Effect of Trajectory Prediction and Stochastic Runway Occupancy Times on Aircraft Delays

This paper examines aircraft delays when arriving flights have been assigned specific times to land by air traffic control. It shows that the benefit from aircraft meeting their prescribed times of arrival with higher precision is moderated if it is not complemented by accurate predictions of runway occupancy times ROTs after touchdown. A single server queueing system is considered, in which the server corresponds to an airport's runway. Landing aircraft are assigned scheduled times of arrival at the runway, which they meet with some stochastic error. The time for an aircraft to clear the runway is included as a random variable in the model. A recursive queueing model is formulated and expressions are derived for the mean and variance of aircraft delays. The analysis then focuses on the effect of unpunctual arrivals on delays and how that is affected by dispersion in aircraft ROTs. A simplified situation, in which model parameters are the same for all aircraft, shows that dispersion in ROTs causes significant losses in throughput efficiency, substantially reducing the benefit of precise trajectory adherence in many cases.