A deep-learning estimate of the decadal trends in the Southern Ocean carbon storage

Uptake of atmospheric carbon by the ocean, especially at high latitudes, plays a crucial role in mitigating anthropogenic emissions. This study develops a machine-learning model to estimate dissolved inorganic carbon (DIC) concentrations in the Southern Ocean up to 4 km depth using only surface data. The model reveals that DIC decreased in the 1990s and 2000s but increased, particularly in the upper ocean, since the 2010s. Zonal variations suggest differing mechanisms across the Southern Ocean, with near-surface DIC decrease potentially enhancing CO2 uptake, while weakened deep-layer connectivity may reduce the ocean's carbon storage capacity.

Varvara E. Zemskova, Tai-Long He, Zirui Wan, Nicolas Grisouard