Abstract
Objectives: For a long time, landslides, as geological hazards, have been a significant factor influencing the ecological environment, landform evolution, and even the pattern of human settlement in the upper Yellow River basin. Studying the spatial distribution patterns and temporal evolution mechanisms of active landslides can help provide a scientific basis for the management and prevention of landslide hazards in the upper Yellow River. Simultaneously, it promotes the engineered application of radar remote sensing technology in landslide research, offering new ideas and methods for monitoring and analyzing similar geological hazards. Methods: First, based on 358 ascending and descending Sentinel-1 images spanning from January 2017 to July 2023, the small baseline subset interferometric synthetic aperture radar (SBAS-InSAR) technique was employed to obtain the surface deformation rates and time series in the study area. Subsequently, active landslides were interpreted and mapped by integrating InSAR velocities, optical imagery, and topographic data. Finally, a three-dimensional deformation model constrained by topographic factors was utilized to obtain the true deformation field of the Jungong landslide. Additionally, independent component analysis (ICA) and hierarchical clustering (HC) methods were introduced to investigate the spatio-temporal evolution characteristics of the Jungong landslide. Results: There are 124 active landslides developed in the Maqin section of the upper Yellow River, primarily concentrated on the slopes on both sides of the main stream, with more landslides occurring on the right bank than on the left. As a typical case, the Jungong landslide has undergone considerable surface deformation over the past few years, with horizontal movement being the dominant type of sliding. The surface deformation of the Jungong landslide exhibits spatially uneven distribution and significant variability in temporal evolution, which are related to the formation process and geological background of the landslide. Based on the distribution results of independent components, the landslides are clustered into three active zones, revealing different deformation patterns and risk levels. Conclusions: The concentrated occurrence of active landslides in the Maqin section of the upper Yellow River is related to the local landform changes caused by the long-term undercutting and erosion of the Yellow River. Specific areas of the Jungong landslide exhibit accelerated deformation, posing a risk of localized collapse, and thus require long term monitoring.