Success and failure of contrail models: a flight-by-flight investigation using satellite observations

A realistic quantification of aviation’s net global climate impact depends on how well models represent aviation-induced aerosols (e.g., soot and sulfate) and their dual role: contributing to net warming through the formation of persistent ice clouds (contrails) and contributing to cooling by altering the microphysical properties of existing liquid clouds. Here, we focus on the warming pathway. Persistent contrails are estimated to produce warming over a year comparable to the warming from aviation CO₂ accumulated over several decades [1] and may account for ~2% of the total anthropogenic surface temperature increase since pre-industrial times [2]. Given their importance, contrails must be modelled both accurately and efficiently to support operational mitigation and to track aviation’s climate impact.

The Contrail Cirrus Prediction (CoCiP) tool is a widely used Lagrangian model that predicts contrail formation and evolution on a flight-by-flight basis. CoCiP is integrated into the Non-CO₂ Aviation Effects Tracking System (NEATS), which supports compliance with recent European reporting requirements for non-CO₂ aviation effects. Despite its broad adoption, CoCiP has been shown to underestimate lifetime-integrated optical depth relative to higher-fidelity models [3], motivating further evaluation against observations.

We analyze ~500 flights from 2025 that flew through the UK and surrounding region (approximately 48°N-63°N, 20°W-4°E) that have been contrail-matched using detections from the Earth Cloud Aerosol and Radiation Explorer (EarthCARE) mission. For each flight, we run CoCiP and compare its output at the advected waypoint closest to the satellite-detected contrail at the detection time. We find that CoCiP fails to predict a contrail for roughly half of the cases. For ~20% of flights, contrail formation is not expected based on the Schmidt–Appleman criterion, which depends on both atmospheric and aircraft characteristics. In other cases, flights satisfy this criterion but occur in ice-subsaturated regions according to the ERA5 reanalysis dataset, again leading CoCiP to predict no persistent contrail. These false negatives are therefore not solely model-driven, but also reflect uncertainties in the meteorological inputs, highlighting the need to disentangle error sources to robustly diagnose and address failures in a modeling chain.

References

[1] Lee, D.S. et al.: The contribution of global aviation to anthropogenic climate forcing for 2000 to 2018, Atmospheric Environment, Volume 244, 117834, ISSN 1352-2310, 2021, https://doi.org/10.1016/j.atmosenv.2020.117834.

[2] IPCC, 2021: Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 3−32, doi:10.1017/9781009157896.001.

[3] Akhtar Martínez, C. et al.: Zero-dimensional contrail models could underpredict lifetime optical depth, Atmos. Chem. Phys., 25, 12875–12891, 2025, https://doi.org/10.5194/acp-25-12875-2025.