Emission hotspots also displayed pronounced spatial heterogeneity. Overall, fertilizer application and irrigation electricity consumption were the dominant contributors to GHG emissions from most crops, accounting for an average of 77.5%. Under the combined effects of multiple mitigation measures, the three major plains demonstrated considerable potential for emission reductions. Notably, the southern Northeast Plain (NE), the northern North China Plain (NC), and the eastern Middle and Lower Yangtze River Plain (MLY) exhibited particularly high mitigation potential across various crop types. Optimized management of fertilizer, water, and energy represents a central pathway for reducing emissions. Nevertheless, mitigation strategies need to be tailored to specific crops and regional contexts. These findings provide new insights for the development of more effective agricultural mitigation policies.