Insights from GO-SHIP hydrography data into the thermodynamic consistency of CO2 system measurements in seawater

Due to advances in technology, routine seawater pH measurements of excellent repeatability are becoming increasingly common for studying the ocean CO2 system. However, the accuracy of pH measurements has come into question due to a widespread observation, from a large number of carefully calibrated state-of-the-art CO2 measurements on various cruises, of there being a significant pH-dependent discrepancy between pH that was measured spectrophotometrically and pH calculated from concurrent measurements of total dissolved inorganic carbon (C-T) and total alkalinity (A(T)), using a thermodynamic model of seawater acid-base systems. From an analysis of four recent GO-SHIP repeat hydrography datasets, we show that a combination of small systematic errors in the dissociation constants of carbonic acid (K-1 and K-2), the total boron-salinity ratio, and in C-T and A(T) measurements are likely responsible for some, but not all of the observed pH-dependent discrepancy. The residual discrepancy can only be fully accounted for if there exists a small, but meaningful amount (similar to 4 mu mol kg(-1)) of an unidentified and typically neglected contribution to measured A(T), likely from organic bases, that is widespread in the open ocean. A combination of these errors could achieve consistency between measured and calculated pH, without requiring that any of the shipboard measurements be significantly in error. Future research should focus on establishing the existence of organic alkalinity in the open ocean and constraining the uncertainty in both CO2 measurements and in the constants used in CO2 calculations.