Population exposure to extreme heat in China: Frequency, intensity, duration and temporal trends

Abstract Research on population exposure to extreme heat is hindered by the limited spatial coverage of weather station and single exposure characteristic. In this study, a random forest regression model was developed to estimate monthly mean maximum temperature and extreme temperature. A cross-regional statistics in mean maximum temperature and extreme temperature was created to calculate a threshold which was used to reflect extreme temperature events. The threshold was used to develop the frequency, intensity and duration in extreme heat exposure for mean maximum temperature and extreme temperature, and quantified their spatiotemporal trends across residential areas in China in summer, 2001–2013. Results show that the risk of extreme heat was the highest in East China and was lower in Northeast and Northwest. The frequency of extreme heat exposure increased in most areas for mean maximum temperature, decreased in northern areas and increased in southern areas for extreme temperature. The intensity of extreme heat exposure increased in East, Central, South, and Southwest China for both mean maximum temperature and extreme temperature. The duration of extreme heat exposure increased nationwide for mean maximum temperature, and decreased in northern areas for extreme temperature. Frequency, intensity and duration of extreme heat exposure increased significantly, accompanied by high frequency, intense intensity and long-lasting in East, Central, Southwest, and South China. Overall, the results identify the high-risk hotspots over China in summer, 2000–2013.