Assessing Historical Landslide Risk Management Based on Trigger Magnitude and Consequences: A Case Study from the Rokko Mountains, Kobe, Japan

Abstract Landslides are a common hazard in mountainous regions, and many countries have implemented landslide risk management to mitigate their negative impacts. Assessing the effectiveness of those measures is important to improve technical and political decision-making and to enhance the selection and implementation of effective landslide risk management strategies. Here, we assessed effectiveness in landslide risk management based on the magnitude of rainfall characteristics that triggered landslides (inducing factor) and landslide consequences in the Rokko mountains, Kobe, Japan. The number of check dams was used as an indicator of progress in landslide risk management. For fatal landslide events in 1938 and 1967, rainfall characteristics that triggered landslides were estimated using the three-layer tank model, and their magnitude was quantified by the return period (RP). We then compared these rainfall magnitudes and landslide consequences (i.e., fatalities and completely collapsed houses) between the two events. The RP of the first tank storage layer value, which indicates rainfall characteristics triggering shallow landslides, was higher at landslides in 1967 than in 1938, whereas landslide consequences were less in 1967 than in 1938. 218 units of check dams were intensively constructed by landslide risk management from 1938 to 1967 and reduced the damage from landslides in 1967 that were triggered by higher magnitude rainfall than in 1938. This study also highlighted the importance of focusing on the magnitude of the inducing factor and landslide consequences to assess the effectiveness of landslide risk management at a local scale.