Intercomparison of carbonate chemistry measurements on a cruise in northwestern European shelf seas

Abstract

Four carbonate system variables were measured in surface waters during a cruise aimed at investigating ocean acidification impacts traversing northwestern European shelf seas in the summer of 2011. High-resolution surface water data were collected for partial pressure of carbon dioxide (pCO2; using two independent instruments) and pH using the total pH scale (pHT), in addition to discrete measurements of total alkalinity and dissolved inorganic carbon. We thus overdetermined the carbonate system (four measured variables, two degrees of freedom), which allowed us to evaluate the level of agreement between the variables on a cruise whose main aim was not intercomparison, and thus where conditions were more representative of normal working conditions. Calculations of carbonate system variables from other measurements generally compared well with direct observations of the same variables (Pearson's correlation coefficient always greater than or equal to 0.94; mean residuals were similar to the respective accuracies of the measurements). We therefore conclude that four of the independent data sets of carbonate chemistry variables were of high quality. A diurnal cycle with a maximum amplitude of 41 μatm was observed in the difference between the pCO2 values obtained by the two independent analytical pCO2 systems, and this was partly attributed to irregular seawater flows to the equilibrator and partly to biological activity inside the seawater supply and one of the equilibrators. We discuss how these issues can be addressed to improve carbonate chemistry data quality on future research cruises.