Probabilistic Modeling of Spatio-temporally Varying Natural Hazard and Risk

大地震、强风和大雪会造成人员伤亡，破坏建筑物和基础设施系统，并中断业务。结构设计规范用于确保结构在自然荷载作用下的破坏概率低于可容忍水平。然而，公众对极端自然事件下多个建筑物倒塌的反应表明，仅考虑单个结构的安全性的设计是重要的，但对于空间分布的建筑物组合和空间扩展的基础设施来说，可能不符合最佳的社会利益。气候和构造特征的不均匀性导致自然灾害和作用于房屋和建筑物的荷载具有时空变化和相关性。风险的概率量化影响着风险评估和设计规范的更新。需要评估灾害对建筑物组合或空间分布式基础设施系统（如输电塔线系统）的影响，以成功实施减轻和减少灾难性损失的政策。该评估促进卓越的设计规范制定，优化资源分配，以减少自然灾害，维持我们的生活水平，并提高社会弹性。

Large earthquakes, strong winds and heavy snowfalls can cause fatalities, damage to structures and infrastructure systems, and interrupt business. The structural design codes are used to ensure that the failure probability of a structure due to natural loads is below a tolerable level. However, public reaction to collapse of multiple buildings under extreme natural events indicates that design by considering safety of a single structure alone is important, but may not be in the best societal interest for a portfolio of spatially distributed buildings and for infrastructure with spatially extended footprint. Inhomogeneity in climate and tectonic characteristics results in the natural hazards and loads acting on houses and buildings being spatio-temporally varying and correlated. Probabilistic quantification of the hazards influences the assessed risk and the design code updating. The impact of hazards on the portfolio of buildings or spatially distributed infrastructure systems, such as a power transmission tower-line system, needs to be appraised for successful implementation of policies to mitigate and reduce catastrophic losses. The appraisal promotes excellence in design code making, and optimal resource allocation to reduce natural disasters, maintain our standard of living, and improve societal resilience. The proposed research program deals with probabilistic modeling of spatio-temporally varying hazards and risk due to extreme wind, snow and earthquake events. More specifically, the program is to: A) map wind and snow hazards for Canada and Canadian design practice, and develop spatial correlation models for extreme winds and ground snow loads; B) Model and simulate tri-directional (3D) seismic ground motions fields conditioned on actual ground motion records and matching target spatial correlation and coherency; C) Estimate the risk to spatially distributed buildings and infrastructure under geographically varying natural hazards; and D) calibrate codified design criteria based on reliability and risk of a group of structures. The novelties of the proposal include the development of new wind and snow hazard maps for Canada using advanced techniques, the advancement of new models and algorithms for simulating conditional 3D ground motion fields with realistic spatial correlation and coherency, the quantitative reliability and risk assessment of groups of structures or spatially distributed infrastructure system under realistic temporally and geographically varying natural hazards, and the incorporation of advanced decision theory under uncertainty for making structural design codes. The funding requested for the research program is to support and train graduate students who will contribute to the society's effort to reduce natural disasters, and to build and maintain a resilient society under the threat of spatio-temporally varying natural hazards.