Slip surface, volume and evolution of active landslide groups in Gongjue County, eastern Tibetan Plateau from 15-year InSAR observations

Abstract

Landslides stand as a prevalent geological risk in mountainous areas, presenting substantial danger to human habitation. The slip surface (SSF), volume, type and evolution of landslides constitute crucial information from which to understand landslide mechanisms and assess landslide risk. However, current methods for obtaining this information, relying primarily on field surveys, are usually time-consuming, labor-intensive and costly, and are more applicable to individual landslides than large-scale landslide groups. To tackle these challenges, we present a novel method utilizing multi-orbit Synthetic Aperture Radar (SAR) data to deduce the SSF, volume and type of active landslides. In this method, the SSF of landslides over a wide area is determined from three-dimensional deformation fields by assuming that the most authentic direction of the landslide movement aligns parallel to the SSF, on the basis of which the volume and type of active landslides can also be inferred. This approach was utilized with landslide groups in Gongjue County (LGGC), situated in the eastern Tibetan Plateau, which pose grave peril to community members and critical construction along the upstream/downstream of the Jinsha River. Firstly, SAR images were gathered and interferometrically processed from four separate platforms, spanning the period from July 2007 to August 2022. Then, three-dimensional displacement time series were inverted based on Interferometric Synthetic Aperture Radar (InSAR) observations and a topography-constrained model, from which the SSF, volume and type were determined using our proposed method. Finally, the Tikhonov regularization method was applied to reconstruct 15-year displacement time series along the sliding surface, and potential driving factors of landslide motion were identified. Results indicate that 53 landslides were detected in the LGGC region, of which ∼70 % were active and complex landslides with maximum cumulative displacement along the sliding surface reaching 1.5 m over the past ∼15 years. In addition, the deepest SSF of these landslides was found to reach 114 m, with volumes ranging from 1.66 × 10⁵ m³ to 1.72 × 10⁸ m³. Independent in-situ measurements validate the reliability of the SSF obtained in this study. More particularly, we found that the 2018 failure of the Baige landslide (approximately 50 km from LGCC) had caused persistent acceleration to those wading landslides, highlighting the prolonged impact of external factors on landslide evolution. These insights provide a deeper understanding of landslide dynamics and mechanisms, which is crucial when implementing early warning systems and forecasting future failure events.