CAASD's Future Air Traffic Timetable Estimator: A Micro-Econometric Approach

The aviation community has a rich collection of tools that simulate the operational flows of the National Airspace System (NAS). In nearly all cases, operational flows of aircraft in the NAS begin with a schedule (i.e., Future Demand Generator (FDG), as it is commonly called) generated outside the simulation model. In the past, the schedule has been derived by translating FAA's Terminal Area Forecast (TAF) into flights by growing Official Airline Guide (OAG) through an allocation mechanism called Fratar algorithm that maintains forecast integrity at the terminal levels. While this tool has worked well, there are some limitations. For example, it is likely that implied network operational flows may be different from those observed and/or expected. The top-down approach underlying FDG, in other words, is not capable of forecasting route-specific growth in operational flows. The challenge is to move from a generic traffic count at a specific terminal to a schedule of flights that includes a "when" and a "where" dimension. Modeled NAS operational performance is highly dependent on the characteristics of the forecasted operations; hence it is critical that the traffic schedule be created correctly. Under NAS-wide system modeling effort at CAASD, we have built a framework which attempts to fill in the gaps using a bottom-up, demand-driven micro-econometric approach. Our ultimate goal was to produce schedule of flights that is linked with origin and destination (O&D) operations via passenger route choice. It will thus be in sync with the Official Airline Guide (OAG), but not driven by it. Our method is comprised of six basic steps, beginning with estimation and forecasts of traveler demand between O&D city pairs, and culminating with the creation of a forecasted schedule that incorporates all major aspects of passenger demand. The result is Future Air Traffic Timetable Estimator for 292 airports in the NAS.