Susceptibility to Gully Erosion: Applying Random Forest (RF) and Frequency Ratio (FR) Approaches to a Small Catchment in Ethiopia

Soil erosion by gullies in Ethiopia is causing environmental and socioeconomic problems. A sound soil water management plan requires accurately predicted gully hotspot areas. Hence, this study develops a susceptibility map (GESM) using frequency ratio (FR) random forest (RF) algorithms. total of 56 were surveyed, their extents derived digitizing Google Earth imagery. Literature review multicollinearity test resulted 14 variables for the final analysis. Model prediction potential was evaluated area under curve (AUC) method. Results showed that best accuracy FR RF models obtained top four most important predictor factors: drainage density, elevation, land use, groundwater table. The notion table one factors novel significant quantifiable finding critical to design effective watershed plans. from separate variable importance analyses cover Nitisols density Vertisols as leading determining locations. Factors such texture, stream power index, convergence slope length, profile curvatures found have little significance formation studied catchment.