ABSTRACT

Increasing intensity and frequency of hurricane events underscores the need for efficient and inclusive evacuation plans, particularly for carless and disabled populations. Hurricane evacuation intrinsically involves both social and spatial processes. People’s decision to evacuate spreads over social networks; once their decisions are made, they flee through spatial transportation networks. This article describes a novel effort to integrate socio-spatial networks into an agent-based evacuation simulation model, taking the Florida Keys in the USA as a study area. In the model, households, as agents, were synthesized from Census data, then connected by a ‘home-workplace-neighborhood’ social network, and registered to a spatial road network. A threshold decision model was used to simulate social contagion of households’ decision to evacuate. The resulting travel demands were input into the TRANSIMS platform to generate on-road traffic. The model analyzed scenarios of automobile-only and public transit-assisted evacuation. The results show that the simulated traffic under the automobile-only scenario aligns with the observed traffic dynamics, which validates our socio-spatially integrated model. Adding public transportation capacity significantly reduces the traffic load and evacuation time, and provides a practical, accessible, and equitable route to safety for low mobility populations.

摘要

飓风事件的强度和频率不断增加，凸显了制定有效和包容性疏散计划的必要性，尤其是针对无车和残疾人士的疏散计划。飓风疏散本质上涉及社会和空间过程。人们撤离的决定通过社交网络传播；一旦做出决定，他们就会逃离空间交通网络。本文以美国佛罗里达群岛为研究区域，描述了将社会空间网络集成到基于代理的疏散模拟模型中的新尝试。在该模型中，作为代理的家庭由人口普查数据合成，然后通过“家庭-工作场所-社区”社交网络连接，并注册到空间道路网络。阈值决策模型用于模拟家庭疏散决策的社会传染性。由此产生的出行需求被输入到 TRANSIMS 平台中以生成道路交通。该模型分析了仅汽车和公共交通辅助疏散的情景。结果表明，仅汽车场景下的模拟交通与观察到的交通动态一致，这验证了我们的社会空间整合模型。增加公共交通容量可显着减少交通负荷和疏散时间，并为行动不便的人群提供一条实用、方便且公平的安全路线。这验证了我们的社会空间整合模型。增加公共交通容量可显着减少交通负荷和疏散时间，并为行动不便的人群提供一条实用、方便且公平的安全路线。这验证了我们的社会空间整合模型。增加公共交通容量可显着减少交通负荷和疏散时间，并为行动不便的人群提供一条实用、方便且公平的安全路线。