Landslide susceptibility analysis using bivariate and multivariate models in Chalus County

Articles in Press

Document Type : Original Article

Authors

Semnan University

Abstract

In mountainous regions, mass movement and landslide are regarded as a significant erosional process. Steep slope areas are geologically and ecologically sensitive and fragile, and the occurrence of landslides makes this area more vulnerable to many hazards. The primary goal of this research is to create a landslide susceptibility map using bivariate statistical models (frequency ratio and information value) and a multivariate model (logistic regression), and selecting the appropriate model for this task in Chalus County. The first step in this research is to prepare distribution map of the landslides and determine their location on the map. This was done using Google Earth imagery and field surveys. The maps of condiioning factors were prepared from different sources and entered into the GIS environment. frequency ratio, information value, and logistic regression as used to create landslide susceptibility map. The ROC curve and the area under the curve were used to assess the accuracy of the models. The AUC for frequency ratio, information value, and logistic regression were 0.705, 0.721, and 0.811. The area of the susceptibility classes were calculated for susceptibility map created using regression logistic. Results show that approximately 44 % of the study area is at high and very high susceptibility to landslides. The calculated area indicates that the region is highly susceptible to landslides, and paying attention to this phenomenon is very necessary.

Keywords

Main Subjects

Journal of Hydrogeomorphology

Articles in Press, Accepted Manuscript
Available Online from 17 September 2025

Files

History

  • Receive Date: 16 July 2025
  • Revise Date: 16 September 2025
  • Accept Date: 17 September 2025

Share

How to cite

Statistics