Item

Abstract

Multi-temporal interferometric synthetic aperture radar (MT-InSAR) is a widely used technique for monitoring subtle ground instabilities, with a precision ranging from centimeters to millimeters. Traditional MT-InSAR analysis often employs low-pass temporal filtering to suppress stochastic noise and extract deformation features from measurements. However, these approaches may hinder the detection and estimation of transient slope instabilities triggered by external factors. In this study, we propose a methodology for characterizing transient deformation of reservoir bank slopes under the coupling effects of rainfall and reservoir water level (RWL) changes. Following MT-InSAR analysis, slope kinematics is analyzed using time series decomposition and independent component analysis (ICA) to separate trends from seasonality. Although triggers of slope instability exhibit similar periods of increase and decrease, they are not entirely synchronized. Therefore, the seasonal components of surface deformation along with environmental triggers are analyzed using a wavelet transform to determine the time-lag between them. This helps define an effective priori search windows for constraining transient behaviors in landslide kinematics. A constrained least-squares optimization is finally applied to extract step-like kinematics features. Our methodology’s effectiveness is evaluated using a dataset comprising 102 TerraSAR-X (TSX) images in High-resolution Spotlight (HS) mode, 197 Sentinel-1 images, 10 ALOS-1 and 12 ALOS-2 Stripmap images, collected from 2006 to 2022 across Badong County, located along the Yangtze River. After the successful validation against in-situ measurements and comparison with conventional post-processing strategies, we apply our method to generate a map of hazardous sliding areas and evaluate the regional-scale slope instability of slow-moving landslides across Badong County in the Three Gorges Area (TGA).