High-resolution Simulation Of China's Regional Climate Change Under Two RCP Scenarios

A regional climate model(Reg CM4)is used to project climate change over China in the 21 st century under the RCP4.5 and RCP8.5 pathways.The driving global climate model is CSIRO-Mk3.6.0(hereafter referred to simply as CSIRO),and the simulation((hereafter referred to as Cd R)is run at a grid spacing of 25 km.Firstly,validating the model performances in the reference period from 1986 to 2005.Followed,comparing the future changes projected by CSIRO and Cd R under the two emission scenarios in the middle of the 21 st century(2046~2065)and the end of the21 st century(2080~2099).The results show follow:(1)The two models have good performances for mean surface air temperature over China and the correlation coefficients are all above 0.9.Compared with CSIRO,Cd R can simulate the small-scale temperature distribution characteristics because of the fine land surface information,and improving the temperature simulation in China by reducing the cold bias in December–January–February(DJF)and warm bias in June–July–August(JJA)of CSIRO.However,the simulation of precipitation is relatively poor,and the location of the maximum precipitation center cannot be accurately simulated.The topographic effect is too strong in Cd R,with large overestimations found over Mountain Taihang and Yanshan in North China and the mountains around the Sichuan Basin,which leads to a low correlation between Cd R and observation.Cd R can roughly reflect the interannual variability of air temperature and precipitation,but the simulation of temperature and precipitation trend by Cd R is relatively poor.(2)We found substantial warming in the future as simulated by both models,more pronounced in DJF compared to JJA,and under RCP8.5 the temperature will increase more than 5°C in the end of the 21 st century.The warming show different spatial pattern and to a less extent magnitude between CSIRO and Cd R.Precipitation change show a general increase in DJF and a mixture of increase-decrease in JJA.Substantial differences are found in JJA for the precipitation change between the two models and the change of precipitation is also not consistent under different scenarios.The paper further emphasizes the uncertainties in climate change projection over the region.(3)In the simulation of the select eight extreme climate indices,in general,simulation performances of Cd R in temperature extreme climate index is better than CSIRO,the correlation coefficients of Cd R simulation results are all over 0.9.The simulation ability of the two models for the precipitation extreme climate index is not as good as the temperature extreme climate index.The temperature extreme index has a good spatial consistency with the mean air temperature,which further indicates that the south of China is warmer than the north.Extreme warm events are more likely happen in the south,and extreme cold events are more likely happen in the north.The precipitation extreme climate index can basically reflect the characteristics of extreme precipitation over China that extreme precipitation in the southeast is greater than in the northwest,and extreme precipitation is closely related to the spatial distribution of mean precipitation.(4)For the projection of extreme events,compared with the RCP4.5 in the late 21 st century,under RCP8.5,due to more intense temperature rise,extreme high and low temperature are heating more.Longer summer days and shorter frost days lead to lower frequency of extreme cold events and higher frequency of extreme warm events.The overall performance of extreme precipitation events over China is increasing.In the south of the Yangtze River,there will be an increase in extreme precipitation events.In the northwest,CSIRO and Cd R have different performances that show an inconsistent in precipitation.