Rainfall-induced shallow slope instability is a significant global hazard, often triggered by water infiltration that affects soil stability and involves dynamic changes in the hydraulic behavior of unsaturated soils. This study employs a hydro-mechanical coupled analysis model to assess the impact of rainfall on slope stability, focusing on the dynamic hydraulic behavior of unsaturated soils. By simulating the soil water content and slope stability under four different rainfall scenarios based on observational data and historical thresholds, this study reveals that higher rainfall intensity significantly increases the soil water content, leading to reduced slope stability. The results show a strong correlation between the soil water content and slope stability, with a 20 mm/h rainfall intensity threshold emerging as a reliable predictor of potential slope instability. This study contributes to a deeper understanding of slope stability dynamics and emphasizes the importance of proactive risk management in response to changing rainfall patterns while also validating current management practices and providing essential insight for improving early warning systems to effectively mitigate landslide risk.

Machine learning algorithms are commonly employed in landslide susceptibility assessments. Recently, algorithms that utilize artificial intelligence have come into prominence. This study attempts to adapt the most fundamental framework of deep learning and introduces fuzzy theory concepts to analyze landslide susceptibility while updating the network parameters with trial-and-error methods. The final analysis results will compare with those of logistic regression (LR). In order to assess the ability of the model to identify landslides in a more objective way, two typhoon events were used as a training event and a validation event, respectively. The results of the analysis show that the area under the curve (AUC) of the fuzzy neural network (FNN) for the training event is 0.915, but the AUC for the validation event drops to 0.746. Although the results of the FNN for training events were better than those of LR, they did not differ much from those of LR in predicting future events. The reason for this is that the difference between the landslide distributions of the training and validation events is too large, making the model biased in its identification. Overall, FNN is still a recommended method for analyzing landslide potential and can be used as a reference for LR.