Presenter: Nicolas Mayot (University of East Anglia)

Description:
The Southern Ocean plays a major role in both the global oceanic uptake of anthropogenic CO₂ and its interannual variations. The origin and magnitude of the interannual variability in the Southern Ocean CO₂ fluxes is debated. Observation-based estimates suggest a large variability (+/- 0.11 PgC/yr) while Global Ocean Biogeochemistry Models (GOBMs) simulate almost no variability. Studying the air-sea fluxes of O₂ can provide independent information that help resolve this data-model inconsistency. Owing to its higher sensitivity to changes in mixing and the lack of masking by the large anthropogenic CO₂ flux, analyses of O₂ variability can help to elucidate the potential presence of model biases in mixed layer dynamics. Here, we used 26 years (1994-2019) of monthly O₂ fluxes from 9 GOBMs. These model outputs were compared to air-sea O₂ fluxes inferred from an atmospheric inversion that used precise measurements of atmospheric O₂ and CO₂ levels. The 26-year time series of air-sea O₂ fluxes from the GOBM ensemble and the atmospheric inversion were significantly positively correlated to each other. This could be linked to the Southern Annular Mode and its influence on air-sea heat flux forcing that induced large-scale changes in observed wintertime Mixed Layer Depth (MLD). However, the amplitude of the interannual variability in air-sea O₂ fluxes was three times higher in the atmospheric inversion than in GOBMs. We show, using one dimensional bulk mixed layer model experiments, that one GOBM simulated lower than expected air-sea fluxes of O₂ south of the subantarctic front because its subsurface vertical gradients in oxygen saturation was lower than observed. It is possible that this was induced by the general overestimation of the mean wintertime MLD by the GOBM. Implications of these results for the variability in air-sea fluxes of CO₂ will be discussed.