During the last two decades, as the air transport industry has changed rapidly, traditional data sources have struggled to keep up and continue to reflect the true state of the industry. This paper describes an alternative way of forecasting air transport demand that addresses these data issues and presents alternative data sources. It explains the econometric system approach developed by MKmetric which includes a new mathematical framework that reflects the consumers’ decision process and overcomes some of the shortfalls of traditional models.

The system approach developed is based on the classical 4 step procedure, namely generation, distribution, modal split and assignment. However it models the interdependent choices such as modal choice, airport and route choice, access and egress simultaneously. The model mirrors consumer behaviour using logistic functions whereby the mathematical form is flexible so that thresholds can be included. The elasticities of consumers to characteristics of the alternatives (e.g. time, cost, frequency, alliance) allow measurement of demand effects from changes in the transport system, e.g. services, infrastructure.

The approach uses dynamic, rather than fixed, airport catchment areas. These are a function of the attractiveness of the alternative transport opportunities available for each trip. Demand forecasts for an airport or a route consider the competitors endogenously. In addition, new methods are proposed to compute network impedance, which reflects the attractiveness of each alternative based on the infrastructure networks of all modes. The full potential of this approach is demonstrated by some examples.