Over the past decade (2013–2024), the ocean has absorbed approximately 2.9 ± 0.4 Gt C per year, or ˜ 30% of anthropogenic fossil fuel emissions. Changes in air-sea fluxes reflect the superposition of interannual to centennial variability in the natural carbon cycle and the response to antropogenic forcing. On interannual to decadal timescales, and thus far, natural climate variability modes appear to be the primary drivers of interannual fluctuations in the ocean carbon sink. Among these modes, ENSO (El Niño–Southern Oscillation) plays a dominant role. The year 2023 saw the combination of such an episode in the Equatorial Pacific and warm anomalies in extratropical and subpolar regions. The expected strengthening of the ocean carbon sink in response to the El Niño episode (decreased CO2 outgassing) was offset by excessive outgassing in extratropical and subpolar regions, particularly in the North Atlantic. This situation appears unprecedented and illustrates the interactions between the natural cycle and anthropogenic forcing (in this case, the warm ocean surface temperature anomaly). To understand contemporary and future fluctuations of air-sea CO2 fluxes, it is therefore important to understand their natural variability.

The Holocene is a period characterized by a relatively stable mean climate and patterns of climate variability similar to those observed today. This study takes advantage of this relative stability of the Holocene’s mean climate to investigate the natural variability of air-sea CO2 fluxes on interannual and decadal timescales. The aim is to establish a baseline against which to evaluate contemporary and future variability.

The first objective will be to confirm and quantify the role played by regional climate variability modes (e.g., ENSO, Equatorial Pacific) in the variability of air-sea CO2 fluxes during the Holocene. The next step will be to explore the existence of compensatory phenomena similar to those observed in 2023. Depending on the results obtained, the approach will subsequently be applied to historical and future periods.