Total Alkalinity Measurements
TALK and pH were measured using an automated potentiometric titration system developed at the University of Miami (hereafter designated as MATS). MATS is described by Millero et al. (1993a). It consisted of two parts: a Metrohm model 665 Dosimat titrator and a pH meter (Orion, Model 720A) which are interfaced with a PC. A water-jacketed, fixed-volume (~200 mL), closed Plexiglass sample cell, of greater volume than but otherwise similar to those used by Bradshaw and Brewer (1988), was used to increase the precision of the measurements. The cell, titrant burette, and sample cell were theromstatted at 25 0.05C using a constant temperature bath (Neslab, Model RTE 221). A Lab Windows/CVI program was used to run the titrators, record the volume of titrant added, and record the measured electromagnetic frequency (emf) of the electrodes through RS232 serial interfaces. The electrodes for measuring the emf during the titration consisted of a ROSS glass pH electrode (Orion, Model 810100) and a double-junction Ag/AgCl reference electrode (Orion, Model 900200).
Seawater samples were titrated by adding enough HCl to exceed the carbonic acid endpoint of the titration. During a typical titration, the emf readings were recorded until stable (0.05 mV). Normally, at this point, a fixed volume of acid would be added; however, the MATS were designed to add enough acid to increase the voltage by a pre-assigned increment (13 mV). This was done to give an even distribution of data points over the course of a full titration, which consists of 25 data points and takes about 20 minutes. With two systems, approximately 7 hours was required to run a 31-bottle station cast. As noted in Sections 3.1 and 3.2, 4-L Niskin sampling bottles were employed on the rosette, which limited the amount of sample available for the carbonate system analysts to one 500-mL bottle. Hence there was not enough sample water to complete duplicate alkalinity analyses from the same bottle or to draw duplicate samples from the same sampling bottle.
The titrant (acid) used throughout the cruises was prepared, standardized, and stored in 500-mL borosilicate glass bottles for use in the field. A single 55-gal batch of 0.25-m HCl acid was prepared by dilution of concentrated HCl (AR Select Mallinckrodt). The acid was prepared in 0.45m NaCl to yield a total ionic strength similar to that of seawater salinity 35.0 (I = ~0.7 M). The acid was standardized by coulometry (Taylor and Smith 1959; Marinenko and Taylor 1968). The acid molality was also checked by titration on seawaters with known alkalinities, and subsamples were sent to the laboratory of A. Dickson at SIO for an independent laboratory determination of the molality. The calibrated molality of the acid used for the North Atlantic WOCE Sections was 0.24892 0.00003 m HCl.
The consistency of the method was checked for each cast using low-nutrient surface seawater, and the accuracy of the method was checked by analyzing CRM Batches 33 (1996), 36, and 37 (1997) and comparing the analyzed values with the certified TALK in the same manner as for TCO2 (see also Section 3.2 for batch data). The mean differences between at-sea measurements and the certified TALK values are given in Table 6. The TALK of each batch was also determined in the laboratory by weight titrations, which were found to agree with the certified values to 2 µmol/kg. In addition, the pH of the CRM batches was determined in the laboratory spectrophotometrically according to the methods of Clayton and Byrne (1993) prior to the cruise. The at-sea titration pH measurements were also compared with the pre-cruise spectrophotometric values, and the reader is referred to Millero et al. (1999) for further details.
Table 6. The mean analytical difference between analyzed and certified TALK for the MATS on WOCE Section AR24 (1996), and Sections A24, A20, and A22 (1997)
| Section | Cells | n | CRM TALK µmol | Measured TALK µmol/kg | ΔTALK µmol/kg |
|---|---|---|---|---|---|
| AR24 | 2, 19, 17 | 59 | 2234.9 | 2233.3 | −1.6 |
| A24 | 2, 18, 12 | 148 | 2283.9 | 2283.3 | −0.6 |
| A20 | 2, 18, 12 | 96 | 2314.1 | 2217.1 | 3.0 |
| A22 | 2, 12 | 65 | 2314.1 | 2215.4 | 1.3 |
The mean differences between the at-sea measurements and the certified TALK were within 3.0 µmol/kg (Table 6). Hence the measured and certified TALK were in good agreement. For pH and TCO2, the corresponding results were 0.021 and 9 µmol/kg, respectively, with the larger deviation in pH attributable to the non-Nernstian behavior of the electrodes near a pH of 8 (Millero et al. 1993b).
The at-sea sample alkalinity titrations were corrected using the results for the CRM. For TALK, the CALFAC used to correct the at sea measurements was
CALFAC = CRM (certified value)/(at-sea value) ,
and for pH the CALFAC was
pH = pH (CRM) / pH (at-sea) .
Duplicate samples were usually taken for each station in the same manner as for TCO2 (surface and deep) and analyzed to determine and monitor the precision of the MATS. The average difference between replicates was 1.0, 1.1, and 1.1 µmol/kg for sections A24, A20, and A22, respectively, which demonstrated the high precision of the MATS throughout the study. A preliminary description of the major trends in the data and the behavior of alkalinity over time in the North Atlantic is given by Millero et al. (1999).
