A stepwise clustered ensemble temperature downscaling (SCETD) approach for assessing the future changes of temperature extremes over Yangtze River Basin

In this study, a stepwise clustered ensemble temperature downscaling (SCETD) model has been proposed based on a stepwise cluster analysis method (SCA) and used for the first time to study the extreme temperature variations at 12 stations in the Yangtze River Basin (YRB). The results show that the proposed model can be used to generate reliable forecasts of daily maximum, minimum, and mean temperatures based on CMIP6 data (R 2 > 0.96). Temperatures in the Yangtze River Basin are likely to maintain a significant upward trend (Tmean is 0.082–0.23 °C/decade under SSP245 and 0.094–0.46 °C/decade under SSP585) in the future. In terms of extreme temperature events, extreme warm events may happen more frequently and extreme cold events to decrease in the future under SSP245 and SSP585. A more significant increase of FD and TNn can be observed in the Qinghai-Tibet Plateau region, while the TXx and SU will grow more considerably in the central Yangtze River region under both scenarios. Semi-arid regions (Qinghai-Tibet Plateau) are more sensitive to climate change. The changes of SU and TXx are significantly smaller than FD and TNn under both SSPs scenarios. The obtained results will be of great significance for responding to international climate goals, formulating emission reduction policies, adopting adaptation measures and international cooperation.