强降雨条件下滑坡体中优先流特征与滑坡触发机理研究

In mountainous regions in China, the frequently occurred earthquakes and extreme rainfall can trigger a large number of landslide disasters. Landslide material commonly has heterogeneous hydraulic and mechanical properties, and the infiltration may trigger preferential flow conducting through cracks and macropores. Preferential flow can cause fast response of the pore water pressure in deeper soils, and decrease the shear strength in slip surface. Such changes will affect both hydrology and slope stability. The commonly-used landslide models neglect the impact of preferential flow, which are not suitable to the heterogeneous landslide material. The accuracy of landslide predication highly relies on the quantification of the impacts of preferential flow on slope stability. The project chose a landslide-prone slope located in the Yindongzi valley in Dujiangyan County, in the central part of Sichuan Province. A hydro-mechanical model the was developed for this area by including the impact of preferential flow. The numerical modelling approach was then integrated with in situ measurement for early-warning of rainfall-triggered landslides. The model itself couples a dual-permeability model with a soil mechanics model, and the numerical results were validated by the observed data of displacement, soil moisture and pore water pressure, and the impact of preferential flow on slope stability was quantified. Moreover, the statistic approach was implemented to calculate the extreme rainfall under different return periods, in which the rainfall thresholds and probability of landslide occurrence were also analyzed. The proposed early-warning system can be used in mountainous regions to provide technique support for preventing and reducing landslide hazards.

在我国山区，地震与极端降雨频发，常常诱发大量次生滑坡灾害。受岩土异质性影响，降雨入渗可通过裂隙和大孔隙以优先流形式快速流动，造成深层土壤含水量和孔隙水压力快速响应，降低潜在滑动面的抗剪强度和稳定性。目前通用的滑坡预警模型忽略了优先流的影响，不适用于异质性显著的滑坡体。因此，定量研究优先流对滑坡触发的影响是提高灾害预警准确度的关键。本研究选取四川省都江堰市虹口乡的潜在滑坡体，将双重孔隙模型与边坡稳定性分析模型耦合，利用野外监测数据验证分析模型结果，研究优先流影响下的渗流模式和滑坡触发机理。另外，利用水文统计方法分析历史降雨数据，计算不同重现期的可能最大暴雨，分析极端降雨条件下滑坡灾害的触发概率和降雨阈值。最后，结合降雨预报进行实时滑坡灾害预警，服务于都江堰市虹口乡的安全保障。本研究构建的滑坡灾害预警体系可以推广至我国广大山区，为滑坡灾害预警及防灾减灾提供技术支持。

在强降雨条件下，优先流可能通过大裂缝、裂隙、生物孔隙等多种大孔隙通道向下快速传播，其速度将远远大于土壤基质，造成深层土壤及地下水的快速水文响应及溶质快速运移。另外，强降雨触发的优先流可增大深层土壤的孔隙水压力和土壤含水量，两者都可减小潜在滑动面上的抗剪强度。本研究将双重渗透模型与无限边坡稳定分析方法结合，用于模拟强降雨条件下优先流对土壤水动力过程及滑坡触发的影响。通过简单算例，模型首先和HYDRUS-1D数值软件的计算结果对比以验证本文构建模型的准确性。随后，模型将用于四川省都江堰市滑坡风险区的小规模人工降雨滑坡实验的观测数据中。分析表明，在汶川地震扰动区，岩土介质粗颗粒较多，具有砂砾土体沉积特征，因此在强降雨条件下的土壤含水量及孔隙水压力快速响应和优先流相关。在受地震影响的滑坡泥石流物源区，降雨入渗导致的次生滑坡受非均质土壤的水力学特性的影响较大。直接将双重渗透边坡水文模型与边坡稳定分析方法相耦合，可以量化土体非均质性对边坡稳定性的影响，为滑坡泥石流灾害预警提供更可靠的结果。