Identifying confidence intervals for aviation climate effect mitigation potentials – using algorithmic climate change function

One option for quantitative assessment of climate effects of single aircraft trajectories relies on spatially and temporally resolved climate change functions (CCFs) and their algorithmic version (aCCFs). Such aCCFs estimate radiative forcing or temperature change of aircraft emissions depending on their time and location of emission. However, the confidence of these estimates is limited by uncertainties arising at several levels: estimates of aircraft emissions, model representations and the probabilistic nature of numerical weather prediction (NWP) forecasts, and the analysis and representation of atmospheric processes when modelling climate effects and associated uncertainties. Aviation’s climate effects originate from perturbations of atmospheric concentrations in the upper troposphere and lower stratosphere (UT/LS), the principal region where contrails, NOₓ, and other aerosols exert their influence. Consequently, our uncertainty framework explicitly incorporates information from both observational techniques and numerical simulation models representative for this atmospheric layer.

In this study we present an integrated workflow that combines uncertainties from four principal sources: emissions, NWP forecast spread and skill, representation of atmospheric processes in atmosphere-climate models, and reduced complexity from regressions. In the numerical workflow as explored in our study, each uncertainty source is described either by a probability distribution (normal, log‑normal, or empirically derived) or by an ensemble of realizations. Through Monte‑Carlo sampling we propagate these uncertainties across the physical relationships that couple emissions to climate effect estimates, producing a probabilistic estimate of the net climate‑impact reduction and its confidence interval.

Application of this proposed uncertainty workflow to a set of city‑pair routes demonstrates how uncertainties can be represented with confidence levels, and with the help of hypothesis test, we can evaluate the robustness of individual proposed alternative aircraft trajectories. This numerical workflow allows balancing in fuel consumption, operational cost, and reduction in climate effects in a mathematical and statistical way. Ultimately, such type of workflow is designed for integration into automated flight‑planning and decision‑support systems. Limitations include a possible underestimation or overestimation of uncertainty values and the current lack of systematic observational validation of the aCCFs. Future work will aim to improve scientific understanding on non-CO₂ climate effects, and to integrate prevailing uncertainties in an overall decision-making-process.