【摘要】 利用一个基于奇异值分解(SVD)的后处理方法来改进中国地区冬季地表气温长期变化的预报。本文使用的数据包括1969～2001年共33年的中国台站的地表气温的观测资料以及加拿大气候变化研究中心Historical Forecasting Project(HFP)项目的超级集合动力预报结果。本文的后处理方法是基于SVD分解中前三个模态和下垫面海表温度的历史观测资料之间的回归关系建立的。研究结果说明,纯动力的模式预报结果不能够合理的预报出观测资料中的地表气温的线性趋势。经过该方法处理后的预报结果对地表气温的长期变化的预报在线性趋势的幅度和空间分布上都有了非常显著的改进。更多还原

【Abstract】 A SVD based dynamical-statistical post-processing approach is applied to the surface air temperatures over China in wintertime during period from 1969 ～2001 aimed to improve the forecast of long-term climate trends in the observations.Both observations and the ensemble forecasts from four atmospheric general circulation models(GCMs) in the Canadian Historical Forecasting Project(HFP2) are used.The statistical post-processing uses the relationship between the predicted 500 hPa geopotential height(Z500) and the observed SAT to calibrate the SAT forecasts.Results show that the original ensemble forecasts can not capture the pronounced long-term climate trends in the observations.The post-processing method significantly improves the magnitudes and distribution of the ensemble forecast results and can help to improve the seasonal forecast skill in winter in China.更多还原