Collaborative Research: A Data-centric Uncertainty-informed Framework for Resilience Analytics of Critical Infrastructure Under Extreme Climate Events

美国的关键基础设施和依赖其服务的社区越来越容易受到气候风险的影响，造成广泛的影响，随后往往需要进行漫长和昂贵的恢复工作。从根本上需要在地方和区域范围内建立可扩展和准确的自然灾害风险预测模型，以更好地评估和管理我国基础设施的复原力。这项研究的结果预计将有助于政策制定者和基础设施运营商在各种不确定的未来情景下描述基础设施的弹性，并确定最佳的适应或缓解战略，从而在系统中实现最大的弹性增益。此外，该项目还具有通过真正的跨学科、研究综合教育计划教育下一代学者进行灾害建模的其他积极社会影响的巨大潜力，致力于增加劳动力培训的多样性，并将成果广泛传播给科学界和利益相关者。本研究项目旨在通过建立一个这是一个多元化、以数据为中心和不确定性为基础的框架，可有效地描述随机灾害下的多维基础设施复原力以及合理的基础设施演变（由于适应或缓解战略）和气候变化情景。这将通过实现以下三个关键目标来实现：（1）建立一个准确的多范式飓风风险模型，（2）建立一个准确的预测框架，用于关键基础设施的弹性分析，基于多维贝叶斯算法，以及（3）利用随机分析的最新进展-基于多项式混沌代理-充分表征与多维弹性模型相关的不确定性，并实现计算效率高的基于MIMO的敏感性分析。该项目的成功实施将在复原力建模方面取得重大突破，通过对基础设施和社区复原力进行可扩展、准确和多维度的评估，并严格有效地考虑不确定性。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The United States' critical infrastructure and the communities that rely on their services are increasingly prone to climatic risks, with widespread impacts that are often followed by lengthy and costly restoration efforts. There is a fundamental need for scalable and accurate prediction models of natural hazard risks at local and regional scales to better assess and manage the resilience of our nation's infrastructure. The outcome of this research is expected to help policy makers and infrastructure operators characterize infrastructure resilience under various uncertain future scenarios and identify the optimal adaptation or mitigation strategies that result in maximum resilience gain in the system. In addition, this project possesses great potential for other positive societal impacts by educating the next generation of scholars in hazard modeling through a truly interdisciplinary, research-integrated educational program, a commitment to increased diversity in workforce training and broad dissemination of the results to scientific communities and stakeholders.This research project aims to advance the theory and practice of resilience engineering through establishing a pluralistic, data-centric and uncertainty-informed framework to efficiently characterize the multi-dimensional infrastructure resilience under stochastic hazards as well as plausible infrastructure evolution (due to adaptation or mitigation strategies) and climate change scenarios. This will be done through implementing the three key objectives of: (1) creating an accurate and multi-paradigm hurricane risk model, (2) establishing an accurate predictive framework for resilience analytics of critical infrastructure, based on a multi-dimensional Bayesian algorithm, and (3) leveraging recent advancements in stochastic analysis - based on Polynomial Chaos surrogates - to both fully characterize the uncertainties associated with the multi-dimensional resilience model, and implement computationally efficient scenario-based sensitivity analysis. Successful implementation of this project will yield a significant breakthrough in resilience modeling by enabling a scalable, accurate, and multi-dimensional assessment of infrastructure and community resilience; with rigorously and efficiently accounting for uncertainties.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.