Search

The early warning of landslides is crucial for risk management and reduction. However, it is challenging due to the complex nature of landslide behaviours and failure mechanisms. Generally, landslides undergo a certain period of evolution from deformation to failure, and own obvious three-stage deformation phases of creep slope failure. Most of landslides own obvious sudden characteristics, such as rockslide, avalanche, and loess flowslide, which are object of this study. In this paper, an early warning systems for regional landslides and a comprehensive warning model are established, taking into account deformation rate threshold and the improved tangent angle as the early warning parameters. A four-level early warning criterion is proposed. Once the early warning parameters exceeded default thresholds, relevant local authorities and scientists could immediately receive the warning information released by the 3D WebGIS-based platform. It is the 11 times our early warning system successfully forecasted landslides in four different regions since its implementation in 2017. Here, we present three typical cases of successful early warning from these three different regions and timely evacuation in advance. It could take a reference and applicable for other cases globally.