国家气候中心大气环流模式冬季模式误差特征分析

本文利用1982—2010年国家气候中心第二代月动力延伸预测系统中大气环流模式(BCC_AGCM)的回报资料和美国国家环境预测中心和美国国家大气研究中心(NCEP/NCAR)再分析资料,分析了该模式对冬季气温预测的误差特征及其与外强迫的联系.结果表明模式能够在整体上较好地反映出欧亚区域冬季气温的变化趋势,能抓住东亚冬季风区气温年际变化的主要空间模态,对东亚冬季风区冬季气温具有一定的预报能力.预报误差的空间分布和时间演变特征的结果表明,误差在陆地大于海洋,高纬地区大于低纬地区,同时与海拔高度也有密切关系.预报误差的主要模态与一些关键区域的海温和海冰存在显著的相关性,表明模式对外强迫异常的响应能力存在缺陷.这为结合模式对关键区海温和海冰异常的响应能力,有针对性地改进模式对东亚冬季风区冬季气温的预测能力提供了依据.

Characteristics of forecast errors in the National Climate Center atmospheric general circulation model in winter

By using the National Climate Center atmospheric general circulation model (BCC_AGCM) for the second generation monthly dynamic extended range system and the NCEP/NCAR reanalysis data accumulated during 1982-2010, the model forecast errors in winter temperature prediction and their relationship with external forcing factors are analyzed. Result indicates that the model can well reproduce the variation trends of winter temperature over Eurasia region, and the main interannual variability pattern of winter temperature over East Asian monsoon domain can be successfully presented. It reveals that the model has acceptable performances in winter temperature prediction over East Asian monsoon domain, while the forecast errors are still considerable. In spatial distribution, the forecast errors over lands are larger than those over oceans. Meanwhile, the errors at high latitudes are larger than at low latitudes, and they are also closely related to altitude. The principal components of forecast errors have significant association with the sea surface temperature and the sea ice concentration over some key regions, it is shown that the response capability to external forcings is insufficient. This may provide a valuable reference for improving the prediction skill of winter temperature for East Asian monsoon domain, in combination with the model response to external forcings.
Keywords：monthly forecastforecast errorssea surface temperaturethe Arctic sea ice concentration