A stepwise-clustered heat stress downscaling approach to analyze future variations of heat stress in East China

In this study, the fifth generation ECMWF reanalysis (ERA5) reanalysis datasets and three global climate models (GCMs) were selected as the inputs of the stepwise-clustered heat stress downscaling (SCHSD) method to simulate the future heat stress indices in East China. The heat stress indices included the Heat Index (HI), Humidex (HUMIDEX) and the simplified Wet Bulb Globe Temperature (sWBGT). Three GCMs (i.e., CanESM5, INM-CM4-8 and MPI-ESM1-2-HR) under two Shared Socioeconomic Pathway (SSP245 and SSP585) were input into the SCHSD model to develop downscaled climate projections. To verify the SCHSD model, the reproduction results from three GCMs during the period of 1990–2014 were compared to daily observational data (i.e., HI, HUMIDEX and sWBGT). The verification results suggest that the coefficient of determination (R2) of the stations in the northern part of East China mostly exceeds 0.8, while the R2 of the stations in the southern part of East China is mostly in the range of 0.6 to 0.8. The projection results suggest that the future heat stress in East China would generally maintain an upward trend from 2021 to 2100. The largest change in heat stress is projected in June under the period of 2021–2100. The results also show that the highest average of days in the danger category is 78.9 in summer under the SSP585 scenario, accounting for 86% of the total summer (June, July and August).