Spatial prediction of landslide susceptibility in Taleghan basin, Iran

摘要: Identifying landslide-susceptible zones is warranted to prevent and mitigate associated hazards in mountainous regions, where a landslide destructive type of erosion. A susceptibility map was developed for the Taleghan basin based on frequency ratio (FR), logistic regression (LR), maximum entropy (MaxEnt), support vector machine (SVM) with radial base (RBF), sigmoid (SIG), linear (LN), polynomial (PL) kernel functions. To this end, an inventory 166 locations prepared partitioned into 70% 30% train validate models, respectively. Subsequently, models were designed 13 factors including elevation, slope degree, aspect, distance stream, Stream Power Index, Topographic Wetness Transport fault, lithology, soil texture, land use, road precipitation. The performance methods assessed using area under receiver operating characteristic curve, Seed Cell Area Index (SCAI), precision index (P). Moreover, statistical measures sensitivity, specificity, accuracy calculated. Friedman test also applied confirm significant differences among seven employed research. validation results showed that MaxEnt had curve (0.812). obtained LR, FR, PL-SVM, SIG-SVM, LN-SVM, RBF-SVM 0.807, 0.732, 0.679, 0.663, 0.643 0.660, P better LR models. trend changes SCAI values, from low- high-susceptibility, indicated best performance. Decision makers can effectively use findings present study financial human costs resulting landslides.