基于多气候因子的长江流域长期降水预测研究

Long-term precipitation prediction based on multiple climatic factors in Changjiang River Basin

水资源供需矛盾是当前中国社会发展面对的一大难题,而长期降水预测对合理分配利用水资源起到重要作用。针对长江流域复杂的降水机制,选取冬季海温、北半球500 hPa位势高度和北极海冰3种气候因子,分析其与长江流域主汛期降水相互联系的关键区,在此基础上采用奇异值分解(SVD)方法构建降水初步预测模型和预测订正模型,并将二者组合得到优化预测模型(SSVDF),然后对该模型预测效果进行检验。结果表明:采用多因子综合预测的方法较单因子预测效果更好,SVD订正方法可以显著消除原有统计模型的预测误差,优化后的SSVDF模型能够有效预测长江子流域主汛期的降水量,并能较好预测长江流域主汛期降水异常的空间分布。

The contradiction between supply and demand of water resources is a big problem in China's social development,and long-term precipitation prediction plays an important role in the rational allocation and utilization of water resources.In view of the complex mechanism of precipitation in the Changjiang River Basin,we analyzed the key relationship between precipitation in the main flood season of the Changjiang River Basin and three climatic factors,namely,winter sea surface temperature,500 hPa geopotential height in the northern hemisphere and Arctic sea ice.On this basis,the singular value decomposition(SVD)method was used to construct a preliminary precipitation prediction model and prediction correction model.The optimized prediction model(SSVDF)was obtained by combining the two models,and its effect was verified.The results showed that the prediction method based on multiple climatic factors was more effective than prediction method based on single factor.SVD correction method could significantly eliminate the prediction error of the original statistical model.The SSVDF model could effectively predict the rainfall precipitation in the river system of the Changjiang River Basin during the main flood season,and it could also predict the spatial distribution of precipitation anomalies in the basin.