Spectrophotometric pH(T) measurements can routinely be obtained with an imp
recision on the order of +/- 0.0005 pH(T) units. This level of imprecision
is equivalent to an imprecision in total hydrogen ion concentration of +/-
0.1% (where pH(T) = -log[H+], and [H+](T) congruent to [H+] + [HSO4-]). At
this level of precision, pH(T) measurements provide an important tool in qu
ality control assessments of other CO2-system parameters (CO2 fugacity, tot
al inorganic carbon, total alkalinity). CO2, fugacities and total alkalinit
ies calculated using measured pH(T) and total inorganic carbon, for the lar
ge data set considered in this work, have relative precisions on the order
of 0.15 and 0.1%, respectively. The precision of total inorganic carbon cal
culated via pH(T), and salinity-normalized-alkalinity is on the order of 0.
02% or better. In this work, using the NOAA 1992 boreal autumn Equatorial P
acific (EqPac) dataset, it is shown that CO2-system variables calculated vi
a pH(T) can be used to enhance both the precision and accuracy of directly
measured parameters. Through the procedures described in this work signific
ant problems were revealed in the initial version of the 1992 NOAA dataset,
and the dataset was greatly improved. Additionally, in this work, we revis
it CO2-system thermodynamic consistency issues in view of changes in the pH
(T), values assigned to tris seawater buffers and consequential changes in
the calibration of sulfonephthalein pH(T) indicators. As the principal resu
lt of a +0.0047 increase in the pK of meta cresol purple, CO2 fugacity calc
ulations and measurements are in very good agreement for the NOAA 1992 bore
al autumn EqPac dataset. We note, as well, that due to a reassignment of th
e titrant acid concentration used in the NOAA 1992 boreal autumn dataset. m
easured total alkalinities are in good agreement with total alkalinities ca
lculated from total carbon and pH(T). (C) 1999 Elsevier Science Ltd. All ri
ghts reserved.