Spatial heterogeneity and impact factors of soil organic carbon in farmland in the dry loess plateau of China based on the OPGD model

Abstract Farmland carbon pools are pivotal to the terrestrial carbon cycle. Studying soil organic carbon density (SOCD) spatial heterogeneity is crucial for uncovering regional carbon balance mechanisms and optimizing sequestration strategies. Focusing on county-scale SOCD heterogeneity, this study aids sustainable farmland management, carbon sequestration, and spatial carbon pool modeling. Based on the 2020 farmland soil organic carbon (SOC) analysis survey data. The SOCD and spatial pattern of farmland in Chengcheng County, Shaanxi Province were estimated and analysed by combining GIS and GPS technologies. Optimal Parameter Geodetector (OPGD) was used to detect spatial heterogeneity and quantitatively attribute the influencing factors of SOCD in farmland in Chengcheng. The results showed that the SOCD of farmland in Chengcheng, ranged from 1.10 to 1.79 kg/m², with an average of 1.39 kg/m². The SOC stock was 151.96 × 10⁴ t. Both values were lower than the national average and much lower than those in European and American countries. Between the north-west and south-east of the plough layer, Chengcheng's farmland's SOCD clearly distinguishes two distinct sections. The average farmland SOCD in the north-west is lower than that in the south-east by 0.27 kg/m². The explanatory power of temperature, population density and DEM all exceeded 0.3, and the explanatory power of the interaction between them exceeded 0.6. This implies that the combined influences of nature and human activity are responsible for the spatial patterns of SOCD in farming at the county scale in the Weibei dry loess plateau.