Multi-scale and large ensemble perspectives on extreme heat stress in Japanese urban areas under global warming

Large ensemble climate projection datasets serve as powerful tools for discussing future projections of various extreme weather events. A Japanese climate research group has developed a high-resolution (5-km) regional climate model dataset, known as d4PDF, which consists of 60-year, 12-member ensemble simulations under three climate conditions: the historical climate experiment (1951–2010), the +2°C global warming experiment (2031–2090), and the +4°C global warming experiment (2051–2110). The 5 km horizontal resolution enables realistic representations of Japan’s complex land use and topography, while also incorporating an urban model to capture urban climate characteristics effectively. Using this dataset, we investigated the worst-case scenario of future heat stress in Japanese urban areas. Specifically, we focused on extreme heat stress events in the area, defined by daily maximum WBGT (WBGTX) and daily maximum wet-bulb temperature (TwetX) exceeding the 90th percentile thresholds (WBGTX90, TwetX90). Our analysis revealed that under global warming, the daily maximum values of WBGTX90 and TwetX90 increase, but the magnitude of increase is smaller compared to the rise in mean temperature. In some regions, the trends of WBGTX90 and TwetX90 differ from those of extreme daily maximum temperature events (TX90), where TX90 shows a larger increase than the mean temperature rise. To understand the atmospheric circulation changes influencing these differences, we compared composite sea-level pressure patterns associated with WBGTX90, TwetX90, and TX90 events. The results revealed distinct circulation patterns among the three indices, suggesting that the circulation patterns identified for each index influence differences in local temperature increases through region-scale surface wind anomalies. These findings emphasize that even urban-scale extreme events must be considered within a multi-scale framework when assessing future climate changes, as their projected changes are influenced by large-scale atmospheric circulation variations and associated uncertainties. This highlights the importance of a multi-scale approach in evaluating future urban climate dynamics.