Publication Date

Spring 5-3-2026

Document Type

Student Project

First Advisor

Luo, Wei

Degree Name

B.S. (Bachelor of Science)

Department

Department of Geology and Environmental Geosciences| Department of Geographic and Atmospheric Sciences

Abstract

Landslides frequently occur in mountainous regions and remain one of the major natural hazards threatening human safety, infrastructure, and regional development. In areas with steep terrain and intense rainfall, landslide disasters often cause serious economic losses and environmental damage. Landslide susceptibility mapping (LSM) has been widely used to identify potential landslide prone areas and to support disaster prevention, land management, and hazard mitigation. The accuracy of LSM is strongly influenced by the selection of mapping units.

This study focuses on I-lan County, Taiwan, and applies a newly derived slope unit dataset to evaluate its influence on landslide susceptibility mapping. Using the same analytical framework as a previous study, four preparatory factors, including slope, aspect, total flux, and lithology, were analyzed to produce the Landslide Susceptibility Index (LSI). The results were evaluated using success rate curves and Area Under the Curve (AUC). Compared with the original model, the new slope unit model shows a more coherent spatial distribution of high susceptibility areas and better agreement with observed landslide locations. These findings suggest that the newly derived slope units provide a more reliable result for landslide hazard assessment and can offer valuable support for future disaster prevention and land management decisions.

Download

Share

COinS