Methodology:

Collection of Seawater Samples

Seawaters were collected during the STERNA expedition of the RRS James Clark Ross between October and December 1992 - RRS James Clark Ross JR19921026 (JR02, JRXX7).

For surface transects, samples were taken from the ship's nontoxic seawater supply (intake at 8 m). Samples for the vertical profiles were collected from a conductivity-temperature-depth (CTD) rosette sampler, using 10-L Go-Flo bottles. Sample collection was in 1-L volumetric flasks, overfilled at least three times to aid exclusion of air bubbles. Following the addition of 200-microlitre saturated HgCl2 solution to arrest microbial action, flasks were sealed with ground glass stoppers, inverted to disperse the HgCl2 and temperature equilibrated at 25.0 +/- 0.05 degrees C prior to N2O analysis. In preliminary work, water samples treated in this way showed no significant changes in N2O concentration during several days storage. Using a similar sampling protocol, Elkins [1980] reports dissolved N2O stabilities of at least 1.5 months.

N2O Analysis

Analyses of N2O in air and seawater were by gas chromatography with electron-capture detection (Shimadzu GC-8AIE, 10mCi 63Ni). For seawaters, analysis was always within 48 hours of collection using a single-phase equilibration technique. The analytical system is described in detail by Upstill-Goddard et al. (1996 - Simultaneous high-precision measurements of methane and nitrous oxide in water and seawater by single phase equilibration gas chromatography, Deep Sea Research, Part I.), where full details of method calibration are also presented. In brief, chromatographic separation of N2O was at 65 degrees C on Porapak-Q (80-100 mesh, 2 m x 2 mm stainless steel column), with prior removal of H2O vapor and CO2 on precolumns of magnesium perchlorate and carbosorb, respectively. The detector temperature was 320 degrees C. Carrier gas was ultrahigh purity (UHP) nitrogen at a flow rate of 15 mL min-1.

The method returns gas partial pressures in the samples at the equilibrium temperature. Dissolved N2O concentrations in situ were calculated from the solubility tables of Weiss and Price [1980], for the equilibrium temperature (~25°C) and in situ salinity. Percent saturation of seawater with N2O was determined by ratioing the N2O partial pressures corrected to in situ temperature to the values determined in atmospheric air samples (mean 313 +/- 5 ppbv).

Data collection:

Analyses of N2O in air and seawater were by gas chromatography with electron-capture detection (Shimadzu GC-8AIE, 10mCi 63Ni).

Data quality:

Primary method calibration was through a commercially prepared, certified N2O/CH4 mixture in an UHP N2 matrix (Air Products) with a certified accuracy of +/-1% or better. For routine analytical calibrations a series of four secondary standards (in the linear range of the detector) were used, prepared from the primary mixtures by pressure dilution [Upstill-Goddard et al., 1990]. Consequently, the best estimate for overall accuracy of the N2O secondary standards was +/-2%. The variability of the method during this study was generally better than 2-3% (mean = 0.97%, n=44), based on replicate analyses of standards, atmospheric air, and compressed air. In support of this calibration, the measured atmospheric mixing ratio of 313 +/- 5 parts per billion by volume (ppbv) shows less than 1% deviation from the interhemispheric corrected global mean and at worst provides an overestimate of 1.3% relative to 1992 data for the South Pole [Swanson et al., 1992].