Re-evaluation of carbonic acid dissociation constants across conditions and the implications for ocean acidification

With the increasing threat of ocean acidification and the important role of the oceans in the global carbon cycle, highly precise, accurate, and intercomparable determination of inorganic carbon system parameters is required. Thermodynamic relationships enable the system to be fully constrained using a combination of direct measurements and calculations. However, calculations are complicated by many formulations for dissociation constants (over 120 possible combinations). To address these important issues of uncertainty and comparability, we evaluated the various combinations of constants and their (dis)agreement with direct measurements over a range of temperature (−1.9–40 °C), practical salinity (15–39) and pCO2 (150–1200 μatm). The results demonstrate that differences between the calculations and measurements are significantly larger than measurement uncertainties, meaning the oft-stated paradigm that one only needs to measure two parameters and the others can be calculated does not apply for climate quality ocean acidification research. The uncertainties in calculated pHt prevent climate quality pHt from being calculated from total alkalinity (TA) and dissolved inorganic carbon (DIC) and should be directly measured instead. However, climate quality TA and DIC can often be calculated using measured pH and DIC or TA respectively. Calculations are notably biased at medium-to-high pCO2 values (∼500–800 μatm) implying models underestimate future ocean acidification. Uncertainty in the dissociation constants leads to significant uncertainty in the depth of the aragonite saturation horizon (>500 m in the Southern Ocean) and must be considered when studying calcium carbonate cycling. Significant improvements in the precision of the thermodynamic constants are required to improve pHt calculations. •pHt should be measured directly rather than calculated from TA and DIC.•Uncertainty in the constants contributes up to 680 m of uncertainty in the aragonite saturation horizon depth.•Calculated pHt is worst at pCO2 ∼500 – 800 μatm, suggesting an underestimation of future ocean acidification in models.•Evaluation of an unidentified or organic component of TA was inconclusive.