述评库岸崩滑机理研究方法的研究现状、存在问题和发展方向。监测法具有速度快、自动化程度高等优势， 但易受到天气、环境等因素的影响。调查法具有灵活性强且适用范围广等特点，但调查法不能全面地获取研究区 灾害信息。监测法和调查法都能为模型试验和数值模拟提供校正数据，但是二者均难以观测库岸崩滑发生、发展 的过程。模型试验不仅能对下垫面形状进行情形假定，还能够在实验室条件下对崩滑过程进行精确观测，但模型 试验必须考虑模型与原型的相似问题。现场试验法不用重塑下垫面地形，易于对复杂的水土流失现场进行模拟， 但在野外往往难以进行精细的模拟和观测。数值法能对较大尺度的库岸崩滑进行统计分析或者预测，但其计算 结果须通过监测、调查或者试验的结果来验证。因此，应根据项目的研究目标和研究内容确定相应的研究方法。

Reservoir-bank avalanche is a common hydrogeological problem whether in China or other countries, and analysis of the mechanism of collapse on reservoir bank is an important issue in the field of disaster prevention and mitigation. In virtue of many advantages, e.g., high efficiency and automation, in-site monitoring has become a powerful tool in evaluating reservoir-bank collapse. Nevertheless, in-site monitoring is also subject to some external factors related to remote sensing technology, such as weather conditions, environment, and other factors. Field investigation has the characteristics of flexibility and maneuverability. However, a field investigation is a sampling method based on the observation in the representative area, and the hazard information can not be comprehensively obtained via field investigation in the study area. Common merit for the in-site monitoring and field investigation is that data observed with the methods mentioned above may be used to calibrate the results of the model experiment and numerical simulation, but in-site monitoring or field investigation are not applicable to track the process of a bank collapse. Different from the monitoring or investigating method, an experiment of reservoir-bank collapse under closely monitored or controlled experimental conditions focuses on the regularities of bank collapse from a micro perspective. The model experiment may be conducted in a laboratory far away from the study area, and the experimental scenario may be freely designed if needed. Yet the researchers should consider the similarity of the cumulative effect while designing the downscaled model experiment for reservoir bank failure. Maybe it is easy to simulate and observe the complicated topographical conditions of bank collapse in the field experiment because no change exists in the scale of the underlying surface and properties of erosion material. Usually making accurate observations and simulations is relatively difficult in the field. Numerical simulations have been widely used to analyze and predict the reservoir-bank avalanche from a macro perspective all over the world, whereas the result of the numerical simulation has to be verified with that obtained from monitoring, investigation, or experiment. In summary, each research method presents its own characteristic set of advantages and limitations. Scientists may use an appropriate analysis route according to the objectives and contents of a specific project.Shortly, improved methods for studying the mechanism of avalanches on reservoir banks may be presented through the following aspects: (1) To consider the coupling effects of multiple impact factors in the process of reservoir-bank avalanches. Rainfall, wave, and fluctuation of reservoir water level are the main external factors causing the deformation of landslides. However, how the reservoir-bank avalanche goes coupled with multiple factors is still at the exploratory stage. Hence, to make a further study of the above contents is surely necessary. (2) To study the mechanism of avalanches on reservoir banks with multi-methods. Usually, a single research method may be particularly vulnerable to weather, environment, and other factors, and some methods can not be used to obtain detailed disaster information required by on-site command quickly and comprehensively. In addition, limitations exist in considering all the stress conditions and multi-dimensional stress states of the reservoir slopes. Hence, future research may be focused on combining the strengths and weaknesses of different kinds of research methods to study the formation mechanism and protection measures of reservoir-bank avalanche.