The concentrations of dissolved atmospheric chlorofluorocarbons, CFC-11 and CFC-12, were measured by shipboard electron-capture gas chromatography via methods similar to those described by Bullister and Weiss (1988). The measurements were done by the University of Miami group, P.I. Dr. Rana A. Fine, under the analytical direction of Kevin F. Sullivan. A total of 1628 water samples were analyzed from 79 of 124 stations and included eight pairs of duplicate water samples. The mean values of duplicate analyses are reported in the data file and are assigned a quality byte (flag) of 6.
Several times during the cruise, problems with the analytical system required extensive downtime. If samples were drawn but not analyzed, and the downtime exceeded 8 hours, some or all of these samples were discarded and fresh samples were drawn on the current station. This situation occurred after stations 16 and 24. In accordance with WHP protocol, the value for these samples has been reported as -999.900 and they have been assigned a quality byte (flag) of 5.
Occasionally after a routine analysis, the CFC values were clearly inappropriate based on the depth at which the Niskin was tripped. Other measured quantities showed unusual results on some of these occasions. Rather than discard these data, we are reporting their values and have assigned a data quality byte (flag) of 4.
The concentrations of the CFCs in air and water were calculated using external gaseous standards calibrated to the SIO86 scale. The gaseous and aqueous analyses were first corrected for any signal due to the analytical system using a weighted average of the four surrounding appropriate blank analyses. The average gaseous blank value was 7.32e-6 picomole (pM) for CFC-12 and 3.16e-5 pM for CFC-11. The average aqueous blank value was 1.44e-5 pM for CFC-12 and 9.94e-5 pM for CFC-11.
The temporal variation of the detector was compensated for by calculating a normalization factor for each analysis. Equations that closely resemble straight lines were fit to groupings of normalized standard analyses to yield calibration curves. These calibration curves were applied to the sample analyses to result in the concentrations of the CFCs.
After the water concentrations were calculated, a final correction was applied. This correction was estimated from the samples collected in waters that were very likely free of CFCs and was to compensate for any trace CFCs originating from the sampling bottles and/or handling. The bottle blanks decreased during the cruise; therefore, different bottle blanks for each Niskin were estimated for sequential ranges of stations. For the 36 Niskins during the entire cruise the applied bottle blanks averaged 0.0016 pM/kg for CFC-12 and 0.0033 pM/kg for CFC-11. If the bottle blank was greater than the measured concentration, a negative concentration is reported in the data file.
The precision of the water analyses can be estimated from the results of duplicate syringes drawn on the same Niskin. For eight pairs of duplicate syringes the average percent standard deviation for all the pairs was 1.66% for CFC-12 and 0.68% for CFC-11. For the samples greater than 0.1 pM/kg, the average percent standard deviation was 1.28% for CFC-12 (n=4) and 0.33% for CFC-11 (n=5).