Forecasting Air Traffic

by Ronald E. Miller,

Serial Information: Journal of the Aero-Space Transport Division, 1965, Vol. 91, Issue 2, Pg. 1-12

Document Type: Journal Paper


Two methods of forecasting passenger traffic between selected city-pairs in the United States air transport network are presented. Background and underlying assumptions are evaluated for each model. One method is a gravity model of spatial interactions involving total projected traffic at each city and air distances between them. The other model uses projections from a linear regression model fitted to an observed historical series of traffic on each route as a proportion of total national traffic. A specific set of forecasts was made, using both models. The results are compared by means of several different statistical measures. The linear regression model is judged to be the more accurate for prediction purposes.

Subject Headings: Traffic models | Forecasting | Air traffic | Regression analysis | Model accuracy | Passengers | Transportation networks | Routing (transportation) | United States

Services: Buy this book/Buy this article


Return to search