Estimation of maximum-likelihood discrete-choice models of the runway configuration selection process

Abstract

The runway configuration is the subset of the runways at an airport that are used for arrivals and departures at any time. Many factors, including weather (wind and visibility), expected arrival and departure demand, environmental considerations such as noise abatement procedures, and coordination of flows with neighboring airports, govern the choice of runway configuration. This paper develops a statistical model to characterize this process using empirical observations. In particular, we demonstrate how a maximum-likelihood discrete-choice model of the runway configuration process can be estimated using aggregate traffic count and other archived data at an airport, that are available over 15 minute intervals. We show that the estimated discrete-choice model not only identifies the influence of various factors in decision-making, but also provides significantly better predictions of runway configuration changes than a baseline model based on the frequency of occurrence of different configurations. The approach is illustrated using data from Newark (EWR) and LaGuardia (LGA) airports.