Abstract We process 69 Sentinel-1A satellite images using SBAS-InSAR from July 2018 to August 2019, and obtain surface deformation in Shuicheng district, Liupanshui city as dynamic evaluation factors to address the lack of dynamic feature data in traditional landslide susceptibility studies. The results show that, by fusing ten static evaluation factors and InSAR deformation feature data as dynamic evaluation factors, in a weighted information model coupled with analytic hierarchy process(AHP) and information volume method, the model performance improves approximately 13.3% compared to using only static feature data. The area under the ROC curves is 0.756 02 and 0.888 68, respectively. To assess zoning accuracy, we overlay historical disaster sites on two types of zoning maps. Compared to scenarios without the inclusion of deformation features, the introduction of deformation features corrected approximately 12.44% of misclassified areas, significantly enhancing zoning reliability.
XIAO Haiping,WAN Junhui,CHEN Lanlan et al. Landslide Susceptibility Assessment by Fusing InSAR Deformation eatures under the Support of Weighted Information Volume[J]. jgg, 2024, 44(7): 718-724.
XIAO Haiping,WAN Junhui,CHEN Lanlan et al. Landslide Susceptibility Assessment by Fusing InSAR Deformation eatures under the Support of Weighted Information Volume[J]. jgg, 2024, 44(7): 718-724.