Evaluation And Projection For Extreme Temperature Events In China By CMIP5 Global Climate Models

Based on the daily temperature data(daily mean temperature, daily maximum temperature, daily minimum temperature) derived from CN05.1 observation and outputs experiments of global climate models(GCM) participated in Fifth Assessment Report of IPCC, sixteen extreme temperature indices are calculated. And the observation and multi-mode ensemble of temporal and spatial distribution of average temperature were compared in China. The ability of global climate models of CMIP5 to simulate mean temperatures climate and extreme temperatures climate in China was assessed. The time series tend of extreme temperature events under RCP2.6, RCP4.5 and RCP8.5 scenarios have been estimated, and the space distribution characteristics of extreme temperature events under RCP4.5 scenarios in early of 21 st century(2016-2035), middle of 21 st century(2046-2065) and the end of the 21 st century(2081-2100) also have been projected in this research. The main conclusions are as follows:The CMIP5 models can well simulate both the temporal variation and the spatial distribution of temperature in China. Most models can reflect the observed interannual variability, multi-mode ensemble mean(MME) lost the interannual variation signal.Multi-mode ensemble is better than most of signal modes. As seen from MME, the distribution of simulated seasonal and annual mean temperature shows similar pattern with that of the observation. MME can reproduce the characteristics of greater heating rate of the northern regions than the southern regions, but the simulation results is lower,the deviation in summer is less than the one in winter.Based on the 16 extreme temperature indices proposed by ETCCDI, the CMIP5models’ simulation capabilities for extreme temperature events in China were evaluated.In this study, the results shows that models can well simulate the temporal trend of extreme temperature events. Models have limited ability in simulating the spatial distribution of extreme temperature events over China. Some simulation results(TX10p,TN10 p, TN90 p, FD, ID, CSDI) are bigger in national or local, some simulation results(TXn, TNn, TX90 p, GSL, SU, WSDI) are smaller. The simulation ability of extreme temperature events of MME in eastern China is better than the one in westernChina, and the worst one appear in the Tibetan plateau. Each single model has limited ability in simulating the extreme temperature events. MME can simulate the similar spatial distribution of the 20-yr, 50-yr, 100-yr return value of annual maximum and minimum temperature.Under different scenarios in 21 st century, extreme warm events(TXx, TNx, TXn,TNn, TN90 p, TX90 p, GSL, SU, TR, WSDI) are increasing, while extreme cold events(TN10p, TX10 p, FD, ID, CSDI) are decreasing, these change trends are obvious in early of 21 st century, and become weaker in middle and the end of 21 st century.Under RCP4.5 scenario, the change of the spatial distribution of extreme temperature indexes is not very significant in different periods. The frenquencies of extreme warm events are increasing, while the ones of extreme cold events are decreasing, these changes are similar with that of the observation, but the degree of change is inconsistent,the change range in 2046-2065 is larger than 2016-2035, in 2080-2099 is larger than2046-2065.Under RCP4.5 scenario, the 20-yr, 50-yr, 100-yr return value of annual maximum and minimum temperatures may increase significantly.