Carbon Modeling of the Subpolar North Atlantic

It is now well established that the global ocean is on average a CO2 sink of about 2 PgC yr-1 accumulating anthropogenic CO2 with a direct impact on the pH and consequently the marine biota. In recent years, several analyses based on CO2 observations in the North Atlantic, North and Equatorial Pacific, and Southern Ocean, show that the ocean uptake is not steady on interannual to decadal scales and that it might impact the atmospheric CO2 content.

Almost all regional analyses suggest that the main drivers of pCO2 multiannual changes (and air-sea fluxes) are thermodynamic and/or dynamic (e.g. warming, lateral/vertical transport, mixing), except in the Bering Sea where the biological activity mostly explains an observed decrease of pCO2.

In the Subpolar North Atlantic (SPNA) region, observational and modeling studies suggest that, depending on the time period, pCO2 multiannual variations are mainly controlled by warming changes in vertical mixing and/or water transport linked to the North Atlantic Oscillation (NAO). This region experiences large blooms in summer/autumn (CO2 drawdown) leading to a strong pCO2 seasonality. At this scale, the biological effect on pCO2 dominates the sea surface temperature (SST) changes.

In this context, it is challenging to evaluate the role of physical versus biological processes to explain long-term variations of CO2 properties (partial pressure of CO2, pCO2; dissolved inorganic carbon, DIC; total alkalinity, TA; and  pH), and not only in winter, when the primary production (PP) is low and the CO2 signal is generally less variable  from year to year. A complete evaluation of the decadal changes of air-sea CO2 fluxes (i.e., a value integrated over each year) calls for a comprehensive analysis of the changes and trends for all seasons and years that is very difficult to achieve when this is only based on sparse observations.

In this study, we combine several in-situ and satellite observations to evaluate the quality of a physical-biogeochemical model from which we derive and interpret, for all seasons, the decadal changes of carbon properties, air-sea CO2 fluxes, and acidification rates. To investigate this issue, we have selected a region south of Iceland (black triangle in Fig. 1) where observed carbon properties are relatively abundant and the biological activity (phytoplankton biomass and PP) has a strong impact on fisheries and the carbon cycle.

The impact of ocean acidification on the marine environment is a topic of increased interest and has received a lot of attention by the government, funding agencies, and academia. Decadal upward trends in surface ocean pCO2 reduce the seawater pH and decrease the carbonate concentration which is crucial for the formation of hard parts such as skeletons and shells of calcium carbonate for calcifying organisms. Therefore, ocean acidification can have harmful consequences to marine biota.

The impact of pCO2 trends on ocean acidification is illustrated in Fig. 2, which shows time series of model vertical profiles of pH, and saturation states for aragonite (Ωa) and calcite (Ωc). Downward trends of these ocean acidification parameters are clearly evident. Winter surface ocean trends in pCO2, TCO2, pH, Ωa, and Ωc were estimated at +2.83 µatm yr-1, +1.059 µmol kg-1 yr-1, -0.003  yr-1, -0.0095 yr-1, and -0.0152 yr-1, respectively. The summer surface ocean trends derived from the model for the same parameters are somewhat smaller, +1.99 µatm yr-1, +0.708 µmol kg-1 yr-1, -0.0024 yr-1, -0.0071 yr-1, and -0.0115 yr-1, respectively.

Figure 1.

Subpolar North Atlantic and Nordic Seas map showing the annual climatologic surface currents from the 3D model and MODIS climatologic chlorophyll for June. The black triangle shows the location where the 1D ecosystem-carbon cycle model was applied. The thick contour line is the annual mean 4oC SST contour representing the location of the Arctic Front.

Figure 2.

Figure 2. Ocean acidification parameters for 1981-2008. a-c, Model time series of vertical profiles of acidification parameters (a), pH (b), aragonite saturation state (Wa) and (c), calcite  saturation state (Wc).