The necessity of macroscopic directional guidance in pedestrian tsunami evacuation: a zonal MFD-based optimization approach

Tsunami evacuation planning is a critical and urgent social issue that requires consideration of traffic impediments and road blockages caused by earthquakes and flooding. Traditional pedestrian evacuation models have focused primarily on micro-scale environments spanning a few dozen meters within buildings, limiting their applicability to tsunami scenarios involving travel over hundreds of meters. This study aims to (1) extend a cell FD-based dynamic traffic distribution model into a zonal MFD-based model and (2) apply it to pedestrian traffic to construct an optimization model that outputs efficient evacuation directions. Focusing on real local cities in Japan, we compare optimal solutions for different pedestrian evacuation direction scenarios. The results confirm that if pedestrians can preemptively understand multiple safe evacuation routes, rather than relying solely on the nearest facilities, efficient directional guidance that reduces congestion can be achieved.