Monitoring of rainfall-induced landslides at Songmao and Lushan, Taiwan, using IoT and big data-based monitoring system.

This study presents a new landslide monitoring system, which is designed to be cost-effective, robust, flexible, and scalable, aiming for long-term and real-time monitoring and large-scale deployment in harsh and remote environments. This wireless monitoring system consists of the hybrid sensor node (including the GNSS receiver and MEMS accelerometer), the GNSS antenna, the power supply system, and the data center with a big-data infrastructure. The system was deployed at Songmao and Lushan landslide-prone sites, Taiwan, and its effectiveness was also examined through monitoring real landslide events. It was found that continuous monitoring, combining displacement and vibration data, enables us to see how and when the landslide is triggered by the rainfall, together with the potential sliding direction, and the associated aftermath behavior, even after the rainfall stops. The entire landslide movement, from initiation to the subsequent continuous movement, can be captured in detail through the DGNSS analyses. The potential landslide direction is derived based on the directivity from the Horizontal to Vertical Spectral Ratio (HVSR) analyses on the seismic measurements by the 3-axis accelerometers. The monitoring results also agree with other measurements from the borehole inclinometer/extensometer and total station. At these two sites, the groundwater level exhibited a distinct time lag in rising after the rainfall started and the subsequent sudden increase triggered the landslides. On the contrary, the drawdown of the groundwater level was relatively slow, which is due to the anisotropic permeability of the slate, and therefore promotes the continued slope movement even after the rainfall ceased. Water-softening on the slate along the sliding surface, which was just fractured by landslide movement, could be one of the associated underlying mechanisms. [ABSTRACT FROM AUTHOR]