Applying statistical methods to map soil organic carbon of agricultural lands in northeastern coastal areas of China

Soil organic carbon (SOC) is an important indicator to evaluate agricultural soil quality. Precise mapping SOC can help to facilitate soil and environmental management decisions. This study applied multiple stepwise regression (MSR), boosted regression trees (BRT) model, and boosted regression trees hybrid residuals kriging (BRTRK) to map SOC of agricultural lands in Wafangdian City, northeastern China. A 10-fold cross-validation procedure was used to evaluate the performance of the three models. The BRTRK method exhibited the best predictive performance and explained 78% of the total SOC variability. The distribution of SOC was mainly explained by elevation, followed by soil-adjusted vegetation index (SAVI), and topographic wetness index (TWI). We conclude that the BRTRK was the most accurate method in predicting spatial distribution of SOC. In addition, our study indicated that topographic variables as key factors to affect SOC should be considered in future SOC mapping.