The power network has a long transmission span and passes through wide areas with complex topography setting and various human engineering activities. They lead to frequent landslide hazards, which cause serious threats to the safe operation of the power transmission system. Thus, it is of great significance to carry out landslide susceptibility assessment for disaster prevention and mitigation of power network. We, therefore, undertake an extensive analysis and comparison study between different data-driven methods using a case study from China. Several susceptibility mapping results were generated by applying a multivariate statistical method (logistic regression (LR)) and a machine learning technique (random forest (RF)) separately with two different mapping-units and predictor sets of differing configurations. The models’ accuracies, advantages and limitations are summarized and discussed using a range of evaluation criteria, including the confusion matrix, statistical indexes, and the estimation of the area under the receiver operating characteristic curve (AUROC). The outcome showed that machine learning method is well suitable for the landslide susceptibility assessment along transmission network over grid cell units, and the accuracy of susceptibility models is evolving rapidly from statistical-based models toward machine learning techniques. However, the multivariate statistical logistic regression methods perform better when computed over heterogeneous slope terrain units, probably because the number of units is significantly reduced. Besides, the high model predictive performances cannot guarantee a high plausibility and applicability of subsequent landslide susceptibility maps. The selection of mapping unit can produce greater differences on the generated susceptibility maps than that resulting from the selection of modeling methods. The study also provided a practical example for landslide susceptibility assessment along the power transmission network and its potential application in hazard early warning, prevention, and mitigation.

Several countries worldwide are funding large-scale programs to mitigate landslide risk by implementing engineering remedial works. However, the overall effectiveness of such measures is rarely monitored, and they are typically performed at the slope scale without fully exploiting the wide-area capabilities of remote sensing technologies. A multi-scale and multi-source monitoring procedure for evaluating the slope stability and the effectiveness of related remedial works was proposed in this study and applied in the middle section of the Three Gorges Reservoir Area (TGRA), China. The area is highly exposed to landslide hazards, and a massive program of engineering remedial works was recently implemented. Satellite interferometric synthetic aperture radar (InSAR)-based techniques were first exploited at the regional scale with the objective to provide a general overview of the deformative scenario and to highlight localized problems (active landslides or high deformation zones) to be further investigated; then, local-scale field investigation and multi-source ground monitoring data were employed to verify the deforming states of active landslides and to evaluate the effectiveness of the landslide engineering remedial works. The results indicated that, among the 310 mapped landslides in the study area, 52 were identified to be active and in a slow-moving state by satellite InSAR; Among the 58 controlled landslides, 9 of them were suspected to be active in a slow-moving state and require further concern. Particular attention was paid to two controlled landslides that were found in a continuously and progressively deforming state. We observed that the regional-scale program of slope stabilization was highly successful; however, the variation of the surrounding environmental setting could have led to landslide reactivation or partial invalidation of the landslide remedial works. The proposed multi-scale and multi-source monitoring framework is low-cost, easy to perform, and very straightforward to communicate to citizens and authorities. It can be easily implemented with very wide areas to assess the slope stability and to investigate the effectiveness of large-scale governmental risk mitigation programs, identifying precursor signals that could allow for intervention before reaching critical conditions.

The power grid is one of the most important lifeline projects in modern society. However, the complex transmission network was fixed to the ground through millions of transmission towers, which were inevitably being affected by various types of geohazard. The failure of local facilities may further cause large-scale power outages, leading serious social impact and economic loss. From this perspective, this study aims to develop a territorial landslide early warning system (Te-LEWS) for the high-voltage transmission line coverage area by introducing a novel method, combining landslide inventory and susceptibility maps, rainfall thresholds and real time rainfall forecast, transmission tower vulnerability analysis and GIS-based dynamic alert system. To this objective, the power grid system consisting of over 130,000 high-voltage transmission towers and covering an area of 7 provinces in China was selected as study objective to conducted susceptibility mapping with different classification methods (information value, random forest and support vector). The rainfall threshold of each county was calculated through analyzing a 7 consecutive day rainfall data for the major historical landslide event. Instead of an ordinary landslide risk assessment practice within the transmission line coverage area, this study mainly focuses on the landslide risk over transmission towers and tries to generate an risk assessment result over a specific risk bearing element with linear distribution characteristic, in this case the electricity transmission lines. With real-time predicted rainfall value as input variable, a dynamic landslide warning system was established on a pixel basis, to identify the transmission towers that are potentially vulnerable to landslide disasters. The performance of the proposed Te-LEWS system were validated through the historical rainfall data and the landslides events from 2015-2019, to gain a comprehensive evaluation on its warning accuracy. Results suggest that the system has a high warning success rate and the false alarm was significantly reduced. In such case, the proposed To-LEWS would greatly support the grid authorities in reducing disaster risks and retrieving huge economic loss. The study shed a new light on the risk analysis method of a specific linear distributed risk bearing element towards geohazard, to demonstrate its potential over wide areas, an application to a huge area in China was shown and discussed.Keywords: Landslides; GIS; early warning system; disaster risk reduction; high-voltage transmission line