Assessment of CMIP3 climate models and projected climate changes in northern Xinjiang
CMIP3气候模式对北疆气候变化模拟评估及未来情景预估

Based on the observed precipitation and temperature data of 44 meteorological stations for the period 1961-2000 in northern Xinjiang(Beijiang),the projected changes of precipitation and temperature in this region in the 20th century of 20 models from the CMIP3(phase 3 of the Coupled Model Inter-comparison Project) dataset are evaluated.The results show that large differences can be found in the simulations of precipitation and temperature.For the climate mean state,there are 5 models with better results in temperature,namely,the GISS_E_R model,GISS_E_H model,NCAR_PCM model,UKMO_HADCM model and MRI_CGCM2 model,and 2 models with better results in precipitation.On annual scale,the amplitudes of most models are bigger than observed data,all the models can reproduce the annual cycle of the temperature,and the MPI_ECHAM5 model is superior to the remaining models.The annual cycle of precipitation is poorly captured by all the models.On monthly scale,most models show a relatively high correlation coefficient with the observed data,but in terms of precipitation,some models show a negative correlation coefficient,and the amplitude of Root Mean Square Error of each model is relatively small in the monthly temperature.In conclusion,the simulations of temperature are much better than those of precipitation by climate models.Finally,the MPI_ECHAM5 model is selected to project the temperatures of northern Xinjiang region for the period 2011-2050,and the result shows that temperature will increase consistently under the three emission scenarios.