Identification Of Landslide Displacement In Alpine Valley Region Based On D-InSAR Data Analysis

Landslide monitoring technology is one of the most important technologies on landslide disaster reduction. The current traditional monitoring is against known and separate landslides, and be put in a lot of manpower, material and financial resources. There is a phenomenon said " the hazard under monitoring is not becoming a disaster, the disaster is taking place where is not be monitored" in the actual work of landslide monitoring. In other words, the landslide disasters often happen and develop beyond the scope of the investigation and evaluation of existing landslides and the danger zone which people expected. Early moving identifying in the Southwest ravine reservoir area is one of the most effective means in geological disaster prevention. D-InSAR monitoring technique has the ability of area monitoring, which is of great significance in the identification and monitoring displacement of landslide.However, there are some problems in practical engineering applications. For example, the influence of topography, th accuracy and precision of satellite data and data processing, and surface deformation caused by the non-rock&soil displacement, and so on. These interfering factors influence the identification of landslide displacement monitoring.Considering the fact, an experiment region in the reservoir of Wudongde hydropower station is about to be studied, using programmed SAR remote sensing data for many years experimental study. Based on the D-InSAR analysis method and the characteristics of radar data, an analysis method of the SAR data trusted zone, the detectable landsliding displacement and the error estimation has been established. Based on the factor analysis of differential interferometry result, a data filtering method of distinction landslide moving has been established. By the study of mass moving detecting, identifying landsliding, dynamical moving tendency analyzing, a relationship of the landsliding moving and the SAR data time-spatial changing had been established. And finally, a method of early landsliding moving identifying has been found in the ravine reservoir area. The known sliding Jinpingzi landslide, Tuozhizuo landslide and other tested new identifying moving landslide, prove the feasibility of this method.1. The deep analysis of imaging the principles and processes of deformation monitoring through InSAR and D-InSAR techniques were studied. The SBAS technology is be involved in D-InSAR data processing process.2. By contrast plains, we summed up the characteristics of the terrain factors and SAR data in ravine region. The effect of topography on SAR data for the backward scattering coefficient and the coherence between the images in a particular incidence angle was studies. The research shows different terrain factors have different effects on the scattering intensity and coherence. In the further study, we integrated the perspective of the sensor, terrain slope and aspect as a parameter, local incidence angle. Through the study on the relationship between local incidence angle and SAR data, we found the impact law of the backward scattering coefficient and the coherence. Based on above research, a method of credible regional filtering was proposed. It improved the credibility of the effective monitoring data through the credible region division and the credible regional filtering.Using the displacements of the region with stable backscattering coefficient (River) and permanent scatterers (CR), the rock and soil mass displacement error in southwest dry hot river valley area is determined. Using the error range setting, we set the non-rock&soil displacement threshold,4cm-6cm. Based on the application of the error criteria to filter, we can identify the landslide displacement (more than4cm) in SAR observation period.3. Utilizeing the slope unit as new distinguishing criterion, we can analysis and filter SAR data belongs to the different slope units, to make an important supplement for the D-InSAR data processing. According to the characteristics of the regional integrity and long time creep in the landslide moving process, we proposed the discriminant criterion of spatial continuity&time continuity in the landslide monitoring in the process of recognition of D-InSAR monitoring result. On the basis of the above research has we established a set of the technical process of D-InSAR processing for landslide displacement identification in alpine valley region. Through the analysis and site investigation of Jinpingzi landslide monitoring and other landslide identificated by SAR monitoring, the method and the data analysis process was proved to be used for the regional landslide with large displacement identification and dynamic monitoring.