Study On Extraction Of Landslide Deformation Information And Landslide Evaluation Method Based On SBAS-InSAR

Landslide geological disasters pose a serious threat to the lives and property safety of the people and can also cause immeasurable consequences for the local ecology,resources,and infrastructure.As a newly developed remote sensing monitoring technology,In SAR technology has been widely used in the field of geological disaster monitoring due to its wide monitoring range,abundant data,and no monitoring environment restrictions.Its results can provide important data references for landslide disaster identification,investigation,and evaluation.A comprehensive analysis of various datasets and survey materials to summarize the spatiotemporal distribution pattern of landslides and the possibility of landslides occurring in the time series is of great significance for the prevention and control of landslide disasters through landslide sensitivity and risk evaluation studies.However,landslide sensitivity evaluation usually use static environmental factors for construction,while the geographic environment changes over time,thus ignoring the dynamic characteristics of landslide sequences and reducing the reliability of landslide sensitivity mapping.The landslide risk evaluation is often affected by the uncertainty of model input parameters.When using models for time dimension analysis,the output error will gradually accumulate over time,causing the forecast field to gradually deviate from the observed field,reducing the reliability of risk evaluation.In response to the above problems,this study takes the jurisdiction of Haidong City as the research object,and based on SBAS-In SAR technology,it uses surface deformation information and optical remote sensing images to carry out early landslide identification work,and explores the fusion of SBAS-In SAR monitoring results with landslide sensitivity evaluation and risk evaluation methods.The main work contents and research conclusions are as follows:(1)The surface deformation information of Haidong City’s jurisdiction in the study area was extracted based on SBAS-In SAR technology.The annual average deformation rate in the study area ranges from-131.5 to 92.1 mm/a,and 25 potential landslides were identified in combination with the annual average deformation rate results and optical remote sensing images.Meanwhile,using two typical landslides as examples,the optical remote sensing interpretation characteristics and disaster formation mechanisms were analyzed by integrating deformation rate and optical remote sensing image information.(2)The landslide sensitivity evaluation method for Haidong City’s jurisdiction integrating SBAS-In SAR was explored.Based on nine traditional evaluation factors,including elevation,slope,aspect,lithology,planar curvature,fault distance,river distance,road distance,and normalized vegetation index,the SBAS-In SAR annual average deformation rate was incorporated into the evaluation factors.The information value-logistic regression model was used for landslide sensitivity prediction.The results show that landslides in the study area are mainly distributed in low-altitude areas on both sides of valleys and high-altitude areas with intense tectonic activity and broken rock and soil bodies.Deformation rate,vegetation coverage,and elevation are the main factors affecting the distribution of landslide disasters in the study area.Compared to landslide sensitivity mapping based on traditional evaluation factors,the sensitivity mapping integrating SBAS-In SAR has a higher prediction efficiency.(3)The method of integrating SBAS-In SAR results with the TRIGRS model for landslide risk evaluation was explored.Based on data assimilation methods,the particle filtering algorithm was used to introduce SBAS-In SAR monitoring data into the TRIGRS model,correcting the model’s safety factor and updating the internal friction angle parameter in real-time.The Gaojiawan landslide,a typical geological disaster along the railway line in the region,was used as an example for landslide risk evaluation.The results show that as more observation data is added,the model’s prediction trajectory is effectively corrected,the uncertainty of model parameters is reduced,and the accumulated error of the model is released.This can provide a more reliable theoretical basis and reference for local landslide risk evaluation.