Suitability of Suction Stress–Based Rainfall Intensity–Duration Paradigm for Predicting Tropical Storm–Induced Multiple Slope Failures

Abstract A case study analyzing three landslides in Pohang, South Korea, during tropical storm Typhoon Hinnamnor in September 2022 is conducted to assess the suitability of the recently developed suction stress–based rainfall intensity–duration (SS-ID) paradigm for predicting multiple slope instabilities. Postfailure field investigation and laboratory testing were carried out to determine field hydrologic and geotechnical settings and soil properties. This information was used to determine these slopes’ critical suction stress profiles where the effective stress reaches the failure state or the local factor of safety reaches unity. Hydromechanical numerical modeling was then conducted under various rainfall intensities to identify the slopes’ corresponding durations for slope failure and thus the rainfall intensity–duration (ID) threshold curves of the slopes. The precipitation records during and before Typhoon Hinnamnor at Pohang were used to determine rainfall episodes and slopes’ rainfall intensity–duration data. The determined rainfall intensity–duration data at the three sites for Typhoon Hinnamnor were all above the corresponding predetermined ID threshold curves, correctly and novelly predicting the occurrence of the observed landslides. Further independent physical simulations of transient two-dimensional hydromechanical processes at the three failed slopes were conducted for independent confirmation. Excellent matches in predictions of the slope failure locations and timings are demonstrated, further physically validating the suitability and effectiveness of the SS-ID paradigm for multiple slope instability prediction. Because the suction stress–based ID curves are characteristic of a given region and they can be determined a priori, the SS-ID paradigm provides a practical way to use live rainfall data to forecast multiple slope instability in a region.