Author(s): Joseph Kim; Ioan Nistor
Linked Author(s): Ioan Nistor
Keywords: Tsunami Evacuation Agent-Based Model
Abstract: Background Swift evacuation is often the sole viable strategy to mitigate loss-of-life during tsunamis for coastal communities (Shibayama et al., 2013). Traditional hard countermeasures such as sea walls and breakwaters, though effective in delaying wave impacts, are economically and socially challenging to implement comprehensively. Instead, soft measures like evacuation drills and early-warning systems are increasingly recognized as crucial aspects of community preparedness. Evacuation drills provide controlled settings to study human behaviour, enhance preparedness, and generate data for validating tsunami evacuation models, such as agent-based models (ABMs). Zushi City, located in Kanagawa Prefecture, Japan, is a high-risk coastal community due to its susceptibility to near-field tsunamis with arrival times as short as 9–11 minutes. The July 2024 tsunami evacuation drill in Zushi provided an opportunity to gather empirical data on evacuation dynamics, including pedestrian speeds, routes, and elevation gains, to inform both local planning and broader tsunami resilience strategies. Novelty This study integrates GPS tracking, drone imagery, 360-degree video recording, and participant feedback to analyze tsunami evacuation behaviour in Zushi City, Japan. Unlike previous studies that often focus on post-event surveys or macroscopic datasets with limited spatial-temporal resolutions, this approach provides granular insights into evacuation dynamics. The inclusion of unguided return phase data distinguishes the impact of drill settings on pedestrian speeds, while extreme weather conditions during the drill highlighted previously underexplored variables such as heat and humidity. The study offers critical data for validating ABMs, which have suffered from a lack of real-world data to support a robust validation process. Objectives The primary aim of this study was to analyze evacuation behaviour during Zushi City’s tsunami evacuation drill to enhance local disaster risk reduction efforts. Specific objectives included: 1. Tracking evacuation routes and speeds using GPS devices. 2. Evaluating the influence of guided versus unguided evacuation phases. 3. Capturing participant responses through post-drill questionnaires. 4. Identifying bottlenecks and opportunities to improve tsunami evacuation strategies. Methods On July 4,2024, the tsunami evacuation drill was conducted under extreme heat conditions, with apparent temperatures reaching 37°C. Participants were primarily beach house workers tasked with leading tourists during an actual tsunami. NEO-6M GPS modules were used to track movements, supplemented by drone footage and 360-degree video recordings. The drill included a guided evacuation phase, where participants were escorted to safety, and an unguided return phase back to the starting points. Data on pedestrian speeds, routes, and elevations were analyzed alongside participant feedback collected via questionnaires. This mixed-methods approach allowed for a comprehensive evaluation of evacuation performance and participant experiences. Results Average evacuation speeds during the drill ranged from 0.88 m/s to 1.02 m/s, significantly slower than reported in other studies. Factors contributing to the reduced speeds included the limiting influence of evacuation leaders and extreme heat conditions. During the unguided return phase, speeds increased slightly (0.89–1.12 m/s), supporting the hypothesis that guides constrained movement during the evacuation phase. Despite the improvements, all observed speeds were below typical walking speeds in similar drills, demonstrating the significant influence of environmental variables such as weather. Participant feedback revealed concerns about narrow streets, steep routes, and the accessibility of evacuation paths for individuals with mobility impairments. The drill’s durations, even under ideal conditions without milling times or water exit delays, approached or exceeded the wave arrival times for Zushi, indicating a high potential risk during an actual tsunami. Conclusions This study highlights critical challenges and opportunities in tsunami evacuation planning for Zushi City. The findings emphasize the importance of equipping evacuation leaders with adequate training, improving route accessibility, and incorporating environmental factors like extreme heat into evacuation models. Participant feedback underscores the need for clearer signage, alternative routes for vulnerable populations, and enhanced communication strategies during drills. By providing high-resolution, field-based data, this study contributes to the validation of tsunami evacuation ABMs and supports the development of more effective emergency management plans for coastal communities. References Shibayama, T., Esteban, M., Nistor, I., Takagi, H., Thao, N. D., Matsumaru, R., Mikami, T., Aranguiz, R., Jayaratne, R., & Ohira, K. (2013). Classification of Tsunami and Evacuation Areas. Natural Hazards, 67 (2), 365–386. https: //doi. org/10.1007/s11069-013-0567-4
Year: 2025