Summers J Kevin, Lamper Andrea, McMillion Courtney, Harwell Linda
Gulf Ecosystem Monitoring and Modeling Division U.S. Environmental Protection Agency Gulf Breeze FL USA.
Geohealth. 2022 Oct 1;6(10):e2022GH000660. doi: 10.1029/2022GH000660. eCollection 2022 Oct.
Users can apply three processes to develop confidence in decision-making tools like models and indices-validation, verification, and observation. The utility of the Cumulative Resilience Screening Index (CRSI) was demonstrated by combining the processes of verification and observation using real-world natural hazard events (i.e., hurricanes, inland flooding, and wildfires). The ability of CRSI to determine the counties most vulnerable to hazards and least likely to recover quickly from natural hazards is demonstrated using these natural hazard events from outside the original index construction data set. Using Hurricane Harvey and Hurricane Michael, the counties in Texas and Florida/Georgia, respectively, experiencing the most damage and the most extended recovery intervals were determined accurately. Similarly, the most vulnerable and least recoverable counties were correctly identified as those associated with the Great Louisiana Flood of 2016. Finally, three different types of wildfires in California were examined to determine the likelihood of recovery and the strength of pre-event planning. All models and indices developed for use by decision-makers should consider undertaking this verification or a similar validation operation to enhance user confidence.
用户可以应用三个过程来增强对模型和指数等决策工具的信心——验证、核实和观察。通过使用现实世界中的自然灾害事件（即飓风、内陆洪水和野火）将核实和观察过程相结合，证明了累积恢复力筛选指数（CRSI）的效用。利用原始指数构建数据集之外的这些自然灾害事件，证明了CRSI确定最易受灾害影响且最不可能迅速从自然灾害中恢复的县的能力。使用哈维飓风和迈克尔飓风，分别准确确定了德克萨斯州和佛罗里达州/佐治亚州遭受破坏最严重和恢复间隔最长的县。同样，最脆弱和最不易恢复的县被正确识别为与2016年路易斯安那大洪水相关的县。最后，对加利福尼亚州的三种不同类型的野火进行了研究，以确定恢复的可能性和事件前规划的力度。为决策者开发的所有模型和指数都应考虑进行这种核实或类似的验证操作，以增强用户信心。
