Revealing considerable emissions reduction potential in flight operations: A real-time emission perspective

Abstract

Accurately estimating aviation emissions is crucial for developing strategies to mitigate environmental impacts. However, aircraft’s unique four-dimensional operation characteristics and limitations in aircraft-to-ground communication technology pose challenges for precise and dynamic emissions estimation. This study proposed a hybrid machine learning approach for real-time aviation emissions estimation based on four-dimensional flight trajectories. Using Qingdao Airport as a case study, we generated a high-resolution emission inventory and explored the emission reduction potential in flight operations. Results demonstrated that: (1) our approach achieved higher prediction accuracy for real-time fuel flow rates and emissions than conventional models, owing to a comprehensively considering the nonlinear relationships between aircraft performance and trajectories under complex operating conditions; (2) emission estimation results from our approach were 22 %−78 % higher than the ICAO reference values; (3) an emission reduction potential of 5 %−65 % was identified in flight operations. Finally, four policy recommendations were developed to reduce emissions from airport flight activities.