Abstract

A check of the calibration of 10 batches of International Association for the Physical Sciences of the Ocean standard seawater (P120-P129) against the defined KCl standard (practical salinity scale 1978) showed differences of less than 0.001 in salinity (equivalent to circa 0.00003 in K15) during storage periods of up to 96 weeks. An experimental batch stored in borosilicate bottles showed no significant difference from the seawater stored in glass ampoules over a storage period of 158 weeks.

1. Introduction

As large-scale international projects (e.g., the World Ocean Circulation Experiment, Global Ocean Observing System, Joint Global Ocean Flux Study, Climate Variability and Predictability Study) occupy increasingly important roles in the production of oceanographic data, greater demands are made on the instrumentation used. However, the quality of those data depends on the calibration accuracy of the instruments and so increasing demands are also made on the standards and reference materials. This is particularly true of salinity data for which the need to define their accuracy is essential for an understanding of oceanic profiles.

The reliability of salinity determinations depends on many factors, but one factor common to all laboratory determinations is the International Association for the Physical Sciences of the Ocean (IAPSO) standard seawater, which is used to calibrate salinometers. Questions are frequently asked about the reliability of standard Seawater, its stability and how it should be stored. This paper summarises some of the data and experiences acquired by the IAPSO Standard seawater, its stability, and how it should be stored. This paper summarises some of the data and experiences acquired by the IAPSO Standard Seawater Service over the past 25 years.

2. History of calibration of IAPSO Standard Seawater

A detailed history of standard seawater can be found in Culkin and Smed (1979), but it is worth mentioning here the major changes in calibration that have taken place. When it was first introduced at the end of the nineteenth century, standard seawater was calibrated in chlorinity by titration with a silver nitrate solution, using potassium chloride as a reference standard. A subsequent change in the definition of chlorinity (Jacobsen and Knudsen 1940) led to the adoption of high-purity silver as the primary reference standard, but, in practice, each batch of standard seawater was still calibrated by silver nitrate titration.

In 1978 the practical salinity scale (UNESCO 1980), involving a fundamental change in the definition of salinity, was adopted and potassium chloride again became the reference standard, this time in conductivity. Salinity was defined in terms of electrical conductivity ratio, at 15°C and 1 atmosphere, relative to a KCl solution containing 32.4356 g kg-1 (corrected for buoyancy). Note that the 15°C mentioned in the salinity definition is on the temperature scale (IPTS-68), which was in operation at the time. If salinity is eventually redefined to take into account the new temperature scale introduced in 1990 (ITS-90), a small change in the defined concentration of KCl will be necessary, but this will not affect the value of K15 for any given seawater. In the meantime, since 1981 the conductivity of each batch of standard seawater has been compared with that of a defined KCl standard in accordance with the definition, where the ratio is K15.

3. Early comparisons of batches of standard seawater

In the years when standard seawater was certified in chlorinity and used as a standard for chlorinity titrations, the calibration was carried out by a time-consuming combined gravimetric, potentiometric titration used only by the IAPSO Standard Seawater Service. The only independent check appears to have been that carried out on batches prepared between 1969 and 1974 in preparation for the transfer of the Service from Copenhagen to the Institute of Oceanographic Sciences, Wormley, England (Hermann & Culkin, 1972). The agreement between the two laboratories (std dev 4 - 6 X 10-4 in salinity) confirmed the reliability of the calibrations but revealed nothing about the stability of standard seawater, as neither reaction with the glass ampoules nor bacterial contamination (that had been encountered previously) was likely to alter the chlorinity.

From the late 1950s, chlorinity titration was gradually replaced by measurement of electrical conductivity for the determination of salinity, but IAPSO standard seawater, although being used as a conductivity standard, continued to be certified in chlorinity. Early comparisons of batches of standard seawater prepared between 1937 and 1978 (Park 1964; Millero et al. 1977; Poisson et al. 1978; Mantyla 1980) revealed variations in the chlorinity/ conductivity relationship and a need for the standard to be calibrated in electrical conductivity.

Significantly, from the point of view of stability, three batches (P49 - P51) were found (Millero et al. 1977, Poisson et al. 1978) to have anomalously high conductivities, which were attributed (F. Hermann 1976, personal communication) to bacterial contamination, possibly combined with oil pollution. Recently, Mantyla (1987) and Takatsuki et al. (1991) reported that agreement between batches had improved since the adoption of a defined KCl solution as reference.

4. Comparisons by the Standard Seawater Service

Although the investigations mentioned above were of high quality, they did not give an absolute measure of change in conductivity, as the comparisons were made relative to older batches of standard seawater, which themselves may have changed. The practical salinity scale 1978, however, provided a means of checking the calibration and stability of standard seawater against a reproducible KCl standard. Since batch P91, produced in 1981, all batches of standard seawater (i.e., the working standard) have been calibrated in conductivity relative to this defined KCl standard and labeled with the appropriate K15. Details of the procedures for preparing the standard KCl solutions and for calibrating new batches of standard seawater have been published (Culkin 1986) and need not be repeated here. During the past few years, however, the calibration of the previous two or three batches were checked at the same time that a new batch was being calibrated, and the results of these measurements are shown in Table 1.

Table 1. Changes in K15 values of IAPSO standard seawaters after storage.