Mechanism-Derived Rainfall Thresholds for Shallow Slope Failure: Infiltration-Controlled Instability under Variable Rainfall Patterns

Jyun-Kai Yang, Ya-Sin Yang, Hsin-Fu Yeh^*
Department of Resources Engineering, National Cheng Kung University, Tainan, Taiwan
* Corresponding Author: Hsin-Fu Yeh. Email: email
(This article belongs to the Special Issue: Advanced Numerical and Data-Driven Modeling for Geoengineering and Subsurface Systems)

Computer Modeling in Engineering & Sciences https://doi.org/10.32604/cmes.2026.086972

Received 08 June 2026; Accepted 31 August 2026; Published online 16 September 2026

Download PDF

Abstract

Traditional rainfall intensity–duration (I–D) thresholds for shallow landslides are predominantly empirical and lack explicit linkage to internal slope hydrological processes, limiting their reliability under variable rainfall conditions. This study establishes a physically based early warning framework by integrating critical suction stress–depth profiles with rainfall I–D thresholds derived from limit equilibrium analysis and unified effective stress theory. Requiring only rainfall data, the framework does not depend on real-time subsurface monitoring and thus remains applicable in data-limited regions. Antecedent rainfall effects are incorporated through an antecedent rainfall duration estimation method, enabling a physically interpretable definition of the initial slope infiltration state. Coupled transient infiltration–stability simulations reveal that infiltration amount, rather than total rainfall depth, is the dominant control on failure initiation. Variations in rainfall temporal distribution produce distinct infiltration responses: earlier intensity concentrations promote more effective infiltration and earlier destabilization, whereas later-stage intensity peaks enhance surface saturation and runoff, limiting infiltration and delaying failure onset. The proposed framework advances existing physically based rainfall threshold studies by explicitly incorporating antecedent rainfall conditions and rainfall temporal distribution into rainfall threshold analysis. Based on these findings, the rainfall pattern associated with the earliest failure timing under comparable conditions is identified as a conservative basis for early warning. Additional analyses considering incomplete rainfall records indicate that prediction performance depends on rainfall update intervals, historical data completeness, and specification of antecedent rainfall conditions. Even under partial data loss, reliable pre-failure warning remains achievable. The results provide a physically interpretable and operationally robust basis for rainfall-triggered slope failure prediction.