Interpreting Measurements of Total Alkalinity in Marine and Estuarine Waters in the Presence of Proton-binding Organic Matter

<p> <p id="x-x-abspara0010"> Total alkalinity ( <em> A </em> _T ) is one of four measurable cornerstone parameters for characterizing the marine carbonate system, yet its measurement by standard titration methods is subject to systematic misinterpretations in the presence of uncharacterized dissolved organic molecules in ocean and estuarine waters. A consequence of these misinterpretations may be the lack of thermodynamic consistency that is routinely observed among measured and calculated parameters of the carbonate system. In this work, a numerical model is used to illustrate (a) how proton-binding dissolved organic molecules influence the reported results of total alkalinity titrations in marine and estuarine settings and (b) how errors in interpretations of reported <em> A </em> _T values can then propagate through carbonate system calculations, thus distorting biogeochemical interpretations of calculated parameters. We examine five distinct approaches for alkalinity measurement by titration. Ideally, the difference between the measured (reported) <em> A </em> _T and the conventional (thermodynamic) definition of inorganic alkalinity ( <em> A </em> _inorg ) would be zero. However, in the presence of titratable organic matter, our model results show consistent non-zero differences that vary with the chemical properties of the organic matter. For all five titration approaches, the differences between reported <em> A </em> _T and <em> A </em> _inorg are greatest when the negative logarithm of the organic acid dissociation constant (p <em> K </em> _org ) is between approximately 5 and 7. The differences between reported <em> A </em> _T and <em> A </em> _inorg also display previously undescribed variation among measurement approaches, most significantly when p <em> K </em> _org is between approximately 3 and 6 (typical of carboxylic acid groups). The measurement approaches that are most effective at limiting the unfavorable influence of these relatively low-p <em> K </em> organic acids on <em> A </em> _T are closed-cell titrations and single-step titrations that are terminated at a relatively high pH. For calculated carbonate system parameters relevant to in situ conditions (e.g., pH, <em> p </em> CO ₂ , calcium carbonate mineral saturation states), errors resulting from the presence of proton-binding organics are largest when calculations are based on the input pair of directly measured dissolved inorganic carbon ( <em> C </em> _T ) and directly measured <em> A </em> _T , and can vary in magnitude depending on the titration approach that is used to obtain <em> A </em> _T . The modeling results presented in this work emphasize the importance of (a) determining <em> A </em> _T in a manner that accounts for the ubiquity of organic alkalinity in marine and estuarine waters and (b) working toward a clearer understanding of the phenomena underlying the routine lack of internal consistency between measured versus calculated carbonate system parameters. Total alkalinity measurements should begin to incorporate either implicit or explicit evaluations of the titration characteristics of the natural organic carbon present in each sample. To that end, we recommend use of secondary titrations to directly measure organic alkalinity (sample-by-sample), characterization of relationships between total dissolved organic carbon concentrations and organic alkalinity (on local to regional scales), and/or exploration of novel curve-fitting procedures to infer the behavior of organic functional groups from titration data. </p></p>