基于SVM分类的淮河流域夏季降水预测模型

采用1959--2009年逐月74项大气环流特征量序列、500hPa月平均高度场和月平均海温场，计算与预报对象淮河流域夏季降水量相关系数，选取预测因子；用主分量分析方法组合预测因子。用支持向量分类机方法分别建立山东淮河流域、河南淮河流域、江苏淮河流域、安徽淮河流域共4个区域夏季降水短期气候预测模型。对2007--2009年夏季降水量SVM分类预测，4个区域的训练集回预测正确率为85％～99％，平均训练集回预测正确率91％；预测结果误差最大不超过1级，绝对值平均为0．4级。结果表明，该模型具有较强的预测能力和推广前景，可在气候预测业务中使用。

Prediction model for summer precipitation in the huaihe river basinbased on support vector machine

Based on Support Vector Machine(SVM),four short-range summer precipitation prediction models were established for four areas in the Huaihe River basin,respectively.Using the monthly data of 74 circumfluent eigen values,the monthly data of sea surface temperature,the monthly data of 500 hPa height from 1959 to 2009,forecast factors were chosen.Combination of the forcast factors was done by using Principal component analysis.A categorical prediction was performed on summer precipitation data from 2007 to 2009...