Appendix Table 4. Summaries of Titration Alkalinity Data for Replicate Sea Water Samples Tables 4a to 4e summarize the laboratory ALK measurements by potentiometric titration of replicate samples collected on the five expedition legs of 1991, presented in the following order: Table 4a. TUNES Leg 1 WOCE Line P17C Table 4b. TUNES Leg 2 WOCE Lines P17S,P16S Table 4c. TUNES Leg 3 WOCE Line P16C Table 4d. Meteor 15 WOCE Line A9 Table 4e. Meteor 18 WOCE Line A1E Comments on each column in the table follow: LEG to SAMPLE The identifying information for each Niskin bottle sampled at STN DATE sea. ANALYSIS The date of potentiometric titration of an aliquot of sea water DATE taken from a replicate sample bottle. After the aliquot for DIC extraction was removed from a bottle, the bottle was immediately resealed and stored at room temperature for short periods of time or refrigerated for longer periods of time until titration. TITR The system used for titration is identified: G, the gravimetric SYST system; V, the volumetric system. Sometimes aliquots from the same bottle were titrated on both systems. SAMPLE The label number for the individual replicate sample bottle BOTTLE filled at sea. The prefix "S" indicates a standard 500 ml Corning borosilicate glass bottle with a hollow stopper (individually lapped to the bottle joint with carborundum grit) equipped with an aluminum bale type fixture for holding the stopper in its sealed position. The prefix "R" indicates a 500 ml borosilicate glass bottle equipped with a standard taper (24/40 size) joint and solid stopper with a Rodawiss plastic knurled nut system for holding the stopper. TRIAL This number indicates whether more than one titration was made on water from a replicate sample bottle. Sometimes titrations were deleted before reaching the data base at all, for example if the titration cell was filled improperly. Quite often, one or more of multiple trials are flagged in the data set: usually multiple trials were made because of a problem with the first trial. FLAG The flag code identifies trials that are rejected from further consideration due to identifiable experimental reasons. The flags used have the following meanings: F, evidence indicated that the sample had no mercuric chloride added (see text for further discussion); X, the titrator malfunctioned for some definable reason; EX, a general flag for some other experimental rejection, such as a very poor fit of the titration data. Appendix Table 4. Summaries of Titration Alkalinity Data for Replicate Sea Water Samples (cont.) S.I.O The ALK in micromoles per kilogram of sea water of hydrogen ion TRIAL equivalent is listed for an individual trial titration. TRIAL The trial difference for two trials as discussed above, if both DELTA trials are unflagged. BOTTLE The average bottle ALK for unflagged trials. In most cases ALK (single trials), the S.I.O. TRIAL and the BOTTLE ALK are identical. BOTTLE The difference in ALK between two bottles collected from a DELTA single Niskin bottle is listed. "NISKIN" Here is listed the average S.I.O. laboratory ALK result for AVG water from a particular Niskin bottle sampled at sea. If two titration systems were used (see TUNES Leg 3, station 222, bottle number R4569), the averages for both systems are listed. {INSTITUTION} The shipboard ALK result from a shipboard titration. ALK The institution abbreviations indicate the shipboard CO2 analytical group operating on that expedition leg, as follows: WHOI, the group of Dr. Catherine Goyet, P.I., of Woods Hole Oceanographic Institution, on TUNES Leg 1; CDRG, the group of Dr. Charles Keeling, P.I., of Scripps Institution of Oceanography, on TUNES Leg 3. The TUNES Leg 1 data is considered to be final, as reported to the Carbon Dioxide Information and Analysis Center. The TUNES Leg 3 data is NOT FINAL pending completion of a report in preparation. The remaining legs had no shipboard ALK measured by the U.S. CO2 group on board. No inference should be made from the listed data as to the imprecision of the data. Here we choose to list data to more places than are significant. {INSTITUTION} The difference between the shipboard ALK result and the - S.I.O. average laboratory replicate sample ALK. Note again that for the case discussed above, the differences for both systems are listed. NOTE: At the end of the Table 4 section for each leg is a note Dilution Factor detailing the multiplicative dilution factor applied to account for the addition of the HgCl2 poison solution to the S.I.O. replicate sample bottles. The nominally 500 ml bottles actually hold about 585 ml. Thus for the addition of 0.1 ml of saturated aqueous HgCl2, a dilution factor of 1.00017 is calculated. The shipboard data listed in the last two columns have also been appropriately corrected for dilution by each shipboard analysis group.