Emissions from land-based transport, aviation, and shipping contribute significantly to climate change. Besides CO₂, these emissions include short-lived compounds that affect air quality but are also climatically relevant. We use a global chemistry-climate model to show that the climate effects of these non-CO₂ emissions are substantial across all transport sectors both now and in the future. In sum, the non-CO₂ impacts result in a cooling, which offsets the positive climate forcing from transport-induced CO₂ by around 80% at present and between 25 and 60% in different scenarios for 2050. The trade-off that air pollutants mitigate global warming is strongly reduced in a future scenario with low anthropogenic emissions, where even small remaining amounts of non-CO₂ compounds cause significant cooling as they are released in a very clean atmosphere. Our findings emphasize the need to take non-CO₂ effects into account when assessing climate protection strategies for the transport sectors.

Aviation emissions of aerosol particles and aerosol precursor gases alter the Earth's radiation budget via both direct and indirect aerosol effects, resulting in a significant climate effect. Current estimates of aviation-induced climate effects are based on coarse-resolution global aerosol-climate models, which are not able to resolve the microphysical processes at the aircraft plume scale. This results in large uncertainties in the aviation-induced impact on aerosol number and size, which are key quantities for estimating the aerosol indirect effect, especially for low-level liquid-phase clouds. In this work, a double-box aircraft exhaust plume model is developed to explicitly simulate the aerosol microphysics inside a dispersing aircraft exhaust plume, together with a simplified representation of the vortex regime (which begins ĝ1/4 10 s after the aircraft emissions and captures the dynamics of aerosol particle interactions with contrail ice particles). The aircraft exhaust plume model is used to quantify the aviation-induced aerosol number concentration at the end of the dispersion regime (ĝ1/446 h) and the results are compared with the result obtained by the instantaneous dispersion approach commonly applied by the global models. The difference between the plume approach (simulated using two boxes) and the instantaneous dispersion approach (simulated by a single box) is defined as the plume correction: for typical cruise conditions over the North Atlantic and typical aviation emission parameters, the plume correction for aviation-induced particle number concentration ranges between -15 % and -4.2 %, depending on the presence or absence of the contrail ice in the vortex regime, respectively. A tendency-based process analysis shows that the negative value of the plume correction is due to the higher efficiency of coagulation and nucleation processes in the plume approach, leading to lower total particle number concentrations compared to the instantaneous dispersion approach. Sensitivity studies over different regions highlight the role of background conditions for the plume microphysics, with the plume correction varying between -12 % for Europe and -42 % for China in a scenario with contrail ice in the vortex regime. Parametric studies performed on various aviation emission parameters used to initialise the plume model demonstrate the high relevance of contrail ice in the vortex regime to considerably reduce the aviation-induced aerosol number concentration in the plume approach. Moreover, the parametric studies show a large sensitivity towards aviation fuel sulfur content, driving sulfur dioxide (SO2) emissions and the sulfuric acid (H2SO4) formation, which in turn is a primary driver for the nucleation process. Thanks to its flexible configuration and minor additional computational costs, the plume model presented here can readily be applied in coarse-resolution global aerosol-climate models or used as offline parametrisation to quantify the climate effects of aviation-induced aerosol particles.

The characteristics of aviation-induced aerosol, its processing, and effects on cirrus clouds and climate are still associated with large uncertainties. Properties of aviation-induced aerosol, however, are crucially needed for the assessment of aviation’s climate impacts today and in the future. We identified more than 1100 aircraft plume encounters during passenger aircraft flights of the IAGOS-CARIBIC Flying Laboratory from July 2018 to March 2020. The aerosol properties inside aircraft plumes were similar, independent of the altitude (i.e., upper troposphere, tropopause region, and lowermost stratosphere). The exhaust aerosol was found to be mostly externally mixed compared to the internally mixed background aerosol, even at a plume age of 1 to 3 h. No enhancement of accumulation mode particles (diameter >250 nm) could be detected inside the aircraft plumes. Particle number emission indices (EIs) deduced from the observations in aged plumes are in the same range as values reported from engine certifications. This finding, together with the observed external mixing state inside the plumes, indicates that the aviation exhaust aerosol almost remains in its emission state during plume expansion. It also reveals that the particle number EIs used in global models are within the range of the EIs measured in aged plumes.

We report on an inconsistency in the latitudinal distribution of aviation emissions between the data products of phases 5 and 6 of the Coupled Model Intercomparison Project (CMIP). Emissions in the CMIP6 data occur at higher latitudes than in the CMIP5 data for all scenarios, years, and emitted species. A comparative simulation with the chemistry-climate model ECHAM/MESSy Atmospheric Chemistry (EMAC) reveals that the difference in nitrogen oxide emission distribution leads to reduced overall ozone changes due to aviation in the CMIP6 scenarios because in those scenarios the distribution of emissions is partly shifted towards the chemically less active higher latitudes. The radiative forcing associated with aviation ozone is 7.6% higher, and the decrease in methane lifetime is 5.7% larger for the year 2015 when using the CMIP5 latitudinal distribution of emissions compared to when using the CMIP6 distribution. We do not find a statistically significant difference in the radiative forcing associated with aviation aerosol emissions. In total, future studies investigating the effects of aviation emissions on ozone and climate should consider the inconsistency reported here.

Aviation is seeking for ways to reduce its climate impact caused by CO₂ emissions and non-CO₂ effects. Operational measures which change overall flight altitude have the potential to reduce climate impact of individual effects, comprising CO₂ but in particular non-CO₂ effects. We study the impact of changes of flight altitude, specifically aircraft flying 2000 feet higher and lower, with a set of global models comprising chemistry-transport, chemistry-climate and general circulation models integrating distinct aviation emission inventories representing such alternative flight altitudes, estimating changes in climate impact of aviation by quantifying radiative forcing and induced temperature change. We find in our sensitivity study that flying lower leads to a reduction of radiative forcing of non-CO₂ effects together with slightly increased CO₂ emissions and impacts, when cruise speed is not modified. Flying higher increases radiative forcing of non-CO₂ effects by about 10%, together with a slight decrease of CO₂ emissions and impacts. Overall, flying lower decreases aviation-induced temperature change by about 20%, as a decrease of non-CO₂ impacts by about 30% dominates over slightly increasing CO₂ impacts assuming a sustained emissions scenario. Those estimates are connected with a large but unquantified uncertainty. To improve the understanding of mechanisms controlling the aviation climate impact, we study the geographical distributions of aviation-induced modifications in the atmosphere, together with changes in global radiative forcing and suggest further efforts in order to reduce long standing uncertainties.