Sampling and Analytical Methodology:
Water samples for CFC and SF6 measurements were collected at all stations, including demi-stations, using the 30-liter Niskin bottle rosette. The CFC/SF6 sample was the first sample drawn from the Niskin bottle. The samples were collected in a 500 cc glass stoppered bottle through a PVC tube connected to the Niskin bottle drain valve. Bubbles were cleared from the tube with water flowing and the tube was inserted to the bottom of the 500 cc glass bottle. The glass bottle was placed in a wide mouth plastic jar that extended above the opening of the glass bottle by 4 cm. The overflow water collected in the jar, filling and overflowing the jar and covering the glass bottle opening. The flow continued for 3 overflow volumes of the glass bottle (1.5 liters) and the glass stopper was then inserted underwater preventing air from being trapped in the sample. The jar was then capped and the sample stored at room temperature until measurement, which was within 8 hours of collection. The measurements were carried out using a dual purge and trap system interfaced to a dual ECD (electron capture detector) HP6890 gas chromatograph. When a water sample is introduced into the system, it is split into two aliquots, a 20 cc aliquot for CFC measurement and a 350 cc aliquot for SF6 measurement. The aliquots are transferred to appropriately sized sparging chambers and stripped with ultra high purity nitrogen, which transports the extracted gases to cold traps (Unibeads-2s for CFCs, Carboxan-1000 for SF6) cooled to -80°C. The traps are then heated to 110°C for CFCs and 165°C for SF6 and flushed into the gas chromatograph where CFCs are separated with a Porasil-B pre-column and a Carbograph 1AC main column and SF6 is separated with a pre-column and main column of Molecular Sieve 5A. The gases are detected by the ECDs. The ECDs are calibrated by running gas standards with known concentrations of CFCs and SF6. Duplicate samples are collected and measured to provide a combined precision for sampling and measurement. The average differences from these duplicate measurements was the larger of 0.006 pmol/kg or 0.6% for CFC-11, 0.004 pmol/kg or 0.8% for CFC-12, 0.002 pmol/kg or 1.6% for CFC-113 and 0.01 fmol/kg or 1.5% for SF6.