BP人工神经元网络在春季降水量预报中的应用

Verification for the Forecasting Ability of HLAFS NWP over Low Latitude Plateau in the Flood Season

通过普查北半球５００ｈＰａ、１００ｈＰａ、北太平洋海温与山东省春季降水量的相关，选取相关信度达到０．０５的相关区作为预报因子，利用ＢＰ（反向传播）人工神经元网络建立山东省春季降水量预报模型，并投入业务运用。结果表明，ＢＰ人工神经元网络具有较好的预报效果。

In order to make full use of HLAFS NWP,the objective verification and the efficiency analyses are conducted for the forecasting ability over low latitude plateau in the flood season in 1996.The results show that forecasts by HLAFS model have important reference value for the circulation evolution and the physical elements.