Rainfall-induced shallow landslides are common in many mountainous countries. Highly concentrated precipitation triggers landslides and debris flows in worldwide. Every year, several shallow slides and debris flows occur in Busan, South Korea during heavy rainfall. To reduce and prevent associated damage, we developed a matrix-based approach of rainfall threshold and landslide susceptibility. This study collects landslide inventories which consist of information of 260 landslide location, 35 landslide event times, and corresponding rainfall intensities and durations. A rainfall threshold warning levels were established using rainfall intensities and durations associated with 35 historical shallow slides, and categorised as null (< 5%), watch (5-20%), attention (20-50%), and alarm (> 50%). We used a back propagation ANN machine-learning algorithm to explore the effects of 14 causative factors on shallow slide distribution. The area under the curve was used to assess accuracy, and accuracy was found to be 86.12%. The derived rainfall threshold warning levels were assigned in rows and the susceptibility classes were used in columns of the matrix. The combined result represents the shallow slide hazards to rainfall threshold warning levels with a varying likelihood to shallow slide initiation. The results aim to raise awareness towards landslide hazards and to support regional decision for the land-use planning.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.