Predicting sediment deposition rate in check-dams using machine learning techniques and high-resolution DEMs

Sediments accumulated in check dams are a valuable measure to estimate soil erosion rates. Here, geographic information systems (GIS) and three machine learning techniques (MARS-multivariate adaptive regression splines, RF-random forest SVM-support vector machine) were used, for the first time, predict sediment deposition rate (SR) check-dams located six watersheds SW Spain. There, 160 dry-stone (~ 77.8 km−2), sediments during period that varied from 11 23 years. The SR was estimated former research using topographical method high-resolution Digital Elevation Model (DEM) (average of 0.14 m3 ha−1 year−1). Nine environmental-topographic parameters calculated employed as predictors SR. ability MARS, RF SVM evaluated by five-fold cross-validation, considering entire area (ALL), on hillslope (HILL) valley-bottoms (VALLEY), well catchments (B, C D) with highest number dams. accuracy models assessed relative root mean square error (RRMSE) absolute (MAE). results revealed able higher more stable than MARS. This is evident datasets ALL, VALLEY D, where errors prediction exhibited MARS 44 77% 37 62% (MAE) those achieved SVM, but it also held HILL B difference RRMSE MAE 7–10% 12–17%, respectively.