Landslide deformation is a complex spatiotemporal process that often evolves across multiple spatial and temporal scales. Traditional field monitoring methods are usually limited in spatial coverage and are difficult to deploy in remote or hazardous terrains. With the rapid development of satellite-based Interferometric Synthetic Aperture Radar (InSAR)

it has become possible to detect subtle ground deformation over large areas with high precision. This study proposes a multi-scale characterization framework for landslide deformation using time-series InSAR data derived from Sentinel-1 and ALOS-2 satellites. By integrating C-band and L-band observations

the proposed approach captures both short-term surface displacement and long-term deep-seated deformation patterns. A typical slow-moving landslide in a mountainous region is selected as the study case. Results show that Sentinel-1 data provide dense temporal sampling for identifying deformation acceleration stages

while ALOS-2 data offer stronger penetration and coherence for long-term trend analysis in vegetated areas. The combined use of multi-frequency InSAR data improves the reliability and interpretability of landslide deformation monitoring. This study demonstrates that multi-scale InSAR characterization is an effective tool for understanding landslide dynamics and supporting early warning and risk management.