Integrating Landscape Pattern Metrics to Map Spatial Distribution of Farmland Soil Organic Carbon on Lower Liaohe Plain of Northeast China

Accurate digital mapping of farmland soil organic carbon (SOC) contributes to sustainable agricultural development and climate change mitigation. Farmland landscape pattern has changed greatly under anthropogenic influence, which should be considered an environmental variable characterize the impact human activities on SOC. In this study, we verified feasibility integrating patterns in SOC prediction Lower Liaohe Plain. Specifically, ten variables (climate, topographic, variables) were selected for with Random Forest (RF) Support Vector Machines (SVMs). The effectiveness metrics was by establishing different combinations: (1) natural variables, (2) variables. results confirmed that improved accuracy compared R2 RF SVM increased 20.63% 20.75%, respectively. performed better than smaller error. Ranking importance showed temperature precipitation most important Aggregation Index (AI) contributed more elevation, becoming variable. Mean Contiguity (CONTIG-MN) Landscape Contagion (CONTAG) also other topographic We conclude can improve support sequestration optimizing management policies.