MEEI-LING LINChen, Yen ChenYen ChenChenLiu, Te ChuTe ChuLiu2024-03-012024-03-012021-01-01https://scholars.lib.ntu.edu.tw/handle/123456789/640057The frequency and extent of landslide hazard might increase due to the increasing trend of severe rainfall events caused by climate change as suggested by many researchers. Among different types of landslide, the shallow landslides occurred frequently and extensively, and often triggered debris flow. In this research, the susceptibility analysis was conducted for debris flow basins in a study area in central Taiwan. Inventory of the shallow landslide was screened and edited considering the corresponding grid size of analysis unit. The causal factors were selected and tested for independency and significance, and the discriminant analysis was conducted for assessment of shallow landslide susceptibility. The constructed susceptibility model was validated with a separate group of data, and both the fitting and validation groups provided consistent accuracy and good results. The backward prediction was conducted using the historical inventory data, and a comparable accuracy was achieved, which suggested the robustness of the susceptibility model. Finally, a regional susceptibility map was established accordingly.conference paper2-s2.0-85184351653https://api.elsevier.com/content/abstract/scopus_id/85184351653