Analysis of FMS-generated trajectory prediction accuracy and sensitivity

The effectiveness and reliability of air traffic Conflict Detection and Resolution (CD&R) is greatly affected by the uncertainty in the future position of the aircraft. On the Eurocontrol-led ERASMUS1 program, we have conducted a study of the accuracy and sensitivity to disturbances of airborne Trajectory Predictions (TP). As future ATM concepts are typically based on information sharing between the air and ground, our analysis is focused on the factors that can be potentially improved by air-ground communication. For instance, ground systems may have access to the most up to date weather forecasts from the NOAA/NWS, yet have little information about the aircraft state and intent or local atmospheric conditions. Aircraft have accurate measurements of the local wind and temperature during the flight but have limited access to the latest weather forecasts. This paper analyzes the sources of errors that impact the TP performance, and the accuracy of the FMS predicted trajectory when compared against the flown trajectory. We also discuss the suitable metrics for this analysis and the potential use of shared intent data as a way to improve the ground based TP.