Tsunami evacuation planning for Camana, Peru: insights from agent-based modeling

Perez J.; Moya L.; Ramirez J.; Gonzales E.; Mas E.; Adriano B.; Koshimura S. · Natural Hazards (2026) · DOI: 10.1007/s11069-026-08244-0

Resumen

Coastal communities within the Ring of Fire, such as Camana, Peru, face significant risks from tsunamis due to their proximity to tectonic plate boundaries. This study employs agent-based modeling (ABM) integrated with reinforcement learning (RL) to assess and optimize tsunami evacuation strategies in the Camana resort area. Simulations focused on pedestrian evacuation under varying population densities, exploring the impacts of infrastructure modifications such as additional pathways and vertical evacuation structures. The results highlight that constructing additional pathways to evacuation zones and incorporating vertical evacuation structures significantly improve survival rates. Six vertical evacuation structures are recommended for a population of approximately 4500, while a larger transient population of 20,300 would require up to 14 structures. Additionally, this study emphasizes the importance of public education and rapid response in reducing evacuation delays, as shorter departure times significantly improve outcomes. By leveraging scientific tools, the proposed strategies offer a practical roadmap for decision-makers to improve tsunami preparedness and resilience in high-risk coastal areas. © The Author(s) 2026.

Instituciones: Pontificia Universidad Católica del Perú; Universidad Nacional de San Agustín de Arequipa