摘要:The future climate change information plays key role for planning adaptation and mitigation strat- egy. In this study, the combination of the widely em- ployed statistical downscaling model (SDSM) and two CMIP5 models. namely MPI-ESM-LR and CNRM-CM5, was used to generate the future pro- jection of maximum and minimum temperature (Tmax and Tmin) under RCP8.5 and RCP 2.6 emis- sion scenarios within a period of 2011 to 2100 overthe Haihe Basin. The historical ground observations (daily maximum and minimum temperature) during 1971-2000 was employed to calibrate the SDSM models. Results showed that:(1) The SDSM model had a good ability to reproduce the daily and monthly mean Tmax and Tmin in the basin; (2) For the his- torical reproduction of Tmax and Tmin, the perfor- mance of CNRM-CM5 was a little worse than that of MPI-ESM-LR. (3) The change in annual mean Tmax and Tmin under the two scenarios for all eval- uation periods will increase and magnitude of Tmax will be higher than Tmin. (4) The increase in magni- tude for the weather stations in the mountains and along the coastline will be remarkably obvious. (5) The future annual Tmax and Tmin will keep a signif- icant upward trend under RCP8.5 scenarios over the whole projection period and the magnitude will be 0.37 C and 0.39 C per decade, respectively; the fu- ture annual Tmax and Tmin will increase in 2020s and then decrease in 2050s and 2070s, and the mag- nitude will be 0.01 C and 0.01C per decade, respec- tively. The related results could provide an insight into the mitigation measure for adverse effect of fu- ture climate change on the regional ecological envi- ronment.