Experiment on Objective Forecast Methods for The Low Visibility of Guangzhou Baiyun Airport

Using GFS model data and AWOS automatic observation data of Guangzhou Baiyun airport, two visibility objective forecast models were established by multiple stepwise regression method and BP neural network method. Two models were utilized to predict a Baiyun airport’s low visibility weather in 2015. The results showed that both models can predict the low visibility weather 24 h in advance when the visibility drops to 2000 m. For the forecasting of foggy weather, both two models can predicted 12 h in advance, and the BP neural network model can forecast the variation tendency of low visibility while the multiple stepwise regression model has a lower vacancy forecasting rate. Therefore, combining the two models has the potential to improving the accuracy of fog forecasting in Baiyun airport.