多时次资料的EOF迭代在云南夏季气候预测中的应用

EOF Iteration Scheme with Multitime Historical Data and Its Application to Summer Climatic Prediction over Yunnan Province

多时次资料的EOF迭代预测方法能较好地应用历史资料中的信息 ,并能将经验法则、观测事实和气候规律等引入到实际的短期气候预测过程中 ,特别是随着我们对影响短期气候变化的物理过程和因子认识的深入 ,这种预测方法将会得到更加有效地改进和实际应用。本文考虑 4个区域月平均海表水温的多时次历史资料 ,基于EOF迭代方案 ,建立了云南夏季气候变化的一种多时次EOF迭代预测模型。在对云南分 5个区域 ,每个区域 16个气象观测站点的夏季降水和气温趋势的预测中 ,该模型对云南 1995— 1999年的夏季 6～ 8月总雨量和平均气温趋势预测检验的最新业务标准评分平均分别达到 79.6 %和 87.0 %。该评分成绩表明 ,这种容纳多时次资料 。

EOF iteration scheme with multitime historical data can make better use of the information in historical data. The forecast method will be improved well and used widely when it introduce the empirical law, the observational fact and the climatic law etc into actual short range climatic prediction, especially when we have a better understanding of the physical process and factors which affect the climate changes. By using the monthly mean sea surface temperature data in 4 regions, a forecast model based on EOF iteration scheme for the summer climatic prediction over Yunnan province is established. The verifications of the summer precipitation and temperature forecasts for Yunnan province show that the operational forecast average scores of this model can reach 79.6% and 87.0% in 1995-1999, respectively. These scores show that the physical statistical forecast method with multitime historical data based on EOF iteration scheme is an effective path of short range climatic prediction.