InSAR-based method for monitoring the long-time evolutions and spatial-temporal distributions of unstable slopes with the impact of water-level fluctuation: A case study in the Xiluodu reservoir

Slope instability in reservoir areas is one of the greatest concerns for public and governments worldwide. The long-time evolution of the unstable slopes in reservoir areas after the impoundments remain poorly understood. Among the technologies applied to detecting and monitoring unstable slopes, interferometry synthetic aperture radar (InSAR) technology is one of the most effective. However, owing to the slopes in a reservoir suffering from periodic water-level fluctuations, InSAR technology has shortcomings for monitoring unstable slopes with large displacement for a long time. In this study, a post-processing method of InSAR and a practical strategy were proposed to monitor the long-time evolution of the unstable slopes in reservoir areas both spatially and temporally. The strategy can capture the evolution trends of the unstable slopes based on the variations in their deformation areas with reservoir water-level fluctuations for a long time. The case study was carried out in the Xiluodu reservoir area of China. The monitoring data of the unstable slopes before and after reservoir impoundment were extracted from multi-source SAR data, including those from ALOS-1, ENVISAT ASAR, and Sentinel-1. It was found that there were 40 unstable slopes before the reservoir impoundment and 152 unstable slopes after the reservoir impoundment. Of the 152 slopes, 142 slopes (94%) began deforming in the first 3 years after reservoir impoundment, 15 slopes have been experiencing continuous deformation, and 137 slopes underwent intermittent deformation. The areas of 79 slopes (52%) increased during the water-level drawdown periods and decreased during the rising periods. The areas of 22 slopes (14%) demonstrated the opposite trend, and the areas of 51 slopes (34%) increased independently of the water-level fluctuations. In terms of evolution trends in the deformation area, 30% of slopes experienced a decreasing trend, 15% were under a constant deformation trend, and 55% showed an increasing trend. The distribution of the unstable slopes varied from the dam site to the end of the reservoir, and most of them favorited the areas where slope angle and height were 20–35° and 30–300 m, respectively. The soil slopes composed of the previous landslide deposits were prone to deformation, followed by the dip to overdip layered rock slopes.