Future ocean hypercapnia driven by anthropogenic amplification of the natural CO2 cycle

Elevated carbon dioxide concentrations in seawater (hypercapnia) can induce neurological, physiological and behavioural deficiencies in marine animals. Prediction of the onset and evolution of hypercapnia in the ocean requires a good understanding of annual oceanic carbon dioxide variability, but relevant global observational data are sparse. Here we diagnose global ocean patterns of monthly carbon variability based on observations that allow us to examine the evolution of surface ocean CO₂ levels over the entire annual cycle under increasing atmospheric CO₂ concentrations. We find that some oceanic regions undergo an up to 10-fold amplification of the natural cycle of CO₂ by 2100, if atmospheric carbon dioxide concentrations continue to rise throughout this century (RCP8.5). Projections from a suite of Earth System Climate Models are broadly consistent with the findings from our data based approach. Our predicted amplification in the annual CO₂ cycle displays distinct global patterns that may expose major fisheries in the Southern, Pacific and North Atlantic Oceans to high CO₂ events many decades earlier than expected from average atmospheric CO₂ concentrations. We suggest that these ocean 'CO₂ hotspots' evolve as a combination of the strong seasonal dynamics of CO₂ and the long-term effective storage of anthropogenic CO₂ that lowers the buffer capacity in those regions, causing a non-linear CO₂ amplification over the annual cycle. The onset of ocean hypercapnia events (pCO₂ >1000 µatm) is forecast for atmospheric CO₂ concentrations that exceed 650 ppm, with hypercapnia spreading to up to one half of the surface ocean by the year 2100 under a high-emissions scenario (RCP8.5) with potential implications for fisheries over the coming century.