A probabilistic decision-making model for runway configuration planning under stochastic wind conditions

Local meteorological conditions such as wind direction, cloud ceiling heights, and visibility directly influence the safety and efficiency of an airport. The safety and efficiency, generally contradictory, are associated with the selection of a particular runway configuration that is a part of job of airport tower controllers. When the wind speed and directions exceed the maximum allowable limit defined by the FAA operational procedures, controllers must change runway configurations. During this configuration change, the status of all arriving and departing aircrafts are turned into airborne or ground holding, respectively, until the new flight paths and the new taxi paths are clarified. This situation greatly influences the airport operational efficiency with its associated cost of delays, fuel burn and emissions. Despite its prominence, only a few studies that discuss the timing of decision-making for the configuration change can be found. In this study, an innovative decision-making approach is proposed for runway configuration planning under stochastic wind conditions. The goal of the decision-making approach is to maximize the airport throughput and minimize the airborne delay in terminal area, given a sequence of wind forecast data. The optimization techniques of dynamic programming and backwards induction are used to solve the probabilistic optimality equation. Based on the simulation results of JFK test case, the proposed approach is shown to reduce the average delay time and improve the throughput of runway systems without violating operational procedures.