Landslide vulnerability analysis using frequency ratio (FR) model: a study on Bandarban district, Bangladesh (2407.20239v1)

Abstract: This study assesses landslide vulnerability in the Chittagong Hill Tracts (CHT), specifically focusing on Bandarban district in Southeast Bangladesh. By employing a multidisciplinary approach, thirteen factors influencing landslides were examined, including terrain features, land use, and environmental variables. Utilizing the FR model and integrating various datasets such as DEM, satellite images, and rainfall data, landslide susceptibility mapping was conducted. The analysis revealed that steep slopes, high elevations, specific aspects, and curvature contribute significantly to landslide susceptibility. Factors like erosion, soil saturation, drainage density, and human activities were also identified as key contributors. The study underscored the impact of land use changes and highlighted the stabilizing effect of vegetation cover. The resulting Landslide Susceptibility Map (LSM) categorized the area into five susceptibility zones. The model demonstrated a prediction accuracy of 76.47%, indicating its effectiveness in forecasting landslide occurrences. Additionally, the study identified significant changes in the study area over three decades, emphasizing the influence of human activities on slope instability. These findings offer valuable insights for policymakers and land-use planners, emphasizing the importance of proactive measures to mitigate landslide risks and ensure community safety. Incorporating these insights into policy frameworks