Item

Abstract

Landslide is a geological disaster with extremely destructive effects, resulting in huge casualties and economic losses in China. The Loess Plateau is widely covered by several to tens of meters of loess, and the underlying bedrock with good water barrier properties. Due to the frequent rains, the soil body is easy to flow or slide along weak structural surfaces. There are lots of typical loess landslides in the Loess Plateau and they seriously affect the lives of residents, so studying landslide deformation in Loess Plateau is of great significance for both society and geological expert. Interferometric Synthetic aperture radar (InSAR), with the advantages of wide monitoring range, high density, high accuracy, and not affected by weather conditions, has become the most effective technical means for regional surface deformation monitoring and landslide identification. In this paper we perform landslide deformation survey based on the small baseline Subset InSAR (SBAS-InSAR) method in Tianshui, which is located on the Loess Plateau, using 594 interferograms from the Sentinel-1 satellite ranging from January, 2017 to December, 2022. SBAS-InSAR time series analysis connects independent SAR images based on certain spatial baseline and time baseline thresholds, and finally gets time series and velocity of the Loess Plateau. The locations of landslides from National Geological Disaster Survey Database provided by China Geological Survey are compared with our results to verify the applicability of SBAS-InSAR technology in the Loess Plateau. The comparison results show that SBAS-InSAR technique using sentinel-1 dataset can effectively identify landslides in most areas except for the areas covered by forest. The results of velocity map and landslide maps can be used for landslide identify and assessment in the Loess Plateau.