【作者】 李丹；

【导师】 吴浩；

【作者基本信息】 华中师范大学 ， 地图学与地理信息系统， 2021， 硕士

【摘要】 滑坡是一种非常严重的地质灾害,它不仅严重威胁着人类的生命财产安全,同时也对环境、资源等产生巨大的破坏。2020年梅雨季节,受强降雨的影响,湖北省恩施市发育了大量滑坡灾害。其中,沙子坝滑坡垮塌导致居民房屋和屯渝公路受损,水电站引水渠冲毁,滑坡物质进入清江堵塞了河道形成堰塞湖。此次滑坡事故受到了社会各界的关心,为避免和减轻滑坡灾害带来的严重后果,需要对恩施地区开展滑坡形变监测和易发性评价研究,保障人民的生命财产安全。合成孔径雷达干涉测量是一种新型的遥感技术,其在数据获取上具有全天时、全天候、不受天气因素影响的特点,在对地观测上表现出监测范围广,精度高的突出优势,使得利用InSAR技术进行滑坡研究成为可能。然而,如何将InSAR遥感技术手段应用于滑坡灾害的形变监测与特征分析,仍是一个处于探索阶段且亟待解决的问题。同时,时序InSAR重点区域的监测需要根据滑坡灾害发育的实际情况和发生概率进行选择,因此有必要对区域进行滑坡易发性评价,从而更好地服务于滑坡的防灾减灾,最大程度地保障人民的生命财产安全。因此,本文利用时序InSAR技术和信息量法模型开展恩施沙子坝滑坡的形变监测与区域易发性评价研究,主要研究内容如下:(1)GACOS辅助下的鄂西南地区InSAR大气效应改正。使用外部数值气象模型GACOS对鄂西南地区InSAR数据中的大气噪声进行去除,并利用相位标准差变化定量评价时序InSAR大气改正效果。(2)沙子坝滑坡时间序列InSAR形变监测。采用SBAS-InSAR技术对沙子坝滑坡滑前和滑后两个不同状态下的形变速率进行解算,得到滑坡前与滑坡后的时序形变速率和累积形变量,同时,对形变速率和降雨量进行相关性分析。(3)沙子坝滑坡失稳垮塌前后形变演化特征时序分析。利用滑坡体特征点的累积形变量的变化,分析出滑坡发生过程的4个不同阶段。(4)滑坡垮塌的失稳信号探测。根据累积形变量的变化特征,对滑坡垮塌前的加速信号进行提取捕捉。(5)恩施区域滑坡易发性评价。收集整理恩施市地质灾害点384个,选用了 8个滑坡地质灾害易发程度影响因子进行了分析,并利用信息量法,对恩施市进行了滑坡灾害易发程度进行评价。结果表明,使用外部数值气象模型GACOS可以有效去除时序InSAR数据中的大气影响效应,去除前后大气噪明显较少,干涉图的质量得到改善,滑坡时序分析精度得到提高。沙子坝滑坡在滑前,即2018年1月11日至2020年7月11日之间,LOS方向平均形变速率-120mm/year～40mm/year,最大累积形变量达到了300mm。在滑坡发生后,即2020年7月23日至2021年1月31日之间,滑坡体的视线向平均形变速率-90mm/year～70mm/year,最大累积形变量为50mm。滑坡前沙子坝滑坡时间演化分为四个阶段,即缓慢变形阶段、加速阶段1、暂时稳定阶段和加速阶段2。滑坡后时间演化特征表现为2个阶段,即不稳定变形阶段和稳定变形阶段。成功探测到了滑坡失稳加速信号,即在2020年5月12日,滑坡发生的前72天,滑坡体位移速率发生突变,开始加速变形,最终导致滑坡的发生。对恩施市降雨量数据与形变量进行了耦合分析,发现它们具有明显的相关性。恩施市滑坡易发性评价结果表明,恩施市滑坡发育风险较高的地方为清江流域与断层分布区域,尤其是东部地区,沙子坝滑坡位于易发性评价中的极高易发区,验证了易发性评价结果的准确性,根据滑坡易发性评价结果,可筛选出极高易发区,对这些区域进行时序InSAR的重点监测,在一定程度上可提前预知滑坡灾害的发生。更多还原

【Abstract】 Landslide is one of the most serious geological disaster.It not only seriously threatens the safety of human life and property,but also causes huge damage to the environment and resources.In the plum rain season of 2020,affected by heavy rainfall,a large number of landslide disasters were happened in Enshi City,Hubei Province.Among them,the collapse of the Shaziba landslide caused damage to residential houses and the Tun Yu highway,the water diversion channel of the hydropower station was washed away,and the landslide material entered the Qingjiang River and blocked the river channel to form a barrier lake.The landslide accident has attracted the attention of all sectors of society.In order to avoid and reduce the serious consequences of landslide disasters,it is necessary to carry out landslide deformation monitoring and susceptibility evaluation research in Enshi to ensure the safety of people’s lives and properties.Synthetic aperture radar interferometry(InSAR)is a new type of remote sensing technology.It has the characteristics of all-weather,all-weather,and unaffected by weather factors in data acquisition.It shows the outstanding advantages of wide monitoring range and high accuracy in earth observation.This makes it possible to use InSAR technology for landslide research.However,how to apply InSAR remote sensing technology to deformation monitoring and characteristic analysis of landslide hazards is still an exploratory problem that needs to be solved urgently.At the same time,the time-series InSAR monitoring in key area need to be selected according to the actual situation and probability of landslide disaster development.Therefore,it is necessary to evaluate the landslide susceptibility of the area,so as to better serve the disaster prevention and mitigation of landslides,and to ensure maximum protection The safety of people’s lives and properties.Therefore,this paper used the time series InSAR technology and the information method model to carry out the deformation monitoring and regional susceptibility evaluation of the Enshi Shaziba landslide.The main research contents are as follows:(1)InSAR atmospheric effect correction in southwestern Hubei assisted by GACOS.The external numerical weather model GACOS was used to remove atmospheric noise from InSAR data in southwestern Hubei,and the phase standard deviation change was used to quantitatively evaluate the effect of time series InSAR atmospheric correction.(2)InSAR deformation monitoring of Shaziba landslide time series.The SBASInS AR technology is used to calculate the deformation rate of the Shaziba landslide before and after the landslide in two different states,and the time series deformation rate and cumulative deformation before and after the landslide are obtained.At the same time,the deformation rate and rainfall are calculated.Correlation analysis.(3)Time series analysis of deformation evolution characteristics before and after Shaziba landslide collapse.Using the cumulative deformation changes of the characteristic points of the landslide body,four different stages of the occurrence process of the landslide are analyzed.(4)Instability signal detection of landslide collapse.According to the change characteristics of the cumulative deformation,the acceleration signal before the collapse of the landslide is extracted and captured.(5)Evaluation of landslide susceptibility in Enshi area.This work collected and sorted out 384 geo-hazard sites in Enshi City,selected 8 landslide geo-hazard susceptibility factors for analysis,and used the information method to evaluate the landslide hazard susceptibility in Enshi City.The results show that the use of the external numerical meteorological model GACOS can effectively remove the atmospheric effects on the time series InSAR data,the atmospheric noise after removal is significantly less,the quality of the interferogram is improved,and the accuracy of the landslide time series analysis is improved.Before the Shaziba landslide,that is,between January 11,2018 and July 11,2020,the average deformation rate in the LOS direction was-120mm/year～40mm/year,and the maximum cumulative deformation reached 300mm.After the landslide occurs,that is,between July 23,2020 and January 31,2021,the average line-of-sight deformation rate of the landslide body is-90mm/year～70mm/year,and the maximum cumulative deformation is 50mm.The evolution of the Shaziba landslide before the landslide is divided into four stages,namely the slow deformation stage,the acceleration stage 1,the temporary stabilization stage and the acceleration stage 2.The time evolution characteristics of the post-landslide show two stages,namely the unstable deformation stage and the stable deformation stage.Successfully detected the acceleration signal of landslide instability,that is,on May 12,2020,72 days before the occurrence of the landslide,the displacement rate of the landslide body suddenly changed and the deformation began to accelerate,which,eventually led to the occurrence of the landslide.Coupling analys is of rainfall data and shape variables in Enshi City,and found that they have obvious correlation.The evaluation results of landslide susceptibility in Enshi City show that the areas with higher landslide development risk in Enshi City are the Qingjiang River Basin and fault distribution areas,especially in the east.The Shaziba landslide is located in the extremely high-prone area in the susceptibility evaluation,which verifies that the landslide is prone According to the accuracy of the landslide susceptibility evaluation results,extremely high-prone areas can be screened out,and time-series InSAR key monitoring of these areas can be performed to predict the occurrence of landslide disasters in advance.更多还原