资源环境科技发展态势分析平台

Direct observations indicate that the global ocean oxygen inventory is decreasing. Climate models consistently confirm this decline and predict continuing and accelerating ocean deoxygenation. However, current models (1) do not reproduce observed patterns for oxygen changes in the ocean's thermocline; (2) underestimate the temporal variability of oxygen concentrations and air-sea fluxes inferred from time-series observations; and (3) generally simulate only about half the oceanic oxygen loss inferred from observations. We here review current knowledge about the mechanisms and drivers of oxygen changes and their variation with region and depth over the world's oceans. Warming is considered a major driver: in part directly, via solubility effects, and in part indirectly, via changes in circulation, mixing and oxygen respiration. While solubility effects have been quantified and found to dominate deoxygenation near the surface, a quantitative understanding of contributions from other mechanisms is still lacking. Current models may underestimate deoxygenation because of unresolved transport processes, unaccounted for variations in respiratory oxygen demand, or missing biogeochemical feedbacks. Dedicated observational programmes are required to better constrain biological and physical processes and their representation in models to improve our understanding and predictions of patterns and intensity of future oxygen change.