Feasibility test of per-flight contrail avoidance in commercial aviation
Aarón Sonabend, Carl Elkin, Thomas Dean, John M. Dudley, Noman Ali, Jill Blickstein, Erica Brand, Brian Broshears, Sixing Chen, Zebediah Engberg, Mark Galyen, Scott Geraedts, Nita Goyal, Rebecca Grenham, Ulrike Hager, Deborah Hecker, Marco Jany, Kevin McCloskey, Joe Yue-Hei Ng, Brian Norris, Frank Opel, Juliet Rothenberg, Tharun Sankar, Dinesh Sanekommu, Aaron Sarna, Ole Schütt, Marc Shapiro, Rachel Soh, Christopher Van Arsdale, John Platt
Abstract
Contrails, formed by aircraft engines, are a major component of aviation's impact on anthropogenic climate change. Contrail avoidance is a potential option to mitigate this warming effect, however, uncertainties surrounding operational constraints and accurate formation prediction make it unclear whether it is feasible. Here we address this gap with a feasibility test through a randomized controlled trial of contrail avoidance in commercial aviation at the per-flight level. Predictions for regions prone to contrail formation came from a physics-based simulation model and a machine learning model. Participating pilots made altitude adjustments based on contrail formation predictions for flights assigned to the treatment group. Using satellite-based imagery we observed 64% fewer contrails in these flights relative to the control group flights, a statistically significant reduction (p = 0.0331). Our targeted per-flight intervention allowed the airline to track their expected vs actual fuel usage, we found that there is a 2% increase in fuel per adjusted flight. This study demonstrates that per-flight detectable contrail avoidance is feasible in commercial aviation.