위험수준별 강우를 고려한 산사태 민감도 및 취약성 평가 기법 개발

Abstract

Most of landslide have been taken place during rainy season (June-September) by typhoon and local rainfall in Korea which cause a considerable damage to properties and human injury. The main triggering factor of these landslides is an intense rainfall so that analysis of landslide-triggering rainfall is essential. In this study, landslide susceptibility and vulnerability assessment techniques were developed using hazard level-based rainfall. First of all, 218 landslide occurrence data from 1999 to 2013 were collected. The rainfall data were also collected using the previous rainfall data from the Korea Meteorological Administration. The landslide-triggering rainfall was estimated using the collected landslide data and rainfall data. In addition, quantile regression analysis was used to estimate the between hazard level-based rainfall threshold and optimum rainfall threshold. Secondly, the landslide susceptibility map was developed according to hazard level-based rainfall using rainfall and physical-based model for risk level in Umyeon moutain in Seoul. In addition, the landslide susceptibility map was developed using a logistic regression method, which is a statistical-based model using causative factors such as topographic, hydrology, Forest, Soil and geology with landslide occurrence data. Source area based on hazard level-based rainfalls were extracted high potential area using these combined physical-statistical-based landslide susceptibility maps. Thirdly, Source area based on hazard level-based rainfalls were extracted and the propagation of debris flow is analyzed by using the Flow-R model. And, Affected area of debris flow considering hazard level-based rainfall was calculated using propagation of debris flow results. Finally, landslide vulnerability assessment was carried out in umyeon moutain of Seoul using the affected area of debris flow from the analysis of the propagation of debris flow. The development of the landslide susceptibility and vulnerability assessment method using hazard level-based rainfalls can be used not only as the core technology for landslide early warning system but also as the important basic data for decision making in risk management.