Probabilistic Forecasts for Aviation Traffic at FAA's Commercial Terminals

Abstract

Forecasting air travel in uncertain times is a challenging task. The effects of uncertainties accompanied with usual errors emanating from estimation of demand prevent deterministic forecasts from representing an uncertain future. The framework that is presented takes into account uncertainty for which the nature and sources are unknown. A parametric distribution fitting the past growth rates captures the probabilities associated with the outcomes representing the future uncertainty at the airports. This process combined with Monte Carlo simulation is used to generate probabilistic forecasts at the top 50 commercial airports in the United States. The methodology is illustrated by using Hartsfield–Jackson Atlanta International Airport in Georgia. Probabilistic forecasts provide an important tool for investment and personnel planning in uncertain times.