Development of tools for landslide hazards assessment and mitigation

滑坡灾害对包括加拿大在内的世界许多地区的人类发展构成重大制约。由于人口和基础设施的增长以及气候变化的影响，滑坡灾害的重要性不断增加。地质和岩土工程专业人员有减少滑坡损失的任务。为了履行这一职责，从业人员需要工具来预测滑坡的发生和行为。我在UBC的研究小组多年来一直处于量化滑坡运动研究的前沿。世界各地的研究人员和从业人员都在使用我们的方法，但它们需要进一步改进。我建议在几个具体的方向上继续发展我们的技术。 第一个是制定一个系统和客观的方法来校准模型。我们的动态建模方法认识到滑坡过程的复杂性（例如Hungr等人，2014年），并赞成一个简化的方法，基于反分析的全面滑坡的情况下。我们相信这是开发实际可用工具的最佳方法，并且我们已经看到这种方法取得了相当大的成功（例如Hungr等人，2007年）。然而，滑坡行为是一个多维过程。校准需要仔细的描述和多个真实的情况下，一个类型分类能够排序的情况下，根据其运动的物理和选择最佳的校准输入参数的客观方法的发展进行反分析。作为这项工作的一部分，还需要研究模型对输入参数（如数字高程阵列）的敏感性，并制定具体的改进措施，如考虑移动滑坡体的内部刚度，并纳入更广泛的流变学来描述水流深度、速度和加速度之间的关系。这项工作的大部分已经在进行中，作为正在进行的博士学位的一部分。项目（例如Aaron和Hungr，2015）。 第二个专题涉及制定一种新的方法来计算泥石流或泥石流雪崩对由钢网组成的柔性障碍物施加的力。一个硕士项目最近已经完成（阿什伍德和亨格，2015年），但该主题需要更多的数据和更详细的分析。我建议这样做，作为另一个硕士项目的一部分，建立在以前的结果。 一位来自北京清华大学的访问学生明年将与我的团队合作，研究快速滑坡与水库水的相互作用。这对我来说是一个新的课题，但实际上是一个重要的课题。 第四个潜在的方向是继续开发我们的新算法，用于模拟破碎岩石坠落的运动。我的小组正在考虑与上述主题有关的几个国际合作研究机会。

Landslide hazards pose significant constraint to human development in many parts of the world, including Canada. The importance of landslide hazards is continuously increasing, due to growing population and infrastructure and due to the effects of climate change. Geological and Geotechnical Engineering professionals have the task of reducing landslide losses. To fulfill this role, practitioners require tools to predict landslide occurrence and behavior. My research group at UBC has for a number of years been at the forefront of research devoted to quantifying landslide motion. Our methods are in use by researchers and practitioners around the world, but they require further improvement. I propose to continue development of our techniques in several specific directions. The first one is to formulate a systematic and objective approach towards calibration of models. Our approach towards dynamic modeling recognizes the complexity of landslide processes (e.g. Hungr et al., 2014) and favours a simplified approach based on back-analysis of full scale landslide cases. We believe that this is the best method to develop practically-useable tools and we have already seen considerable success with this approach (e.g. Hungr et al., 2007). However, landslide behavior is a multi-dimensional process. Calibration requires careful description and back-analysis of multiple real cases, the development of a typological classification able to sort cases according to the physics of their motion and an objective method of selecting optimal calibrated input parameters. As part of this work, it is also necessary to study the sensitivity of the models to input parameters such as the digital elevation array and to develop specific improvements such as accounting for internal rigidity of a moving landslide mass and incorporating a wider range of rheology to describe the relationships between flow depth, velocity and acceleration. Much of this work is already in progress as part of an on-going Ph.D. project (e.g. Aaron and Hungr, 2015). The second topic involves the development of a new method to calculate the forces exerted by a debris flow or debris avalanche on a flexible barrier, consisting of a steel net. A Master's project has recently been completed (Ashwood and Hungr, 2015), but the topic requires additional data and more detailed analysis. I propose to do this as part of another Master's project, building on the results of the previous. A visiting student from the Tsinghua University in Beijing will work with my team next year on modeling interaction of rapid landslides with reservoir water. This is a new subject for me, but one that is practically important. A fourth potential direction is to continue the development of our new algorithms for modeling the movement of fragmental rock fall. Several opportunities for international collaborative research, connected with the above topics are being considered by my group.