An empirical approach to predict regional organic carbon in deep soils

Deep soil organic carbon (SOC) plays an important role in carbon cycling. Precisely predicting deep SOC at the regional scale is crucial for the accurate assessment of carbon sequestration potential in soils but has been challenging for a century. Herein, we developed a depth distribution function-based empirical approach to predict SOC in deep soils at the regional scale. We validated this approach with a dataset from four regions of the world and examined the application of this approach in China’s Loess Plateau. We found that among the reported depth distribution functions describing vertical patterns of SOC, the negative exponential function performed best in fitting SOC along the soil profile in various regions. Moreover, the parameters (i.e., C e and k ) of the negative exponential function were linearly correlated to surface SOC (0–20 cm) and the changing rates of SOC within the topsoil (0–40 cm). Based on the above relationships, the empirical equations for predicting the negative exponential parameters are established. The validation results from site-specific and regional dataset showed that combining the negative exponential function and such empirical equations can precisely predict SOC concentration in soils down to 500 cm depth. Our study provides a simple, rapid and accurate method for predicting deep soil SOC at the regional scale, which could simplify the assessment of deep soil SOC in various regions.