The expected fall of the airborne fraction

The airborne fraction (AF) denotes the proportion of anthropogenic carbon emissions that remains in the Earth’s atmosphere, driving global warming. Consequently, the AF reflects the combined sink efficiency of land and ocean within the global carbon cycle. Several studies, including the Global Carbon Project, report an remarkably constant AF over the last decades, suggesting no sign of sink saturation. Here we first revisit the AF, scrutinizing both observed and simulated AF over the last decades, and crucially, examine future AF trajectories in emission-driven scenarios simulated in a multi-model ensemble setting (CMIP6) and specially designed experiments. After removing land-carbon cycle variability with temperature-sensitive CO2 inversions, a subtle increasing trend in the observed AF emerges. This trend aligns generally with our analyses of model predictions under (quasi-)exponential emission pathways, akin to the historical period, suggesting a diminishing emission fraction absorbed by sinks. We continue exploring non-exponential emission scenarios featuring plateauing and declining emission rates. Our findings reveal that AF begins to decline with the deceleration of annual emission rates. This phenomenon can be attributed to the decay of old carbon in the atmosphere dominating the annual sink rate, resulting in an AF decrease, despite the increase in atmospheric lifetime of newly added carbon due to sink saturation. Utilizing a simple model coupling global energy-balance and carbon-cycle dynamics, alongside an emulator incorporating a varying system state based on impulse-response functions, we demonstrate that diverse model parameter settings can replicate historical AF trajectories under exponential emission pathways. However, significant divergence occurs once emissions plateau, with all model versions indicating an eventual decline in AF. In conclusion, while past and present trends pointing at a rising AF, an eventual decline is anticipated. The trajectory of future AF during phases of decelerating emission rates remains a key unknown, warranting further research.