Research On Objective Forecasting Method Of Foggy Weather At Maotai Airport

Since the official opening of Guizhou Renhuai Moutai Airport in 2017,the passenger throughput and take-off and landing flights have shown a rapid growth trend,and the flight volume has grown rapidly,but the airport terrain is complex,and the foggy weather has always been one of the important factors leading to abnormal flights at the airport,which has a huge impact on the normal operation and efficiency of flights at Moutai Airport,so improving the accuracy of fog forecasting is of great significance to improving the quality of aviation services at Moutai Airport.In this thesis,the weather characteristics of the fog at Moutai Airport are analyzed by selecting the observation data of ground meteorological elements from January 19,2017 to December 2019,the weather forecast of the airport terminal and the ERA5 reanalysis data,and the fog is forecasted by using the decision tree model,logistic regression algorithm and recurrent neural network,and the accuracy of the airport manual forecast is compared with the three objective forecasting techniques,in order to better improve the accurate forecast rate of fog.Reduce the harm of heavy fog to civil aviation,thereby improving the operational efficiency of Moutai Airport.The main conclusions reached are as follows:(1)Moutai Airport fog weather is frequent,fog is cyclical oscillation changes,seasonal changes are more winter and less summer,and monthly changes are the highest frequency of fog from October to March of the following year.The daily change is the highest frequency of occurrence in the night time period,while the dissipation frequency is the highest in the morning hour,and the short fog at Moutai Airport is.Visibility seasonal variations are highest in summer and lowest in winter,daily variations are highest in the afternoon,and visibility is lowest in the morning.(2)Through the analysis of the meteorological statistical characteristics of fog at Moutai Airport,it is found that temperature,humidity,24-hour precipitation,air pressure,wind field and cloud cover are the main influencing factors.Low temperature,high humidity,precipitation,light wind,southeast wind direction and cloud cover in the early stage are more conducive to the formation and development and maintenance of fog,which paves the way for the establishment of fog objective prediction factor data sets in the later stage.(3)Based on a variety of objective forecasting methods,the winter fog of Moutai Airport in 2019 is forecasted,and compared with the artificial forecast of the airport,it is found that the forecasting effect of the two objective prediction fog methods of Moutai Airport logistic regression and decision tree algorithm is average,while the recurrent neural network based on the massive time series data of Moutai Airport fog has a higher forecast accuracy of Moutai Airport fog,so the objective forecasting method of recurrent neural network significantly improves the accurate forecast rate of Moutai Airport fog and provides a new means for airport fog forecasting.(4)Moutai Airport two typical radiation fog and flat current fog during the occurrence of the ground temperature are low,the wind speed is small,the ground relative humidity is close to saturation,in the later stage of the use of recurrent neural network to predict the two kinds of fog,it can be seen that the recurrent neural network on the radiation fog prediction effect is better than the advection fog,which is related to the fogging nature of the two fogs.