三峡库区缓倾角地层滑坡机理与预报判据研究

Landslides along the Three-Gorges Reservoir bank formed in the gently dipped strata, which represented in the districts of Wanzhou, Yunyang and Zhongxian, is a special category of landslide type. The displacement and evaluation of this kind of landslide is controlled by the sliding surface of a week layer which contains abundant montmorillonite. After the initial impoundment of the reservoir since 2003, large amount of displacement monitoring data of the landslides during the last decade in the reservoir indicated that the reactivation of landslides is variated seasonally including abrupt and progressive movements. Their influencing factors are mainly precipitation and fluctuation of the reservoir level. In terms of the creep theory, the reliability of practical prediction based on the monitoring data of displacement is not satisfied, which gives new requirement of the predictive criteria and models. This project will study the landslide mechanism in gently-dipped Strata. Furthermore choosing three typical active landslides as study cases which will been monitored of their surficial displacement, displacement in boreholes, underground water level, precipitation and elevation of reservoir level in four hydrological years, the project tries to study the predictive criteria and modes according to the combination between the landslide mechanism studies and the monitoring information.

三峡水库以万州、云阳、忠县库段为代表的侏罗系缓倾角地层库岸滑坡是一类特殊的滑坡类型，其滑动面的主要控制因素是富含蒙脱石矿物的软弱夹层，对滑坡的产生及位移的发展起到了地质上的控制作用。自2003年水库蓄水以来，这类库岸滑坡10多年来的变形监测结果表明，在水库水位升降变化和降雨的影响下，库岸滑坡变形呈现波动性特点，既有突发性的，也有渐进性的。在实际预报工作中，以单一位移监测曲线为依据，采用蠕变模型进行滑坡预报经常出现很大偏差，从而对目前的预报模型与判据研究提出了新的要求。本申请项目拟以缓倾角地层滑坡在水库运行条件下的变形破坏机理研究为基础，选择3个正在变形的典型滑坡，开展4个完整水文年的野外综合监测，获得滑坡地表位移、深部滑动面位移、地下水位变化、降雨过程及库水位等综合监测数据，结合室内试验和模拟计算等手段，研究滑坡动态变形规律，探索滑坡机理分析与监测信息相结合的预报判据与模式。

三峡水库万州、云阳、忠县库段存在有特殊的侏罗系缓倾角地层库岸滑坡，在水库水位升降变化和降雨的影响下，库岸滑坡变形呈现波动性特点，在实际预报工作中，以单一位移监测曲线为依据，采用蠕变模型进行滑坡预报经常出现很大偏差，从而对目前的预报模型与判据研究提出了新的要求。按照项目要求，完成了三峡库区缓倾角地层滑坡地质环境与成因规律研究，通过试验揭示了缓倾角地层滑坡滑动带软弱夹层特征与性质。通过监测数据统计分析及数值模拟，完成了滑坡在水库水位升降和降雨诱发条件下变形破坏机理的研究。采用InSAR和机器学习进行滑坡多源信息提取，构建多种算法耦合的预测预报模型，建立了地质信息、诱发因素为一体的滑坡综合预警预报判据。项目成果对三峡库区的防灾减灾和预警预报工作有着重要意义。