A Machine Learning Approach Towards Analyzing Impact of Surface Weather on Expect Departure Clearance Times in Aviation

Abstract

Commercial air travel in the United States has grown significantly in the past decade. While the reasons for air traffic delays can vary, the weather is the largest cause of flight cancellations and delays in the United States. Air Traffic Control centers utilize Traffic Management Initiatives such as Ground Stops and Expect Departure Clearance Times (EDCT) to manage traffic into and out of affected airports. Airline dispatchers and pilots monitor EDCTs to adjust flight blocks and flight schedules to reduce the impact on the airline’s operating network. The use of time-series data mining can be used to assess and quantify the impact of surface weather variables on EDCTs. A major hub airport in the United States, Charlotte Douglas International Airport, was chosen for the model development and assessment, and Vector Autoregression and Recurrent Neural Network models were developed. While both models were assessed to have demonstrated acceptable performance for the assessment, the Vector Autoregression outperformed the Recurrent Neural Network model. Weather variables up to six hours before the prediction time period were used to develop the proposed lasso regularized Vector Autoregression equation. Precipitation values were assessed to be the most significant predictors for EDCT values by the Vector Autoregression and Recurrent Neural Network models.