Modeling Outdoor Heatstroke Morbidity to Improve Urban Climate Adaptation Strategies

Abstract Climate change and urban heat islands intensify heat waves, raising health risks. In 2003, France saw 14,800 heat-related deaths, while Japan’s 2024 heatstroke cases hit a record high since 2008. The Ministry of Environment issues WBGT-based heat alerts, but these have limitations. Heatstroke risks differ between indoor and outdoor environments due to varying thermal conditions and countermeasures, making it essential to assess outdoor risks separately. Indoors people can cope with extreme heat, but outdoors they are directly harmed. Predicting outdoor heatstroke remains difficult because thermal environments vary by location. In this study, we simulated the outdoor thermal environment for 1-km² city blocks daily during summer over nine years using a meteorological model. Morbidity was calculated as heatstroke-related emergency transports divided by the estimated outdoor population for the same area and period. These values were aggregated into a single dataset for the 23 wards using a population-weighted average to construct an outdoor heatstroke risk model. The model showed high accuracy, with a root mean squared error (RMSE) of 1.77 × 10⁻⁶, corresponding to 0.18 per 100,000 people, indicating performance comparable to previous studies.