Quantifying Trajectory Uncertainty Using a Sensitivity-Based Complexity Metric Component

Previous studies on complexity measures have focused on traffic density and the geometry of interacting flights to quantify the difficulty of a particular traffic situation. Within this paper, the strong correlation between trajectory uncertainty and controller workload is used to show that sector complexity is not primarily driven by geometric features, but rather by the quantity of uncertainty that a controller has to face. A complementary complexity component is introduced that can help quantify the amount of uncertainty and therefore improves the assessment of controller workload required to handle a given traffic situation. This complexity component is based on sensitivity analysis, where the magnitude of uncertainty is obtained by directly investigating the impact of parameter variations to the aircraft’s predicted position ahead in time. Therefore, the quantification of uncertainty is based on the inherent dynamics of a given flight and its trajectory, rather than an assumed random error. Numerical results for simulations based on the BADA model and an aircraft intent formalization are presented to illustrate the potential benefit of quantifying traffic complexity by using sensitivity analysis. Furthermore, by following this approach, trajectory uncertainty from an air traffic controller’s perspective for a particular flight may be quantified and estimations can be made as the predictability of trajectories improves with surveillance technologies and data sharing with the aircraft flight management system. Keywords-Trajectory Uncertainty, Complexity Metric, Sector Complexity, Controller Workload, Sensitivity Analysis