Effect of selected calculation routines and dissociation constants on the determination of total carbon dioxide in seawater

Abstract

During the 1989 and 1990 JGOFS North Atlantic Pilot Study a comparison was made between the Coulometric and the acid titration method for determination of total carbon dioxide (TCO 2) in seawater. TCO 2 and alkalinity have been calculated from acid titration using either the modified Gran plot or the curve-fitting routine. Depth profiles showed fair agreement (on average 0.6% or about 12.5 μmol 1 −1) between the TCO 2 calculated from the acid titration method and the TCO 2 measured independently by Coulometry. It is shown that different data processing routines combined with the proper use of dissociation constants can influence the acid titration result considerably. There appears to be a slight offset between calculated and Coulometric data which is smallest when using the combined dissociation constants of Hansson ( Deep-Sea Research, 20, 461–478, 1973) with Goyet and Poisson ( Deep-Sea Research, 36, 1635–1654, 1989). No statistically significant difference could be found between the two used calculation methods (Gran plot and curve-fitting). The agreement between the independent methods of Coulometry and acid titration is encouraging and furthermore independent of depth, for this dataset there is no reason for invoking the existence of interfering (organic) protolytes.