中文摘要：

文中设计了一种基于SVD迭代的短期气候预测模型,通过选择适当的影响因子,建立与预测对象之间的联系,从而实现对气象场序列的预测.分别对华北16个测站和长江中下游地区30个测站1991-2000年近10 a夏季（6-8月）降水做预测试验,平均均方误差分别为0.352和0.312,平均符号相关系数分别为0.575和0.623.此研究表明,基于SVD迭代的气候预测模型是一种非常有效的短期气候预测途径,具有很强的应用价值.

英文摘要：

A new short climatic prediction model based on the singular value decomposition （SVD） iteration was designed, which has nice mathematics and strictly logical reasoning. Taking predictand into prediction model, using iteration computation, taking the last results into next computation, we can acquire better effects with improving precision. Precipitation prediction experiments were separately done for 16 stations in North china and 30 stations in mid-lower reaches area of Yangtze River during 1991-2000. Their average mean square errors are 0. 352 and 0. 312 and the results are very stable. Mean square errors of 9 years are less than 0.5. Only one year＇s is more than 0.5. The mean sign Correlations coefficient between forecasted precipitation and observed rainfall in summer during 1991--2000 are 0. 575 in North China and 0. 623 in mid-low reaches area of Yangtze River. Librations of them in North China during the 10 years are small. Only sign correlation coefficient in 1996 is below 0.5. Others are all over 0.5. But sign correlation coefficients in mid-low reaches of Yangtze River vary obviously. The lowest is only 0.3, in 1992, and the highest is 0.9, in 1998. As the forecasted precipitation abnormal field and distribution of that in 1998 summer are considered, We can see that the model forecasts abnormal of precipitation well, and forecasts also well the distribution of precipitation. But the errors are obvious in quantities of forecasted precipitation compared to that of observed precipitation. There is much to do with the model, Climate characters of meteorological elements in large scale, such as summer precipitation, have obvious difference in spatial distribution. We can forecast better if we divided into sub-regions depending on the discrepancy of climatic characteristics in the region, and predicted in each sub-region. The research shows that the model of SVD iteration is a very effective forecast model and has a strongly applicable value.