Hydrographic Measurements
Water samples were collected in twenty-four 10-L Niskin bottles attached to a general oceanic rosette sampler mounted on a Neil Brown Mark III CTD (MKIIIB) provided by the IfMK (Brown and Morrison 1978). All stations were profiled to within 5 to 10 m of the bottom with the CTD. The Niskin bottles were closed 10 m from the bottom for the chemical samples, and the last two bottles were closed in the mixed layer near the surface at a depth of approximately 10 m. The closing depths for the remaining bottles varied. During two successive stations the bottles were closed at fixed depths, and then for the next two stations, the closing depths were set midway between those of the preceding two profiles (see Fig. 2). This was done to ensure that contour plots of the chemical parameters would not miss any significant water column features. The prescribed WOCE sample order was followed: CFCs, helium (He), oxygen (O2), CO2, nutrients, tritium (H3), and salinity. For stations greater than 3500 m in depth, two separate CTD/rosette casts were made to get adequate coverage (n = 36). For stations less than 3500 m, one CTD/rosette cast of up to 24 bottles was made. Surface currents down to 1000 m, surface temperature, and surface salinity were measured continually during the cruise with a hull-mounted ADCP. Near-surface (approximately 4 m) temperature and salinity were measured continually by the ship's data collection and distribution system (DVS) and integrated with the ship's navigation data (position). The ship's position was updated every two minutes. In between CTD stations, expendable bathythermographs (XBTs) were routinely launched. Over the boundary currents XBTs were launched every half hour, and the XBT launches were supplemented with free-falling expendable current profilers (XCP).
No serious problems were experienced with the CTD/rosette systems during this cruise. Repeated checks on board and several careful verifications using the complete bottle data sets were carried out. Pre-cruise calibrations of pressure and temperature for both CTDs were performed in November 1993 at IfMK's calibration laboratory and are described in Saunders et al. (1991). For these calibrations, a Rosemount Pt25 resistance from Sensoren Instrumente Systeme (SIS, Kiel) was used to convert the CTD temperature output to the international temperature scale of 1990 (ITS90), which is bound at the melting points of water and gallium (0.01°C and 28°C, respectively). Comparisons of the CTD temperature output with three electronic reversing thermometers (SIS, Kiel) were made during the cruise, and no drift in the CTD sensor was observed. Based on these data and the pre-cruise calibration, temperature is thought to be accurate to within ± 2 mK, and pressure good to 2 dbar or better.
Salinity calibrations were made using a bottle salinity measured one to two days after collection on a Guildline Autosal model 8400A that was calibrated with standard seawater (batch P120). The precision of the salinity determination was ± 0.0005, and the drift of the salinometer over the whole cruise did not exceed 0.0005. The accuracy of the bottle data was ± 0.0015 and the CTD data yielded an accuracy of ± 0.002. Bottle oxygen, nitrate, nitrite, and silicate were determined on every bottle closed on the A8 section. Not counting duplicates, some 3700 samples were analyzed. Oxygen was determined by Winkler titration after the technique of Culberson and Huang (1987) using a Metrohm Titrino with the end point determined amperometrically. The calculation of oxygen concentrations followed the procedure outlined in the WOCE Manual of Operations and Methods (Culberson and Williams 1991). Appropriate corrections for sample density, blanks, and volumetric expansion have been included. The precision of the analyses, defined as the mean of the absolute differences from duplicate samples drawn from the same Niskin bottles during the cruise, was ± 1.12 µmol/kg (n = 559). The concentrations of nitrate, nitrite, and silicate dissolved in seawater were determined on samples collected in virgin polystyrene 30 mL vials, on an autoanalyzer (Chemlab AAII) according to procedures given by Grasshoff (1976), Hydes (1984), and Hydes and Hill (1985). Phosphate could not be measured because the sensitivity of the analyzer was too low. Samples were stored at 4°C until analyzed (i.e., within 12 hours of collection). Precision for silicate was ± 0.2 to ± 0.3 µmol/kg (n = 594); and for both nitrite and nitrate, the precision was ± 0.1 to ± 0.1 µmol/kg (n = 594). Preweighed standards were used to prepare the nutrient working standards on board the ship. In the case of silicate, the shipboard standards were compared with a standard solution prepared in a laboratory before the cruise. No significant differences were noted. For accuracy, a nutrient standard solution (from Sagami Chemical Co.) was used as the certified reference material (CRM). New bottles of the standard were opened each week and analyzed. The results were: nitrate 9.76 ± 0.14 µmol/kg (n = 36) and silicate 49.70 ± 0.4 µmol/kg (n = 27). The known concentrations were 10.0 and 50.0 µmol/kg, respectively.
