Feasibility Assessment of a Magnetic Layer Detection Method for Field Applications

The lack of current data on the spatial distribution of soil erosion hinders the ability to predict soil erosion on slopes. To address this problem, we present a simple and sensitive soil erosion measurement method called the magnetic layer detection (MLD) method. This method can measure the changes in soil layer thickness (SLT) at a site featuring a pre-buried magnetic layer (ML) using the relationship between magnetic susceptibility and soil thickness, and then use this information to determine the erosion or accumulation at that point. To verify the adaptability and accuracy of this method, we designed three field tests. First, we conducted a simulated sediment siltation experiment using the erosion pin (EP) method as the control. This experiment demonstrated the accuracy of the MLD method for measuring SLT (R-ad(2) = 0.97, NSE = 88.14%, and RMSE = 1.17 mm). Next, a simulated rainfall experiment with the runoff plot (RP) method as a control was used to demonstrate the effectiveness of the MLD method under water erosion. The results demonstrated that this method can reliably estimate soil erosion on both bare plots (R-ad(2) = 0.83, NSE = 71.78%, and RMSE = 0.56 mm) and sparse grass plots (R-ad(2) = 0.90, NSE = 81.77%, and RMSE = 0.25 mm), with performance that is better than that of the traditional EP method. Finally, a soil scouring experiment, again with the RP method as a control, was designed to verify whether the MLD method could accurately measure the erosion of a slope after the soil was scoured by surface runoff. The MLD method could accurately measure the slope erosion (R-ad(2) = 0.91, NSE = 89.55%, and RMSE = 0.42 mm), with results superior to those of the traditional EP method. The MLD method's results were similar to those from the laser scanner (LS) method, but more accurate and with less associated cost and data processing time. Therefore, the MLD method has potential as a reliable measurement method that can provide useful guidance for elucidating the spatial distribution of soil erosion and predicting slope soil erosion. This approach could be used to obtain raw data to quantify soil erosion on the Loess Plateau of China and beyond.