洪水(心理学)
脆弱性(计算)
风险评估
危害
计算机科学
自然灾害
脆弱性评估
风险分析(工程)
计算机安全
业务
地理
心理学
气象学
心理治疗师
化学
有机化学
心理弹性
作者
Zhu Han,Yasuhiro Mitani,Kohei Kawano,Hisatoshi Taniguchi,Hiromi Honda,Le Yin Meng,Zhiyuan Li
标识
DOI:10.1016/j.ijdrr.2023.104113
摘要
Flooding is a frequent natural hazard, threatening people's lives and properties worldwide. Even though completely preventing damage from flooding is impossible, an appropriate evacuation can help minimize its impact and mitigate human casualties. However, analyzing various flooding information to make quick evacuation judgments is challenging for decision-makers. Therefore, a quantitative flooding risk assessment is needed to support appropriate evacuation decisions. This study introduces the travel time under ideal evacuation conditions to quantify evacuation vulnerability, with Dijkstra's algorithm finding the evacuation route and the Random Forest algorithm predicting the evacuation time. The evacuation vulnerability is then integrated with time-varying flooding hazards and estimated congestion levels to assess evacuation risk. Herein, a case study in Joso, Japan is used as an example to demonstrate the practicality and applicability of the proposed approach. As a result, the prediction model achieved high accuracy with an RMSE of 0.88min and a MAPE of 5.88 %. The proposed approach clearly illustrated the vulnerable areas and times that may hinder an efficient evacuation process. Also, this approach identified and visualized the individual evacuation risk distribution during a given flooding scenario: 49.38 % of buildings were assessed as having the highest-risk: no evacuation ability, while a maximum of 11.86 % and 3.83 % of the buildings were assessed as middle-risk and high-risk. Furthermore, the regional evacuation deadline and priorities were specified to ensure timely and effective evacuation. Overall, this study develops an approach for assessing evacuation vulnerability and risk that could facilitate awareness among the public and provide judgmental bases for evacuation preparedness and decision-making.
科研通智能强力驱动
Strongly Powered by AbleSci AI