عنوان مقاله [English]
The aim of this study was landslide susceptibility modelling using random forest machine learning technique and prioritization of effective factors on landslide occurrence in Rais Ali Delvari Dam Watershed. For this aim, at first landslide inventory map was prepared using extensive field surveys and Iranian Landslides Working Party. In total, of 279 identified landslide locations, 70% was used for modelling purposes and the remaining (30%) was applied for validation of the built model. In the current study, different thematic layers including elevation, slope angle, plan curvature, profile curvature, topographic wetness index (TWI), distance from rivers, drainage density, distance from faults, distance from roads, lithological units, and normalized difference vegetation index (NDVI) were selected. In the next step, random forest algorithm run using package of “randomForest” and according to relationship beteen dependent (landslides) and independent (effective factors) variables in R statistical software and landslide susceptibility map was prepared. Accuracy of the mentioned model was test using the receiver operating characteristic (ROC) curve and also based on 30% of unused landslides in modelling process. Accuracy results indicated that random forest model with an AUC value of 0.983 had an excellent precision. Also, prioritization of effective factors showed that slope angle, elevation, plan curvature, distance from road, and lithological units had the highest effect on landslide occurrence.
So, the prepared landslide susceptibility map could be effective in land use planning and also management of Rais Ali Delvari Dam Watershed.