The aviation industry significantly contributes to anthropogenic climate change through both CO₂ and non-CO₂ emissions. Recent research emphasizes the substantial role of non-CO₂ emissions, which are believed to cause approximately two-thirds of the total warming effect from aviation. These emissions include nitrogen oxides (NOₓ), water vapor, aerosols, and contrail formation. The EU's updated Emissions Trading System (ETS) mandates the inclusion of aviation non-CO₂ effects in a monitoring, reporting, and verification (MRV) framework, necessitating the selection of a suitable climate metric to quantify these effects. The Global Warming Potential (GWP), while widely used in international climate policy, is criticized due to its time horizon dependency and its limitations in capturing the complexities of aviation's non-CO₂ impacts. Aviation non-CO₂ effects are characterized by variable atmospheric lifetimes and efficacies, are sensitive to emission altitude and location, and are subject to considerable uncertainty. The choice of climate metric can inadvertently bias the evaluation of specific aircraft designs or emission species. Therefore, selecting an appropriate metric is vital for stakeholders to ensure effective climate policy implementation and genuine reductions in aviation's climate footprint.

Existing literature extensively documents the climate impacts of aviation non-CO₂ emissions. Studies highlight the importance of NOₓ emissions influencing tropospheric ozone and methane levels (Fuglestvedt et al., 1999; Grewe et al., 2002), the radiative forcing from aircraft NOx (Stevenson et al., 2004; Myhre et al., 2011), aviation water vapor emissions (Wilcox et al., 2012), and the impact of aerosols (Righi et al., 2023; Penner et al., 2018; Gettelman & Chen, 2013; Righi et al., 2013). Contrail cirrus formation is another significant contributor, with research exploring its radiative forcing (Bickel et al., 2020; Burkhardt et al., 2018; Burkhardt & Kärcher, 2011; Schumann et al., 2017; Bock & Burkhardt, 2019; Stuber & Forster, 2007). The literature also addresses the challenges in developing appropriate climate metrics for aviation, including the limitations of GWP (Manne & Richels, 2001; Shine et al., 2007; Allen et al., 2016; Ocko et al., 2017), and the need for metrics that capture regional and temporal variations in climate effects (Lund et al., 2017). Previous work has explored alternative metrics such as the Integrated Global Temperature Change Potential (IGTP) and Average Temperature Response (ATR) (Peters et al., 2011; Dallara et al., 2011), with discussions on their suitability for aviation (Forster et al., 2006; Shine et al., 2005). The use of efficacy-weighted metrics has also been suggested (Ponater et al., 2006).

The study employs the AirClim climate-chemistry response model to evaluate the performance of various climate metrics. AirClim calculates yearly global mean radiative forcing and temperature change values from spatially resolved aviation scenarios. The analysis considers CO₂, water vapor, contrails, and NOₓ-induced changes in ozone and methane. The impact of short-lived species like aerosols is assumed to be qualitatively similar to contrails. The study uses both absolute (e.g., AGWP) and relative (e.g., rATR) forms of the metrics. Several metrics are analyzed, including RF, RFI, GWP, GTP, IGTP, ATR, GWP*, and EGWP*. Four main requirements for a suitable aviation climate metric are defined: neutrality in representing the climate indicator, temporal stability for policy monitoring, compatibility with existing climate policy frameworks, and simplicity in understanding and implementation. The neutrality assessment compares the peak and average temperature responses and climate metric values of various future aircraft fleets generated using Monte Carlo simulations with varying parameters (fuel type, engine design, etc.). Temporal stability is evaluated using CO₂-equivalent trajectories for the full aviation industry based on CORSIA and FP2050 scenarios. Compatibility with existing policy involves the ability to calculate CO₂-equivalent emissions and single values for fleets and flights. Simplicity assesses the ease of understanding and implementing each metric. The study uses the AirClim model with the Shared Socioeconomic Pathway SSP2-4.5 (and variations between SSP1-SSP5) for background emissions scenarios. The GWP* and EGWP* methods, which are flow-based and provide temporal trajectories, are also analyzed; but their suitability as metrics is questioned due to the difficulty in selecting a single representative value.

The analysis reveals significant differences in the performance of various climate metrics. The frequency of 'incorrect fleet pairs' (where the signs of temperature change and climate metric change differ) is used to assess neutrality. Integrated metrics like AGWP, AEGWP, and ATR show less dependence on the temporal emission profile compared to endpoint metrics like RF and AGTP. The EGWP* shows promising low dependence on the time horizon and a low frequency of incorrect fleet pairs. The ATR and AEGWP show generally low error frequencies, particularly for longer time horizons (above 70 years for ATR). Concerning temporal stability, RF, GTP, and ATR metrics exhibit stability for full aviation emissions scenarios. However, the GWP* and EGWP* show significant instability due to their sensitivity to emission rates. For compatibility, RF, GWP, EGWP, GTP, IGTP, and ATR are compatible with existing climate policy, while GWP* and EGWP* are not due to their flow-based nature. Regarding simplicity, RF and GTP are the easiest to understand and implement, while GWP* and EGWP* are the most complex. Overall, the ATR and EGWP emerge as the most suitable metrics, exhibiting high neutrality, stability, compatibility, and relative simplicity when compared against the GWP. The study highlights the sensitivity of GWP, EGWP, and ATR to the chosen time horizon, suggesting that longer time horizons (greater than 70 years) are generally more appropriate. Figure 4 illustrates the GWP, EGWP and ATR responses for various time horizons and emission scenarios, showcasing the influence of efficacy on the relative importance of different emission species.

The findings underscore the critical importance of climate metric selection in the effectiveness of climate policies for aviation. The inherent biases in certain metrics, favoring specific aircraft designs, highlight the need for neutral metrics that let policy decisions drive choices rather than the metric itself. The study's rigorous analysis reveals that the commonly used GWP has limitations compared to the ATR and EGWP. The ATR, being temperature-based, incorporates more climatic processes but introduces additional assumptions and uncertainties. The EGWP offers a compromise, more accurately representing the climate impact while retaining the GWP methodology. The selection of an appropriate time horizon is crucial, with longer horizons (above 70 years) being generally preferred for integrated metrics to capture the full climate response. The use of a short time horizon requires strong justification. The study's limitations, such as using theoretical aircraft fleets and a single climate model, suggest avenues for future research to refine the results. Despite these limitations, the findings strongly support the adoption of ATR and EGWP for assessing aviation's climate impact in policy settings.

This research demonstrates that the choice of climate metric significantly influences the effectiveness of aviation climate policies. The ATR and EGWP emerge as superior alternatives to the GWP, exhibiting higher neutrality, temporal stability, and compatibility with existing policies while remaining relatively simple to implement. A time horizon exceeding 70 years is recommended for these metrics to account for the delayed temperature response of the atmosphere. Future research should explore the advantages and potential disadvantages of EGWP in more detail and focus on improving efficacy estimates for aviation non-CO₂ emissions. The selection process for climate metrics should not be contentious; instead, the tools and methods presented in this study facilitate a thorough assessment of metric performance, enabling informed and effective policy choices.

The study employs theoretical aircraft fleet designs generated via Monte Carlo simulations, potentially omitting certain physically infeasible designs. While the randomization method minimizes bias, using real aircraft designs could improve accuracy. The study's reliance on the AirClim model warrants further validation using alternative climate models to enhance the robustness of the findings. The simplified assumptions concerning the impact of novel aviation fuels on emissions also represent a potential limitation.