SUMMARY: Seawater samples are collected at discrete depths with PVC sampling bottles attached to a CTD/ rosette system. Seawater samples are brought to constant temperature and pH is determined from the millivolt output of an Orion reference electrode calibrated with buffer solutions. The buffer solutions are prepared in a special matrix which mimics the ionic characteristics of seawater.

1. Principle

The pH (-log of the hydrogen ion concentration) is a master variable for describing the acid-base equilibria in seawater. The pH of seawater is dependent upon the in situ chemical, physical and biological conditions, and pH is also one of four variabilities which can be used to describe the carbonate system in seawater. This parameter is therefore useful for examining the exchange of carbon dioxide across the air-sea interface and for the calculation of the saturation state of calcium carbonate in the interior of the ocean. pH is a unique physiochemical quantity because it involves the activity of a single ion which cannot be measured directly. As a consequence, pH is defined operationally in terms of the method by which it is measured. In order for pH measurements to be interpreted correctly the reference scale must be precisely defined. In this procedure, pH is measured electrochemically using a combination electrode. Measurements are made relative to seawater buffers prepared as described.

2. Precautions

Because pH is dependent upon temperature, the temperature of the pH sample must be carefully controlled, or at least accurately measured before pH is determined. In addition, the pH of a seawater sample changes if CO2 is lost or gained on contact with the atmosphere. This is a concern particularly with deep water samples where CO2 concentrations are high. It is therefore important that samples be drawn as soon as possible after arrival of the rosette on deck. Furthermore, contact with the atmosphere should be avoided during sampling and sample bottles should remain tightly sealed until pH measurements are made.

3. Water Sampling

3.1. pH samples are collected from the PVC sampling bottles into a clean 200 ml polyethylene bottle using a Tygon drawing tube to minimize sample contact with the atmosphere.
3.2. The sample bottle is overflowed with at least one volume before the sample bottle is tightly capped.
3.3. The air space with the sample bottle is kept to a minimum in order to minimize CO2 exchange with the head space.

4. Sample Analysis

4.1. pH sample bottles are immersed in a constant temperature water and brought to 25°C.
4.2. After reaching constant temperature, the sample bottle is opened and a clean stir bar is placed in the bottle.
4.3. The bottle is placed on a stir plate and the sample temperature is measured to 0.1°C.
4.4. The potentiometric electrode is placed in the bottle and the mv output is recorded after the reading stabilizes.

5. Electrode calibration and calculation of pH

The potentiometric electrode is calibrated with the seawater buffers 2-amino-2-hydroxymethyl-1,3-propanedion (tris) and 2-amino-2- methyl-1,3-propanediol (bis). These buffers are made as described by Dickson (1991) and pH is computed on the "total" hydrogen scale. The electrode slope and an isoelectric point is computed from the mv outputs when the electrode is immersed in the seawater buffers as described in Grasshoff (1983). The pH of seawater samples is calculated using the electrode slope and isoelectric point and sample temperature.

6. Precision and Accuracy

Because of a variety of problems inherent in electrometric pH measurements, including electrode drift, electromagnetic interference and problems with the reference electrode, the precision of these pH measurements is relatively poor. On average, we obtain a precision of +0.02 pH units on replicate samples. The accuracy of our pH measurements are difficult to evaluate directly because we have no seawater standard for pH measurements. The accuracy is therefore dependent primarily on the accuracy of the seawater buffers that are used for electrode calibration. In order to improve the precision of our time-series pH measurement data, we are currently evaluating the spectrophotometric methods for pH measurements described by Byrne et al. (198_). Although these measurements are currently being made on a regular basis, the methodological details are not finalized and are not described here.

7. Equipment and Supplies

8. References