Trajectory Optimisation in Air Traffic Management: A Method for Reducing CO2 and non-CO2 Climate Impacts under Operational Constraints

Abstract

In addition to CO2 emissions, non-CO2 emissions like NOx, water vapour emissions or the formation of persistent contrails contribute significantly to the total radiative forcing due to aviation. To mitigate these effects, operational measures such as climate-sensitive routing offer high potential for short-term implementation. As non-CO2 effects are highly dependent on the spatial and temporal atmospheric conditions of the emission site, their mitigation requires a comprehensive 4D trajectory optimisation. Avoiding climate-sensitive regions often requires deviations from fuel- or time-optimal routes, resulting in trade-offs between operating costs and climate costs. Existing research primarily focuses on the mitigation potential or the high-fidelity optimisation of single trajectories, frequently neglecting scalability and operational constraints essential for large-scale application. In this thesis, an Evolutionary Algorithm (EA) is developed for 4D trajectory optimisation that integrates climate costs from both CO2 and non-CO2 emissions. The tool is implemented within the Future Air Traffic Simulator (FATS), a research platform designed for the analysis of global air traffic scenarios. The algorithm is designed with a primary focus on operational constraints and scalability. By embedding aircraft-specific performance limitations and Air Traffic Control (ATC) regulations directly into the optimisation process, the algorithm ensures that generated trajectories remain feasible and realistic in today’s air traffic. The algorithm’s modular architecture allows for the seamless integration of additional parameters, such as restricted airspaces or different cost models. This makes the developed EA a robust and scalable research tool for evaluating future sustainable Air Traffic Management (ATM) strategies. The algorithm was validated across three distinct scenarios. First, zero-wind tests confirmed the fundamental functionality, as the EA reliably identified the great-circle route as the optimal solution. Under realistic atmospheric conditions, the algorithm demonstrated its ability to perform effective trade-off analyses, successfully rerouting the trajectory to reduce the climate impact while balancing CO2 and non-CO2 effects against operating costs. By rerouting the aircraft into a cooling Ice-Supersaturated Region, the total climate impact is reduced by 40% (F-ATR50) compared to the reference trajectory optimised for Direct Operating Costs. This substantial mitigation is achieved through a 6% increase in Direct Operating Costs, reflecting a significant but effective trade-off between operating costs and environmental benefits. Finally, a statistical analysis of 30 independent runs confirmed the robustness of the stochastic search process, showing a narrow cost spread of only 1.58%, thereby ensuring the reproducibility of the results in a dynamic 4D environment.