The estimation of extreme wind speeds, their directional variation, and potential future changes is essential for wind-resistant design and is possible using climate models. Accurate evaluations of local topographic winds such as downslope windstorms and gap winds require high-resolution calculation and many ensemble years. However, few climate databases satisfy both requirements and none have been validated for extreme wind speeds.
We assessed directional extreme wind speeds using a massive high-resolution ensemble climate dataset (d4PDF-5km-DS) for Hokkaido Island, Japan. Despite some issues due to limited reproducibility of spring extratropical cyclones, downslope windstorms caused by large mountains are reproduced well, indicating decreasing wind speed under predicted future climate. The findings suggest future climatic conditions may influence extreme topographic wind speeds.