Alternative climate metrics to the Global Warming Potential are more suitable for assessing aviation non-CO₂ effects

Aviation contributes to climate change through both CO₂ and non-CO₂ emissions, with recent studies suggesting non-CO₂ effects account for around two-thirds of aviation-induced warming. Key non-CO₂ contributors include NOₓ, water vapour, aerosols, and contrails. The EU has moved toward including aviation non-CO₂ effects in monitoring, reporting, and verification frameworks and potentially the ETS by 2027, necessitating a climate metric that can translate non-CO₂ effects onto a common scale with CO₂. Selecting such a metric is difficult because aviation non-CO₂ effects have diverse atmospheric lifetimes and efficacies, depend on emission time, altitude, and location, and suffer from high uncertainty. Moreover, different metrics target different climate indicators (RF vs temperature), which can unintentionally bias technology choices or policy outcomes. This study therefore investigates the applicability of existing physical climate metrics to aviation, assessing their neutrality, temporal stability, policy compatibility, and simplicity.

GWP is the most commonly used metric in international climate policy but has long been criticized, particularly for its sensitivity to the choice of time horizon. For aviation, complications arise due to the short-lived nature and location dependence of many non-CO₂ effects and their varied efficacies. Alternative metrics include endpoint metrics such as Radiative Forcing (RF) and Global Temperature change Potential (GTP), and integrated metrics such as Integrated GTP (IGTP) and Average Temperature Response (ATR). Recently, GWP* has been proposed to relate changes in emission rates of short-lived species to CO₂-equivalent trajectories but is argued to be a model rather than a metric for policy applications. Efficacy-weighted variants (EGWP, EGWP*) attempt to incorporate species-specific temperature responses more accurately. Prior work has emphasized that metric choice can embed value judgments and potentially favour particular technologies or species, highlighting the need for neutrality and robustness for aviation policy.

The authors define four requirements for aviation climate metrics: (REQ1) neutrality with respect to emission species and aircraft design changes; (REQ2) temporal stability suitable for monitoring (e.g., annual/quarterly ETS reporting); (REQ3) compatibility with existing policy frameworks (e.g., computing CO₂-equivalents and single values for fleets/flights); and (REQ4) simplicity of understanding and implementation. They assess seven metrics: RF/RFI, GWP, EGWP, GTP, IGTP, ATR, and GWP* (and its efficacy-weighted derivative EGWP*). Calculations require time series of radiative forcing (RF) and global mean temperature change (ΔT), derived using the AirClim climate-chemistry response model. AirClim provides yearly global mean RF and temperature responses for CO₂, water vapour, contrails, and NOₓ-induced short-term ozone, long-term ozone, and methane effects, with SSP2-4.5 as the default background scenario (varying SSP1–SSP5 in multivariate analyses). The study distinguishes absolute and relative metric forms (e.g., AGWP/AEGWP vs relative forms to CO₂; rATR equals iGTP in relative terms). For REQ1 (neutrality), they perform a Monte Carlo analysis of 10,000 theoretical narrowbody fleets (around A320-class) with varied high-level parameters (fuel burn, NOₓ, cruise pressure/altitude proxy, contrail distance modifier, fuel type including Jet-A1, SAF, and hydrogen combustion/fuel-cell) and fleet introduction years (2030–2050). Contrail distance reductions are coupled with fuel burn penalties; fuel-type-specific emission changes follow ranges from a Clean Hydrogen report. Each fleet assumes a 30-year constant production and 35-year aircraft lifetime. They compute each fleet’s peak and average temperature responses and corresponding metric values. Neutrality is quantified as the frequency of incorrect fleet pair orderings: pairs where the sign of differences in metric value and climate outcome (peak/average temperature) do not match. For REQ2 (stability), they evaluate CO₂-equivalent (CO₂-eq) trajectories for two full aviation scenarios: CORSIA (business-as-usual with post-2020 CO₂-offsetting) and FP2050 (Flightpath 2050 targets: 75% CO₂ and 90% NOₓ reduction by 2050), extended to 2200 with 0.5% annual growth after 2100. They compare yearly CO₂-eq emissions across metrics using a 100-year time horizon, alongside fuel use and temperature responses from AirClim. For REQ3 (compatibility), they assess whether metrics can produce both temporal CO₂-eq trajectories and single values suitable for fleets/flights within current policy constructs. For REQ4 (simplicity), they qualitatively assess conceptual transparency and implementation complexity (e.g., need for climate/carbon-cycle models, efficacy inputs, and historical running averages). Metric formulations follow standard definitions: endpoint (RF at t₀+H; GTP as ΔT at t₀+H) and integrated forms (AGWP and AEGWP as ∫RF dt; iAGTP as ∫ΔT dt; ATR as iAGTP/H). EGWP multiplies species GWPs by efficacy factors; GWP* and EGWP* adapt the warming-equivalent method to aviation by using 20-year running averages of RF and species-specific efficacies. Sensitivity to time horizon (5–100 years) is examined for neutrality and species contributions under different temporal emission profiles (pulse, fleet, increasing).

- Neutrality (REQ1): Endpoint metrics (F-RF, F-AGTP/F-ATR endpoint) show strong dependence on time horizon and emission profile shape, leading to higher rates of incorrect fleet pairings. Integrated metrics (F-AGWP, F-AEGWP, F-iAGTP/F-ATR) retain memory of past emissions and are less sensitive to temporal profiles. F-AGWP and F-GWP* are more linear above ~60 years but generally have higher error frequencies than temperature-based or efficacy-weighted metrics. F-EGWP* exhibits low time-horizon dependence with lower error frequencies but is problematic in policy use (see REQ3/REQ2). F-iAGTP/F-ATR and F-AEGWP typically have incorrect-pair frequencies under ~2%, with minima around 50–80 years depending on the temperature objective (e.g., F-iAGTP/F-ATR minima near 70 years for peak T and at 20, 50, 100 years for average T; F-AEGWP minima slightly lower at 55, 15, 45, 80 years). - Temporal stability (REQ2): For full aviation scenarios, S-RFI and S-GTP yield similar total CO₂-eq trajectories and are generally stable, though they can diverge under changing emission rates and differ for pulse vs constant emissions. S-GWP, S-iGTP/S-rATR, and S-EGWP give broadly similar and stable CO₂-eq trajectories for Jet-A1-dominated fleets, with rATR100 closely matching GWP100 totals despite different species contributions. In contrast, S-GWP* and S-EGWP* show strong dependence on emission-rate changes due to their 20-year averaging, leading to large swings, including negative CO₂-eq values (notably 2050–2080 in FP2050), which risk misinterpretation as cooling and overstate changes relative to the temperature response. - Compatibility (REQ3): RF, GWP, EGWP, GTP, IGTP, and ATR can provide both temporal trajectories and single values (for fleets/flights) suitable for ETS-style implementation. GWP* and EGWP* produce time-varying flows rather than single values for a given horizon, making them unsuitable as policy metrics despite technical insight. - Simplicity (REQ4): Endpoint metrics (RF, GTP) are conceptually simplest; integrated metrics (GWP, EGWP, iGTP, ATR) are more complex. RF-based metrics are easier to implement than temperature-based metrics, while EGWP adds complexity via efficacy inputs. GWP*/EGWP* are the least intuitive and require historical data (20-year running averages), complicating implementation. - Metric choice and horizon: ATR and EGWP perform best overall across requirements, offering neutrality and stability with policy compatibility. rATR100 approximates GWP100 totals, facilitating transition; EGWP further aligns RF-based aggregates with temperature-based behavior. Time-horizon sensitivity diminishes at larger horizons; integrated metrics, particularly ATR, are better suited to horizons >~70 years to capture delayed temperature responses. - Anticipated divergence with novel fuels: With changes in relative emission indices (e.g., SAF, hydrogen), total CO₂-eq from ATR vs GWP may diverge; rATR may align more closely with EGWP than with GWP as fleets evolve.

The findings show that metric selection critically shapes how aviation’s climate impact is assessed and managed. Metrics can embed unintended priorities—for example, RF-based metrics may underweight short-lived but impactful effects at long horizons, while endpoint temperature metrics can be overly sensitive to profile shape—thereby biasing aircraft design and operational decisions. Integrated, temperature-aware approaches (ATR/iGTP) and efficacy-weighted RF approaches (EGWP) better preserve neutrality across species and design options and remain stable for industry-scale monitoring. For policy continuity, rATR’s close alignment with GWP100 totals can ease transition, and EGWP—being a derivative of GWP—may offer a pragmatic compromise to better reflect temperature relevance while maintaining familiar methodology. However, as novel propulsion systems and fuels change species contributions, GWP-based totals and ATR totals may diverge, reinforcing the need for metrics that represent temperature outcomes more directly or incorporate efficacies. Ultimately, the appropriate metric and time horizon should reflect policy goals (peak vs average temperature), scenario characteristics, and the need for relative vs absolute metrics. Providing values at larger horizons (>70 years) reduces sensitivity and better represents long-term impacts.

Across neutrality, stability, compatibility, and simplicity, the ATR and EGWP emerge as the most suitable climate metrics for assessing aviation non-CO₂ effects in both aircraft design and policy contexts. rATR100 closely matches GWP100 totals for current fleets, facilitating adoption, while EGWP’s efficacy-weighting improves alignment of RF-based aggregates with temperature outcomes. Time horizons greater than about 70 years are recommended for integrated metrics, particularly ATR, to capture delayed atmospheric temperature adjustment. As novel fuels and propulsion technologies shift relative non-CO₂ contributions, ATR and EGWP are expected to provide more accurate and resilient assessments than standard GWP. Future work should refine efficacy estimates, further evaluate EGWP behavior across contexts, and validate results with additional models and real aircraft designs. Policymakers should select metric/time-horizon combinations consistent with specific climate objectives, leveraging existing tools to analyze performance and anticipate potential pitfalls.

- Efficacy estimates, especially for contrail cirrus, are uncertain and can affect EGWP and EGWP* results. - The fleet dataset is theoretical with randomly sampled design parameters; some configurations may be infeasible, though the symmetric sampling across fleets is assumed to limit bias in pairwise comparisons. - Results rely on the AirClim model and specific background scenarios (SSPs); while chosen for efficiency and flexibility, cross-model validation would strengthen confidence. - Full-aviation scenarios (CORSIA, FP2050) were extended to 2200 with simple growth assumptions after 2100 and were not evaluated for predictive reliability. - GWP* implementation uses parameters (e.g., s=0.75) derived for methane that may not be optimal for aviation non-CO₂ effects, and its flow-based nature complicates single-value comparisons.