Modeling the Impact of Supersonic Civil Aviation Emissions on the Ozone Column Using Gaussian Process Regression

The environmental impact of future commercial supersonic transport (CST) aircraft remains highly sensitive to cruise altitude and the resulting perturbations to stratospheric composition. This work develops an end-to-end modeling framework combining high-fidelity GEOS–Chem High Performance (GCHP) simulations with a Gaussian Process (GP) surrogate to quantify ozone column changes and radiative forcing (RF) across multiple supercruise altitudes. Emissions from a notional SST fleet are regridded onto a common cubed-sphere domain and propagated through GCHP to generate ozone perturbations for 45,000, 48,000, 52,000, and 53,000 ft cruise levels. The resulting fields reveal strong nonlinear altitude dependence, with weak ozone depletion near 45,000 ft and larger, spatially concentrated responses in the lower stratosphere. A GP model trained on the simulated fields achieves an R2 of 0.9992 and an RMSE of 0.045 DU on a spatially masked test set. Information-loss analysis shows that short GCHP simulations retain most of the signal at 45,000 ft but underestimate depletion at 53,000 ft, thus reflecting slower stratospheric adjustment at higher altitudes. RF estimates indicate a competition between sulfate-driven cooling and altitude-dependent ozone forcing. Overall, the developed framework provides a fast and computationally efficient model for exploring altitude sensitivities and enables rapid environmental assessments for SST design and operations.