Using Persistent Scatterer Interferometry for Post-Earthquake Landslide Susceptibility Mapping in Jiuzhaigou

Earthquakes cause a huge number of landslides and alter the regional landslide risk distribution. As result, after significant earthquake, susceptibility maps (LSMs) must be updated. The study goal was to create seismic containing causative variables which are adaptable great changes in Jiuzhaigou earthquake (MS 7.0) perform rapid update LSM by means distributed scatterer interferometric synthetic aperture radar (DS-InSAR) technique. We selected territory County (southwestern China) as region. is world-renowned natural heritage tourist area human ecological value. For mapping, we examined applicability three models (logistic regression, support vector machine, random forest) for mapping offered strategy updating using DS-InSAR. First, logistic forest techniques, were built based on twelve contributing variables. Second, obtained best model parameters Bayesian network search, while five-fold cross-validation validate optimized model. According receiver operating characteristic curve (ROC), SVM RF had excellent prediction capability strong robustness over large areas compared with LR models. Third, surface deformation calculated DS-InSAR technology, data adopted correction matrix. resulted class transition 4.87 percent research practical examples, this method correcting LSMs continuous monitoring effective. Finally, analyze reasons change revised point out help restoration reducing susceptibility. results show that integration InSAR not only improved performance but also adapted it track evolution future susceptibility, including activities.