A Multi-model Research On The Detection And Prediction Of Climate Change In China

Using the CRU monthly data of the surface mean temperature and precipitation, and the output data of 22 AOGCMs participating in Intergovemmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4), the assessment of AOGCM's simulated capability in China for the 20th century climate change have been made according to the analyses of the simulation deviation, time series, the correlation coefficient of space, the time correlation coefficient, etc. And then the simulated climate changes in China for the 21th century are expressed with the Special Report on Emissions Scenarios (SRES) A1B, A2, and B1 emission scenarios.Main conclusions are as follows: AOGCMs have a good performance in simulating the geographical distribution of temperature in China. Simulated temperature deviation is in the range of-5.1℃-0.4℃. The deviations are larger in Southwest and Northwest area, smaller in southern China. The warming in China since 1950s can be reproduced well. The simulated seasonal mean temperature is usually lower than the observation, and the deviation in summer is smaller than in winter. As regard to the simulated capability, it is much better for the minimum temperature than the maximum temperature. For the spatial distribution of precipitation, models can't produce that good enough. Model's space correlation coefficients about precipitation between simulation and observation is mostly in the range of 0.6-0.8. In northern and western China, models make more precipitation than observation, while less in southern China. Judging from the time correlation coefficient between the simulation and the observation for a hundred years, the decision is made that whether a single model is reasonable to participate in the multi-model ensemble mean or not. For the 20th century climate change in China, multi-model ensemble clearly improves the simulation. Compared with the single model, the ensemble mean is better to reproduce present climatic change than each model; and the ensemble mean have a higher correlation with the observation, especially for precipitation. However, it also weakens some useful information that individual model with good performance offeres.In the 21st century, with the SRESA1B, SRESA2, SRESB1 scenarios, temperature will continue to rise in China, and in the period of 2091-2099, temperature will increase by 3.8℃, 4.4℃, and 2.6℃respectively; the linear trend of temperature will be 4.0℃/100a, 4.4℃/100a, and 2.4℃/100a. The warming trend in winter is largest, 4.5℃/100a, 5.1℃/100a, and 2.6℃/100a respectively. The warming is greatest in Northeast China, and smaIlest in South China. Extreme temperature changes are similar to the temperature change in the 21st century. In the period of 2091-2099, precipitation in China will increase by 11%, 12%, and 8% respectively. Precipitation changes are more complex during the 2010-2040; mostly there are show a decreasing trend in many years. However, the precipitation will begin to rise after 2040. For the whole in 21st century, precipitation in winter increases markedly, and the linear trend can reach 22%/100a, 14%/100a, and 8%/100a respectively. Precipitation in the northern China will have a greater increasing than in the South; nevertheless a decreasing will happen in Hua-nan.