石固林, 陈强, 刘先文, 等. 2022. 联合升降轨Sentinel-1A数据监测桃坪乡古滑坡沿坡向的形变速度场[J]. 工程地质学报, 30(4): 1350-1361. doi: 10.13544/j.cnki.jeg.2020-016.
    引用本文: 石固林, 陈强, 刘先文, 等. 2022. 联合升降轨Sentinel-1A数据监测桃坪乡古滑坡沿坡向的形变速度场[J]. 工程地质学报, 30(4): 1350-1361. doi: 10.13544/j.cnki.jeg.2020-016.
    Shi Gulin, Chen Qiang, Liu Xianwen, et al. 2022. Deformation velocity field along aspect direction of an ancient landslide at Taoping Village derived from ascending and descending Sentinel-1A data[J]. Journal of Engineering Geology, 30(4): 1350-1361. doi: 10.13544/j.cnki.jeg.2020-016.
    Citation: Shi Gulin, Chen Qiang, Liu Xianwen, et al. 2022. Deformation velocity field along aspect direction of an ancient landslide at Taoping Village derived from ascending and descending Sentinel-1A data[J]. Journal of Engineering Geology, 30(4): 1350-1361. doi: 10.13544/j.cnki.jeg.2020-016.

    联合升降轨Sentinel-1A数据监测桃坪乡古滑坡沿坡向的形变速度场

    DEFORMATION VELOCITY FIELD ALONG ASPECT DIRECTION OF AN ANCIENT LANDSLIDE AT TAOPING VILLAGE DERIVED FROM ASCENDING AND DESCENDING SENTINEL-1A DATA

    • 摘要: 单一平台或轨道的InSAR(Interferometric Synthetic Aperture Radar,InSAR)技术通常只能获取到滑坡体沿雷达卫星视线(LOS)方向的一维形变信息,无法直接获取到滑坡体沿坡向的形变信息,使得InSAR技术无法充分满足滑坡监测的工程需求。针对这一问题,本文提出联合升降轨Sentinel-1A数据提取滑坡沿坡向形变速度场的模型算法,其基本方法是:首先利用InSAR技术分别获取升轨和降轨沿LOS方向的形变速度场,然后根据坡向和坡度建立滑坡的滑动面坐标系,基于LOS方向与滑动面的几何关系对升降轨对应的LOS向形变速度值进行建模,在滑动面上不存在法线方向形变的合理假设下通过解算独立方程组获得滑坡体沿坡向的形变速度场和沿垂直坡向的形变速度场。本文选取四川省理县桃坪乡古滑坡作为研究对象,并通过模拟实验对模型的解算精度进行验证,结果表明:相比于LOS向形变速度场,重建的沿坡向和沿垂直坡向形变速度场更加清晰地揭露了滑坡体的主要位移模式和危险性较高的坡体区域。同时,模拟实验结果表明,模型算法的解算误差与噪声水平大致呈线性增加的关系,当噪声水平达到10 mm·a-1时,沿坡向形变速度场的解算误差为14.1 mm·a-1,沿垂直坡向形变速度场的解算误差为11.9 mm·a-1,证明了模型算法以及结果的可靠性。

       

      Abstract: InSAR technology is based on a single satellite platform or orbit and can derive only one-dimensional deformation information of the landslide in the direction of the line of sight(LOS) of the radar satellite. It cannot derive the deformation along its aspect direction and cannot meet well the engineering requirements for landslide monitoring. To obtain more direct landslide deformation information of interest,we develop the model and algorithm for extracting the deformation velocity field of landslides in the aspect direction from ascending and descending Sentinel-1A data. Our method includes the following steps:(1)deriving LOS deformation velocity fields,(2)defining the slope-based coordinate system,(3)modeling the LOS deformation according to the configuration between the satellites and surface slopes and (4)solving for the displacements of landslides in the aspect and aspect-perpendicular directions with the assumption of no deformation in the slope surface-normal direction. We conduct a case study of an ancient landslide at Taoping Village,Li County in Sichuan Province and verify the accuracy of the model and algorithm through simulation experiments. The results show that,in contrast with the LOS deformation,the derived deformation in the aspect and aspect-perpendicular directions can display clearer deformation patterns and high-susceptibility areas. According to the results of the simulation experiments,the calculation error of the model and algorithm increases approximately linearly with the noise level. When the noise level reaches 10 mm·a-1,the calculation errors of the deformation velocity field in the aspect and aspect-perpendicular directions are 14.1 and 11.9mm·a-1,respectively. This proves the reliability of our method and results.

       

    /

    返回文章
    返回