Impact of real-time shelter capacity information on tsunami evacuation efficiency

View at publisher
Authors: Tomoyuki Takabatake, Nanami Hasegawa, Sorayasu Iwane
Cited by: 0
References: 58

Abstract

Efficient tsunami evacuation is crucial for reducing casualties; however, evacuees often face shelter capacity constraints and re-evacuation. Although real-time shelter capacity updates and community-based information sharing can improve evacuation efficiency, their effectiveness remains unclear. This study developed an agent-based tsunami evacuation model that incorporated three information-sharing strategies: no information sharing (No-Info), real-time updates provided to all evacuees (Broadcast), and peer-to-peer information exchange through evacuee encounters (Relay). By comparing evacuation outcomes under these different strategies, this study assessed the effectiveness of real-time updates in reducing evacuation time and the extent to which community collaboration compensates for the limitations in official capacity updates. The results showed that Broadcast consistently reduced evacuation time across all study conditions, increasing completion rates by 5–10% compared to No-Info. Relay also improved completion rates, exceeding No-Info by 2–6%, but its effectiveness remained consistently lower than that of Broadcast. Although both strategies enhanced overall evacuation efficiency, the results revealed that they did not benefit all evacuees equally, as some evacuees rerouted prematurely toward smaller, lower-capacity shelters in these strategies. This unintended effect suggests that although real-time shelter capacity information generally enhances evacuation efficiency, it must be carefully designed to prevent suboptimal routing decisions.

Fields of Study & Subjects

Evacuation and Crowd Dynamics Facility Location and Emergency Management Disaster Management and Resilience

Publication details

Journal / source
Georisk Assessment and Management of Risk for Engineered Systems and Geohazards
Pages
1–19
Publication date
13 January 2026
Language
EN