RAPID/Collaborative Research: Households' Immediate Protective Actions and Trade-Off Processes Between Property Security and Life Safety in Response to 2022 Hurricane Ian

近年来，气候变化加剧以及沿海社区人口增加，使居民面临飓风等破坏性极端天气事件的风险不断上升。这些威胁增加了对沿海家庭的风险认知和行为反应进行研究的紧迫性，这些研究可以改善保护生命和财产的政策。尽管取得了几十年的进展，但飓风疏散研究仍然缺乏对家庭在财产保护和生命安全之间权衡选择的充分理解。飓风伊恩是一场4级风暴，在最后一刻增强并改变了路径，影响了具有不同社会人口特征的广泛社区。这次飓风提供了一个独特的机会来研究家庭如何在他们的保护行动决策中处理相互冲突的目标，以及为什么他们推迟撤离甚至拒绝离开。这项快速反应研究拨款（Rapid）项目通过扩展保护行动决策模型（PADM）和收集佛罗里达州两个沿海县和一个内陆县的关键数据，提高了家庭对飓风威胁紧急反应的科学知识。这些成果促进了教学、培训和学习；增加代表性不足群体的参与；并显著改善美国沿海地区的应急管理和社区复原力。本项目解决了目前疏散文献中的四个局限性：1)飓风特征快速变化中的风险沟通挑战；2)基于线性统计方法的保护行动决策预测准确率较低；3)对房屋在极端环境条件下的脆弱性（例如，房屋对飓风级风速的实际强度）的家庭决策理解不足；4)在疏散决策过程中忽略了保护行动方案之间的权衡。为了应对这些研究挑战，该项目收集了关于住房损害的临时数据和关于家庭经历和看法的经验数据。五项任务包括：1)实地调查；2)入户调查；3)比较受访者对专家评估的风险水平的看法；4)传统线性回归分析与机器学习增强逻辑回归模型的比较；5)建立了多种保护行为决策的权衡模型。本项目提高了家庭在应对飓风紧急情况时的防护行动决策和防护行动方案之间权衡的知识状况。该奖项反映了美国国家科学基金会的法定使命，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。

In recent years, increased climate change together with rising populations in coastal communities has exposed residents to escalating risks of destructive extreme weather events such as hurricanes. These threats are increasing the urgency of conducting studies on coastal households’ risk perceptions and behavioral responses that could improve policies to protect lives and property. Despite decades of progress, hurricane evacuation studies still lack an adequate understanding of households’ decisions about the trade-offs between property protection and life safety among alternative protective actions. Hurricane Ian was a Category 4 storm with a last-minute intensification and track-change that affected a wide range of communities with diverse socio-demographic characteristics. This hurricane presents a unique opportunity to study how households dealt with conflicting objectives in their protective action decision making and why they delayed evacuating or even refused to leave. This Grant for Rapid Response Research (RAPID) project enhances scientific knowledge of households’ emergency responses to hurricane threats by extending the Protective Action Decision Model (PADM) and gathering critical data in two coastal counties and one inland county in Florida. The outcomes promote teaching, training, and learning; increase the participation of underrepresented groups; and significantly improve emergency management and community resilience across the US coastal areas.This project addresses four current limitations in the evacuation literature: 1) risk communication challenges amid rapid changes in hurricane characteristics; 2) the low accuracy in predicting protective action decisions based on linear statistical approaches; 3) an inadequate understanding of households’ decisions in the context of buildings’ vulnerabilities to extreme environmental conditions (e.g., the actual strength of houses against hurricane-force wind speeds); and 4) ignoring tradeoffs among protective action alternatives during the evacuation decision-making process. To tackle these research challenges, this project gathers ephemeral data on housing damage and empirical data regarding households’ experiences and perceptions. The five tasks include 1) field investigation; 2) household survey; 3) comparison of respondents’ perceptions against risk levels as evaluated by experts; 4) comparison of conventional linear regression analyses and machine-learning-enhanced logistic regression models; and 5) development of a trade-off model of decision making among multiple protective actions. This project advances the state of knowledge about households’ protective action decision making and tradeoffs among protective action PA alternatives in response to a hurricane emergency.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.