Total Alkalinity Measurements
The TALK concentrations were determined by potentiometric titration of 1153 Niskin samples, 574 from Leg 1 and 579 from Leg 2. Samples from throughout the water column were measured on 39 stations (nominally 36 depths per station) and from surface Niskins only on 41 additional stations. The TALK was measured on an aliquot of seawater taken from the same 500-mL bottle previously analyzed for TCO2. Calibration of the shipboard measurements of TALK reported in this numeric data package depends upon the standardization of the HCl titrants with titrations of primary standard sodium carbonate solutions at SIO. The titration system and its calibration are described in Guenther et al. (1994a), a reprint of which is provided in Appendix A of this report. Adjustments to the TALK calibration scale are likely to be made in the future.
Data quality was assessed at sea by titration of replicate seawater samples, secondary standard bicarbonate solutions prepared at SIO before expedition, and bottles of CRM batch number 13. Aliquots from the replicate seawater samples and the CRMs were titrated after aliquots had been removed for TCO2 measurements.
The short-term repeatability was estimated by analyzing the agreement of pairs of replicate seawater samples titrated simultaneously, using equation (3) in Standard Operating Procedure (SOP) 23 of DOE (1994). One or two pairs usually were measured on each day of analysis. On Leg 1, for 33 pairs, the sample standard deviation, si, of a single measurement was estimated to be 1.56 µmol/kg. On Leg 2, for 30 pairs, si was estimated to be 2.13 µmol/kg.
Two batches of bicarbonate reference materials were titrated during the cruise. Usually four measurements were made per day. Analysis of the results using the normal equation for sample standard deviation yields an estimate of the reproducibility of the measurements over the entire cruise. The si was found to be 2.77 µmol/kg for 75 measurements of batch "A" and 2.03 µmol/kg for 90 measurements of batch "B."
Titrations of CRM samples provided an additional estimate of reproducibility and also an estimate of the accuracy through comparison of the at-sea results with the value certified by the laboratory of A. G. Dickson at SIO. The value for CRM batch 13, certified by titrations in 1996 on archived samples, was 2203.79 µmol/kg. During the cruise 84 titrations of CRM batch 13 were made. After 6 measurements were rejected, the si calculated for 78 measurements was 2.29 µmol/kg. The average TALK for the 78 measurements was 2201.26 µmol/kg, nearly within one standard deviation of the certified value. The TALK measurements of seawater reported here have NOT been adjusted by this difference. Figure 5 is a plot of the difference between the shipboard TALK of CRM batch 13 and the certified value versus time during both legs of the cruise.
For more detailed information on TALK measurements during the R/V John V. Vickers cruise along WOCE section P13 please see Appendix A in this report.