Application of Support Vector Machine and Analytical Hierarchy Process Models in Landslide Susceptibility Assessment

Articles in Press

Document Type : Original Article

Authors

1 Associate professor, Department of Environment, Faculty of Natural Resources, Semnan University, Semnan, Iran.

2 - Assistant professor, Department of Environment, Faculty of Natural Resources, Semnan University, Semnan, Iran

10.22034/hyd.2026.69769.1821

Abstract

Landslide susceptibility mapping is an essential part of landslide risk assessments. Subsequently, it helps to manage landslide loss reduction in an area. The primary objective of this research is to produce landslide susceptibility mapping using analyical hierarchy process and support vector machine models in the north of Tehran. Landslide distribution mapping and determine their location is the first step in this research. The maps of condiioning factors were prepared from different sources and entered into the GIS environment. Finally, the weights calculated from the AHP and SVM methods were imported into the ArcGIS software to create landslide susceptibility maps. The AHP-based prioritization of conditioning factors showed that elevation, slope, and distance to faults have the greatest impact on landslide occurrence in the study area. The ROC curve and Landslide Density Index (LDI) were used to assess the accuracy of the models. The findings indicated that the accuracy of the SVM model was very good, and AHP model was good for the study area. These results indicate that the conditioning factors for landslide occurrence were appropriately selected, resulting in satisfactory accuracy of the models used to generate the susceptibility map. To assess the overall susceptibility in the region the area of the four susceptibility classes were calculated for both models. The results showed that a large portion of the region is highly susceptible to landslides. The high and very high susceptibility classes identified priority areas for focused landslide management and mitigation efforts.

Keywords

Main Subjects

Journal of Hydrogeomorphology

Articles in Press, Accepted Manuscript
Available Online from 21 April 2026

Files

History

  • Receive Date: 19 October 2025
  • Revise Date: 20 April 2026
  • Accept Date: 21 April 2026

Share

How to cite

Statistics