摘要
用预报量与多变量预报因子的均值生成函数 ,通过 Gram- Schmidt正交化使均值生成函数彼此正交 ,同时利用双评分准则确定入选预报模型的均生函数个数 ,最后建立山西省季节降水的预报模型。试验结果表明 ,该模型预报山西省季节降水的效果很好 ,是一种提高短期区域气候预报水平的可行方法。
By using muti-variable mean generation f uncti on of the predictand and its predictor, the mean generation function is orthogon alized by means of the Gram-Schmidt's method. Meanwhile,the number of mean gene ration function has been selected in the forecast model by means of Couple Score Criterion. Finally, the forecast model of seasonal precipitation of Shanxi prov ince is established. The results indicate that the method is favorable in foreca st of seasonal precipitation of Shanxi province, and is feasible for improving t he accuracy of short-term regional climate forecast.
出处
《南京气象学院学报》
CSCD
2000年第3期422-427,共6页
Journal of Nanjing Institute of Meteorology
关键词
多变量
均生函数
正交化
双评分准则
降水预报
muti-variable
mean generation function
or thogonalization
couple score criterion
forecast model
