Abstract. Oxygen is fundamental to ocean biogeochemical processes, with deoxygenation potentially reducing biodiversity, and disrupting biogeochemical cycles. In recent decades, the global ocean oxygen concentration has been decreasing, but this decrease is underestimated in numerical ocean models by as much as 50 %. Mechanisms responsible for this deoxygenation include (i) solubility-driven deoxygenation due to ocean warming, and (ii) changes in the remineralised signal due to either a change in the supply of biological material to depth or a change in circulation leading to change in the residence time of water, and hence the accumulation of the remineralised oxygen deficit, or a combination of both. The magnitude of oxygen change due to each process is currently unclear. Here, we describe and implement a new method to decompose oxygen change into its constituent parts by linking each process to concomitant changes in temperature and dissolved inorganic carbon. Using observations on a repeated section of the North Atlantic at 24.5° N, we show that the consistent oxygen decrease observed since 1992 in the upper 2000 m has been dominated by an increase in remineralisation-related oxygen-consumption. While warming-driven solubility changes have a much smaller impact on the upper ocean in comparison, their impact has trebled in the past twenty years, suggesting they will become an increasingly significant driver of deoxygenation with future warming. Remineralisation-related oxygen consumption peaks at a depth of approximately 600 m, where it is responsible for up to 70 % of the total deoxygenation. While this study does not determine the exact cause of the remineralisation-driven change, little change in primary productivity has been observed in the region, suggesting that a change in ocean circulation is indirectly driving the majority of deoxygenation in the Subtropical North Atlantic, via a non-local change in remineralisation.
Sanders et al. (2026) studied this question.