Sea-air CO2 fluxes in the Southern Ocean: 1990–2009 (7749)
The Southern Ocean has a major influence on the global carbon cycle and is a region where climate change could significantly alter the efficiency of the ocean carbon sink and the global carbon balance. An understanding of the variability in the net sea-air CO2 flux for the region is a key need in establishing the current and future roles of the Southern Ocean as a sink for atmospheric CO2. The assessment of the net sea-air CO2 fluxes for the region is made difficult by sparse data coverage, which has resulted in a variety of observational products, ocean models and atmospheric and oceanic inversions being applied to estimate the fluxes on seasonal through interannual scales. As a part of the Regional Carbon Cycle Assessment Project we provide a first assessment of how the various approaches agree for the Southern Ocean, for the period 1990-2009. When integrated over the entire Southern Ocean (44°S–75°S), the mean of the annual net sea-air flux for 26 ocean models and inversions of −0.42 ± 0.07 PgC yr−1 is consistent with an observational based estimate of −0.27 ± 0.13 PgC yr−1. The agreement between modeling approaches and observations is less straightforward on seasonal and regional scales. The median of ocean biogeochemical models captures the observed seasonal cycle in the 44°S–58°S circumpolar region, while the median of 11 atmospheric inversions shows little seasonal change in the net flux. South of 58° S, neither atmospheric inversions nor ocean biogeochemical models reproduce the phase and amplitude of the observed seasonal sea-air CO2 flux, particularly in the Austral Winter. None of the individual ocean models and atmospheric inversions reproduce both the observed annual mean uptake and the observed seasonal cycle and there is considerable variability in the regional patterns of the net fluxes. This raises concerns about projecting future changes in the Southern Ocean CO2 sink. We also find that trends in the net CO2 flux from the inversions and models are not statistically different from an increase of –0.05 Pg C yr−1 decade−1 that is expected from increasing atmospheric CO2 concentrations. However, resolving long-term trends is difficult due to the large interannual variability and short time frame considered (1990–2009).