Abstract: Landslides are among the most threatening geological hazards globally. Traditional monitoring techniques are often constrained by safety risks associated with sensor installation and low spatiotemporal resolution of deformation data, rendering them ineffective for emergency monitoring of sudden landslides. This study aims to develop a high-precision, reliable dynamic deformation monitoring technology that can provide real-time, scientific, accurate monitoring results and early warnings, thereby ensuring the safety of rescue personnel and facilitating smooth rescue operations. We propose a fusion technology integrating Ground-Based Interferometric Synthetic Aperture Radar (GB-InSAR) and Persistent Scatterer Interferometry (PS-InSAR). By selecting natural Persistent Scatterers (PS) with strong scattering characteristics, such as exposed rocks, a deformation monitoring network is constructed to overcome signal attenuation in vegetated areas. Innovatively, stable environmental reference points combined with real-time temperature monitoring data are introduced to establish a deformation signal separation mechanism, enabling the identification and quantification of temperature-induced influences on deformation measurements. Applied to the emergency monitoring of the massive landslide in Shanyang County, Shaanxi Province, this technology successfully establishes an effective PS monitoring network in vegetated areas, achieving real-time monitoring with sub-millimeter precision on high-risk slopes. It clearly reveals a high correlation between deformation measurements and temperature variations. Based on this mechanism, two instances of anomalous regional deformation are accurately identified and high-risk warnings are issued promptly during periods of relatively stable temperature, ensuring the safety of rescue personnel. This study validates the advancement and reliability of the GB-InSAR and PS-InSAR fusion technology for deformation monitoring and early warning in complex scenarios. The technology overcomes the technical challenge of low coherence in vegetated monitoring zones, enabling dynamic, high-precision deformation monitoring and early warning for landslides. Furthermore, the proposed environmental interference separation mechanism effectively guarantees the scientific rigor of geological hazard monitoring, demonstrating significant value for wider application.