中文摘要：

文章总结近百年来中国气候变化的特点、人类活动对碳循环的影响以及温室气体气候效应的模式研究结果.近百年来,中国年平均温度呈上升趋势,但温度变化具地区性和季节性特征.近50年观测到的冬季增温最为明显,长江中下游地区夏季地区还出现了降温.人类活动被认为是导致全球变暖的重要原因.大气CO2浓度从工业化前的约280ppm增加到了2008年的385.2ppm.20世纪90年代期间,全球碳源为8.0GtC/a（1Gt=10亿吨）,包括化石燃料燃烧产生的碳（6.4±0.4GtC/a）和土地利用变化产生的碳1.6[0.5～2.7]GtC/a.同时大气中增加的碳为 3.2±0.1GtC/a和海洋吸收的碳为 2.2±0.4GtC/a.碳源比碳汇高出2.6[0.9～4.3]GtC/a,这部分目前学术界还不能解释的碳汇被称为＂碳失汇＂.北半球陆地生态系统是寻找＂碳失汇＂的重要方向.目前多数气候模式能够成功再现全球平均气温在过去百年的实际演变.就全球年平均温度在 1880～1999年的变化而言,在自然因子和人为因子的共同强迫作用下,参加IPCC AR4的19个耦合模式集合模拟的变暖趋势为0.67℃/100a,非常接近观测的0.53℃/100a[3]. 多模式集合的结果与观测序列的相关系数可以达到0.87,这种高相关系数主要来自20世纪的变暖趋势.19个耦合模式模拟中国平均气温演变的能力较之模拟全球平均情况要差,与实际观测值之间的相关系数为0.55.这表明对区域尺度的气候变化而言,其情况要比全球平均情况复杂的多,特别是中国地区存在的高浓度气溶胶,能在很大程度上影响中国区域的气候变化.由于气候变化同时受地球系统的自然变率和人为因子的影响,更进一步了解全球碳循环对中国近百年气候变化的影响还依赖于地球气候系统模式对各种自然和人为气候强迫的模拟准确性,特别需要结合观测和模拟减小陆地生态系统碳源汇的不确定性.

英文摘要：

This review paper summarizes features in climate change in China over the past century,anthropogenic changes in global carbon cycle,and the simulated climatic effects of greenhouse gases in climate models. Over the past 100 years, the observed annual mean surface air temperature in China exhibited an overall increase. In the recent 50 years,a warming prevailed in China in winter, whereas a cooling was found in the middle and lower reach of the Yangtze River in summer. Human activities have been considered as the major driver of global warming. Atmospheric CO2 concentrations increased from 280 ppm in the preindustrial time to 385.2 ppm in 2008. For the decade of the 1990s, the average annual emission of CO2 is 8.0Gt C/a, including 6. 4± 0. 4Gt C/a from the combustion of fossil fuels and 1.6[0. 5 -2.7] GtC/a from the changes in land use. The total emission of CO2 is greater than the sum of the annual accumulation of carbon in the atmosphere （3.2±0. 1GtC/a）and the annual uptake by the oceans （2.2±0.4GtC/a）. An additional sink of 2.6 [0.9 - 4.3 ] GtC/a（ the so-called ＂ missing carbon sink＂） is required for balancing the global carbon budget. The biosphere has been considered as the major sink for missing carbon removal. Most global climate models can now reproduce the changes in annual mean surface air temperature over the past decades. With both natural and anthropogenic forcings in climate models,the Intergovernmental Panel on Climate Change （IPCC）Fourth Assessment Report （AR4）multi-model ensemble simulated a warming trend of 0.67℃/100a over 1880- 1999, which is very close to the observed trend of 0.53 ℃/100a in this time period. The simulated time series of temperature has a high correlation with the observed series（with a correlation coefficient of 0.87 ） ,reflecting the warming trend in the 20^th century. The performance of the IPCC AR4 multi-model ensemble in simulating the average temperatures over China was not as good as that in simulating the global mean values; the correl