Generally, the causes of large-scale landslides are related to geological conditions and tectonics,and bedrock location is one of the important factors. Since the investigation of geological structure is usually time-consuming and expensive, many researchers have tried to explore a geological structure by analyzing groundwater flow. This study uses an integrated Rainfall-Infiltration-Slope stability (IRIS) model to simulate the process of
rainfall-runoff and predict the time of occurrence and the scale of large-scale landslides. The simulation results show
that the bedrock location influences runoff, and the end time of interflow offers a better index for detecting the bedrock location when using runoff data. The simulation results also indicate that the soil strength dominates the scale of landslides, and the soil hydraulic characteristics only affect the time of occurrence (or probability) of landslides. Accordingly,large-scale landslides tend to occur in areas which have a lower angle of internal friction, higher water retention,and lower hydraulic conductivity. Moreover, the water contents of the soil are almost same in areas where landslides occur even if under the different rainfall patterns. That is, the water contents of soil can be used as an effective index for predicting landslides.
Key Words : Large scale landslides, rainfall, infiltration, runoff, bedrock, water content.