Optimizing evacuation routes for human mobility during wildfires: A case study of the 2023 McDougall Creek Wildfire
利用GPS数据分析了2023年麦克杜格尔溪野火中的拥堵瓶颈，开发了动态Dijkstra-A*算法，通过多准则成本函数整合距离、拥堵和火灾风险，模拟验证了分区域错峰疏散能有效降低峰值拥堵、提升安全性。
Wildfires present significant challenges to evacuation planning due to their dynamic nature, rapid spread, and the limitations of static routing methods, which often lead to congestion and increased safety risks. This study addresses these issues by analyzing the 2023 McDougall Creek Wildfire (Kelowna, BC) using GPS data, identifying congestion bottlenecks, and developing a dynamic Dijkstra-A* algorithm with a multi-criteria cost function that integrates distance, congestion, and fire risk. Validated through SUMO simulations across two scenario groups with fire origins in the northwest and southwest, we tested four evacuation strategies: simultaneous departure, temporally staggered departure, region-based evacuation with uniform response times, and region-based evacuation with realistic response time variability. Our results demonstrate that region-based, staggered evacuations with realistic response times effectively reduce peak congestion and improve safety compared to simultaneous approaches. This research highlights the potential of GPS-informed behavior and hazard-aware routing to improve adaptive evacuation strategies for climate-driven wildfire events.