Wind Forecast Error and Trajectory Prediction for En-Route Scheduling

Abstract

This work examines wind forecast forecast error as it pertains to aircraft trajectory prediction for scheduling and spacing optimized profile descents. We first perform a statistcal analysis of the RUC-2 wind forecast by comparing the forecast with one year of archived ACARS wind speed observations. This analysis provided insights into the RUC-2 forecast’s tendency to over or under predict wind speed contributions to aircraft trajectory predictions depending on the direction of flight. These results were supported by an additional simuilation study using historical ACARS paths and airspeed profiles wand RUC-2 wind forecasts to predict flight times along multiple routings from west-coast origin airports to KATL. These predictions were then compared to the actual flights times of the ACARS path and used to form flight time error distributions as a function of aircraft routing and prediction horizon. We found that flights originating from more northernly located airports were less susceptiable to error in the RUC-2 forecasts, becuase a significant difference existed bewtween their ground track angle and the prevailing wind direction from the wind forecast. The ultimate contribution of this work is the recognition and evidence that the spatial distribution of wind forecast errors plays a significant role in trajectory prediction error. More importantly, these values must not be seen as static values, rather as continuous parameters that must be reevaluated as a predicted trajectory enters different regions of the national airspace.