Rainfall-Induced Landslide Risk Mitigation - Development and Testing of an Integrated Early Warning System

Landslides often occur without warning, making it challenging to implement timely evacuation and mitigation measures. Therefore, development of real-time early warning systems is crucial for reducing the risks associated with rainfall-induced landslides. This study aims to evaluate the effectiveness of a cost-effective landslide early warning system using internet networks. The system is designed with moisture sensors, acceleration sensors, and pore water pressure sensors, which provide comprehensive monitoring of slope conditions. Slope model tests were conducted using silica sand number seven with a 45° slope angle, under simulated rainfall intensities of 45 mm/h and 70 mm/h. The tests demonstrated that landslide initiation typically began with layered vertical collapses along the surface of the slope, originating at the slope toe and progressively accumulating beneath the failure point. As rainfall continued, both water content and pore water pressure within the soil increased, indicating that the slope was approaching an unstable condition. Once instability occurred, the acceleration sensors recorded slope deformations.

Keywords: Landslide, Early warning system, Rainfall, Internet of things, Slope, Monitoring, Model test.