Integrated remote sensing and geotechnical investigations of the 2022 Jiujiawan mudstone landslide in Xining city, China: from deformation monitoring to failure mechanisms

Mudstone landslides are widespread worldwide, with their initiation and evolution mainly governed by the coupled effects of rock mass microstructure and hydromechanical behavior. However, the linkage between Earth observation (EO)-derived deformation signals and microscale weakening mechanisms remains poorly constrained. Using the catastrophic Jiujiawan landslide (15 September 2022, Xining, China) as a case study, this research aims to reconstruct the spatiotemporal evolution of the landslide and explain the multiscale mechanisms connecting microscale material degradation with macroscale slope failure through evidence-based interpretation. To achieve this, an integrated framework combining multisource EO data, geotechnical investigations, and laboratory analyses was applied. This approach clarifies deformation characteristics and failure mechanisms from coupled macro- and microscale perspectives. At the macroscopic scale, a high-fidelity three-dimensional model was reconstructed using unmanned aerial vehicle photogrammetry. Satellite optical imagery was used to qualitatively trace the pre-failure evolutionary trajectory, complemented by interferometric synthetic aperture radar to quantitatively reconstruct spatiotemporal displacement sequences. At the microscopic scale, X-ray diffraction, scanning electron microscopy, and energy dispersive spectroscopy characterized the mineralogical suite and microstructural fabric. These analyses reveal the properties of Paleogene mudstone intercalated with gypsum layers, showing how pore-fracture networks accelerate rock mass degradation. Synthesis of the results indicates a process-based linkage whereby microscale hydromechanical weakening (clay mineral hydration, argillization, and gypsum dissolution) is progressively manifested as macroscopic creep and acceleration in EO-derived displacement time series, with seismic preconditioning (the 2022 Ms 6.9 Menyuan earthquake) further reducing the slope’s structural threshold through mechanical fatigue, ultimately leading to slope failure under extreme precipitation. This transferable multiscale framework supports mechanism-informed hazard assessment and enables EO‑based early detection and staged early warning for clay‑rich, water‑sensitive slopes that undergo progressive creep under hydromechanical coupling, especially during extreme climatic events.