New Disaster Sheltering Prediction Tool Launched in Honor of Kumamoto Earthquake Victims
April 2026 marks a profound milestone: ten years since the 2016 Kumamoto Earthquakes that devastated our region. As a university deeply rooted in the heart of this community, Kumamoto University has dedicated the past decade not just to recovery, but to harnessing our expertise to ensure a safer, more resilient future.Researchers at the Kumamoto University Graduate School of Science and Technology, led by Associate Professor Hiroshi Okajima from the Faculty of Advanced Science and Technology, have developed a groundbreaking new mathematical model that effectively predicts how the number of people requiring emergency shelter changes over time following a major earthquake. This innovative tool moves beyond abstract estimates, incorporating real-time data on critical infrastructure damage—specifically electricity and roads—and the level of household emergency stockpiling to create a highly accurate picture of a community’s unfolding needs.
The model’s origins are deeply personal. Many members of the research team, including Prof. Okajima, experienced sheltering firsthand after the 2016 quakes. These real-life experiences of the crucial importance and immense difficulty of providing effective support became the driving force behind this work. The new model explicitly simulates how a lack of essential supplies, exacerbated by road closures and power outages, influences a person's decision to seek or remain in a shelter. Simulation results confirm that regions with lower household stockpiles see a rapid, early surge in shelter numbers, while delayed road repairs cause long-term spikes in demand for aid.
Beyond immediate aid distribution and shelter management in post-disaster phases, this tool provides local governments with a robust foundation for proactive planning. By validating the model with historical data and collaborating on practical implementation, the team aims to offer a framework for policymakers to quantitatively evaluate infrastructure and stockpiling initiatives, significantly advancing regional disaster prevention strategies.
Reference
| Authors | 江崎美波、笹本偉留、岡島寛 |
| Title of original paper |
インフラ環境に基づく物資不足率を考慮した避難者数推移予測 モデルの構築 |
| Journal | システム制御情報学会論文誌 2026年2月15日掲載 |
| URL | https://www.iscie.or.jp/pub/journal |