In order to solve the problem that the simulation results of the primary prediction model are not ideal, the principal component analysis (PCA) is used to reduce the dimensions of 14 meteorological factors and extract four comprehensive evaluation indicators. The sparrow search algorithm（SSA） is used to optimize the secondary prediction model of BP neural network and the difference between the results of the primary prediction and the real data is used as the output to train the prediction model and make a prediction method. Apply the model to some regions, and verify the model with corresponding data. The results show that the prediction model based on SSA optimized BP neural network and the new model introduced error have higher accuracy and stronger generalization ability than BP neural network model.