CRUISE REPORT: I08S/I09S
(Updated APR 2008)


A.  HIGHLIGHTS

A.1.  CRUISE SUMMARY INFORMATION

         WOCE section designation  I08S/I09S
                         ExpoCode  316N145_5
                 Chief Scientists  Mike McCartney/WHOI
                                   Thomas Whitworth III/TA&MU
                            Dates  1994 DEC 01 -1995 JAN 19
                             Ship  R/V Knorr
                    Ports of call  Freantle, Australia

            Geographic boundaries  I08S
                                               30°17.83'S
                                   81°51.67'E             110°14'E
                                                64°9'S

                                   I09S
                                               34°49.17'S
                                   110°48.83'E            115°3.83'E
                                               64°51.83'S
                         Stations  147
                Floats & drifters  0
                         Moorings  0

                                Chief Scientists:
         Mike McCartney                       Thomas Whitworth III
         Woods Hole Oceanographic Institute   Department of Oceanography
         Woods Hole MA 02543                  Texas A&M University
         phone: 508-457-2000 ext. 2797        Mail Stop 3146
         Fax: 508-457-2181                    College Station TX 77843-3146
         e-mail: mike@gaff.whoi.edu


A.2.  CRUISE AND DATA INFORMATION (see pdf)


A.3.  List of Principal Investigators

      Name              Institution  Responsibility
      ----------------  -----------  -----------------------------
      Firing, Eric      UH           ADCP
      Wallace, Douglas  BNL          Air chemistry
      Falkner, Kellyy   OSU          Barium
      Toole, John       WHOI         CTD
      Key, Robert       Princeton    Carbon-13
      Quay, Paul        UW           Carbon-14
      Smethie, William  LDEO         Chlorofluorocarbons-all types
      Schlosser, Peter  LDEO         Helium
      Gordon, Arnold    OSU          Nutrients
      Toole, John       WHOI         Nutrients
      Toole, John       WHOI         Oxygen
      Key, Robert       Princeton    Radium-228
      Toole, John       WHOI         Salinity
      Wallace, Douglas  BNL          Total alkalinity AT
      Wallace, Douglas  BNL          Total carbon CT
      Key, Robert       Princeton    Tracer measurements
      Schlosser, Peter  LDEO         Tritium


A.4.  Scientific Programme and Methods

DESCRIPTION OF SCIENTIFIC PROGRAM
The object of this cruise was to occupy a series of CTD-O2 (Conductivity-
Temperature-Depth-Oxygen) stations along two, approximately north-south tracks.
The first track started at 30°S. 95°E and ended at the edge of the ice of
Antarctica at 82°E.  The second track began at the ice edge at 111°E and
proceeded north to the continental shelf of Australia at 115°E.

This collection of high-quality water-property data will help define the pattern
of circulation in the Indian Ocean. At each station measurements of temperature,
salinity, and dissolved-oxygen concentration were made continuously with depth,
and the concentrations of dissolved silica, phosphate, nitrate, and nitrite were
measured at up to 36 discrete levels. In addition, measurements of freon, 
tritium concentrations and CO2 were made at selected levels. The station spacing
ranged from 5 to 40 nautical miles, and all flowerings were made to within 10-20
m of the bottom.  Continuous echo-sounding was maintained along the cruise 
track,as well as ADCP current measurements.

OBSERVATIONS AND SAMPLES
The beginning, bottom and end positions of all the CTD stations occupied on this
cruise are listed in the attached table, with the stations numbered sequentially
through the cruise.  Positions are also shown on the attached chart.  We
anticipate completion of the calibration and editing of the various data by 1
August 1996.  As the hydrographic data for this section are WOCE data, the data
then move through an additional quality-evaluation stage managed by the WOCE
Hydrographic Programme Office (WHPO) in Woods Hole, which is generally expected
to be completed within two years of cruise end and which includes the formal
issuing (by WHPO) of a final ship-based data report about one year after the
cruise end; and a final ship-and shore-based data report about two years after
the cruise end.

As this is the most intensive phase of WOCE, the timing of these reports is
quite approximate due to the heavy workload of the technical groups making the
measurements and doing the quality control assessments.  With that in mind, we
intend to issue to Australia the preliminary version that results from the
calibration and editing phase in mid 1996, and subsequently issue revisions
should the latter WHPO process lead to alterations.  The data will be in digital
form on 9-track magnetic tape, or other suitable media; and the final report
will be printed copy and/or a text file.


A.5.  Major Problems and Goals not Achieved
A.6.  Other Incidents of Note


A.7.  List of Cruise Participants

Name                     Institution             Responsibility
-----------------------  ----------------------  ----------------------------
McCartney, Michael       WHOI                    Co-Chi. Sci., CTD-O2/Rosette
Whitworth, Thomas III    TAMU                    Co-Chi. Sci., CTD-O2/Rosette
Swartz, H.Marshall, Jr.  WHOI                    CTD team leader Watch leader
Rutz, Steven B.          TAMU                    CTD Watch Leader
Goepfert, Laura          WHOI                    CTD Data Analysis
Knapp, George            WHOI                    Water sample processor
Turner, Toshiko          WHOI                    Water sample processor
Hufford, Gwyneth         WHOI                    CTD Watchstander
Bennett, Paul  WHOI      CTD                     Watchstander
Bouchard, George         WHOI                    CTD Watchstander
McKay, Thomas Jason      WHOI                    CTD Watchstander
Primeau, Francois        WHOI                    CTD Watchstander
Jennings, Joseph J.      OSU                     Nutrient Analysis
Mordy, Calvin W.         PMEL                    Nutrient Analysis
Firing, Eric             U Hawaii                ADCP specialist
Hargreaves, Kirk         PMEL                    CFC Analysis
Mathieu, Guy             LDEO                    CFC Analysis
Mathieu, Sally           LDEO                    CFC Analysis
Johnson, Kenneth M.      BNL                     CO2 analysis
Haynes, Charlotte H.     BNL                     CO2 analysis
Haynes, Elizabeth M.     BNL                     CO2 analysis
Wysor, Brian S.          BNL                     CO2 analysis
Brockington, Melinda     U Washington            C14 analysis
Boenisch, Gerhard W.     LDEO                    Helium/Tritium analysis
Ludin, Andrea            LDEO                    Helium/Tritium analysis
Tynan, Cynthia T.        NOAA Marine Mammal Lab  Observations
Cotton, James M.         NOAA Marine Mammal Lab  Observations
Pitman, Robert L., Jr.   NOAA Marine Mammal Lab  Observations
Rowlett, Richard A.      NOAA Marine Mammal Lab  Observations


C.2. EQUIPMENT CONFIGURATION

Equipment used aboard the R/V Knorr for WOCE section I8SI9S was provided by both
Woods Hole Oceanographic Institution CTD Operations (WHOI CTD Ops) and the
Scripps Institute of Oceanography's Shipboard Technical Services/ Ocean Data
Facility (SIO STS/ODF). A total of 147 stations were taken during the cruise.

Two complete sampler frames were provided by ODF, each consisting of a coated
aluminum frame and thirty-six ODF-built 10-liter bottles. For this cruise two
CTDs were usually attached to the frame, one providing real-time data via FSK
telemetry, and another recording internally. Also mounted on the frame were a GO
pylon, independent ocean temperature modules (OTM), a lowered acoustic doppler
current profiler (LADCP) provided by the University of Hawaii, and an Ocean
Instruments System's 12 kHz pinger for bottom-finding. 141 of the 147 CTD
station data came from WHOI CTD 9, a WHOI-modified Neil Brown MK-3b CTD,
sampling at 23.8 Hz, and incorporating a Sensormedics oxygen sensor assembly, a
titanium strain gauge pressure tranducer and a platinum temperature sensor with
a lag of 150 ms.

A General Oceanics (GO) model 1016-36 position pylon was mounted to the 36-
bottle frame to control the firing of the bottles at depth. The 1016 pylon was
driven by a GO 1016-SCI Surface Control Interface (SCI) in the lab, which
provided power and commands down the sea cable, and received status data back.
The SCI was controlled through a dedicated personal computer. Due to SCI
performance problems, the 1016-36 pylon was replaced with two GO 1015-24 pylons
mounted one on top of the other. The 1015-24 pylons were controlled by two GO
1015PM deck units, which provided power and commands down the cable.

One of two Falmouth Scientific CTDs, ICTD1338 and ICTD1344, were placed on the
primary frame in internal-recording mode to acquire comparison data. In
addition, one of two Falmouth Scientific OTMs were placed on the frame to
provide an independent temperature measurement channel in the CTD data stream.

During rough weather a smaller specially-designed stainless steel frame was
used. The frame was built at WHOI and is based on a design from John Bullister's
group at NOAA/PMEL, uses 25 4-liter sample bottles, and is intended to provide
CTD capability in high seas.  Five stations were taken with this frame using a
1015-24 pylon and WHOI CTD 12, a GO-upgraded MK3c CTD sampling at 25.0 Hz, a
Sensormedics oxygen sensor assembly, a titanium pressure transducer, a platinum
temperature sensor with a lag of 200ms, and a fast thermistor.


EQUIPMENT PROBLEMS
Stations 1-3 were test stations. Station 1 used ICTD1338, with the 1016-36 pylon
and SCI.  Numerous problems were encountered including communication
interferences between the fsk ICTD data and the pylon-SCI communication.  It was
also found that the oxygen sensor was not working properly and it was deduced
after the cruise that the SeaCon underwater connectors were failing open-circuit
at various pressures.

Station 2 used CTD9, 1016 SCI and pylon, and again communication problems
developed causing synch errors in the CTD data and unreliable operation of the
pylon.  The oxygen assembly on CTD9 was not secured properly thus not recording
reliable oxygen data.  Station 3 used CTD12 and the 1016-36 pylon and SCI, and
again the cast had communication interference between the SCI and the CTD.
Efforts were made to adjust the telemetry levels to minimize the data
disruption.

For stations 4 and 5, CTD9 was used with the 1016 SCI and pylon, again
communication problems were noted.  During the down cast the pylon was turned 
off and only turned on during the upcast.  The acquisition program was placed in
stand-by when firing bottles because the CTD data had unacceptably high error
rates when the pylon was used.

After station 5, the 1016-36 position pylon was removed from the frame and
replaced with a GO 1015-24 position pylon.  For station 6 through station 29 
only 24 bottles were tripped, as only one 24-position pylon was able to be used.  
For station 30, a second 24-position pylon was stacked underneath the first,
providing the capability to trigger all 36 sample bottles.

On numerous occasions, data reported by the FSI OTM would indicate a data latch-
up, sometimes accompanied by a subsequent restart.  The problem was not solved 
on the cruise, but was later traced to insufficient clearances of the internal
components in the pressure case.

The three GO 1016-36 pylons which were initially tried all failed.  Two failures
were traced to damaged internal power supplies, and one had a broken position-
indicating switch.  All pylons were initially supplied in fully tested and
satisfactory condition, but it was later found that using them with the GO-
supplied SCIs could cause the power supply failures.  We have since stopped 
using the GO-supplied SCIs.  The mechanical failure to the position switch 
caused the pylon to lose it's place, and thus become useless.  As a result, the 
technician first rigged one 1015-24 pylon in place of the 1016-36, and by 
station 30, added another 1015-24, providing sufficient release mechanisms for 
all 36 frame sample bottles.  The Knorr's engine department provided outstanding 
assistance in making the necessary support mounts and modifications to help meet 
the science objectives.

The GO 1015-24 pylons were a source of occasional uncertainty, as it could not
always be determined where a bottle tripped. Sometimes, hydrographic data
indicated that two bottles closed at one stop, and although every effort was
made to maintain, align and clean the pylons, this problem was not entirely
eliminated. They performed better than anticipated, however, going for more than
40 consecutive stations without a mistrip, and allowed the cruise to gather 36
samples per cast.

Early on in the cruise, the tensiometer for the starboard winch failed.  This
forced us to use the port winch for the remainder of the cruise.  In addition,
station 81 was aborted due to winch problems, when a bearing for the tension
block failed.

On stations 50 through 53, the oxygen sensor with CTD9 was found to be operating
erratically.  It was subsequently replaced. CTD9 had been provided with a new
design of pressure compensation for the mineral-oil reservoir behind the sensor.
This was demonstrated to provide smoother pressure compensation and fewer jumps
in the data as the pressure differential equalized across the oxygen sensor
membrane.


AQUISITION AND PROCESSING METHODS
Data from CTD 9 was acquired at 23.8 Hz and with a temperature lag of 150 ms.
Data from CTD 12 was acquired at 25.0 Hz and with a temperature lag of 200 ms.
The temperature lag was checked by comparing density reversals in theta salinity
(TS) plots (Giles and McDonald, 1986).  It was found that the afore mentioned
lags showed the least amount of looping or density reversals.

Data was acquired by an EG&G Mk-III deck unit providing demodulated data to two
personal computers running EG&G version 5.2 rev 2 CTD acquisition software
(EG&G, Oceansoft acquisition manual, 1990), one providing graphical data to
screen and plotter, and the other a running listing output. Bottom approach was
controlled by following the pinger direct and bottom return signals on the ship-
provided PDR trace.

After each station, the CTD data was forwarded to another set of personal
computers running both EG&G CTD post-processing 3.0 software and custom-built
software from WHOI (Millard and Yang, 1993).  The data was first-differenced, 
lag corrected, pressure sorted, and pressure-centered into 2 decibar bins for 
final data quality control and analysis, including fitting to water sample 
salinity and oxygen results.


SUMMARY OF LABORATORY CALIBRATIONS FOR CTDS
The pressure, temperature, and conductivity sensors were calibrated by Maren
Tracy Plueddemann and Marshall Swartz at the Woods Hole Oceanographic
Institution's CTD Calibration Laboratory.


PRESSURE CALIBRATIONS

METHOD/CALIBRATION STANDARDS
The pressure transducers of CTD9, CTD12, ICTD1338, and ICTD1344 were calibrated
in a temperature controlled bath to WHOI's Ruska Model 2480 Dead Weight Tester
(DWT) as described by Millard and Yang (1993) over the range of atmospheric to
6,200 dbars.

The pre-cruise pressure calibration was performed at three different
temperatures, 1.78°C, 14.82°C, and 30.10°C. The calibrations were completed
November 7, 1994. Post-cruise pressure calibrations were performed at only one
temperature point, 1.20°C and were completed April 7, 1995.

       ______________________________________________________________

                                 BIAS         SLOPE       QUADRATIC
        -------------------  ------------  -----------  ------------
        CTD 9
        pre-cruise   1.78°C  -.495103E+01  .100588E+00   .112622E-10
                    14.82°C  -.439017E+01  .100576E+00   .100853E-09
                    30.10°C  -.371797E+01  .100592E+00  -.192585E-09
        post-cruise  1.20°C  -.421198E+01  .100585E+00   .847090E-10

        CTD 12
        pre-cruise   1.78°C  -.405781E+02  .107379E+00   .430549E-09
                    14.82°C  -.399422E+02  .107390E+00   .370115E-09
                    30.10°C  -.392364E+02  .107395E+00   .383934E-09
        post-cruise  1.20°C  -.395154E+02  .107384E+00   .385736E-09

        ICTD 1338
        pre-cruise   1.78°C   .707844E+00  .999402E-01   .131998E-09
                    14.82°C   .674421E+00  .999320E-01   .368154E-09
                    30.10°C   .177411E+00  .999467E-01   .248022E-09
        post-cruise  1.20°C   .152460E+01  .998550E-01   .734740E-09

        ICTD 1344
        pre-cruise   1.78°C   .293056E+01  .999521E-01  -.263500E-09
                    14.82°C   .168364E+01  .999844E-01  -.360033E-09
                    30.10°C   .171705E+01  .999784E-01  -.291289E-09
        post-cruise  1.20°C   .410510E+01  .999568E-01  -.466373E-09
       ______________________________________________________________


TEMPERATURE CALIBRATIONS

METHOD/CALIBRATION STANDARDS
For both the pre and post cruise temperature calibrations an Automated Systems
Laboratory (ASL) F18 temperature bridge with a Rosemount 162-CE SPRT were used
as transfer standards. During the calibration, the CTD was fully immersed in a
well-stirred constant temperature 700-liter salt water bath. The pre-cruise
temperature calibration was completed November 1, 1994 for all instruments
brought on the cruise. The post-cruise temperature calibration was completed
March 17, 1995 on CTD 9. Due to a failure of CTD 12, a post-cruise calibration
could not be performed. The CTD worked fine during the cruise, however during
the post cruise calibration the CTD was unable to synch on the data. Data is
reported to WOCE on the ITS-90 scale, but is processed internally on the IPTS-68
scale for compatibility with the equations for the Practical Salinity Scale of
1978 (PSS-78).

________________________________________________________________________________

CTD PRIMARY PLATINUM TEMPERATURE
                           BIAS           SLOPE        QUADRATIC
        -----------    ------------    -----------   ------------
        CTD 9
        pre-cruise     -.179120E+01    .496261E-03    .385531E-11
        post-cruise    -.179285E+01    .496217E-03    .467567E-11

        CTD 12
        pre-cruise      .621572E+01    .499695E-03    .688332E-12
        post-cruise         N/A            N/A            N/A

        ICTD 1338
        pre-cruise      .198004E-02    .499934E-03   -.483458E-12
        post-cruise     .213918E-02    .499918E-03   -.971791E-12

        ICTD 1344
        pre-cruise     -.452392E-02    .500201E-03   -.330744E-11
        post-cruise    -.643159E-02    .500258E-03   -.404936E-11


OXYGEN TEMPERATURE
                           BIAS          SLOPE         QUADRATIC
        -----------    ------------    -----------   ------------
        CTD 9
        pre-cruise      .717010E-02    .124856E+00   -.381392E-05
        post-cruise     .197632E+00    .123681E+00   -.494725E-05

        CTD 12
        pre-cruise     -.771413E+01    .761267E-03   -.186160E-08
        post-cruise         N/A           N/A            N/A

        ICTD 1338
        pre-cruise          N/A           N/A            N/A
        post-cruise    -.201461E+01    .161598E+00   -.127533E-03

        ICTD 1344
        pre-cruise     -.374508E+01    .153921E+00   -.836036E-04
        post-cruise    -.401615E+01    .159201E+00   -.125456E-03


PRESSURE TEMPERATURE
                                                          S1          S2     T0
        -----------    ------------    -----------    --------------------------
        CTD 9                                         
        pre-cruise      .376241E+02   -.938036E-02    -1.7188E-2   .035381  1.78
        post-cruise     .374444E+02   -.920480E-02

        CTD 12
        pre-cruise      .145943E+03   -.374919E-02     4.1010E-7   .047316  1.78
        post-cruise         N/A            N/A

        (Note: ICTDs do not have a separately reporting temperature channel).
________________________________________________________________________________


CONDUCTIVITY CALIBRATIONS

METHOD/CALIBRATION STANDARDS
A pre-cruise conductivity calibration was performed on CTD 9 and CTD 12. Five
salinity samples were drawn and analyzed on a Guildline Autosal 8100-B
autosalinometer at each temperature point during the temperature calibration.
These values were then converted to conductivity and compared to the values read
by the CTD at the different temperatures (Millard and Yang, 1993).

                 ____________________________________________
 
                  CTD 9
                  pre-cruise     -.113915E-01    .998004E-03
                  post-cruise    -.724614E-02    .998114E-03

                  CTD 12
                  pre-cruise      .278165E-01    .100049E-02
                  post-cruise         N/A            N/A
                 ____________________________________________

For final processing of the data the pre-cruise calibration constants were used
to scale the data for CTD12, ICTD1338, and CTD9.


CTD DATA

SUMMARY OF AT SEA CALIBRATIONS
The pressure of the CTDs at the sea surface was recorded at the beginning of
each station. The on deck pressure was found using by graphing the calculated
pressure prior to the package entering the water. This number was then
subtracted from the pressure bias term for each station.


CONDUCTIVITY CALIBRATION

BASIC FITTING PROCEDURE
The CTD conductivity sensor data was fit to the water sample conductivity as
described in Millard and Yang 1993. The stations were fit as a drift of the
sensor was noted.  


OXYGEN CALIBRATIONS

BASIC FITTING PROCEDURE
The CTD oxygen sensor variables were fit to water sample oxygen data to
determine the six parameters of the oxygen algorithm (Millard and Yang, 1993).
As with conductivity, the stations were fit as a drift in the sensor was noted.


QUALITY CONTROL OF 2DB CTD DATA AND SEA FILES

Stations 3, 8, 31, and 62 had several pressure bins where there was no CTD data.
These bins have been marked as 6's in the *.CTD files. During these stations
there were a lot of synch errors in the raw data that had to be cleaned up and
this resulted in very few good scans in several pressure bins.

For stations 1 and 2, where the oxygen sensors were not working, the CTD values
in the *.CTD and *.SEA files were changed to -9.000 and the quality word to 5.
For CTD9, stations 50-53 the oxygen sensor showed erroneous values. The CTD
oxygen values were again changed to -9.000 and the quality word change to 5 to
reflect the bad sensor. For stations 46 and 47 it was noted that the sensor may
have begun failing, thus the quality word for these oxygen CTD values was
changed to 3 to reflect a questionable oxygen value in both the *.CTD and *.SEA
file.

In the *.SEA files the down trace CTD oxygen value is used, in some cases there
was no pressure bin in the down trace so the oxygen value was taken from the
nearest pressure bin. These values are marked as questionable in the *.SEA
files.


REFERENCES:

Giles, Alan B. and Trevor J. McDonald. 1986. Two methods for the reduction of
    Salinity Spiking of CTDs. Deep Sea Research, Vol 33, no 9. 1253-1274.

Mangum, B.W. and G.T. Furukawa. 1990. Guidelines for Realizing the International
    Temperature Scale of 1990 (ITS-90). NIST Technical Notes 1265.

Millard, R.C. and K. Yang. 1993. CTD Calibration and Processing Methods used at
    Woods Hole Oceanographic Institution. Technical Report No. 93-44, 96 pages.

Oceansoft MKIII/SCTD Acquisition Software Manual. 1990. P/N Manual 10239. EG&G
    Marine Instruments.

Owens, Brechner W. and Robert C. Millard, Jr. 1985. A New Algorithm for CTD
    Oxygen Calibrations. J. Phys. Oc. vol 15.621-631.



CFC-11 and CFC-12 Measurements on WOCE I08S/I09S
(John Bullister)

Specially designed 10 liter water sample bottles were used on the cruise to
reduce CFC contamination. These bottles have the same outer dimensions as
standard 10 liter Niskin bottles, but use a modified end-cap design to minimize
the contact of the water sample with the end-cap O-rings after closing. The O-
rings used in these water sample bottles were vacuum-baked prior to the first
station on the Indian Ocean Expedition. Stainless steel springs covered with a
nylon powder coat were substituted for the internal elastic tubing standardly
used to close Niskin bottles.

CFC samples were drawn from approximately 50% of 4600 water samples collected
during the expedition. Water samples for CFC analysis were usually the first
samples drawn from the 10 liter bottles. Care was taken to co-ordinate the
sampling of CFCs with other samples to minimize the time between the initial
opening of each bottle and the completion of sample drawing. In most cases,
dissolved oxygen, total CO2, alkalinity and pH samples were collected within
several minutes of the initial opening of each bottle. To minimize contact with
air, the CFC samples were drawn directly through the stopcocks of the 10 liter
bottles into 100 ml precision glass syringes equipped with 2-way metal
stopcocks. The syringes were immersed in a holding tank of clean surface
seawater until analysed.

To reduce the possibility of contamination from high levels of CFCs frequently
present in the air inside research vessels, the CFC extraction/analysis system
and syringe holding tank were housed in a modified 20' laboratory van on the aft
deck of the ship.

For air sampling, a ~100 meter length of 3/8" OD Dekaron tubing was run from the
CFC lab van to the bow of the ship. Air was pulled through this line into the
CFC van using an Air Cadet pump. The air was compressed in the pump, with the
downstream pressure held at about 1.5 atm using a back-pressure regulator. A tee
allowed a flow (~100 cc/min) of the compressed air to be directed to the gas
sample valves, while the bulk flow of the air (>7 liters per minute) was vented
through the back pressure regulator.

Concentrations of CFC-11 and CFC-12 in air samples, seawater and gas standards
on the cruise were measured by shipboard electron capture gas chromatography,
using techniques similar to those described by Bullister and Weiss (1988). The
CFC system used was built at the Scripps Institution of Oceanography and had
been used on several Pacific WOCE legs as well as several Indian Ocean WOCE
legs. The SIO system was modified from the Bullister and Weiss (1988) design to
use a fixed volume, variable pressure gas loop injection system. The sample
loops were either pressurized or evacuated to known pressures in order to vary
the amount of gas sample introduced. The sample loop(s) were periodically filled
with CFC-free gas to one atmosphere and analyzed to check for analytical blanks.
The typical analysis time for a seawater, air, standard or blank sample was
about 12 minutes.

The CFC analytical system functioned well during this expedition.

Concentrations of CFC-11 and CFC-12 in air, seawater samples and gas standards
are reported relative to the SIO93 calibration scale (Cunnold, et. al., 1994).
CFC concentrations in air and standard gas are reported in units of mole
fraction CFC in dry gas, and are typically in the parts-per-trillion (ppt)
range. Dissolved CFC concentrations are given in units of picomoles of CFC per
kg seawater (pmol/kg). CFC concentrations in air and seawater samples were
determined by fitting their chromatographic peak areas to multi-point
calibration curves, generated by pressurizing sample loops and injecting known
volumes of gas from a CFC working standard (PMEL cylinder 38415) into the
analytical instrument. The concentrations of CFC-11 and CFC-12 in this working
standard were calibrated versus a primary CFC standard (36743) (Bullister, 1984)
before the cruise and a secondary standard (32386) before and after the cruise.

Full range calibration curves were run several times (approx. every 5 days
during the cruise). Single injections of a fixed volume of standard gas at one
atmosphere were run much more frequently (at intervals of 1 to 2 hours) to
monitor short term changes in detector sensitivity.

As expected, low (~0.015 pmol/kg) but non-zero CFC concentrations were measured
in deep and bottom samples along the northern ends (~32S) of I8S and I9S. Deep
and bottom CFC concentrations increased significantly southward along the
sections. It is likely that most of the deep CFC signals observed on I8S and
I9S, which are strongly correlated with elevated dissolved oxygen and cold
temperatures, are due to deep ventilation processes in this high latitude
region, and not simply blanks due of the sampling and analytical procedures. The
measured levels of CFC in deep water samples on the northern ends I8S and I9S
sections are considerable higher than those found on WOCE sections in the low
latitude Indian Ocean. For example, typical measured deep water CFC measurements
along WOCE section I2 (at about 8S) were ~0.003 pmol/kg for CFC-11 and <0.001
for CFC-12. Since no "zero" CFC water was present anywhere along I8S or I9S, and
later cruises (e.g. I2) showed low CFC blanks for the sampling procedures, no
corrections for 'sampling blanks' have been applied to the reported CFC signals
for I8S and I9S. A few samples (~86 of a total of ~2300) had clearly anomalous
CFC-11 and/or CFC-12 concentrations relative to adjacent samples. These appeared
to occur more or less randomly, and were not clearly associated with other
features in the water column (e.g. elevated oxygen concentrations, salinity or
temperature features, etc.). This suggests that the high values were due to
isolated low-level CFC contamination events. These samples are included in this
report and are flagged as either 3 (questionable) or 4 (bad) measurements. A
total of 32 analyses of CFC-11 were assigned a flag of 3 and 25 analyses of CFC-
12 were assigned a flag of 3. A total of 17 analyses of CFC-11 were assigned a
flag of 4 and 24 CFC-12 samples assigned a flag of 4.

On this expedition, we estimate precisions (1 standard deviation) of about 1% or
0.005 pmol/kg (whichever is greater) for dissolved CFC-11 and 1% or 0.005
pmol/kg (whichever is greater) for dissolved CFC-12 measurements (see listing of
replicate samples given at the end of this report).

In addition to the file of mean CFC concentrations, tables of the following are
included in this report:

  Table 1a. I8SI9S Replicate dissolved CFC-11 analyses 
  Table 1b. I8SI9S Replicate dissolved CFC-12 analyses
  Table 2.  I8SI9S CFC air measurements 
  Table 3.  I8SI9S CFC air measurements interpolated to station locations

A value of -9.0 is used for missing values in the listings.


REFERENCES:

Bullister, J.L. Anthropogenic Chlorofluoromethanes as Tracers of Ocean
    Circulation and Mixing Processes: Measurement and Calibration Techniques
    and Studies in the Greenland and Norwegian Seas, Ph.D. dissertation, Univ.
    Calif. San Diego, 172 pp.

Bullister, J.L. and R.F. Weiss, Determination of CCl3F and CCl2F2 in seawater
    and air. Deep-Sea Research, 35 (5), 839-853, 1988.

Cunnold, D.M., P.J. Fraser, R.F. Weiss, R.G. Prinn, P.G. Simmonds, B.R. Miller,
    F.N. Alyea, and A.J.Crawford. Global trends and annual releases of CCl3F    
    and CCl2F2 estimated from ALE/GAGE and other measurements from July 1978 to     
    June 1991. J. Geophys. Res., 99, 1107-1126, 1994.


TABLE 1a: Replicate F-11 Samples

Stn  Sample  F-11           Stn  Sample  F-11           Stn  Sample  F-11
---  ------  -----          ---  ------  -----          ---  ------  -----
  1     6    0.024           13    21    3.084           35    28    3.668
  1     6    0.029           13    21    3.164           35    28    3.640
  1     6    0.031           15    14    1.568           37    34    4.459
  1    12    0.020           15    14    1.570           37    34    4.348
  1    12    0.032           16    10    0.024           40    24    2.485
  1    14    0.036           16    10    0.020           40    24    2.506
  1    14    0.032           16    15    2.263           40    27    3.556
  1    21    0.161           16    15    2.277           40    27    3.545
  1    21    0.025           16    21    3.435           41     2    0.112
  2     1    0.013           16    21    3.395           41     2    0.110
  2     1    0.029           16    21    3.473           44     1    0.313
  2     8    0.012           17     6    0.030           44     1    0.309
  2     8    0.015           17     6    0.005           44    32    4.450
  2    15    0.028           17    23    3.151           44    32    4.444
  2    15    0.028           17    23    3.164           46    27    2.553
  3     1    0.022           18    13    1.165           46    27    2.562
  3     1    0.014           18    13    1.156           50     2    0.582
  3     7    0.011           18    14    1.551           50     2    0.576
  3     7    0.026           18    14    1.527           55     3    0.652
  3    13    0.017           19    15    2.309           55     3    0.629
  3    13    0.015           19    15    2.340           55    32    5.446
  3    25    0.016           19    21    3.387           55    32    5.451
  3    25    0.018           19    21    3.389           56    19    0.169
  3    31    0.013           21    15    2.770           56    19    0.175
  3    31    0.042           21    15    2.804           56    29    2.721
  4    18    0.201           21    17    3.605           56    29    2.745
  4    18    0.195           21    17    3.573           62     7    0.529
  4    19    0.436           22    13    0.871           62     7    0.550
  4    19    0.422           22    13    0.883           62    33    5.957
  4    25    3.107           22    17    3.649           62    33    5.758
  4    25    3.303           22    17    3.586           75     9    0.062
  4    25    3.326           22    21    3.506           75     9    0.064
  4    31    2.953           22    21    3.496           79     3    0.060
  4    31    2.955           24     7    0.041           79     3    0.057
  7     6    0.009           24     7    0.042           79    34    6.568
  7     6    0.012           24    23    3.409           79    34    6.595
  7    23    2.601           24    23    3.408           82    31    1.461
  7    23    2.541           24    23    3.416           82    31    1.469
  9    14    0.570           29    15    1.962           85     2    1.458
  9    14    0.569           29    15    1.949           85     2    1.416
  9    17    3.310           29    18    3.453           85    21    0.326
  9    17    3.277           29    18    3.232           85    21    0.330
 11    19    3.167           30    21    1.336           85    35    6.081
 11    19    3.206           30    21    1.350           85    35    6.158
 12    15    2.150           34    22    1.868           87     9    0.486
 12    15    2.126           34    22    1.832           87     9    0.489
 13     9    0.023           34    30    4.299           87    29    2.740
 13     9    0.046           34    30    4.317           87    29    2.739


Stn  Sample  F-11           Stn  Sample  F-11           Stn  Sample  F-11
---  ------  -----          ---  ------  -----          ---  ------  -----
 92     5    0.745          122    33    3.905          141    30    3.232
 92     5    0.756          122    33    3.900          141    32    3.073
 92    33    6.201          124    18    1.522          141    32    3.145
 92    33    6.178          124    18    1.524          144    10    0.013
 94     2    1.260          124    21    2.120          144    10    0.013
 94     2    1.246          124    21    2.104          144    33    2.426
 94    33    6.619          124    33    3.725          144    33    2.418
 94    33    6.644          124    33    3.721
 97    34    6.385          126     6    0.049
 97    34    6.434          126     6    0.050
 99     8    0.260          126    34    3.692
 99     8    0.255          126    34    3.673
 99    16    0.145          127    32    3.635
 99    16    0.141          127    32    3.661
100    12    0.565          129     2    0.070
100    12    0.564          129     2    0.073
100    16    4.962          129     2    0.035
100    16    4.953          129    33    3.596
101     6    0.232          129    33    3.575
101     6    0.231          130     1    0.068
101    22    5.964          130     1    0.072
101    22    5.952          130     2    0.072
103    14    0.068          130     2    0.067
103    14    0.070          130     2    0.069
105     5    0.348          130     2    0.067
105     5    0.350          131     5    0.067
105    28    3.033          131     5    0.064
105    28    2.959          131    28    3.621
105    34    5.529          131    28    3.605
105    34    5.514          133     6    0.033
107    35    5.521          133     6    0.037
107    35    5.550          133    34    3.422
111    16    0.086          133    34    3.405
111    16    0.085          135    26    3.394
111    29    2.387          135    26    3.378
111    29    2.427          135    33    3.140
114     6    0.047          135    33    3.136
114     6    0.047          137    21    1.577
116     6    0.040          137    21    1.578
116     6    0.040          137    32    3.307
116    32    4.646          137    32    3.355
116    32    4.636          137    33    3.276
120    14    0.242          137    33    3.252
120    14    0.238          137    34    3.171
120    35    4.067          137    34    3.167
120    35    4.039          139    34    3.225
122    31    3.721          139    34    3.221
122    31    3.733          141    30    3.226



TABLE 1b: Replicate F-12 Samples

Stn  Sample  F-11           Stn  Sample  F-11           Stn  Sample  F-11
---  ------  -----          ---  ------  -----          ---  ------  -----
  1     6    0.057           13    21    1.641           35    28    1.901
  1     6    0.060           13    21    1.677           35    28    1.898
  1     6    0.066           15    14    0.798           37    34    2.271
  1    12    0.059           15    14    0.810           37    34    2.151
  1    12    0.058           16    10    0.005           40    24    1.182
  1    14    0.020           16    10    0.003           40    24    1.197
  1    14    0.045           16    15    1.153           40    27    1.734
  1    21    0.019           16    15    1.142           40    27    1.744
  1    21    0.069           16    21    1.816           41     2    0.053
  2     1   -0.004           16    21    1.783           41     2    0.058
  2     1    0.002           16    21    1.842           44     1    0.153
  2     8   -0.003           17     6    0.015           44     1    0.154
  2     8   -0.008           17     6   -0.002           44    32    2.254
  2    15   -0.001           17    23    1.689           44    32    2.244
  2    15   -0.006           17    23    1.689           46    27    1.206
  3     1   -0.002           18    13    0.577           46    27    1.224
  3     1    0.003           18    13    0.606           50     2    0.276
  3     7   -0.008           18    14    0.811           50     2    0.283
  3     7   -0.004           18    14    0.763           55     3    0.307
  3    13    0.003           19    15    1.158           55     3    0.297
  3    13   -0.001           19    15    1.156           55    32    2.671
  3    25    0.009           19    21    1.816           55    32    2.674
  3    25    0.010           19    21    1.730           56    19    0.076
  3    31    0.010           21    15    1.416           56    19    0.083
  3    31    0.006           21    15    1.418           56    29    1.287
  4    18    0.104           21    17    1.855           56    29    1.299
  4    18    0.105           21    17    1.834           62     7    0.259
  4    19    0.220           22    13    0.439           62     7    0.268
  4    19    0.214           22    13    0.444           62    33    2.909
  4    25    1.529           22    17    1.906           62    33    2.866
  4    25    1.623           22    17    1.842           75     9    0.034
  4    25    1.595           22    21    1.823           75     9    0.038
  4    31    1.462           22    21    1.827           79    34    3.188
  4    31    1.457           24     7    0.019           79    34    3.099
  7     6    0.006           24     7    0.029           82    31    0.683
  7     6    0.008           24    23    1.772           82    31    0.681
  7    23    1.392           24    23    1.761           85     2    0.682
  7    23    1.363           24    23    1.781           85     2    0.665
  9    14    0.283           29    15    0.965           85    21    0.149
  9    14    0.281           29    15    0.954           85    21    0.151
  9    17    1.562           29    18    1.770           85    35    2.872
  9    17    1.547           29    18    1.732           85    35    2.884
 11    19    1.620           30    21    0.646           87     9    0.229
 11    19    1.672           30    21    0.658           87     9    0.228
 12    15    1.071           34    22    0.889           87    29    1.282
 12    15    1.049           34    22    0.895           87    29    1.278
 13     9    0.007           34    30    2.133           92     5    0.351
 13     9    0.021           34    30    2.200           92     5    0.350


Stn  Sample  F-11           Stn  Sample  F-11           Stn  Sample  F-11
---  ------  -----          ---  ------  ----           ---  ------  ------
 92    33    2.981          124    18    0.729          144    10    0.017
 92    33    2.949          124    18    0.728          144    33    1.333
 94     2    0.592          124    21    1.038          144    33    1.304
 94     2    0.567          124    21    1.034
 94    33    3.142          124    33    1.933
 94    33    3.167          124    33    1.898
 97    34    2.999          126     6    0.027
 97    34    3.021          126     6    0.028
 99     8    0.127          126    34    1.873
 99     8    0.123          126    34    1.911
 99    16    0.066          127    32    1.912
 99    16    0.064          127    32    1.903
100    12    0.265          129     2    0.049
100    12    0.262          129     2    0.038
100    16    2.329          129     2    0.017
100    16    2.353          129    33    1.892
101     6    0.116          129    33    1.852
101     6    0.108          130     1    0.043
101    22    2.817          130     1    0.043
101    22    2.871          130     2    0.045
103    14    0.031          130     2    0.042
103    14    0.032          130     2    0.038
105     5    0.164          130     2    0.039
105     5    0.169          131     5    0.043
105    28    1.442          131     5    0.043
105    28    1.421          131    28    1.889
105    34    2.705          131    28    1.854
105    34    2.701          133     6    0.026
107    35    2.716          133     6    0.032
107    35    2.748          133    34    1.784
111    16    0.034          133    34    1.795
111    16    0.035          135    26    1.715
111    29    1.117          135    26    1.705
111    29    1.170          135    33    1.664
114     6    0.027          135    33    1.671
114     6    0.029          137    21    0.757
116     6    0.019          137    21    0.783
116     6    0.020          137    32    1.728
116    32    2.343          137    32    1.794
116    32    2.356          137    33    1.700
120    14    0.119          137    33    1.701
120    14    0.116          139    34    1.673
120    35    2.079          139    34    1.692
120    35    2.105          141    30    1.681
122    31    1.896          141    30    1.656
122    31    1.880          141    32    1.625
122    33    1.993          141    32    1.649
122    33    1.987          144    10    0.015



TABLE 2: CFC Air Measurements:

LEG 1
             Time                             F11     F12
   Date     (hhmm)  Latitude    Longitude     PPT     PPT
---------   ------  ---------   ----------   -----   ------
 5 Dec 94    0258   30 40.7 S   099 46.5 E    -9.0   513.7
 5 Dec 94    0307   30 40.7 S   099 46.5 E    -9.0   513.0
 5 Dec 94    0316   30 40.7 S   099 46.5 E    -9.0   514.3
 5 Dec 94    0325   30 40.7 S   099 46.5 E    -9.0   514.1
 5 Dec 94    0335   30 40.7 S   099 46.5 E    -9.0   514.4
 7 Dec 94    2020   33 06.2 S   094 57.8 E    -9.0   515.4
 7 Dec 94    2029   33 06.2 S   094 57.8 E    -9.0   515.5
 7 Dec 94    2038   33 06.2 S   094 57.8 E    -9.0   512.3
 9 Dec 94    2247   36 50.7 S   095 00.5 E   260.1   516.1
 9 Dec 94    2256   36 50.7 S   095 00.5 E   259.3   513.5
 9 Dec 94    2305   36 50.7 S   095 00.5 E   259.7   513.6
10 Dec 94    1908   38 10.7 S   095 00.7 E   259.5   513.2
10 Dec 94    1917   38 10.7 S   095 00.7 E   259.7   511.7
10 Dec 94    1926   38 10.7 S   095 00.7 E   259.9   510.1
13 Dec 94    1323   43 23.4 S   095 01.0 E   260.5   512.0
13 Dec 94    1332   43 23.4 S   095 01.0 E   260.1   513.7
13 Dec 94    1341   43 23.4 S   095 01.0 E   260.3   509.4
18 Dec 94    1143   50 34.0 S   090 02.0 E   262.4   515.7
18 Dec 94    1152   50 34.0 S   090 02.0 E   260.9   510.8
18 Dec 94    1201   50 34.0 S   090 02.0 E   260.8   513.2
22 Dec 94    1528   55 26.8 S   085 22.8 E   260.5   510.7
22 Dec 94    1537   55 26.8 S   085 22.8 E   261.1   514.6
22 Dec 94    1546   55 26.8 S   085 22.8 E   261.5   512.8
26 Dec 94    1839   61 58.5 S   082 01.0 E   261.0   514.6
26 Dec 94    1847   61 58.5 S   082 01.0 E   259.9   515.0
26 Dec 94    1856   61 58.5 S   082 01.0 E   260.0   514.9

LEG 2
             Time                             F11     F12
   Date     (hhmm)  Latitude    Longitude     PPT     PPT
---------   ------  ---------   ----------   -----   ------
 2 Jan 95    0445   64 51.1 S   110 49.2 E   260.1   513.4
 2 Jan 95    0454   64 51.1 S   110 49.2 E   260.4   512.2
 2 Jan 95    0503   64 51.1 S   110 49.2 E   260.4   513.2
 2 Jan 95    0514   64 51.1 S   110 49.2 E   261.0   513.8
 5 Jan 95    1925   58 07.5 S   115 00.1 E   260.3   512.3
 5 Jan 95    1934   58 07.5 S   115 00.1 E   261.1   512.8
 5 Jan 95    1952   58 07.5 S   115 00.1 E   260.6   514.2
 5 Jan 95    2001   58 07.5 S   115 00.1 E   261.4   512.7
 7 Jan 95    1529   55 00.0 S   115 00.0 E   260.5   514.7
 7 Jan 95    1538   55 00.0 S   115 00.0 E   259.5   513.2
 7 Jan 95    1548   55 00.0 S   115 00.0 E   260.5   512.2
 8 Jan 95    1929   52 36.4 S   114 59.1 E   260.6   513.3
 8 Jan 95    1938   52 36.4 S   114 59.1 E   260.3   514.5
 8 Jan 95    1946   52 36.4 S   114 59.1 E   259.7   514.7
10 Jan 95    1645   49 00.1 S   115 00.2 E   260.7   514.6
10 Jan 95    1653   49 00.1 S   115 00.2 E   259.5   511.9
10 Jan 95    1702   49 00.1 S   115 00.2 E   260.9   516.3
14 Jan 95    1351   41 30.4 S   114 59.8 E   260.4   513.4
14 Jan 95    1400   41 30.4 S   114 59.8 E   259.7   512.0
14 Jan 95    1408   41 30.4 S   114 59.8 E   258.8   511.7


TABLE 3: i8si9s CFC Air values (interpolated to station locations)

STN                                         F11     F12
NBR   Latitude    Longitude       Date      PPT     PPT
---   ---------   ----------   ---------   -----   ------
  1   31 29.3 S   110 13.5 E    2 Dec 94   259.6   513.6
  2   31 13.3 S   106 17.0 E    3 Dec 94   259.6   513.6
  3   30 57.2 S   102 44.7 E    4 Dec 94   259.7   513.6
  4   30 18.0 S   095 00.0 E    5 Dec 94   259.7   513.6
  5   31 18.0 S   095 00.0 E    6 Dec 94   259.7   513.6
  6   32 00.5 S   095 00.3 E    6 Dec 94   259.7   513.6
  7   32 00.2 S   095 00.3 E    7 Dec 94   259.7   513.6
  8   32 30.0 S   095 00.0 E    7 Dec 94   259.7   513.6
  9   33 00.0 S   094 59.7 E    7 Dec 94   259.7   513.6
 10   33 30.0 S   095 00.0 E    7 Dec 94   259.7   513.6
 11   34 00.0 S   095 00.0 E    8 Dec 94   259.7   513.5
 12   34 30.0 S   095 00.0 E    8 Dec 94   259.7   513.5
 13   34 59.7 S   095 00.0 E    8 Dec 94   259.7   513.5
 14   35 29.8 S   095 00.0 E    9 Dec 94   259.7   513.5
 15   35 59.7 S   095 00.2 E    9 Dec 94   259.7   513.5
 16   36 30.0 S   095 00.0 E    9 Dec 94   259.7   513.0
 17   36 59.8 S   095 00.2 E    9 Dec 94   259.7   513.0
 18   37 30.0 S   095 00.0 E   10 Dec 94   259.7   513.0
 19   37 59.8 S   095 00.0 E   10 Dec 94   259.7   513.0
 20   38 29.3 S   095 01.2 E   11 Dec 94   259.7   513.0
 21   38 59.5 S   095 00.2 E   11 Dec 94   259.7   513.0
 22   39 29.8 S   095 00.2 E   11 Dec 94   259.7   513.0
 23   40 00.0 S   094 59.8 E   11 Dec 94   259.9   512.6
 24   40 30.0 S   095 00.0 E   12 Dec 94   260.0   511.7
 25   41 00.3 S   095 00.5 E   12 Dec 94   260.0   511.7
 26   41 30.2 S   094 59.8 E   12 Dec 94   260.0   511.7
 27   41 59.8 S   095 00.0 E   12 Dec 94   260.0   511.7
 28   42 30.2 S   095 00.3 E   13 Dec 94   260.0   511.7
 29   43 00.0 S   095 00.2 E   13 Dec 94   260.0   511.7
 30   43 30.0 S   094 59.8 E   13 Dec 94   260.0   511.7
 31   43 45.0 S   095 00.0 E   13 Dec 94   260.0   511.7
 32   44 00.0 S   095 00.0 E   13 Dec 94   260.0   511.7
 33   44 15.0 S   095 00.0 E   14 Dec 94   260.0   511.7
 34   44 29.8 S   095 01.0 E   14 Dec 94   260.5   512.2
 35   44 59.5 S   095 00.2 E   14 Dec 94   260.5   512.2
 36   45 25.7 S   094 38.3 E   14 Dec 94   260.8   512.5
 37   45 50.2 S   094 16.8 E   15 Dec 94   260.8   512.5
 38   46 16.7 S   093 53.0 E   15 Dec 94   260.8   512.5
 39   46 42.8 S   093 31.5 E   15 Dec 94   260.8   512.5
 40   47 08.8 S   093 09.5 E   16 Dec 94   260.8   512.5
 41   47 33.7 S   092 45.2 E   16 Dec 94   260.8   512.5
 42   47 59.7 S   092 22.2 E   16 Dec 94   260.8   512.5
 43   48 25.3 S   091 59.7 E   17 Dec 94   260.8   512.5
 44   48 51.0 S   091 36.2 E   17 Dec 94   260.8   512.5
 45   49 16.7 S   091 13.0 E   17 Dec 94   260.8   512.5
 46   49 42.0 S   090 49.0 E   17 Dec 94   260.9   512.5
 47   50 07.8 S   090 25.2 E   18 Dec 94   261.2   513.0
 48   50 33.5 S   090 02.3 E   18 Dec 94   261.2   513.0
 49   50 59.2 S   089 36.5 E   19 Dec 94   261.2   513.0
 50   51 25.2 S   089 12.2 E   19 Dec 94   261.2   513.0
 51   51 37.7 S   088 59.5 E   19 Dec 94   261.2   513.0
 52   51 50.2 S   088 45.8 E   19 Dec 94   261.2   513.0
 53   52 15.5 S   088 19.8 E   20 Dec 94   261.2   513.0
 54   52 41.2 S   087 53.7 E   20 Dec 94   261.2   513.0
 55   53 06.3 S   087 27.8 E   20 Dec 94   261.2   513.0
 56   53 31.5 S   087 01.0 E   21 Dec 94   261.2   513.0
 57   53 57.2 S   086 34.0 E   21 Dec 94   261.2   513.0
 58   54 22.3 S   086 07.0 E   22 Dec 94   261.2   513.0
 59   54 47.7 S   085 39.5 E   22 Dec 94   261.2   513.0
 60   55 12.7 S   085 11.3 E   22 Dec 94   261.2   513.0
 61   55 38.2 S   084 43.7 E   23 Dec 94   261.2   513.0
 62   56 03.7 S   084 14.8 E   23 Dec 94   260.9   513.6
 63   56 29.0 S   083 46.3 E   23 Dec 94   260.7   513.8
 64   56 54.2 S   083 17.8 E   24 Dec 94   260.7   513.8
 65   57 19.7 S   082 47.7 E   24 Dec 94   260.7   513.8
 66   57 30.8 S   082 32.3 E   24 Dec 94   260.7   513.8
 67   57 36.8 S   082 24.3 E   24 Dec 94   260.7   513.8
 68   57 55.2 S   082 14.0 E   24 Dec 94   260.7   513.8
 69   58 13.0 S   082 00.0 E   25 Dec 94   260.7   513.8
 70   58 36.7 S   082 00.2 E   25 Dec 94   260.7   513.8
 71   59 00.0 S   082 00.2 E   25 Dec 94   260.7   513.8
 72   59 30.0 S   082 00.0 E   25 Dec 94   260.7   513.8
 73   60 00.0 S   082 00.2 E   25 Dec 94   260.7   513.8
 74   60 28.8 S   082 00.2 E   26 Dec 94   260.7   513.8
 75   61 00.0 S   082 00.0 E   26 Dec 94   260.7   513.8
 76   61 29.5 S   082 00.3 E   26 Dec 94   260.7   513.8
 77   61 58.5 S   082 00.7 E   26 Dec 94   260.7   513.8
 78   62 30.3 S   082 00.3 E   26 Dec 94   260.7   513.8
 79   63 00.2 S   082 00.2 E   27 Dec 94   260.7   513.8
 80   63 30.8 S   081 59.5 E   27 Dec 94   260.7   513.8
 82   64 09.0 S   081 53.5 E   27 Dec 94   260.7   513.8
 83   63 50.5 S   081 54.8 E   28 Dec 94   260.7   513.8
 84   63 15.5 S   082 00.2 E   28 Dec 94   260.7   513.8
 85   64 30.7 S   111 23.8 E    1 Jan 95   260.7   513.1
 86   64 51.8 S   110 49.5 E    2 Jan 95   260.7   513.1
 87   64 05.8 S   112 05.3 E    2 Jan 95   260.7   513.1
 88   63 40.8 S   112 35.7 E    2 Jan 95   260.7   513.1
 89   63 15.8 S   113 12.8 E    2 Jan 95   260.7   513.1
 90   62 51.0 S   113 47.2 E    3 Jan 95   260.7   513.1
 91   62 24.8 S   114 25.7 E    3 Jan 95   260.7   513.1
 92   62 00.2 S   115 00.0 E    3 Jan 95   260.7   513.1
 93   61 30.0 S   115 00.3 E    3 Jan 95   260.7   513.1
 94   61 00.0 S   114 59.8 E    4 Jan 95   260.7   513.1
 95   60 23.8 S   115 00.2 E    4 Jan 95   260.7   513.1
 96   59 47.5 S   115 01.5 E    4 Jan 95   260.6   513.2
 97   59 11.8 S   115 00.0 E    5 Jan 95   260.6   513.2
 98   58 36.0 S   115 00.0 E    5 Jan 95   260.6   513.2
 99   58 00.0 S   115 00.3 E    5 Jan 95   260.6   513.2
100   58 00.0 S   115 00.3 E    6 Jan 95   260.6   513.2
101   57 23.8 S   114 59.7 E    6 Jan 95   260.6   513.2
102   56 48.0 S   115 00.2 E    6 Jan 95   260.6   513.2
103   56 11.7 S   115 00.2 E    6 Jan 95   260.6   513.2
104   55 36.0 S   115 00.2 E    7 Jan 95   260.5   513.5
105   55 00.2 S   115 00.3 E    7 Jan 95   260.2   513.8
106   54 24.0 S   115 00.3 E    7 Jan 95   260.2   513.8
107   53 48.0 S   115 00.0 E    8 Jan 95   260.2   513.8
108   53 12.2 S   115 00.8 E    8 Jan 95   260.2   513.8
109   52 36.0 S   115 00.0 E    8 Jan 95   260.2   513.8
110   52 00.2 S   115 00.3 E    8 Jan 95   260.2   513.8
111   51 30.0 S   115 00.3 E    9 Jan 95   260.3   514.2
112   51 00.2 S   115 00.3 E    9 Jan 95   260.3   514.2
113   50 30.0 S   115 00.5 E    9 Jan 95   260.3   514.2
114   50 00.0 S   115 00.3 E   10 Jan 95   260.3   514.2
115   49 30.0 S   115 00.2 E   10 Jan 95   260.3   514.2
116   49 00.0 S   115 00.3 E   10 Jan 95   260.3   514.2
117   48 29.7 S   115 00.3 E   10 Jan 95   260.3   514.2
118   48 00.0 S   115 00.3 E   11 Jan 95   260.3   514.2
119   47 30.0 S   115 00.0 E   11 Jan 95   260.1   513.6
120   47 00.2 S   115 00.0 E   11 Jan 95   260.1   513.6
121   46 30.0 S   115 00.2 E   11 Jan 95   260.0   513.3
122   45 59.8 S   115 00.7 E   12 Jan 95   260.0   513.3
123   45 29.8 S   115 00.3 E   12 Jan 95   260.0   513.3
124   45 00.0 S   114 59.8 E   12 Jan 95   260.0   513.3
125   44 29.8 S   115 00.2 E   12 Jan 95   260.0   513.3
126   43 59.8 S   115 00.2 E   13 Jan 95   260.0   513.3
127   43 29.8 S   115 00.2 E   13 Jan 95   260.0   513.3
128   43 00.0 S   115 00.0 E   13 Jan 95   260.0   513.3
129   42 29.7 S   115 00.2 E   14 Jan 95   260.0   513.3
130   42 00.0 S   115 00.0 E   14 Jan 95   260.0   513.3
131   41 30.3 S   114 59.8 E   14 Jan 95   260.0   513.3
132   40 53.7 S   115 00.2 E   14 Jan 95   260.0   513.3
133   40 18.0 S   115 00.0 E   15 Jan 95   260.0   513.3
134   39 41.8 S   115 00.0 E   15 Jan 95   260.0   513.3
135   39 05.8 S   115 00.0 E   15 Jan 95   260.0   513.3
136   38 29.8 S   115 00.0 E   15 Jan 95   260.0   513.3
137   38 00.0 S   114 59.8 E   16 Jan 95   260.0   513.3
138   37 29.8 S   115 00.0 E   16 Jan 95   260.0   513.3
139   37 00.0 S   115 00.0 E   16 Jan 95   260.0   513.3
140   36 29.8 S   115 00.0 E   17 Jan 95   260.0   513.3
141   36 00.0 S   115 00.0 E   17 Jan 95   260.0   513.3
142   35 39.0 S   114 59.7 E   17 Jan 95   260.0   513.3
143   35 38.8 S   115 00.7 E   17 Jan 95   260.0   513.3
144   35 31.0 S   114 59.7 E   17 Jan 95   260.0   513.3
145   35 12.0 S   115 00.0 E   18 Jan 95   260.0   513.4
146   34 57.8 S   115 00.2 E   18 Jan 95   260.0   513.4
147   34 49.2 S   114 59.8 E   18 Jan 95   260.0   513.4








_____________________________________________________________________________________________________________
_____________________________________________________________________________________________________________









                                  APPENDIX A:

                        REPRINT OF PERTINENT LITERATURE




Johnson K.M., A.G. Dickson, G. Eischeid, C. Goyet, P.R. Guenther, R.M. Key,
K. Lee, E.R. Lewis, F.J. Millero, D. Purkerson, C.L. Sabine, R.G. Schottle,
D.W.R. Wallace, R.J. Wilke, and C.D. Winn. 2002. Carbon Dioxide, Hydrographic 
and Chemical Data Obtained During the Nine R/V Knorr Cruises Comprising the 
Indian Ocean CO2 Survey (WOCE Sections I8SI9S, I9N, I8NI5E, I3, I5WI4, I7N, I1, 
I10, and I2; December 1, 1994 -January 22, 1996), Ed. A. Kozyr. ORNL/CDIAC-138, 
NDP-080. Carbon Dioxide Information Analysis Center, Oak Ridge National 
Laboratory, U.S. Department of Energy, Oak Ridge, Tennessee.





ORNL/CDIAC-138
NDP-080


           CARBON  DIOXIDE, HYDROGRAPHIC,  AND  CHEMICAL  DATA  OBTAINED
      DURING  THE  NINE R/V KNORR  CRUISES COMPRISING  THE  INDIAN OCEAN  CO2
    SURVEY (WOCE  SECTIONS I8SI9S, I9N, I8NI5E, I3, I5WI4, I7N, I1, I10, and I2;
                       DECEMBER 1, 1994 JANUARY 19, 1996)


                                 Contributed by

        Kenneth M. Johnson,1   Andrew G. Dickson,2       Greg Eischeid,3   
        Catherine Goyet,4      Peter R. Guenther,2       Robert M. Key,5  
        Kitack Lee,6           Ernest R. Lewis,7         Frank J. Millero,6   
        David Purkerson,6      Christopher L. Sabine,8   Rolf G. 
        Schottle,9             Douglas W.R. Wallace,10   Richard J. Wilke,7 and 
                               Christopher D. Winn,11


 1  Department of Applied Science, Brookhaven National Laboratory, Upton, NY,
    U.S.A.  Retired, now at P.O. Box 483, Wyoming, RI, U.S.A.
 2  Scripps Institution of Oceanography, University of California, La Jolla, CA,
     U.S.A.
 3  Woods Hole Oceanographic Institute, Woods Hole, MA, U.S.A.
 4  University of Perpignan, Perpignan, France
 5  Department of Geosciences, Princeton University, Princeton, NJ, U.S.A.
 6  Rosenstiel School of Marine and Atmospheric Science, University of Miami,
    Miami, FL, U.S.A.
 7  Department of Applied Science, Brookhaven National Laboratory, Upton, NY,
    U.S.A.
 8  Pacific Marine Environmental Laboratory, NOAA, Seattle, WA, U.S.A.
 9  Department of Oceanography, University of Hawaii, Honolulu, HI, U.S.A.
10  Institute for Marine Sciences, Kiel, Germany
11  Hawaii Pacific University, Kaneohe, HI, U.S.A.




                                  Prepared by
                                Alexander Kozyr
                    Carbon Dioxide Information Analysis Center
                          Oak Ridge National Laboratory
                          Oak Ridge, Tennessee, U.S.A.

                          Date Published: October 2002

                                Prepared for the
                         Environmental Sciences Division
                  Office of Biological and Environmental Research
                            U.S. Department of Energy
              Budget Activity Numbers KP 12 04 01 0 and KP 12 02 03 0

                                Prepared by the
                    Carbon Dioxide Information Analysis Center
                          OAK RIDGE NATIONAL LABORATORY
                         Oak Ridge, Tennessee 37831-6335
                                   managed by
                                UT-BATTELLE, LLC
                                    for the
                            U.S. DEPARTMENT OF ENERGY
                          under contract DE-AC05-00OR22725



                                    CONTENTS

LIST OF FIGURES
LIST OF TABLES
ACRONYMS
ABSTRACT
PART 1:  OVERVIEW
1.  BACKGROUND INFORMATION
2.  DESCRIPTION OF THE EXPEDITION
     2.1  R/V Knorr: Technical Details and History
     2.2  The Indian Ocean CO2 Survey Cruises Information
     2.3  Brief Cruise Summary
3.  DESCRIPTION OF VARIABLES AND METHODS
     3.1  Hydrographic Measurements
          3.1.1 SIO/ODF Methods and Instrumentation
          3.1.2 WHOI Methods and Instrumentation
          3.1.3 Underway Measurements
     3.2  Total Carbon Dioxide Measurements
     3.3  Total Alkalinity Measurements
     3.4  Carbon Data Synthesis and Analysis
     3.5  Radiocarbon Measurements
4.  DATA CHECKS AND PROCESSING PERFORMED BY CDIAC
5.  HOW TO OBTAIN THE  DATA AND DOCUMENTATION
6.  REFERENCES





LIST OF FIGURES (see PDF report for figures)

Figure

1  The cruise track during the R/V Knorr expeditions in the Indian Ocean along
   WOCE Sections I8SI9S, I9N, I8NI5E, I3, I5WI4, I7N, I1, and I2
2  Sampling depths at all hydrographic stations occupied during the R/V Knorr
   Indian Ocean survey along WOCE Section I9N
3  Example of ODV station mode plot: measurements vs depth for Stations
   172 174 of Section I9N
4  Distribution of the TCO2 and TALK in seawater along WOCE Section I9N
5  Property-property plots for all stations occupied during the R/V Knorr
   cruise along WOCE Section I9N



LIST OF TABLES

Table

 1  Technical characteristics of R/V Knorr
 2  Dates, ports of call, expedition codes (EXPOCODEs), and names of chief
    scientists during Indian Ocean CO2 survey cruises
 3  WOCE measurement programs and responsible institutions during
    Indian Ocean CO2 survey cruises
 4  Principal investigators and senior at-sea personnel responsible for the
    WOCE measurement programs during Indian Ocean CO2 survey cruises
 5  Personnel responsible for carbonate system parameter measurements,
    number of CTD stations, and number of TCO2 and TALK analyses made
    during Indian Ocean CO2 survey cruises
 6  Required WHP accuracy for deep water analyses
 7  The short-term precision of the nutrient analyses for Indian Ocean Section
    I2
 8  Certified salinity, TALK, and TCO2 for CRM supplied for Indian
    Ocean CO2 survey
 9  Precision of discrete TCO2 analyses during Indian Ocean CO2 survey
10  Mean difference and standard deviation of the differences between at-sea
    TCO2 by coulometry and on-shore TCO2 by manometry on aliquots of the same
    sample from Indian Ocean CO2 survey, and mean replicate precision of the
    manometric analyses
11  Mean analytical difference (TALK) between analyzed and certified TALK
    for CRM used during Indian Ocean CO2 survey
12  Mean analytical difference (TALK) between analyzed and certified TALK
    for each section during Indian Ocean CO2 survey
13  Final count of carbonate system parameter (CSP) analyses during Indian Ocean
    CO2 survey
14  Content, size, and format of data files



ACRONYMS

A/D      analog-to-digital
ADCP     acoustic Doppler current profiler
ALACE    autonomous Lagrangian circulation explorer
BOD      biological oxygen demand
BNL      Brookhaven National Laboratory
14C      radiocarbon
CALFAC   calibration factor
CDIAC    Carbon Dioxide Information Analysis Center
CFC      chlorofluorocarbon
CO2      carbon dioxide
CTD      conductivity, temperature, and depth sensor
CRM      certified reference material
d.f.     degree of freedom
DIW      deionized water
DOE      U.S. Department of Energy
EEZ      Exclusive Economic Zone
emf      electro-magnetic fields
EXPOCODE expedition code
FSI      Falmouth Scientific Instruments
fCO2     fugacity of CO2
FTP      file transfer protocol
GO       General Oceanics
GMT      Greenwich mean time
GPS      global positioning system
Hcl      hydrochloric acid
IAPSO    International Association for the Physical Sciences of the Ocean
IMET     Improved METeorology
I/O      input-output
JGOFS    Joint Global Ocean Flux Study
kn       knots
LADCP    lowered ADCP
LDEO     Lamont-Doherty Earth Observatory
MATS     Miami University alkalinity titration systems
NBIS     Neil Brown Instrument system
NCSU     North Carolina State University
NDP      numeric data package
NOAA     National Oceanic and Atmospheric Administration
nm       nautical mile
NSF      National Science Foundation
ODF      Ocean Data Facility
ONR      Office of Naval Research
OSU      Oregon State University
PC       personal computer
PI       principal investigator
POC      particulate organic carbon
PMEL     Pacific Marine Environmental Laboratory
PU       Princeton University
QA       quality assurance
QC       quality control
R/V      research vessel
RSMAS    Rosenstiel School of Marine and Atmospheric Sciences
SIO      Scripps Institution of Oceanography
SOMMA    single-operator multiparameter metabolic analyzer
SSW      standard seawater
TAMU     Texas A&M University
TALK     total alkalinity
TCO2     total carbon dioxide
TD       to-deliver
UH       University of Hawaii
UM       University of Miami
UW       University of Washington
VFC      voltage to frequency converter
WHOI     Woods Hole Oceanographic Institution
WHPO     WOCE Hydrographic Program Office
WOCE     World Ocean Circulation Experiment
WHP      WOCE Hydrographic Program




                                   ABSTRACT

Johnson K. M.,  A. G. Dickson,  G. Eischeid,  C. Goyet, P. R. Guenther,
        R. M. Key, K. Lee,  E. R. Lewis,  F. J. Millero,  D. Purkerson,
        C. L. Sabine,  R. G. Schottle,  D. W. R. Wallace,  R. J. Wilke,
        and C. D. Winn.  2002.  Carbon Dioxide, Hydrographic and Chemi-
        cal Data Obtained During the Nine  R/V Knorr Cruises Comprising
        the Indian Ocean CO2 Survey  (WOCE Sections I8SI9S, I9N, I8NI5E,
        I3, I5WI4, I7N, I1, I10, and I2;  December 1, 1994 - January 22,
        1996),  Ed. A. Kozyr.  ORNL/CDIAC-138, NDP-080.  Carbon Dioxide
        Information Analysis Center, Oak Ridge National Laboratory, U.S.
        Department of Energy, Oak Ridge, Tennessee.

This document describes the procedures and methods used to measure total carbon
dioxide (TCO2) and total alkalinity (TALK) at hydrographic stations taken during
the R/V Knorr Indian Ocean cruises (Sections I8SI9S, I9N, I8NI5E, I3, I5WI4,
I7N, I1, I10, and I2) in 1994 1996.  The measurements were conducted as part of
the World Ocean Circulation Experiment (WOCE).  The expedition began in
Fremantle, Australia, on December 1, 1994, and ended in Mombasa, Kenya, on 
January 22, 1996.  During the nine cruises, 12 WOCE sections were occupied.

Total carbon dioxide was extracted from water samples and measured using single-
operator multiparameter metabolic analyzers (SOMMAs) coupled to coulometers.
The overall precision and accuracy of the analyses was ±1.20 µmol/kg.  The
second carbonate system parameter, TALK, was determined by potentiometric
titration.  The precision of the measurements determined from 962 analyses of
certified reference material was ±4.2 µmol/kg (REFERENCE).  This work was
supported by grants from the National Science Foundation, the U.S. Department
of Energy, and the National Oceanographic and Atmospheric Administration.

The R/V Knorr Indian Ocean data set is available as a numeric data package (NDP)
from the Carbon Dioxide Information Analysis Center (CDIAC).  The NDP consists
of 18 oceanographic data files, two FORTRAN 77 data retrieval routine files, a
readme file, and this printed documentation, which describes the contents and
format of all files as well as the procedures and methods used to obtain the
data.  Instructions for accessing the data are provided.

Keywords: carbon dioxide; TCO2; total alkalinity; coulometry; gas
           chromatography; World Ocean Circulation Experiment; Indian
           Ocean; hydrographic measurements; carbon cycle.



1.  BACKGROUND INFORMATION

The World Ocean Circulation Experiment (WOCE) Hydrographic Program (WHP) was a
major component of the World Climate Research Program.  The primary WOCE goal
was to understand the general circulation of the global ocean well enough to be
able to model its present state and predict its evolution in relation to long-
term changes in the atmosphere.  The impetus for the carbon system measurements
arose from concern over the rising atmospheric concentrations of carbon dioxide
(CO2).  Increasing atmospheric CO2 may intensify the earth's natural greenhouse
effect and alter the global climate.

The carbon measurements, which were carried out on the U.S. WOCE Indian Ocean
cruises, were supported as a core component of the Joint Global Ocean Flux
Study (JGOFS).  This coordinated effort received support in the United States
from the U.S. Department of Energy (DOE), the National Oceanic and Atmospheric
Administration (NOAA) and the National Science Foundation (NSF).  Goals were to
estimate the meridional transport of inorganic carbon in a manner analogous to
the estimates of oceanic heat transport (Bryden and Hall 1980; Brewer, Goyet,
and Drysen 1989; Holfort et al. 1998; Roemmich and Wunsch 1985) and to build a
database suitable for carbon-cycle modeling and the estimation of anthropogenic
CO2 in the oceans.  The global data set includes approximately 23,000 stations.
Wallace (2001) recently reviewed the goals, conduct, and initial findings of
the survey.

This report discusses the CO2 science team effort to sample the entire Indian
Ocean for inorganic carbon (Fig.1).  The total CO2 (TCO2) and total alkalinity
(TALK) were measured in the water column and the fugacity of CO2 (fCO2) in the
surface waters [see Sabine and Key (1998) for a description of the fCO2 methods
and data].  The TCO2 analytical systems were furnished and set up by Brookhaven
National Laboratory under the supervision of D.W.R. Wallace and K.M.
Johnson, and the alkalinity titrators were furnished and set up by the
University of Miami under the supervision of F.J. Millero.  During the survey,
certified reference material (CRM) was used to ensure measurement accuracy.  All
shipboard measurements followed standard operating procedures (DOE 1994).  This
report focuses on TCO2 and TALK measurements.  Because the team shared equipment
throughout all nine cruises and so much material, including quality assessments
of the data, has already appeared in the refereed literature, it will be
limited to a brief summary.  Published documentation appears in appendices.


2.  DESCRIPTION OF THE EXPEDITION

2.1.  R/V Knorr: Technical Details and History

The R/V Knorr, built in 1969 by the Defoe Shipbuilding Company in Bay City,
Michigan, is owned by the U.S. Navy.  It was turned over to the Woods Hole
Oceanographic Institution in 1971 for operation under a charter agreement with
the Office of Naval Research.  It was named for E.R. Knorr, a hydrographic
engineer and cartographer who in 1860 held the title of Senior Civilian and
Chief Engineer Cartographer of the U.S. Navy Office.  Its original length and
beam were 245 and 46 ft, respectively.  Beginning on February 6, 1989, it
underwent a major midlife retrofit or "jumbo-izing" at the McDermott Shipyard
in Amelia, Louisiana.  A midsection was added to the ship to stretch its length
by 34 ft, to 279 ft, and fore and aft azimuthing propulsion systems were added
to make it one of the most maneuverable and stable ships in the oceanographic
fleet. By the time it was returned to the Woods Hole Oceanographic Institution
in late 1991, the retrofit had taken 32 months.  The P6 Section was the vessel's
first scientific cruise after the retrofitting.  The R/V Knorr was designed for
a wide range of oceanographic operations and possesses antiroll tanks and a
strengthened bow for duty in icy waters.  Like its sister ship, the R/V
Melville, it is used for ocean research and routinely carries scientists from
many different countries.  Table 1 provides a list of technical characteristics
of the R/V Knorr, while Table 2 provides individual cruise information,
parameters measured, and responsible personnel with their institutional
affiliations.


2.2.  The Indian Ocean CO2 Survey Cruises Information

      Ship name:            R/V Knorr
      Cruise/Leg:           WOCE Sections I8SI9S, I9N, I8NI5E, I3, I5WI4, I7N,
                            I1, I10, and I2
      Ports of call:        Fremantle Australia (start), and
                            Mombasa, Kenya (end)
      Dates:                December 1, 1994 - January 22, 1996
      TALK instrumentation: F.J. Millero, RSMAS
      TCO2 instrumentation: D.W.R. Wallace and K.M. Johnson, Brookhaven
                            National Laboratory (BNL)
      Reference material:   A.D. Dickson, SIO
      Funding support:      DOE, NSF
      Chief scientist:      See Table 2


TABLE 2: Dates, ports of call, expedition codes (EXPOCODEs), and names of chief
         scientists during Indian Ocean CO2 survey cruises
_________________________________________________________________________________________

 Section  Start     Finish    From        To          EXPOCODE       Chief
          date      date                                             Scientist
 -------  --------  --------  ----------  ----------  ------------   -------------------
 I8SI9S   12/01/94  01/19/95  Fremantle   Fremantle   316N145_5      M. McCartney (WHOI)
 I9N      01/24/95  03/06/95  Fremantle   Colombo     316N145_6      A. Gordon (LDEO)
 I8NI5E   03/10/95  04/16/95  Colombo     Fremantle   316N145_7      L. Talley (SIO)
 I3       04/20/95  06/07/95  Fremantle   Port Louis  316N145_8      W. Nowlin (TAMU)
 I5WI4    06/11/95  07/11/95  Port Louis  Port Louis  316N145_9      J. Toole (WHOI)
 I7N      07/15/95  08/24/95  Port Louis  Muscat      316N145_10     D. Olson (RSMAS)
 I1       08/29/95  10/18/95  Muscat      Singapore   316N145_11,12  J. Morrison (NCSU)
 Dry Dock 10/19/95  11/05/95  Dampier
 I10      11/06/95  11/24/95  Dampier     Singapore   316N145_13     N. Bray (SIO)
 I2       11/28/95  01/22/96  Singapore   Mombasa     316N145_14,15  G. Johnson (PMEL)
_________________________________________________________________________________________

      Participating Institutions:
      LDEO   Lamont-Doherty Earth Observatory
      NCSU   North Carolina State University
      PMEL   Pacific Marine Environmental Laboratory
      RSMAS  Rosenstiel School of Marine and Atmospheric Science
      SIO    Scripps Institution of Oceanography
      WHOI   Woods Hole Oceanographic Institution


The extent and nature of the complete measurement program and the responsible
institutions for each cruise are summarized in Table 3.


TABLE 3: WOCE measurement programs and responsible institutions during
         Indian Ocean CO2 survey cruises
________________________________________________________________________

 Program                        Section/Cruise
              I8SI9S  I9N   I8NI5E  I3    I5WI4  I7N   I1    I10   I2
 -----------  ------  ----  ------  ----  -----  ----  ---   ----  ----
                             Responsible institution(a)
 -----------  ---------------------------------------------------------
 CTD/Rosette  WHOI    ODF   ODF     ODF   ODF    ODF   WHOI  ODF   WHOI
 BTL Oxygen   WHIO    ODF   ODF     ODF   ODF    ODF   WHOI  ODF   WHOI
 BTL Salts    WHOI    ODF   ODF     ODF   ODF    ODF   WHOI  ODF   WHOI
 Nutrients    OSU     ODF   ODF     ODF   ODF    ODF   OSU   ODF   OSU
 CFCs         LDEO    UM    LDEO    SIO   UW     UM    UW    UM    PMEL
 He/Tr        LDEO    WHOI  WHOI    WHOI  WHOI   UM    WHOI  WHOI  WHOI
 Deep He/Tr                         LDEO  LDEO         UM    WHOI  LDEO
 14C          UW      PU    PU      PU    PU     PU    PU    PU    PU
 ADCP         UH      UH    UH      OSU   UH     UH    SIO   SIO   UH
 TCO2, TALK   BNL     PU    UH      RSMAS BNL    UH    SIO   SIO   UH
________________________________________________________________________
(a) Participating institutions: 
    BNL    Brookhaven National Laboratory 
    LDEO   Lamont-Doherty Earth Observatory 
    NCSU   North Carolina State University
    PMEL   Pacific Marine Environmental Laboratory
    ODF    Ocean Data Facility (SIO)
    OSU    Oregon State University
    PU     Princeton University
    RSMAS  Rosenstiel School of Marine and Atmospheric Science (UM)
    SIO    Scripps Institution of Oceanography, Univ. of California, San Diego
    TAMU   Texas A&M University
    UH     University of Hawaii
    UM     University of Miami
    UW     University of Washington
    WHOI   Woods Hole Oceanographic Institute


The principal investigators (PIs) and the senior technical staff for the WOCE
measurements program are summarized in Table 4.


TABLE 4: Principal investigators and senior at-sea personnel responsible for
         WOCE measurement programs during Indian Ocean CO2 survey cruises

__________________________________________________________________________________

 Program        Responsible personnel (Institution)
 -------------  -----------------------------------------------------------------
 CTD/Rosette    James Swift (SIO/ODF), John Toole (WHOI),  Frank Delahoyde
                (SIO/ODF), Carl Mattson (SIO/ODF), Marshall Swartz (WHOI), Laura
                 Goepfert (WHOI)
 Bottle oxygen  James Swift (SIO/ODF), John Toole (WHOI), George Knapp (WHOI),
                John Boaz (SIO/ODF)
 Bottle salts   James Swift (SIO/ODF), John Toole (WHOI), George Knapp (WHOI)
 Nutrients      Louis Gordon (OSU), James Swift (SIO/ODF), Marie-Claude Beaupre
                (ODF), Joe Jennings (OSU)
 CFCs           John Bullister (PMEL), Rana Fine (RSMAS), William Smethie (LDEO),
                Mark Warner (UW), Ray Weiss (SIO), Kevin Sullivan (RSMAS),
                Frederick A. Van Woy (SIO)
 He/Tr          William Jenkins (WHOI), Peter Schlosser (LDEO), Zafer Top 
                (RSMAS), Peter Landry (WHOI)
 14C            Robert Key (PU)
 ADCP           Teri Chereskin (SIO), Peter Hacker (UH), Eric Firing (UH), Mike
                Kosro (OSU)
 TCO2, TALK     See Table 5
__________________________________________________________________________________

Table 5 contains a summary of the personnel responsible for the discrete carbon-
ate system measurements.


TABLE 5: Personnel responsible for carbonate system parameter measurements,
         number of CTD stations, and number of TCO2 and TALK analyses made
         during Indian Ocean CO2 survey cruises
         ______________________________________________________________________

          Section  Institution  PI(s)       Group       Stations  TCO2    TALK
                                            Leader        (No.)   (No.)   (No.)
          -------  -----------  ----------  ----------  --------  -----   ----
          I8SI9S   BNL          D. Wallace  K. Johnson    147     2184    1910
                                K. Johnson
          I9N      PU           R. Key      C. Sabine     131     2511    2504
                                C. Sabine
          I8NI5E   UH           C. Winn     C. Winn       166     2419    2421
          I3       RSMAS        F. Millero  D. Purkerson  120     1734    1810
          I5WI4    BNL          D. Wallace  R. Wilke      136     1991    1831
                                K. Johnson
          I7N      UH           C. Winn     R. Schottle   156     2235    2577
          I1       WHOI         C. Goyet    G. Eischeid   158     2400    2387
          I10      PU           R. Key      C. Sabine      61      927     926
                                C. Sabine
          I2       UH           C. Winn     R. Schottle   168     2562    2562
          --------------------------------------------------------------------
          Total                                          1244    18963   18928
         ______________________________________________________________________



2.3.  Brief Cruise Summary

Unlike other CO2 survey cruises where a single institution was responsible for
all phases of the work, these cruises were a group effort in which the
measurement groups used the same ship and  instrumentation for a 14-month
period.  BNL supplied two single-operator multiparameter metabolic analyzers
(SOMMA) systems [S/N 004(I) and 006(II)] that were certified at BNL.  A complete
back-up system (S/N 023) was supplied by WHOI.  The alkalinity titrators were
supplied by RSMAS.  Preparation began with a 4-day workshop held in September
1994 at RSMAS under the direction of and in the laboratory of F.J. Millero.
Cruise participants and group leaders from BNL, LDEO, SIO, RSMAS, PU, WHOI, and
UH were instructed in the use of the alkalinity titrators by F.J. Millero and
D. Campbell and in the use of the SOMMA-coulometer systems by K.M. Johnson and
R.W. Wilke.  The day after Thanksgiving the BNL and RSMAS TCO2 groups left for
Australia.  Setup of the alkalinity and coulometric titration systems began on
November 28, 1994.  The I8SI9S cruise began on December 1, 1994.

The first of the nine cruises on the R/V Knorr was the longest continuous cruise
during the survey.  It occupied a series of CTD stations along two north-south
tracks essentially proceeding from Australia to the ice edge (I8S) along 90°E
and then back again to Australia (I9S) at approximately 110°E.  Station 
spacing ranged from 5 to 40 nautical miles (nm).  Testing and selection of the 
best of the available titration systems and components was completed during I8S.  
The alkalinity and especially the coulometric titration systems benefited from 
This "shake-out" period.  Components damaged during transit were identified and
repaired or replaced.  By the beginning of the I9S, operations were more or less
routine.  Except for one approximately 12-h period when high winds of ~60 knots
(kn) made sampling impossible, work proceeded pretty much on schedule during the
50-day cruise.  During the cruise the ability of a team of four marine mammal
and bird observers onboard from PMEL, under the direction of C. Tynan, to remain
in the cold weather and identify whales that were little more than blips on the
horizon amazed all participants of the expedition.  Both Christmas and New Year
holidays were celebrated aboard the ship.  The fine Christmas dinner was
highlighted by the appearance of three humpback whales, who put on a spectacular
display, jumping and passing under and about the ship.  The ship docked in
Fremantle, to the relief of the CO2 team members, on January 19, 1995, after 147
stations were occupied.  Measurement crews were exchanged, and the new team
brought along some badly needed spare parts and components.

The ship departed Fremantle for I9N on January 24 with A. Gordon as Chief
Scientist and a CO2 measurement group from PU.  This section was basically a
northward continuation of I8S.  The weather was perfect during all 43 days of
the cruise.  The participants celebrated the equator crossing on February 14.
This cruise ended on March 5 in Colombo, Sri Lanka, with 131 stations logged.
During the stopover, the carrier gas supply for the coulometric titrators was
shifted from bottled high purity nitrogen to a calibration gas generator (Peak
Scientific), which supplied CO2-free carrier gas for the remaining of the
cruises.

I8NI5E began in Colombo on March 10 with L. Talley as chief scientist and a CO2
measurement group from UH on board.  No problems were noted for the sampling
program, and the weather remained excellent for most of this leg.  The ship
track proceeded southward from Sri Lanka along 88°E to 24°S, then angled
southeastward to the junction of the Ninety-East Ridge and Broken Ridge.  Next,
the ship followed a 1987 section along approximately 32°S.  This zonal section
included the Central Indian Basin, and crossed the northward flow of deep water
just west of Australia.  Due to the good weather, some extra sampling was
carried out, and by the time the ship docked in Fremantle on April 15, 166
stations had been occupied.  On station 296, the rosette accidentally hit bottom
at 3630 m, but the cast was successfully completed.  A postcruise inspection
showed no apparent damage to the equipment.  This cruise included sampling for
particulate organic carbon (POC) in the surface waters near the equator.  POC
samples were also taken at 65 stations for 13C/12C analyses.  Between April 15
and 23, measurement crews were exchanged and spare parts inventories were
updated.

On April 23, the R/V Knorr departed Fremantle for section I3 with W. Nowlin as
chief scientist and a CO2 measurement group from RSMAS.  The ship had to detour
almost immediately back to Fremantle for a medical emergency.  The injured
analyst was able to rejoin the ship in Port Louis, Mauritius.  In addition to
the CTD work, this cruise included the deployment of current meters, drifters,
and autonomous Lagrangian circulation explorer (ALACE) floats.  The cruise track
ran along 20°S from Australia to Mauritius to Madagascar, crossing the West
Australian Basin, Ninety-East Ridge, Central Indian Basin, and Central Indian
Ridge before veering southward to 22  S around Rodrigues Island.  After this, it
proceeded to the east coast of Mauritius, where a 2-day port stop was made in
Port Louis.  Returning to sea, the ship continued sampling westward along 20°S
from the continental shelf to Madagascar.  Weather was characterized by
southeasterly winds of 10-20 kn, mostly sunny skies, occasional rain squalls,
and 4-6 ft swells with slightly higher winds and seas in mid-May.  The Knorr
returned to Port Louis, Mauritius, on June 5 with 120 stations logged.

The next cruise, I4I5W, began on June 11 with J. Toole as chief scientist and a
CO2 measurement group from BNL on board.  This leg focused on major circulation
features of the southwest region of the Indian Ocean, including the region where
the Agulhas Current originates and where dense waters filtering through 
fractures in the Southwest Indian Ridge form a northward deep boundary current
east of Madagascar.  The cruise track formed a closed box to aid in deducing the
absolute circulation.  A stop was made in Durban, South Africa, on June 21 to
pick up a replacement drum of CTD wires.  Attempts were also made to repair the
ship's bow thruster, which had failed very early in the leg; although the repair
was not successful, the lack of a bow thruster had no effect on the scientific
work.  The R/V Knorr departed Durban on June 22 and began I5W including 
reoccupation of stations where data had been taken in 1987.  Bad weather was
experienced on June 30 when wind gusts of 40-50 kn and high seas slowed winch
operations.  As the ship moved across the Madagascar Basin toward port, station
spacing was decreased to 20 nm.  When the ship arrived in port on July 11, 136
stations had been occupied 20 more than planned.

After four days in port, the R/V Knorr departed on I7N with D. Olson as chief
scientist and a CO2 measurement group from UH.  The director of the U.S. WOCE
office, Piers Chapman, was aboard and served as a salt analyst during the
section.  I7N was designed to define the water mass properties and transports
across the Mascarene Basin and to measure water mass properties and baroclinic
structure on a short section across the Amirante Passage, located between the
Mascarene and Somali Basins.  It included a cross-equatorial section and a
reoccupation of stations previously sampled to confirm water mass flows.  This
work included sampling along 65°E in the central Arabian Basin.  The 
concluding phase of the cruise was a deep line of stations up the center of the 
Gulf of Oman.  The last station of this phase was in the Strait of Hormuz, and 
it identified inflows of Arabian (Persian) Gulf water into the Arabian Basin.  
The cruise terminated on August 24 in Muscat, Oman, with 156 stations occupied.

After a 5-day layover, the R/V Knorr departed Muscat on I1 with J. Morrison as
chief scientist and a CO2 measurement group from WHOI.  I1 was the northernmost
Indian Ocean section.  It enclosed the Arabian Sea and Bay of Bengal, which are
important sources of salt and fresh water, respectively.  The Knorr proceeded
from Muscat to the southern end of the Red Sea and then to the coast of Somali,
where the zonal section started at a nominal latitude of 8°N.  The section
crossed the Arabian Sea, in part to study the carbon transport in and out of the
Arabian Sea, and ended on the continental shelf of India.  After a brief port
stop in Colombo, Sri Lanka, on September 28-30, the leg continued from the Sri
Lankan shelf across the Bay of Bengal to the Myanmar continental shelf.  CTD
problems caused considerable difficulty for the scientific party and resulted in
a somewhat noisy hydrographic data set compared to data obtained from the other
sections.  After the last station on the Myanmar shelf, the Knorr deadheaded to
Singapore, arriving on October 16 with 158 stations logged.  I1 was not only the
northernmost section, it was clearly the most adventurous.  ALACE float 
deployments had to be canceled in the territorial waters of India because the
Indian observer on board refused to allow them, and then the threat of pirates
caused the cancellation of a planned section across the Gulf of Aden.  In the
vicinity of Colombo, the ship had to be escorted by four Sri Lankan gunboats,
and planned stops at stations over the Trincomalee Canyon could not be taken
because of the threat of attack by the Tamil Tigers.  Nevertheless, the Knorr
was able to coordinate scientific activities and physical oceanographic
measurements with the nearby R/V Meteor (F. Schott, chief scientist) in an area
of German current meter moorings near Socotra.  Sampling during I1 enabled
comparison of bottle and TCO2 data with earlier JGOFS results and Meteor Pegasus
and Knorr lowered acoustic Doppler current profiler (LADCP) horizontal 
velocities.  From Singapore, the Knorr proceeded to Dampier, Australia, where it
was placed in dry dock from October 19 until November 5.

With the R/V Knorr back in the water, the I10 CO2 measurement group from PU
arrived.  This group was required to do some additional work not normally part 
of the crew exchange routine.  During the dry dock period, the CO2 
instrumentation had been depowered, and the measurement group had to repower and 
check the instrumentation.  Some minor repairs were required for the coulometric
titrators, including the replacement of one or two solenoid valves (the only
valves replaced during the cruises).  In addition, the sample pipettes and
coolant lines were dismounted and cleaned of algal growth.

The R/V Knorr departed Dampier, Australia, on November 11 with N. Bray as chief
scientist.  WOCE Section I10 was set to run from Shark Bay, Western Australia,
to the Indonesian Exclusive Economic Zone (EEZ) 120 nm south of Sunda Strait.
However, constraints imposed by the Indonesian government caused the endpoint to
be moved from the Sunda Strait to near central Java.  The Knorr was not granted
permission to enter the EEZ of Indonesia, and concluding stations had to be
taken along the boundary of the EEZ.  These restrictions prevented full
resolution of the South Java current.  Throughout the Indian Ocean survey,
bottle casts were normally made to within 5-20 m of the bottom; however, on I10
four stations over the Java Trench this could not be done.  Instead, the casts
were made to the maximum CTD depth of 6000 m.  The quality of the bottle data
was considered to be excellent throughout with very few mis-trips.  ALACE floats
were also released during this cruise.  A festive Thanksgiving was celebrated
aboard the ship, and after the last station (1075), the Knorr steamed to
Singapore, arriving on November 28, with 61 stations logged.

The R/V Knorr departed Singapore on December 2 for the last Indian Ocean WOCE
section, I2, with G. Johnson as chief scientist and the UH CO2 measurement group
aboard.  Again, clearance for work in the Indonesian EEZ was not available, and
after a 3-day steam, work commenced with a reoccupation of the final station of
the I10 Section (station 1075).  The Knorr skirted the Indonesian EEZ and moved
westward, crossing the Ninety-East Ridge and the Chagos-Laccadive Ridge.  The
ship continued at approximately 8°S until it made a brief port call in Diego
Garcia from December 28-30.  At this point, the chief scientist departed the
ship and was replaced by Bruce Warren, accompanied by two Kenyan observers.  The
Knorr returned to the 8°S line, passing the crest of the Central Indian Ridge
and then the Mascarene Plateau before it turned southwestward and crossed the
Amirante Passage on the way to the northern tip of Madagascar.  Rounding the
tip, the ship headed northwest toward Africa, making a dogleg to avoid the
Tanzanian EEZ.  After completing the final Indian Ocean Survey station 1244, it
proceeded to Mombasa, arriving on January 22, 1996, with 168 stations logged.

For inorganic carbon, the principal analytical problems for the cruise centered
on the breakage of glass components in the alkalinity titrators; resupply;
accumulation of bubbles in the acid lines of the alkalinity titrators; damaged
coulometric cathode electrodes; algal growth in the sample lines, baths,
pipettes, and alkalinity cells; wide swings in laboratory temperature (19-33 C),
and the failure of the TCO2 glassware drying oven.  Fortunately, glassware
drying oven was repaired.  Temperature swings (21-29 C) were also noted for the
salinometer and nutrient laboratories.  The most vexing problem for the
inorganic carbon analysts was the failure of the refrigerated baths used by both
the alkalinity and coulometric titration systems.  The baths had to be
constantly jury-rigged so that one bath did the work of two, repaired by ship's
technicians when possible, or replaced when possible.  The two groups used
almost 12 different baths, and by the time the work ended, not one could be
considered in reliable condition.  Some were never repaired, while others were
repaired and used for the North Atlantic survey in 1997.


3.  DESCRIPTION OF VARIABLES AND METHODS

3.1.  Hydrographic Measurements

During the survey, responsibility for hydrographic and bottle data was divided
between ODF and WHOI.  Each of these groups uses or may use different
procedures.  Hence, the hydrographic measurements are described in separate
sections.  Because the greater number of the cruises were made under the
auspices of SIO/ODF, the bulk of the methods description is provided in Sect.
3.1.1.  Information specific to WHOI is given in Sect. 3.1.2; in this section
however, the discussion is limited to significant differences between the
SIO/ODF and WHOI operations or methods.  Unless otherwise stated in Sect. 3.1.2,
material presented in Sect. 3.1.1 applies to all cruises.  Sect. 3.1.3 contains
a brief description of the underway measurements common to all cruises.


3.1.1  SIO/ODF Methods and Instrumentation

Hydrographic measurements consisted of salinity, dissolved oxygen, and nutrient
(nitrite, nitrate, phosphate, and silicate) samples collected from Niskin
bottles filled during CTD/rosette casts, and temperature, pressure, salinity,
and dissolved oxygen from the CTD.  At 5- to 40-nm intervals, depending on the
topography, hydrographic casts were made to within 5 20 m of the bottom with a
36-bottle Rosette frame belonging to ODF.  This unit consisted of a 36-bottle
frame, thirty six 10-L bottles, and a 1016 General Oceanics (GO) 36-place pylon.
The GO pylon was used in conjunction with an ODF-built deck unit and power
supply.  The underwater components comprising the CTD included an ODF-modified
Neil Brown Instrument Systems (NBIS) Mark III CTD with conductivity, pressure,
oxygen, and temperature sensors.  The underwater package also consisted of a
SeaTech transmissometer, an LADCP, a Sensormedics dissolved oxygen sensor, a
Falmouth Scientific Instruments (FSI) secondary PRT sensor, a Benthos altimeter,
and a Benthos pinger.  The CTD was mounted horizontally along the bottom of the
frame, while the LADCP was vertically mounted inside the bottle rings.  The
system was suspended from and powered by a three- conductor 0.322-in.
electromechanical cable.  The Rosette was deployed from the starboard side using
either the port side Markey CTD or the starboard side Almon Johnson winch.
Standard CTD practices (i.e., soaking the conductivity and O2 sensors in
distilled water between casts and protecting the sensors against sunlight and
wind by storing the rosette in the hanger between casts) were observed
throughout the cruises.  Regular CTD maintenance included the replacement of O-
rings when needed, bottle inspections, and a regular cleaning of the
transmissometer windows.  At the beginning of each station the time, position,
and bottom depth were logged.  The CTD sensors were powered and control was
transferred to the CTD acquisition and control system in the ship's laboratory.
The CTD was lowered to within 10 m of the bottom if bottom returns were
adequate.  Continuous profiles of horizontal velocity from the sea surface to
the bottom were made for most CTD/rosette casts using the LADCP.

The CTD's control and acquisition system displayed real-time data [pressure,
depth, tem- perature, salinity (conductivity), oxygen, and density] on the video
display of a SunSPARC LX computer.  A video recorder was provided for real-time
analog backup.  The Sun computer system included a color display, a keyboard, a
trackball, a 2.5-GB disk, 18 RS-232 ports, and an 8-mm cartridge tape.  Two
additional Sun systems were networked for display, backup, and processing.  Two
HP 1200 C color ink-jet printers provided hard copy.  The ODF data acquisition
software not only acquired the CTD data but also processed it so that the real-
time data included preliminary sensor corrections and calibration models for
pressure, temperature, and conductivity.  The sampling depths were selected
using down-cast data.  Bottles were tripped on the up-cast.  Bottles on the
rosette were identified with a serial number and the pylon tripping sequence, 1-
36, where the first (deepest) bottle tripped was no. 1.  For shallow-depth
stations, fewer than 36 bottles were closed.

After the CTD was on deck, the acquisition system, the CTD, the pylon, and video
recording were turned off and the sensor protective measures were completed
before sampling began.  If a full suite of samples was drawn, the sampling order
was CFCs, 3He, O2, TCO2, TALK, 14C, 3H, nutrients, and salinity.  Only salinity,
O2, and nutrients were measured at every station.  A deck log was kept to
document the sampling sequence and to note anomalies (e.g., status of bottle
valves, leaks, etc.).  One member of the sampling crew was designated the
"sample cop," and it was his or her responsibility to maintain this log and to
ensure that the sampling order was followed.  Oxygen sampling included
measurement of the temperature, which proved useful for determining leaking or
mis-tripped bottles.  Following the cruises, WHP quality flags were assigned
according to the WOCE Operations Manual (Joyce and Corry 1994) to each measured
quantity.

The principal ODF CTD (no. 1) was calibrated for pressure and temperature at the
ODF Calibration Facility (La Jolla, Calif.) in December 1994 prior to the five
consecutive WOCE Indian Ocean sections beginning with I9N and ending with I7N.
The CTD was also calibrated postcruise in September 1995 prior to the I10
cruise.  Pre- and postcruise laboratory calibrations were used to generate
tables of corrections, which were applied by the CTD data.  At sea, bottle
salinity and oxygen data were to calibrate or check the CTD sensors.  Additional
details concerning calibration and the CTD data processing can be obtained from
the chief scientists' cruise reports at the WHPO web site: http://whpo.uscd.edu/.

Bottle salinity samples were collected in 200-mL Kimax high alumina borosilicate
bottles, sealed with custom-made plastic insert thimbles and Nalgene screw caps.
Salinity was determined after equilibration in a temperature-controlled 
laboratory, usually within 8-20 h of collection.  Salinity was measured with two
ODF-modified Guildline Autosal Model 8400A salinometers, normally at 21 or 
24°C, depending on the prevailing temperature of the salinometer laboratory.  
The salinometers included interfaces for computer-aided measurements (e.g.,
acquiring the measurements, checking for consistency, logging results, and
prompting the analyst).  The salinometers were standardized with International
Association for the Physical Sciences of the Ocean (IAPSO) Standard Seawater
(SSW) Batches P-124, P-126, or P-128 using at least one fresh vial per cast
(usually 36 samples).  The accuracy of the determination was normally 0.002
relative to the SSW batch used.  PSS-78 was then calculated for each sample
(UNESCO 1981).  On some stations (e.g., on Section I5EI8N), bottle salinity
exhibited small offsets (0.002 0.004) compared to the corresponding CTD results
and bottle salinity from nearby stations, and corrections of this magnitude need
to be applied to the bottle salinity.  Errors of this magnitude have no
practical effect on the calculated TCO2 or TALK values.  Hence, bottle salinity
is sufficiently accurate to express inorganic carbon results in µmol/kg.

Bottle oxygen was determined by rinsing 125-mL iodine flasks twice and then
filling to overflowing (3x-bottle volume) with a draw tube.  Sample temperature
was measured immediately with a thermometer imbedded in the draw tube.  The
Winkler reagents were added; and the flask was stoppered, shaken, and then
shaken again 20 min later to ensure that the dissolved O2 was completely fixed.
Oxygen was determined within 4 h of collection using a whole-bottle modified
Winkler titration following the technique of Carpenter (1965) and incorporating
the modifications of Culberson et al. (1991) on an SIO/ODF-designed automated
oxygen titrator.  A Dosimat 665 burette driver fitted with a 1.0-mL burette was
used to dispense thiosulfate solution (50 g/L).  Standards prepared from
preweighed potassium iodate (0.012N) were run each time the automated titrator
was used, and reagent blanks were determined by analyzing distilled water.  The
final oxygen results were converted to µmol/kg using the in situ temperature.
Bottle volumes were precalibrated at SIO.  Laboratory temperature stability
during the sections was considered poor, varying from 22 to 28°C over short 
time periods; and therefore, portable fans were used by ODF analysts to maintain
temperature.

Phosphate, nitrate, nitrite, and silicate samples were collected in 45-mL high-
density polypropylene, narrow-mouth, screw-capped centrifuge tubes which were
cleaned with 10% hydrochloric acid (HCl) and then rinsed three times with sample
before filling.  The samples were analyzed on an ODF-modified four-channel
Technicon AutoAnalyzer II, usually within 1 h of the cast, in a temperature-
controlled laboratory.  If the samples were stored for longer than 1 h prior to
analysis, they were stored at 2 6°C (for no more than 4 h).  The AutoAnalyzer
incorporates the method of Armstrong, Stearns, and Strickland (1967) for
silicate, this same method as modified for nitrate and nitrite, and the method
of Bernhardt and Wilhelms (1967) for phosphate.  The last method is described by
Gordon and coworkers (Atlas et al. 1971; Hager et al. 1972; and Gordon et al.
1992).  Standards were analyzed at the beginning and end of each group of sample
analyses, with a set of secondary intermediate concentrations prepared by
diluting preweighed primary standards.  Replicates were also drawn at each
station for measurement of short-term precision.  For reagent blanks, deionized
water (DIW) from a Barnstead Nanopure deionizer fed from the ship's potable
water supply was analyzed.  An aliquot of deep seawater was run with each set of 
samples as a substandard.  The primary standard for silicate was Na2SiF6; and
for nitrate, nitrite, and phosphate the standards were KNO3, NaNO2, and KH2PO4,
respectively.  Chemical purity ranged from 99.97% (NaNO2) to 99.999% (KNO3).

Most hydrographic data sets met or exceeded the WHP requirements.  Some
exceptions for silicate were noted when differences between overlapping stations
on I3 (Station 548) and I4I5W (Stations 705 and 574) approached 3%; these
silicate data (Stations 702-707) were corrected by adding 3% to the original
results.  Instrument problems also caused difficulties for the nitrite and
silicate analyses on many of the I2 cruise stations.  Silicate problems were
noted at some 30% of these stations, with errors typically being on the order of
2 4%.  This required considerable post- cruise evaluation and workup before the
desired between-station precision for deep water values of 1% was attained.
However, users of the I2 silicate data are urged to use caution or to contact
the analysts for assistance.  Because of the difficulties with the nutrient
analyses on the I2 cruise, the post-cruise I2 precision is given in Table 7 as a
"worst case" for comparison with the WHP standards shown in Table 6.  Short-term
precision is the absolute mean difference between replicates analyzed within a
sample run; the standard deviation of the differences is also shown. The authors
know of no remaining CTD problems, that would affect the quality of the 
carbonate system data.


TABLE 6: Required WHP accuracy for deep water analyses
         ___________________________________________

          Parameter  Required accuracy
          ---------  ------------------------------
          Salinity   0.002 relative to SSW analysed
          Oxygen     1% (2 µmol/kg)
          Nitrate    1% (0.3 0.4 µmol/L)
          Phosphate  1% (0.02 0.03 µmol/L)
          Silicate   1% (1 5 µmol/L)
         ___________________________________________


TABLE 7: The short-term precision of the nutrient analyses for Indian Ocean
         Section I2
         ______________________________________

          Parameter     Difference  ± St. Dev.
                         (µmol/L)
          ------------  ----------  ----------
          Nitrate         0.123       0.093
          Phosphate       0.015       0.009
          Silicic Acid    0.440       0.260
         ______________________________________


3.1.2  WHOI Methods and Instrumentations

Unless otherwise stated procedures are as described in Sect. 3.1.1, above.  For
the hydrographic work on I8SI9S, I1, and I2, the R/V Knorr was outfitted with
equipment belonging to both WHOI and SIO/ODF.  For the I8SI9S section a NBIS CTD
was used.  For I1, four CTDs were available.  The primary sensors were two new
FSI CTDs belonging to WHOI with a Sensormedics oxygen sensors, a titanium
pressure transducer, and a temperature monitor.  The secondary sensors were two
NBIS Mark-III CTDs (WHOI Nos. 9 and 12) also with a Sensormedics oxygen sensor,
a titanium pressure transducer, and a temperature monitor. The MKIII CTDs
experienced failures early during I1 (Stations 858 and 864), and the bulk of the
hydrography was carried out using the FSI (Nos. 1338 and 1344) CTDs.  Usually,
the frame was set up with the two CTDs - one configured to send data up the wire
and one configured to record data internally.  Electrical modifications had to
be made to the CTDs and the deck controllers before CTD data dropouts were
eliminated and the confirmation of bottle closure from the pylon was restored.

For the CTDs, a FSI DT-1050 deck unit was initially used to demodulate the data,
but this unit was replaced for most of the cruise with an EG&G MK-III deck unit.
These units fed serial data to two personal computers (PCs) running EG&G CTD
acquisition software, with one displaying graphical output and the other a
running data listing.  After each station, the CTD data were forwarded to
another set of PCs running EG&G postprocessing and software modified by WHOI
(Millard and Yang 1993) in which the data were centered into 2 dbar bins for
data quality control, which included fitting to bottle salinity and oxygen
results.

The CTDs were calibrated before and after the cruise for temperature and
pressure at WHOI by M. Swartz and M. Plueddemann.  Both calibrations were
consistent, but the data set for I1 was considered to be only of fair quality
because noise levels in the data set are somewhat larger than typical for other
CTDs.  For example, this data set has a salt noise level of 0.002 which is 2
times larger than the norm.  Residuals between the bottle and profile data range
from 0.001 to 0.004.  For a detailed discussion of the CTD calibration and
problems experienced at sea during I1, consult the chief scientist's cruise
report on the WHPO web site.

For I2, WHOI CTD No. 9, a WHOI-modified NBIS MK-IIIb, was used.  The CTD
incorporated a Sensormedics oxygen sensor, titanium pressure transducer, and
temperature sensor, which were calibrated in November 1995 immediately before
the cruise.  On most stations, one of the FSI CTDs was used in the memory mode
and downloaded after station sampling to provide independent or backup CTD
traces.  An FSI Ocean Temperature Module was also attached to the MK-III and
CTDs.  The Mark-III CTD data were acquired using an NBIS Mark-III deck
unit/display that provided demodulated data to two PCs, as described for the
Section I1 cruise.  A PC was also devoted to recovering the data from the FSI
CTDs.  Post-cruise calibration, including dunk tests of the CTDs, was completed
in April and May of 1996 in the WHOI calibration laboratory.  The procedure of
Millard and Yang (1993) was used to correct the pressure temperature sensor
calibration post-cruise to eliminate down/up pressure historesis.  Multiple
regression fits of the CTD data to the bottle data were used to calibrate the
oxygen and conductivity sensors.  See the chief scientist's report on the WHPO
web site for further details.

Bottle salinity samples were collected in 200-mL glass bottles with removable
polyethylene inserts and caps.  Then they were removed to a temperature-
controlled van at 23 C and analyzed on a Guildline Autosal Model 8400B
salinometer (WHOI No. 11).  The salinometer was standardized once a day using
IAPSO SSW (128, dated July 18, 1995).  The accuracy was ~0.002.  A complete
description of the WHOI measurement techniques is given by Knapp, Stalcup, and
Stanley (1990).

Bottle oxygen was determined according to procedures given by Knapp, Stalcup,
and Stanley (1990).  WHOI used a modified Winkler technique similar to that
described by Strickland and Parsons (1972).  The oxygen reagents and bi-iodate
standard were prepared at WHOI in August 1994.  There was no evidence that the
reagents or standard deteriorated during the 17 months they were aboard the
Knorr.  Standardization of the thiosulphate titrant was made daily.  The
accuracy of the method was 0.5%, or approximately 1.0 µmol/kg.

The nutrients were analyzed as described in Sect. 3.1.1 (see also Gordon et al.
1994).


3.1.3.  Underway Measurements

Navigational data (heading, speed, time, date, and position) were acquired from
the ship's Magnavox MX global positioning system (GPS) receiver via RS-232 and
logged automatically at 1-min intervals on a SunSPARC station.  Underway
bathymetry was logged manually at 5-min intervals from the hull-mounted 12-kHz
echo sounder and a Raytheon recorder corrected according to methods described by
Carter (1980).  These data were merged with the navigation data to provide a
time-series of underway position, course, speed, and bathymetry data that were
used for all station positions, depths, and vertical sections.  The Improved
METeorology (IMET) sensors logged meteorological data which  included air
temperature, barometric pressure, relative humidity, sea surface temperature,
and wind speed and direction at 1-min intervals.  Underway shipboard
measurements were made throughout the work to document the horizontal velocity
structure along the cruise tracks using a 150-kHz hull-mounted acoustic Doppler
current profiler (ADCP) manufactured by RD Instruments.  The ADCP was mounted at
a depth of 5 m below the sea surface.  Underway chemical measurements in
water and air included salinity, pCO2 (PU and SIO), pN2O (SIO), and CH4 (SIO).
Two different systems were used for pCO2; the PU group used a rotating disk
equilibrator and infrared detector, while the Scripps group used a shower type
equilibrator and gas chromatograph for the detection of CO2.  The pCO2
measurements, including a comparison of the shower and disk equilibrator
results, were described by Sabine and Key (1998).

A thermosalinograph (manufactured at FSI) was mounted on the bow approximately
3 m below the surface for underway salinity, which was calibrated against
surface CTD and bottle salinity values after the cruise (Sabine and Key 1998).
The CFC groups periodically analyzed air for CFCs using sampling lines from the
bow and stern of the ship.


3.2.  Total Carbon Dioxide Measurements

TCO2 was determined on 18,963 samples using two automated single-operator
multiparameter metabolic analyzers (SOMMA) with coulometric detection of the CO2
extracted from acidified samples.  A description of the SOMMA-coulometry system
and its calibration can be found in Johnson et al. 1987; Johnson and Wallace
1992; and Johnson et al. 1993.  A schematic diagram of the SOMMA analytical
sequence and a complete description of the sampling and analytical methods used
for discrete TCO2 on the Indian Ocean WOCE sections appear in Appendix B
(Johnson et al. 1998).  Further details concerning the coulometric titration can
be found in Huffman (1977) and Johnson, King, and Sieburth (1985).  The
measurements for the Indian Ocean Survey were made on two systems provided by
BNL (S/Ns 004 and 006) and a backup by WHOI (S/N 023).

TCO2 samples were collected from approximately every other station [~ 60 nm
intervals, 50% of the stations (Fig. 2)] in 300-mL glass biological oxygen
demand (BOD) bottles.  They were immediately poisoned with 200 µL of a 50%
saturated solution of HgCl2, thermally equilibrated at 20°C for at least 1 h,
and analyzed within 24 h of collection (DOE Handbook of Methods 1994).  Certified 
reference material (CRM) samples were routinely analyzed, usually at 
the beginning and end of the coulometer cell lifetime, according to DOE (1994).  
As an additional check of internal consistency, duplicate samples were usually
collected on each cast at the surface and from the bottom waters.  These
duplicates were analyzed on the same system within the run of cast samples from
which they originated, but the analyses were separated in time usually by ~3 h.
Periodically, replicate samples were also drawn for shipboard analysis at sea
using coulometry and for later analysis on shore at SIO by manometry.  The
latter samples, typically designated as the "Keeling samples," consisted of two
500-mL replicate samples collected at two depths (four samples total per
station).  These were analyzed only if both replicates survived the storage and
the return journey to SIO.

Seawater introduced from an automated "to-deliver" (TD) pipette into a stripping
chamber was acidified, and the resultant CO2 from continuous gas extraction was
dried and coulometrically titrated on a model 5011 UIC coulometer.  The 
coulometer was adjusted to give a maximum titration current of 50 mA, and it was
run in the counts mode [the number of pulses or counts generated by the
coulometer's voltage-to-frequency converter (VFC)] during the time the titration
was displayed and acquired by the computer.  In the coulometer cell, the acid
(hydroxyethylcarbamic acid) formed from the reaction of CO2 and ethanolamine was
titrated coulometrically (electrolytic generation of OH-) with photometric
endpoint detection.  The product of the time and the current passed through the
cell during the titration was related by Faraday's constant to the number of
moles of OH- generated and thus to the moles of CO2 that reacted with
ethanolamine to form the acid.  The age of each titration cell was logged from
its birth (time that electrical current was applied to the cell) until its death
(time when the current was turned off).  The age was measured from birth
(chronological age) and in mass of carbon (mgC) titrated since birth (carbon
age).  The systems were controlled with PCs equipped with RS232 serial ports for
the coulometer and the barometer, a 24-line digital input/output (I/O) card for
the solid state relays and valves, and an analog-to-digital (A/D) card for the
temperature, conductivity, and pressure sensors.  These sensors monitored the
temperature of the sample pipette, gas sample loops, and, in some cases, the
coulometer cell.  The controlling software was written in GWBASIC Version 3.20
(Microsoft Corp., Redmond, Wash.), and the instruments were driven from an
options menu appearing on the PC monitor.

The TD volume (Vcal) of the sample pipettes was determined gravimetrically prior
to the cruise and periodically during the cruise by collecting aliquots of
deionized water dispensed from the pipette into pre-weighed serum bottles which
were sealed and re-weighed on shore. The apparent weight of water collected
(Wair), corrected to the mass in vacuo (Mvac), was divided by the density of the
calibration fluid at the calibration temperature to give Vcal.  The sample
volume (Vt) at the pipette temperature was calculated from the expression

                     V(t) = V(cal) [1 + a(v) (t - t(cal))] ,

where av is the coefficient of volumetric expansion for Pyrex-type glass
(1 X 10(^-5)/°C), and t is the temperature of the pipette at the time of a
measurement.  Vcal for the Indian Ocean CO2 survey cruises and a chronology of
the pipette volume determinations appear in Appendix B.

The coulometers were electronically calibrated at BNL prior to the cruises and
recalibrated periodically during the cruises (Sections I8SI9S and I5WI4) to
check the factory calibration as described in Johnson et al. (1993) and DOE
(1994).  The results for the electronic intercepts (Intec) and slopes (Slopeec)
are given in Appendix B.  For all titrations, the micromoles of carbon titrated
(M) was

      M = [Counts/4824.45 - (Blank x T(t)) - (Int(ec) x T(i))]/Slope(ec) ,

where 4824.45 (counts/µmol) was the scaling factor obtained from the factory
calibration, T(t) was the length of the titration in minutes, Blank is the
system blank in µmol/min, and T(i) the time of continuous current flow in
minutes.

The SOMMA-coulometry systems were calibrated daily with pure CO2 (calibration
gas) by titrating the mass of CO2 contained in two stainless steel gas sample
loops of known volume and by analyzing CRM samples supplied by Dr. Andrew
Dickson of the SIO.  The ratio of the calculated (known) mass of CO2 contained
in the gas sample loops to the mass determined coulometrically was the CALFAC
(~1.004).  A complete history of the calibration results appears in Appendix B.
For water and CRM samples, TCO2 concentration in µmol/kg was

                  TCO2 = M x CALFAC x [1 / (V(t) x ñ)] x d(Hg) ,

where p is the density of seawater in g/mL at the analytical t and S calculated
from the equation of state given by Millero and Poisson (1981), and d(Hg) is the
correction for sample dilution with bichloride solution (for the cruises d(Hg) =
1.000666).

System 006 was equipped with a conductance cell (Model SBE-4, Sea-Bird
Electronics, Bellevue, Wash.) for the determination of salinity as described by
Johnson et al. (1993).  Whenever possible, SOMMA and CTD salinities were
compared to identify mis-trips or other anomalies, but the bottle salinities
(furnished by the chief scientist) have been used to calculate p throughout.

Three CRM batches were used for the Indian Ocean Survey.  The certified TCO2
concentrations were determined by vacuum-extraction/manometry in the laboratory
of C.D. Keeling at SIO and are given in Table 8.


TABLE 8: Certified salinity, TALK, and TCO2 for CRM supplied for Indian Ocean
         CO2 survey
         _________________________________________________

          Batch  Salinity  TCO2 (µmol/kg)  TALK (µmol/kg)
          -----  --------  --------------  --------------
           23     33.483      1993.10         2212.70
           26     33.258      1978.34         2176.60
           27     33.209      1988.10         2214.90
         _________________________________________________


Optimal cell and platinum electrode configurations, according to criteria given
in Appendix B, were selected on the first section (I8S) and were used on all
subsequent cruises.

The quality control-quality assurance (QC-QA) of the coulometric TCO2 determina-
tions was assessed from analyses of 983 CRM samples during the nine Indian Ocean
CO2 survey cruises.  For both coulometric titration systems (004 and 006) the
average TCO2 (measurement minus CRM value) for the whole survey was 0.86 µmol/kg
and the standard deviation was ±1.21 µmol/kg.  A cruise-by-cruise breakdown of
the accuracy and precision of the CRM analyses is given in Appendix B.

The small mean difference between the analyzed and certified TCO2 and the very
high precision (1.21 µmol/kg) of the differences indicates that the two systems
gave very accurate and virtually identical results over the entire survey (see
also Fig. 6 in Appendix B).

The second phase of the QC-QA procedure was an assessment of sample precision,
which is presented in Table 9.  The sample precision was determined from
duplicate samples analyzed on each system during sections I8SI9S at the
beginning of the survey and I4I5W about half way through the survey.  The pooled
standard deviation (Sp2), shown in Table 9, is the square root of the pooled
variance according to Youden (1951) where K is the number of samples with one
replicate analyzed on each system, n is the total number of replicates analyzed
from K samples, and n - K is the degree of freedom (d.f.) for the calculation.
Precision was calculated this way because TCO2 was analyzed on two different
systems, and an estimate of sample precision independent of the analytical
system was required.  Hence Sp2 is the most conservative estimate of precision
and includes all sources of random and systematic error (bias).  Bias between
systems would increase the imprecision of the measurements, but the excellent
agreement between the Sp2 values for natural seawater samples (Table 9) and the
high precision of the CRM differences confirms the virtually uniform response,
accuracy, and high precision of both systems during the survey.  This finding
confirms that the precision of the TCO2 analyses during the Indian Ocean CO2
survey was ±1.20 µmol/kg.


TABLE 9: Precision of discrete TCO2 analyses during Indian Ocean CO2 survey
                        _____________________________

                         Section  Sp2  (K,  n,  d.f)
                         -------  ------------------
                         I8SI9S   1.26 (15, 30, 15)
                         I4I5W    0.91 (21, 42, 21)
                         CRM      1.21
                        _____________________________


The next phase of the QC-QA procedure was the comparison of replicate samples
analyzed at sea and in the shore-based laboratory.  Samples from every cruise
were analyzed at sea by continuous gas extraction/coulometry, and later, after
storage, duplicate samples were analyzed on shore by vacuum extraction/manome-
try.  The results of the analyses are summarized in Table 10.

TABLE 10: Mean Difference [TCO2(S-SIO)] and standard deviation of the dif-
          ferences [S.D.(S-SIO)] between at-sea TCO2 by coulometry and on-
          shore TCO2 by manometry on aliquots of the same sample from Indian 
          Ocean CO2 survey, and the mean replicate precision [S.D.(SIO)] of 
          the manometric analyses
          __________________________________________________________________

           Section  Pairs Analyzed  TCO2(S-SIO)  S.D.(S-SIO)  S.D.(SIO) (a)
                         (n)         (µmol/kg)    (µmol/kg)    (µmol/kg)
           -------  --------------  -----------  -----------  -------------
           I8SI9S         23           -4.14        1.80          0.82
           I9N            24           -1.96        1.67          0.80
           I8NI5E         17           -4.80        2.87          1.31
           I3             29           -3.29        1.26          0.82
           I4I5W          16           -2.95        1.40          1.30
           I7N            13           -5.37        1.92          1.40
           I1             26           -5.59        1.38          1.05
           I10             8           -4.94        1.52          1.28
           I2             10           -4.42        1.50          0.83
           n             166            9           9             9
           ----------------------------------------------------------------
           Mean                        -4.16        1.70          1.07
           S.D.                         1.21        0.49          0.25
          __________________________________________________________________
           (a) Each on-shore TCO2 by manometry is always the mean of two
               analyses (see text).


In general, the reproducibility and the uniformity of the data as a whole, and
specifically, the high precision of the manometric analyses shown in Table 10,
indicate that the collection and return of the "Keeling samples" was
successfully performed by each of the measurement groups.  Poor sampling or
storage techniques would probably have been manifested in a much higher
imprecision for the on-shore replicate analyses and in the differences between
the at-sea and on-shore analyses.  However, the negative mean difference (4.16 ±
1.21, n = 9) for the Indian Ocean sections was greater than the mean difference
for WOCE sections in other oceans (-1.36 ± 1.37 µmol/kg, n = 22).  The accuracy
of the CRM analyses, the tendency for the coulometric analyses to give slightly
lower results, and the reproducibility of the at-sea and on-shore differences
are similar everywhere, but the magnitude of the Indian Ocean difference is
clearly the largest observed to date.  Even if the consistent and slightly
negative difference for the CRM is taken into account (-0.86 µmol/kg), the at-
sea coulometric measurements are approximately 2 µmol/kg lower than the
manometric method.  A suite of samples from the 1997 North Atlantic sections
remains to be analyzed.  Until these analyses are completed and a thorough
statistical evaluation of the entire CO2 survey data set is made, the
explanation of the at-sea and on-shore differences, including those found for
the Indian Ocean, is not possible.

An additional step in the QA-QC was also undertaken.  Inspection of Fig. 1 shows
points where the cruise tracks cross or nearly cross.  The agreement between
TCO2 measurements made at these crossover locations (± 100 km) on different
cruises was examined by assuming that the temporal and spatial variations in
deep-ocean TCO2 are small relative to the measurement accuracy and precision.
Hence, deep ocean waters should have the same TCO2 at different times in the
absence of internal vertical motion, and because deep ocean motion probably
occurs along constant density surfaces (isopycnals), the comparisons of TCO2
measurements were made with reference to density and not depth.  Appendixes B
and D (Johnson at al. 1998 and Sabine et al. 1999) give a complete description
of the statistical procedures used to make the crossover comparisons.  Briefly,
crossover points were selected for comparison of water samples collected below
2500 m.  A smooth curve was fit through the TCO2 data as a function of the
density anomaly referenced to 3000 dbar (sigma3) using Cleveland's LOESS
smoother (Cleveland and Devlin 1988).  A separate fit was performed for the data
collected at each of the two intersecting crossover points, but the same tension
parameter was used for all of the crossover points so that the smoothing
function was consistently applied to all crossover locations.  The difference
between the two smoothed curves was evaluated at 50 evenly spaced points
covering the density range where the two data sets overlapped.  A mean and
standard deviation for the 50 comparisons was calculated for each crossover
point.  For TCO2, differences never exceeded 3 µmol/kg, and the overall mean and
standard deviation of the differences was -0.78 ± 1.74 µmol/kg.  The latter
differences were consistent with the overall precision of the CRM analyses (±
1.2 µmol/kg).

Tables 8 10 show an internally consistent TCO2 data set for the Indian Ocean
with excellent accuracy with respect to the CRM certified values, consistently
good precision, no analytical bias between the coulometric titration systems,
and crossover agreement to within the precision of the method.  However, the
agreement between the at-sea and on-shore analyses is not as good as for earlier
WOCE sections from other oceans (i.e., the Pacific and the South Atlantic).
Based on the accuracy of the CRM analyses and the high precision of the sample
analyses, the TCO2 data were not corrected in any way and were deemed to meet
survey criteria for accuracy and precision.


3.3.  Total Alkalinity Measurements

Total alkalinity was measured on 18,928 samples using two closed-cell automated
potentiometric titration systems (hereafter designated as MATS) developed at the
University of Miami.  The MATS are described by Millero et al. (1993) and by
Millero et al. (1998).  The latter reprinted in Appendix C of this document,
completely describes the Indian Ocean Survey TALK measurements and results.
Briefly, the MATS consisted of three parts: a water-jacketed, fixed-volume
(about 200 mL determined to ± 0.05 mL) closed Plexiglass sample cell, a Metrohm
model 665 Dosimat titrator, and a pH meter (Orion, Model 720A), the last two
controlled by a PC.  The titration cell was similar to those used by Bradshaw
and Brewer (1988), but had a greater volume to improve the precision of the
measurements.  The cell was equipped with flush-mounted fill and drain valves to
increase the reproducibility of the cell volume.  The cell, titrant burette, and
sample container were held at a temperature of 25 ± 0.01°C using a constant
temperature bath (e.g., Neslab, Model RTE 221).

A Lab Windows C program was used to run the titrators, record the volume of
titrant added, and record the measured electromagnetic fields (emf) of the
electrodes through RS232 serial interfaces.  Two electrodes were used in each
cell: a ROSS glass pH electrode (Orion, Model 810100) and a double-junction
Ag/AgCl reference electrode (Orion, Model 900200).  The specific electrodes used
during the Indian Ocean survey were selected after careful screening for non-
Nernstian behavior.  Only those electrodes which gave TCO2 results in good
agreement with TCO2, as determined coulometrically, were used (Sect. 3.2).

Seawater samples were titrated by adding increments of HCl until the carbonic
acid endpoint of the titration was exceeded.  During a titration, the emf
readings were monitored until they were stable (± 0.09 mV).  Sufficient volume
of acid was added to increase the emf by preassigned increment (~13 mV) in order
to give an even distribution of data points over the course of a full titration,
which consists of 25 data points.  A single titration takes about 20 min.  A
FORTRAN computer program based on those developed by Dickson (1981) and by
Johansson and Wedborg (1982) was used to calculate the carbonate parameters.
The pH and pK of the acids used in the program are on the seawater scale, and
the dissociation constants for carbonic acid were taken from Dickson and Millero
(1987).  For further details see Appendix C and DOE (1994).

The titrant (acid) used throughout the cruises was prepared prior to the cruise,
standardized, and stored in 500-mL borosilicate glass bottles for use in the
field.  The 0.25-M HCl acid solution was prepared by dilution of 1-M HCl in
0.45-M NaCl to yield a solution with total ionic strength similar to that of
seawater of salinity 35.0 (I = 0.7 M).  The acid was standardized by coulometry
(Taylor and Smith 1959; Marinenko and Taylor 1968), and was also checked by
independent titration in A. Dickson's laboratory at SIO.  The independent
determinations agreed to ± 0.0001 M, which corresponds to an uncertainty in TALK
of ~ 1 µmol/kg.  The Dosimat titrator burettes were calibrated with Milli-Q
water at 25°C to ± 0.0005 mL.

While CRM samples were available to the TCO2 analysts from the beginning of the
measurement program in 1990, the Indian Ocean cruises were the first to have a
certified alkalinity standard as well.  Hence, the accuracy of the method was
checked in the laboratory by analyzing CRM samples from batches 23, 24, 26, 27,
29, and 30 and comparing the analyzed values with the certified TALK determined
by A. Dickson at SIO (in the same manner as for TCO2).  These results are
summarized in Table 11 (see also Appendix C).  The mean difference between the
MATS measurements in the laboratory and the certified TALK values was -0.8
µmol/kg for CRM samples with a concentration range approximately one-half as
large as the range of a typical seawater profile.  The excellent agreement
indicated that the CRM concept for alkalinity was valid and that the methodology
for TALK was ready for the Indian Ocean survey.  The results for the at-sea
measurements of the CRM samples have been extracted from Table 2 of Appendix C,
summarized, and are given in Table 12.


TABLE 11: Mean analytical difference (TALK) between analyzed and certified
          TALK for CRM used during Indian Ocean CO2 survey
          ____________________________________________________________________

           Batch  Salinity   Certified values    MATS mean TALK   delta TALK
                               TCO2     TALK        (µmol/kg)    (MATS - CRM)
                            (µmol/kg) (µmol/kg)
           -----  --------  --------- ---------  --------------  ------------
            23     33.483    1993.10   2212.7        2213.7          1.0
            24     33.264    1987.53   2215.5        2215.8          0.3
            26     33.258    1978.34   2176.6        2175.1         -1.5
            27     33.209    1988.10   2214.9        2214.3         -0.6
            29     33.701    1902.33   2184.8        2182.3         -2.5
            30     33.420    1988.78   2201.9        2200.5         -1.4
           Range    0.492      90.77     38            40.7          3.5
           Mean                                                     -0.8
          ____________________________________________________________________


The analytical differences are for the most part within the precision of the
measurements (~ 2-5 µmol/kg) except for the I7N Section.  The larger at-sea
differences were attributed to operator error or procedures and to uncertainties
in the volume of cells, especially after repairs due to leakage, breakage, or
repositioning the electrodes after changing the inner filling solutions.
Variations between different MATS systems used on a single cruise were corrected
using the adjustments required to reproduce the values assigned for the CRM (see
Table 11).  The at-sea sample titrations were corrected using the results of the
at-sea CRM analyses.  For TALK, the calibration factor (CF) used to correct the
at sea measurements was

                 CF = TALK (meas., CRM) - CRM (certified value),

and the corrected TALK (TALKc) was

                (TALKc) = TALK (meas., Spl) x [ CRM / (CRM + CF)],

where CRM was the certified TALK and Spl was the measured sample TALK.

The overall precision of TALK determinations during the Indian Ocean survey was
± 4.2 µmol/kg.  The precision of the potentiometric pH and TCO2 measurements are
given in Table 3 of Appendix C.


TABLE 12: Mean analytical difference (TALK) between analyzed and certified
          TALK for each section during Indian Ocean CO2 survey
          _____________________________________________________________

           Batch  Section  Certified  MATS mean   S.D. (n)    ∆ TALK
                             TALK       TALK     (µmol/kg)  (MATS-CRM)
                           (µmol/kg)  (µmol/kg)              (µmol/kg)
           -----  -------  ---------  ---------  ---------  ----------
            23    I8SI9S    2212.7     2221.5     5.1 (49)     8.8
            23    I9N       2212.7     2216.2     3.3 (138)    3.5
            23    I8NI5E    2212.7     2211.6     4.9 (80)    -1.1
            23    I3        2212.7     2215.4     1.4 (65)     2.7
            26    I3        2176.6     2178.0     1.2 (30)     1.4
            26    I5WI4     2176.6     2182.6     3.8 (79)     6.0
            26    I7N       2176.6     2184.0     5.7 (59)     7.4
            27    I7N       2214.9     2221.5     3.1 (8)      6.6
            23    I7N       2212.7     2222.4     7.4 (10)     9.7
            27    I1        2214.9     2219.4     3.9 (244)    4.5
            27    I10       2214.9     2212.9     4.0 (62)    -2.0
            27    I2        2214.9     2219.4     4.5 (67)     4.5
            n                                     891         12
          _____________________________________________________________


TALK was also checked at the crossover locations of two cruises in the same way
as TCO2.  The agreement between the corrected TALK measurements made at the
crossover locations (± 100 km) on different cruises was examined by assuming
that the temporal and spatial variations of the deep-ocean TALK were small
relative to measurement accuracy and precision.  Hence, deep ocean waters should
have the same TALK at different times in the absence of internal vertical
motion, and because deep ocean motion probably occurs along constant-density
surfaces (isopycnals), the comparisons of TALK measurements were made with
reference to density and not depth.  Appendixes C and D give a description of 
the statistical procedures used to make the crossover comparisons.  For water
samples collected below 2500 m, a smooth curve was fit through the TALK data as
a function of the density anomaly referenced to 3000 dbar (sigma3) using
Cleveland's LOESS smoother (Cleveland and Devlin 1988).  A separate fit was
performed on the data collected at each of the two intersecting crossover
points, with the same tension parameter being used for all of the crossovers so
that the smoothing function was consistently applied.  The difference between
the two smoothed curves was evaluated at 50 evenly-spaced points covering the
density range where the two data sets overlapped.  Mean and standard deviations
for the differences at the 50 points were calculated for each crossover point.
For TALK, differences never exceeded 6 µmol/kg, and the overall mean and
standard deviation of the differences was 2.1 ± 2.1 µmol/kg.  The latter were
consistent with the overall precision of the CRM analyses (± 4 µmol/kg).

Table 13 is a final summation of the inorganic carbon analytical work completed
during the Indian Ocean CO2 survey from 1994 to 1996.


TABLE 13: Final count of carbonate system parameter (CSP) analyses during
          Indian Ocean CO2 survey
                      _______________________________________

                                   No. of CSP determinations
                       Parameters   Discrete   CRM   Total
                       ----------   --------  -----  ------
                        TCO2         18,963     983  19,946
                        TALK         18,928     949  19,877
                        Total        37,891   1,932  39,823
                      ______________________________________


3.4.  Carbon Data Synthesis and Analysis

In accordance with one of the stated goals of the program, an evaluation of the
data set with respect to estimated anthropogenic CO2 distributions in the Indian
Ocean has been completed and published by Sabine et al. (1999) (see Appendix D).
The document is appended to this report as Appendix D.  Additional crossover
comparisons of the survey data with data gathered in the 1980s and in 1993 by
French scientists are included.  Briefly, the sequestering of anthropogenic CO2
has been estimated by comparing the Indian Ocean survey results with the Indian
Ocean GEOSECS expedition data from 1977 to 1978.  Although CRM samples were not
available for evaluating the earlier data, statistical methods were used to fit
these data and correct for calibration offsets so that they could be compared
with the current survey data.  The data analysis was complicated by regions of
pronounced denitrification (Arabian basin) and other regional variations that
had to be considered and quantified.  In summary, the estimate of the
anthropogenic inventory was relatively small in the Indian and Southern Oceans,
with anthropogenic carbon uptake lower by a factor of 2 compared to that of the
Atlantic Ocean.  Importantly, discrepancies between model and data-based
estimates were found especially for the Southern Ocean where carbon uptake
appears to have been traditionally overestimated by the extant circulation
models. (See Appendix D for further details.)  The initial data synthesis work
indicates that the survey data will provide an important baseline with respect
to future studies and that the spatial distribution of anthropogenic carbon can
be an important tool for understanding model-based carbon uptake estimates and
the response of models to atmospheric increases in CO2.


3.5.  Radiocarbon Measurements

Full information on the radiocarbon measurement method, instrumentation, and
results can be found in Appendix E of this document.



4.  DATA CHECKS AND PROCESSING PERFORMED BY CDIAC

An important part of the numeric data packaging process at the Carbon Dioxide
Information Analysis Center (CDIAC) involves the quality assurance (QA) of data
before  distribution.  Data received at CDIAC are rarely in a condition that
would permit immediate distribution, regardless of the source.  To guarantee
data of the highest possible quality, CDIAC  conducts extensive QA reviews that
involve examining the data for completeness, reasonableness, and accuracy.  The
QA process is a critical component in the value-added concept of supplying
accurate, usable data for researchers.

The following information summarizes the data processing and QA checks performed
by CDIAC on the data obtained during the R/V Knorr cruise along WOCE Sections
I8SI9S, I9N, I8NI5E, I3, I5WI4, I7N, I1, I10, and I2  in the Indian Ocean.

1.  The final carbon-related data were provided to CDIAC by the ocean carbon
    measurement PIs listed in Table 5.  The final hydrographic and chemical
    measurements and the station information files were provided by the WOCE
    Hydrographic Program Office (WHPO) after quality evaluation.  A FORTRAN 90
    retrieval code was written and used to merge and reformat all data files.

2.  Every measured parameter for each station was plotted vs depth (pressure) to
    identify questionable outliers using the Ocean Data View (ODV) software
    (Schlitzer 2001) Station Mode (Fig. 3).

3.  The section plots for every parameter were generated using the ODV's Section
    Mode in order to map a general distribution of each property along all 
    Indian Ocean sections (Fig. 4).

4.  To identify "noisy" data and possible systematic, methodological errors,
    property-property plots for all parameters were generated (Fig. 5), 
    carefully examined, and compared with plots from previous expeditions in the 
    Indian Ocean.

5.  All variables were checked for values exceeding physical limits, such as
    sampling depth values that are greater than the given bottom depths.

6.  Dates, times, and coordinates were checked for bogus values (e.g., values of
    MONTH < 1 or > 12; DAY < 1 or > 31; YEAR < 1994 or > 1996; TIME < 0000 or >
    2400; LATITUDE <  70.000 or > 60.000; LONGITUDE < 19.000 or > 119.000.

7.  Station locations (latitudes and longitudes) and sampling times were 
    examined for consistency with maps and cruise information supplied by PIs.

8.  The designation for missing values, given as  9.0 in the original files, was
    changed to  999.9 for the consistency with other oceanographic data sets.



6.  REFERENCES

Armstrong, F.A.J., C.R. Stearns, and J.D.H. Strickland.  1967.  The
    measurement of upwelling and subsequent biological processes by means of 
    the Technicon Autoanalyzer and associated equipment.  Deep-Sea Research 
    14:381-9.

Atlas, E.L., S.W. Hager, L.I. Gordon, and P.K. Park.  1971.  A Practical
    Manual for Use of the Technicon AutoAnalyzer in Seawater Nutrient Analyses
    (revised).  Technical Report 215, Reference 71-22, Oregon State University,
    Department of Oceanography, Oreg.

Bernhardt, H. and A. Wilhelms.  1967.  The continuous determination of low
    level iron, soluble phosphate and total phosphate with the AutoAnalyzer.
    Technicon Symposia 1:385-9.

Bradshaw, A.L. and P.G. Brewer.  1988.  High precision measurements of
    alkalinity and total carbon dioxide in seawater by potentiometric
    titration: 1. Presence of unknown protolyte (s). Marine Chemistry 28:69-86.

Brewer, P.G., C. Goyet, and D. Dyrssen.  1989.  Carbon dioxide transport by
    ocean currents at 25° N latitude in the Atlantic Ocean.  Science 246:477-79.

Bryden, H.L., and M.M. Hall.  1980.  Heat transport by ocean currents across
    25° N latitude in the North Atlantic Ocean.  Science 207:884.

Carpenter, J.H.  1965.  The Chesapeake Bay Institute technique for the Winkler
    dissolved oxygen method.  Limnology and Oceanography 10:141-3.

Carter, D.J.T.  1980.  Computerized Version of Echo-sounding Correction Tales
    (3rd Ed.). Marine Information and Advisory Service, Institute of
    Oceanographic Sciences, Wormley, Godalming, Surrey, U.K.

Cleveland, W.S. and S.J. Devlin.  1988.  Locally-weighted regression: an
    approach to regression analysis by local fitting.  Journal of American
    Statistical Association 83:596-610.

Culberson, C.H., G. Knapp, M.  Stalcup, R.T. Williams, and F. Zemlyak.  1991.
    A comparison of methods for the determination of dissolved oxygen in
    seawater.  WHP Office Report, WHPO 91-2.  WOCE Hydrographic Program Office,
    Woods Hole, Mass. U.S.A.

Dickson, A.G.  1981.  An exact definition of total alkalinity and a procedure
    for the estimation of alkalinity and total CO2 from titration data.  Deep-
    Sea Research 28:609-23.

Dickson, A.G. and F.J. Millero.  1987.  A comparison of the equilibrium
    constants for the dissociation of carbonic acid in seawater media.  Deep-
    Sea Research 34:1733-43.

DOE (U.S. Department of Energy).  1994.  Handbook of Methods for the Analysis
    of the Various Parameters of the Carbon Dioxide System in Seawater. Version
    2.0. ORNL/CDIAC-74.  A.G. Dickson and C. Goyet (eds.). Carbon Dioxide
    Information Analysis Center, Oak RidgeNational Laboratory, Oak Ridge,
    Tenn., U.S.A.

Gordon, L.I., J.C. Jennings, Jr., A.A. Ross, and J.M. Krest.  1992.  A
    suggested protocol for continuous flow automated analysis of seawater
    nutrients (phosphate, nitrate, nitrite and silicic acid) in the WOCE
    Hydrographic Program and the Joint Global Ocean Fluxes Study. Grp. Tech.
    Rpt. 92-1.  Chemical Oceanography Group, Oregon State University, College
    of Oceanography, Oregon, U.S.A.

Gordon, L.I., J.C. Jennings, Jr., A.A. Ross, and J.M. Krest.  1994.  A
    suggested protocol for continuous flow automated analysis of seawater
    nutrients (phosphate, nitrate, nitrite and silicic acid) in the WOCE
    Hydrographic Program and the Joint Global Ocean Fluxes Study.  In WOCE
    Operations Manual.  WHP Office Report WHPO 91-1. WOCE Report No. 68/91.
    Revision 1. Woods Hole, Mass., U.S.A.

Hager, S.W., E.L. Atlas, L.I. Gordon, A.W. Mantyla, and P.K. Park.  1972.
    A comparison at sea of manual and autoanalyzer analyses of phosphate,
    nitrate, and silicate.  Limnology and Oceanography 17:931-7.

Huffman, E.W.D., Jr.  1977.  Performance of a new automatic carbon dioxide
    coulometer. Microchemical Journal 22:567-73.

Johansson, O., and M. Wedborg.  1982.  On the evaluation of potentiometric
    titrations of seawater with hydrochloric acid.  Oceanology Acta 5:209-18.

Johnson, K.M., A.E. King, and J.McN. Sieburth.  1985.  Coulometric TCO2
    analyses for marine studies:  An introduction.  Marine Chemistry 16:61-82.

Johnson, K.M., P.J. Williams, and L. Brandstroem, and J.McN. Sieburth.
    1987.  Coulometric TCO2 analysis for marine studies:  Automation and
    calibration.  Marine Chemistry 21:117-33.

Johnson, K.M., and D.W.R. Wallace.  1992.  The single-operator multipara-
    meter metabolic analyzer for total carbon dioxide with coulometric
    detection.  DOE Research Summary No. 19.  Carbon Dioxide Information
    Analysis Center, Oak Ridge National Laboratory, Tenn., U.S.A.

Johnson, K.M., K.D. Wills, D.B. Butler, W.K.Johnson, and C.S. Wong.
    1993.  Coulometric total carbon dioxide analysis for marine studies:
    Maximizing the performance of an automated gas extraction system and
    coulometric detector. Marine Chemistry 44:167-87.

Johnson, K.M., A.G. Dickson, G. Eischeid, C. Goyet, P.R. Guenther, R.M.
    Key, F.J.Millero, D. Purkerson, C.L. Sabine, R.G. Schotle, D.W.R.
    Wallace, R.J. Wilke, and  C.D. Winn. 1998.  Coulometric total carbon
    dioxide analysis for marine studies: Assessment of the quality of total
    inorganic carbon measurements made during the U.S. Indian Ocean CO2 Survey
    1994 1996.  Marine Chemistry 63:21-37.

Joyce, T., and C. Corry.  1994.  Requirements for WOCE Hydrographic Programme
    Data Reporting.  Report WHPO 90-1, WOCE Report No. 67/91, WOCE Hydrographic
    Programme Office, Woods Hole, Mass. U.S.A. pp. 52-55.  Unpublished
    Manuscript.

Knapp, G.P., M.C. Stalcup, and R.J. Stanley.  1990.  Automated oxygen and
    salinity determination.  WHOI Technical Report No. WHOI-90-35.  Woods Hole
    Oceanographic Institution, Woods Hole, Mass., U.S.A.

Marinenko, G. and J.K. Taylor.  1968.  Electrochemical equivalents of benzoic
    and oxalic acid. Analytical Chemistry 40:1645-51.

Millard, R.C., Jr.  1982.  CTD calibration and data processing techniques at
    WHOI using the practical salinity scale.  Proceedings Int. STD Conference
    and Workshop, Mar. Tech. Soc., La Jolla, Calif., p. 19.

Millard, R.C. and K. Yang.  1993.  CTD calibration and processing methods used
    at Woods Hole Oceanographic Institution.  Woods Hole Oceanographic
    Institution Technical Report. WHOI 93-44.  Woods Hole Oceanographic
    Institution, Woods Hole, Mass., U.S.A.

Millero, F.J., and A. Poisson.  1981.  International one-atmosphere equation
    of state for seawater. Deep-Sea Research 28:625-29.

Millero, F.J., J.Z. Zhang, K. Lee, and D.M. Campbell.  1993.  Titration
    alkalinity of seawater. Marine Chemistry 44:153-60.

Millero, F.J., A.G. Dickson, G. Eischeid, C. Goyet, P.R. Guenther, K.M.
    Johnson, K. Lee, E. Lewis, D. Purkerson, C.L. Sabine, R. Key, R.G.
    Schottle, D.R.W. Wallace, and C.D. Winn.  1998.  Total alkalinity
    measurements in the Indian Ocean during the WOCE hydrographic program CO2
    survey cruises 1994 1996.  Marine Chemistry 63:9-20.

Roemmich, D., and C. Wunsch.  1985.  Two transatlantic sections: Meridional
    circulation and heat flux in the subtropical North Atlantic Ocean.  Deep-
    Sea Research 32:619-64.

Sabine, C.L. and R.M. Key.  1998.  Surface water and atmospheric underway
    carbon data obtained during the world ocean circulation experiment Indian
    Ocean survey cruises (R/V Knorr, December 1994 January 1996).  ORNL/CDIAC-
    103, NDP-064, Carbon Dioxide Information Analysis Center, Oak Ridge
    National Laboratory, Oak Ridge, Tenn., U.S.A.

Sabine, C.L., R.M. Key, K.M.. Johnson, F.J. Millero, J.L. Sarmiento, D.R.W. 
    Wallace, and C.D. Winn.  1999.  Anthropogenic CO2 inventory of the
    Indian Ocean. Global Biogeochemical Cycles 13:179-98.

Schlitzer, R.  2001.  Ocean Data View.  http://www.awi-bremerhaven.de/GEO/ODV.
    Online publication.  Alfred-Wegener-Institute for Polar and Marine
    Research. Bremerhaven, Germany.

Taylor, J.K. and S.W. Smith.  1959.  Precise coulometric titration of acids
    and bases.  Journal of Research of the National Bureau of Standards
    63A:153-9.

UNESCO.  1981.  Background papers and supporting data on the practical salinity
    scale, 1978. UNESCO Technical Papers in Marine Science, No. 37: p. 144.

Wallace, D.W.R.  2002.  Storage and transport of excess CO2 in the oceans:
    The JGOFS/WOCE Global CO2 survey.  In J. Church, G. Siedler, and J. Gould
    (eds.). Ocean Circulation and Climate,  Academic Press, (in press).

Youden, W.J.  1951.  Statistical Methods for Chemists.  Wiley, New York.



List of CO2 measurement group members participating in the Indian Ocean CO2
Survey aboard the R/V Knorr in 1994 1996 (CO2 group leaders for each section 
are given in Table 4 in the text)

             ________________________________________________________

              Section  Name                  Sponsoring  Affiliation
                                             institute
              -------  --------------------  ----------  -----------
              I8SI9S   Haynes, Charlotte H    BNL         WDNR
                        Haynes, Elizabeth M   BNL         RU
                        Wysor, Brian S.       BNL         SHC
              I9N      Dorety, Art            PU          PU
                        Kozyr, Alex           PU          ORNL/CDIAC
                        Suntharalingam, Parv  PU          PU
              I8NI5E   Parks, Justine         UH          SIO
                        Popp, Brian           UH          UH
                        Schottle, R.          UH          UH
              I3       Aicher, Jennifer       RSMAS       RSMAS
                        Edwards, Christopher  RSMAS       RSMAS
                        Krenisky, Joann       RSMAS       RSMAS
              I4I5W    Lewis, Ernie           BNL         BNL
                        Pikanowski, Linda     BNL         SHML
                        Zotz, Michelle        BNL         BNL
              I7N      Adams, Angela          UH          UH
                        Angeley, Kelly        UH
                        Phillips, Jennifer    UH          UHH
              I1       Amaoka, Toshitaka      WHOI        GSEESHU
                        Okuda, Kozo           WHOI        GSEESHU
                        Ording, Philip        WHOI        WHOI
              I10      Boehme, Sue            PU          RU
                        Markham, Marion       PU          PU
                        Mcdonald, Gerard      PU          PU
              I2       Admas, Angela          UH          UH
                        Cipolla, Cathy        UH          GSOURI
                        Phillips, Jennifer    UH          UHH
             ________________________________________________________



Participating institutions:

BNL         Brookhaven National Laboratory
ORNL/CDIAC  Oak Ridge National Laboratory/Carbon Dioxide Information Analysis
            Center
GSEESHU     Graduate School of Environmental and Earth Science, Hokkaido
            University
GSOURI      Graduate School of Oceanography, University of Rhode Island
PU          Princeton University
RSMAS       Rosenstiel School of Marine and Atmospheric Science, University of 
            Miami
RU          Rutgers University
SHC         South Hampton College
SHML        Sandy Hook Marine Laboratory
SIO         Scripps Institution of Oceanography
UH          University of Hawaii, Honolulu
UHH         University of Hawaii at Hilo
WDNR        Wisconsin Department of Natural Resources
WHOI        Woods Hole Oceanographic Institution












                                  APPENDIX B:

                        REPRINT OF PERTINENT LITERATURE




Johnson, K.M. , A.G. Dickson, G. Eischeid, C. Goyet, P. Guenther, R.M. Key, 
F.J. Millero, D. Purkerson, C.L. Sabine, R.G. Schottle, D.W.R. Wallace, R.J. 
Wilke and C.D. Winn, Coulometric total carbon dioxide analysis for marine 
studies: assessment of the quality of total inorganic carbon measurements
made during the US Indian Ocean CO2 Survey 1994-1996, Marine Chemistry 63:21-37.





Marine Chemistry 
63(1998) 21-37


           COULOMETRIC TOTAL CARBON DIOXIDE ANALYSIS FOR MARINE STUDIES:
          ASSESSMENT OF THE QUALITY OF TOTAL INORGANIC CARBON MEASUREMENTS
                MADE DURING THE US INDIAN OCEAN CO2 SURVEY 1994-1996

          Kenneth M. Johnson(a)*, Andrew G. Dickson(b), Greg Eischeid(c),
  Catherine Goyet(c), Peter Guentherb(b), Robert M. Key(d), Frank J. Millero(e), 
        David Purkerson(e), Christopher L. Sabine(d), Rolf G. Schottle(f),
      Douglas W. R. Wallace(a), Richard J. Wilke(a) and Christopher D. Winn(f)

(a) Department of Applied Science, Brookhaven National Laboratory, Upton, NY 
    11973, USA
(b) Scripps Institution of Oceanography, University of California, San Diego, La 
    Jolla San Diego, CA 92093, USA
(c) Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA
(d) Geology Department, Princeton University, Princeton, NJ 08544, USA
(e) Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, 
    Miami, FL 33149, USA
(f) Department of Oceanography, University of Hawaii, Honolulu, HI 96822, USA

*Corresponding author. Tel.: +1-516-344-5668; Fax: +1-516-344-3246

Received 8 January 1998; accepted 6 May 1998. Available online 8 December 1998.
0304-4203/98/$ - see front matter (c) 1998 Elsevier Science B.V. 
All rights reserved. 
PII: S0304-42039800048-6





ABSTRACT

Two single-operator multiparameter metabolic analyzers (SOMMA)-coulometry 
systems (I and II) for total carbon dioxide (TCO2) were placed on board the R/V 
Knorr for the US component of the Indian Ocean CO2 Survey in conjunction with 
the World Ocean Circulation Experiment-WOCE Hydrographic Program (WHP). The 
systems were used by six different measurement groups on 10 WHP Cruises 
beginning in December 1994 and ending in January 1996. A total of 18,828 
individual samples were analyzed for TCO2 during the survey. This paper assesses 
the analytical quality of these data and the effect of several key factors on 
instrument performance. Data quality is assessed from the accuracy and precision 
of certified reference material (CRM) analyses from three different CRM batches. 
The precision of the method was 1.2 µmol/kg. The mean and standard deviation of 
the differences between the known TCO2 for the CRM (certified value) and the CRM 
TCO2 determined by SOMMA-coulometry were -0.91±0.58 (n=470) and -1.01±0.44 
(n=513) µmol/kg for systems I and II, respectively, representing an accuracy of 
0.05% for both systems. Measurements of TCO2 made on 12 crossover stations 
during the survey agreed to within 3 µmol/kg with an overall mean and standard 
deviation of the differences of -0.78±1.74 µmol/kg (n=600). The crossover 
results are therefore consistent with the precision of the CRM analyses. After 
14 months of nearly continuous use, the accurate and the virtually identical 
performance statistics for the two systems indicate that the cooperative survey 
effort was extraordinarily successful and will yield a high quality data set 
capable of fulfilling the objectives of the survey.


AUTHOR KEYWORDS: total carbon dioxide (TCO2); single-operator multiparameter 
                 metabolic analyzers (SOMMA) coulometry; marine studies

INDEX TERMS:     reproductive toxicity; boron




ARTICLE OUTLINE

1. Introduction
2. Materials and methods
2.1. Preparations
2.2. Selection of cell assemblies
2.3. At-sea operations
2.4. Calculation of results
2.5. Assessment of analytical accuracy
2.6. Data distribution
3. Results
3.1. To-deliver pipette volume
3.2. CRM analyses and system accuracy
3.3. System repeatability and precision during the survey
4. Discussion
5. Crossover analysis
6. Conclusions
Acknowledgements
References




1. INTRODUCTION

Between 1990 and 1997 an international effort was made to determine the global 
oceanic distribution of inorganic carbon in conjunction with the World Ocean 
Circulation Experiment (WOCE) Hydrographic Programme (WHP). This effort is 
referred to as the Global Survey of CO2 in the oceans, and it is an integral 
part of the Joint Global Ocean Flux Study (JGOFS). The goals of this survey are 
to:

1. Accurately determine the oceanic distribution of dissolved inorganic carbon,

2. Quantify the uptake of anthropogenic carbon dioxide by the oceans to better 
   predict future atmospheric carbon dioxide levels,

3. Provide a global description of anthropogenic carbon dioxide in the oceans to 
   aid development of a 3-dimensional model of the oceanic carbon cycle,

4. Characterize the transport of carbon dioxide between the ocean and the 
   atmosphere and the large scale (e.g., meridional) transports of carbon 
   dioxide within the ocean.

The survey has acquired a global data set of profile measurements of dissolved 
carbon dioxide parameters on both zonal and meridional oceanographic transects 
throughout the world's oceans. With reference to program goals, Bates et al. 
(1996) found that for mixed layer waters the average rate of increase in CO2 
concentration due to the uptake of anthropogenic CO2 was 1.7 µmol/kg/yr (<0.1%). 
This rate of increase establishes a natural target for the accuracy of the TCO2 
measurements. The distribution of this 'excess' CO2 signal is not uniform 
spatially, and it is masked by variability in CO2 concentrations arising from 
natural biological and physicochemical processes. Hence, the goals of the 
program imply that measurements must be extremely accurate (0.1% or better) and 
spatially extensive. A large part of the US contribution to this survey has been 
conducted by a team of investigators supported by the US Department of Energy 
(DOE). This team has developed certified reference materials (Dickson, 1990), 
instrumentation (Johnson and Wallace, 1992), a set of standard operating 
procedures (DOE, 1994) and, to a large extent, shared a common approach to the 
measurement program.

This paper presents the DOE team effort which sampled the Indian Ocean for 
inorganic carbon during the course of approximately 1 year. All the measurements 
were made aboard a single research vessel during sequential cruises which 
allowed the investigators to share equipment and procedures to an unprecedented 
extent. This paper concentrates on estimating the accuracy of the shipboard 
determinations of the total dissolved inorganic carbon concentration of 
seawater. This parameter was established at the onset of the survey as the 
primary carbonate system parameter because its concentration should change in 
response to anthropogenic CO2 uptake and it had the highest potential for 
measurement accuracy. Our results highlight some factors which affect the 
accuracy of this measurement. The Indian Ocean Survey aboard the R/V Knorr 
encompassed the cruise legs shown in Fig. 1 in the sequence given in Table 1. 
Fig. 1 also gives the location of the crossover points (cruise track 
intersections) where comparisons of the reproducibility of the TCO2 analyses 
were made. The six survey groups measured two water column carbonate system 
parameters, total dissolved carbon dioxide (TCO2) and total alkalinity (TA), and 
assisted with the operation of an underway pCO2 (surface) system. This paper 
focuses on TCO2 by coulometry, while the total alkalinity (TA) and partial 
pressure of CO2 (pCO2) measurements are the subject of companion papers and 
reports (Millero et al., 1998; Sabine and Key, 1998).


Fig. 1.  The cruise tracks for the nine legs of the US Indian Ocean WOCE Survey 
         1994-1996. Crossover points between the various legs are marked with a 
         square and numbered. These intersection points and crossovers are 
         referred to in Table 4.


TABLE 1: Approximate dates and ports of call for the 9 legs of the Indian Ocean 
         CO2 Survey, and the measurement groups responsible for the 
         determination of the carbonate system parameters
______________________________________________________________________________

 Leg            Dates          From        To          Group         Duration 
           Start     End                                             (days) 
 --------  --------  --------  ----------  ----------  ------------  --------
 I8SI9S    12r1r94   1r19r95   Fremantle   Fremantle   BNL             50 
 I9N       1r24r95   3r6r95    Fremantle   Colombo     Princeton U.    42 
 I8NI5E    3r10r95   4r16r95   Colombo     Fremantle   U. of Hawaii    38 
 I3        4r20r95   6r7r95    Fremantle   Port Louis  U. of Miami     49 
 I5WI4     6r11r95   7r11r95   Port Louis  Port Louis  BNL             31 
 I7N       7r15r95   8r24r95   Port Louis  Matrah      U. of Hawaii    41 
 I1        8r29r95   10r18r95  Matrah      Singapore   WHOI            51 
 Dry Dock  10r19r95  11r5r95   Singapore                               17 
 I10       11r6r95   11r24r95  Singapore   Singapore   Princeton U.    19 
 I2        11r28r95  1r19r96   Singapore   Mombasa     U. of Hawaii    53 
______________________________________________________________________________
 Abbreviations: BNL, Brookhaven National Laboratory; U, University; 
                WHOI, Woods Hole Oceanographic Institution. 



2. MATERIALS AND METHODS

2.1. PREPARATIONS

The total carbon dioxide concentration (TCO2) was determined using two single-
operator multiparameter metabolic analyzers (SOMMA) each connected to a Model 
5011 coulometer (UIC, Joliet, IL 60434). Descriptions of the SOMMA-coulometer 
system and its calibration can be found in the works of Johnson (1995), Johnson 
and Wallace (1992), and Johnson et al. (1987) and Johnson et al. (1993). A 
schematic diagram of the SOMMA is shown in Fig. 2, and further details 
concerning the coulometric titration can be found in the works of Huffman (1977) 
and Johnson et al. (1985). Briefly, seawater fills an automated to-deliver 
sample pipette. The contents of the pipette are pneumatically injected into a 
stripping chamber containing approximately 1.2 cm3 of 8.5% (v/v) phosphoric 
acid, and the resultant CO2 is extracted, dried, and coulometrically titrated. 
Calibration is performed by titrating known masses of pure CO2 and checked by 
analyzing certified reference material (CRM). The coulometers were adjusted to 
give a maximum titration current of 50 mA, and they were run in the counts mode 
(the number of pulses or counts generated by the coulometer's voltage to 
frequency converter during the titration is displayed). In the coulometer cell, 
the acid (hydroxyethylcarbamic acid) formed from the reaction of CO2 and 
ethanolamine is titrated coulometrically (electrolytic generation of OH¯) with 
photometric endpoint detection. The systems were equipped with conductance cells 
(Model SBE-4, Sea-Bird Electronics, Bellevue, WA) for measuring salinity as 
described by Johnson et al. (1993).


Fig. 2.  SOMMA-coulometer system schematic. Carbon dioxide extracted from a 
         water sample (I) or from volume-calibrated gas sample loops filled at a 
         known pressure and temperature is degassed from the stripper (IV), 
         dried (V), and coulometrically titrated (VI). The water sample is 
         pneumatically injected from the pipette (II) into the stripper, and the 
         pure CO2 contained in the gas loops is delivered to the stripper from 
         an 8-port chromatography valve (VII) equipped with pressure and 
         temperature sensors. Salinity is measured using a conductance cell 
         (III) integrated into the SOMMA chassis. The pipette and conductance 
         cell are thermostatted and equipped with temperature sensors.


The DOE supported the construction of nine SOMMA-coulometer systems for the US 
CO2 Survey Measurement Groups in the early 1990's (Johnson and Wallace, 1992), 
and two of these systems from the DOE instrument pool were set up aboard the R/V 
Knorr in Fremantle, Australia on November 28, 1994. Before they were shipped to 
Australia, the temperature sensors were calibrated, the glassware was chemically 
cleaned and gravimetrically calibrated, the gas sample loop volumes were 
calibrated according to the procedure of Wilke et al. (1993), the coulometers 
were electronically calibrated (Johnson et al., 1993; DOE, 1994), and system 
accuracy was verified with CRM at Brookhaven National Laboratory (BNL). The same 
two systems (hereafter called I and II) were used by all measuring groups. A 
backup system (from Woods Hole Oceanographic Institution) was onboard but was 
not used. Pre-cruise preparations also included a training session for 
participants at the University of Miami in September 19-23, 1994.

Referring to Fig. 2, the analytical gases included UHP nitrogen (99.998%) for 
carrier and pneumatic gases, compressed air for the headspace gas, and 
analytical grade CO2 (99.995%) from Scott Speciality Gases (South Plainfield, 
NJ) for the calibration gas. The survey began with the use of compressed gases, 
but prior to leg I8N in April 1995, a N2 generator (TOC Model 1500, Peak 
Scientific, Chicago, IL) was placed into service. The generator provided N2 
(99.9995%, hydrocarbons<0.1 ppm, CO2<1.0 ppm) for carrier and pneumatic gases to 
both systems for the remainder of the survey. Unless otherwise stated, all other 
reagents remain as described by Johnson et al. (1993).

The BNL measurement group supplied 7 side-arm type glass titration cells (UIC, 
PN 200-034), 7 silver electrodes (PN 101-033), and 5 rubber cell caps (PN 192-
005). A platinum electrode (PN 101-034), temperature sensor (PN LM34CH, National 
Semiconductor, Santa Clara, CA), and a teflon inlet tube were mounted in each 
cap. Together, the cell and cap comprise the cell assembly shown in Fig. 3. For 
this paper, each cell assembly is assigned an 'age' or lifetime which is 
measured in minutes (chronological age) or by the mass of carbon titrated in mg 
C (carbon age) from the time when current is first applied to the assembly (cell 
birth) until the current is turned off (cell death). The software continuously 
records the chronological and carbon ages.


Fig. 3.  The titration cell assembly and the cathodic and anodic half reactions 
         for the coulometric titration of the H+ from the acid formed by the 
         reaction of CO2 and ethanolamine.


2.2. SELECTION OF CELL ASSEMBLIES

The performance of individual cell assemblies (Fig. 3) varies widely (K.M. 
Johnson, unpublished data). Unacceptable assemblies exhibit high blanks, 
prolonged blank determinations (>2 h), reduced accuracy or precision, or become 
noisy early in their lifetime. Acceptable assemblies stabilize quickly (within 
60 min) and function well for periods exceeding 24 h. Cell behavior will be 
discussed elsewhere, but our experience suggests several factors play a role: 
quality of the reagents; quality (purity) of the carrier gases; damage to the 
platinum electrode; and perhaps the porosity of the cell frit. Therefore, a 
systematic effort was made at the beginning of leg I8SI9S to select 
satisfactorily performing cell assemblies using pretested reagents and carrier 
gas sources known to be uncontaminated. During this first leg, the assemblies on 
hand were evaluated for conformance to the following empirical criteria.

(1) Cell assemblies should attain a blank of ≤0.005 µmol C/min within 90 min of 
    cell birth. Satisfactory assemblies usually exhibit a 15-25% decline in the 
    blank with each successive determination.

(2) The gas calibration factor, which is the ratio of an accurately known mass 
    of CO2 to the mass of this gas determined coulometrically, should be 
    1.004±0.0015 (recoveries of 99.6%).

(3) Titrations of CO2 extracted from gas sample loops (gas calibration) or 
    pipettes of 20 cm3 (sample analysis) should take 9-12 min.

(4) Cell assemblies, which repeatedly exhibit titrations longer than 20 min (no 
    endpoint) before their carbon age reached 30 mg C titrated, were considered 
    defective. An occasional failure to attain an endpoint after the carbon age 
    exceeds 30 mg C was interpreted to mean that the cell frit required cleaning 
    with 6 N HNO3 and retesting.

Based on these criteria, three assemblies (2 primary and a third as backup) were 
found to be acceptable during the first leg, and these assemblies were used 
throughout the survey (at the midpoint of the survey an additional assembly was 
placed into service).


2.3. AT-SEA OPERATIONS

The following TCO2 sampling and measurement practices were followed throughout 
the survey.

(1) The daily sequence of analytical operations for each system as described in 
    the SOMMA operator's manual (Johnson, 1995) consisted of changing the cells 
    and drying agents, determining the blank, running test seawater samples, 
    calibrating the system using pure CO2 (gas calibration), analyzing samples, 
    and analyzing certified reference material (CRM) at the beginning and end of 
    the cell lifetime.

(2) A complete deep vertical profile for TCO2 and TA consisted of 36 samples. A 
    lesser number of samples were drawn at shallower stations. Complete profiles 
    were taken at every other station, and if time permitted, additional 
    truncated profiles (0-1000 m) were taken. TCO2 samples always coincided with 
    14C samples. Samples were drawn from 10-l Niskin bottles according to DOE 
    (1994).

(3) Samples for TCO2 were collected in 300 cm3 BOD-type glass bottles. They were 
    poisoned with a saturated HgCl(2) solution (200-400 µl) upon collection. The 
    appropriate correction factors for dilution were applied by the measurement 
    groups according to DOE (1994).

(4) Sample bottles were rinsed and then allowed to overflow by at least 1/2 
    volume before poisoning. Prior to April 1995, a glass stopper was inserted 
    into the full BOD bottle. After April 1995, a headspace of approximately 4 
    cm3 was created before poisoning and stoppering. This was done in a 
    reproducible manner by squeezing the filling tube shut before withdrawing it 
    from the bottle. This change was made to ensure that no HgCl(2) was 
    displaced by the stopper, and to allow for water expansion. The gas-liquid 
    phase ratio was approximately 1.3%. A correction (±0.5 µmol/kg) for the 
    reequilibration of the liquid with the gas phase was applied by the 
    measurement groups according to DOE (1994).

(5) To estimate sample precision, duplicate samples were normally collected at 
    surface, mid depth, and at the deepest depth. The duplicate analyses were 
    interspersed with the analysis of the other profile samples with a minimum 
    of 2 h and up to 12 h between duplicate analyses. Because the duplicate 
    analyses were separated in time, these data could potentially detect drift 
    (decreased precision) as the cell aged. Every effort was made to run each 
    station profile on a single cell assembly, and to limit the cell lifetime to 
    ≤35 mg C.

(6) Although salinity was determined by the SOMMA-coulometer systems, post-
    cruise sample density was calculated using bottle salinities supplied by the 
    chief scientists. However, SOMMA-based salinities were often compared to the 
    real-time CTD salinities to spot bottle mistrips during the taking of the 
    vertical profiles. The agreement between SOMMA-based and CTD salinities was 
    ±0.02 or better.

(7) To monitor the volume of the SOMMA pipettes, they were periodically filled 
    with deionized water at known temperatures, and their output collected in 
    preweighed serum bottles. The bottles were sealed immediately and stored 
    until they were reweighed at BNL on a model R300S (Sartorius, Göttingen, 
    Germany) balance. The mass of water corrected for buoyancy was used to 
    calculate the to-deliver pipette volume (V(cal), Eq. 3) according to DOE 
    (1994).

(8) After use, cells were cleaned with deionized water followed by an acetone 
    rinse of the glass frit. Before reuse, they were dried at 55°C for at least 
    12 h. Cell caps and the platinum electrodes were thoroughly washed with 
    deionized water and dried at 55°C for at least 6 h before reuse.

(9) Duplicate samples from approximately 3000 m and 20 m were regularly 
    collected for shore-based reference analyses of TCO2 by vacuum 
    extraction/manometry by C.D. Keeling at the Scripps Institution of 
    Oceanography (SIO). Between 2 and 5% of the samples analyzed at sea will be 
    analyzed at SIO and reported elsewhere.


2.4. CALCULATION OF RESULTS

For the coulometric determination, the mass of carbon titrated from CO2 
extracted from the gas sample loops or a water sample in µmol of carbon is given 
by M according to:

      M=[Counts/4824.45-(Blank x t(t))-(Int(ec) x t(i))]/Slope(ec),     (1)

where Counts is the coulometer display, i.e., the number of pulses accumulated 
by the coulometer's voltage to frequency circuit (VFC); 4824.45 (counts/µmol) is 
a scaling factor derived from the factory calibration of the VFC and the value 
of the Faraday (96,485.309 C/mol); Blank is the system blank in µmol/min; t(t) 
is the length of the titration in minutes; Int(ec) is the intercept from the 
electronic calibration of the coulometer; ti is the duration (min) of continuous 
current flow, and Slope(ec) is the slope from electronic calibration (Johnson et 
al., 1993; DOE, 1994). Electronic calibration serves as a check of the factory 
calibration. If the coulometer was perfectly calibrated, the slope and intercept 
would be 1 and 0, respectively. Typically, minor deviations from the theoretical 
slope (0.998-0.999) and intercept (0.001-0.01) are observed. The water sample 
TCO2 concentration in µmol/kg is calculated from:

               TCO2 = M x Calibration Factor x 1/(V(t)p))D+∆TCO(2),     (2)

where VT is the sample volume (to-deliver volume of the SOMMA pipette) 
calculated from:

                          V(T)=V(cal)[1+a(v)(T-T(cal)],                 (3)

and T is the analytical temperature; V(cal) is the calibrated volume of the 
pipette at the calibration temperature, T(cal); av is the coefficient of 
volumetric expansion for Pyrex glass (1.0 x 10(^-5)/deg). In Eq. 2, Calibration 
Factor is the gas calibration factor (see Eq. 4); p is the density of seawater 
from the seawater equation of state (Millero and Poisson, 1981) at the sample 
salinity and T; D is the correction due to dilution of the sample with HgCl(2) 
preservative; ∆TCO2 is the correction for the repartitioning of CO2 into the 
sample headspace according to DOE (1994). Note that correction factors D and 
∆TCO2 (Eq. 2) are not incorporated into the SOMMA software and were applied post 
cruise by the individual measurement groups.

The gas calibration factor (Calibration Factor) is the ratio of:

                                   M(calc)/M,                           (4)

where M(calc) is the mass of CO2 contained in the gas sample loop calculated 
according to DOE (1994), and M is the coulometric determination of that same 
mass from Eq. 1.


2.5. ASSESSMENT OF ANALYTICAL ACCURACY

Analytical accuracy was assessed by analyzing certified reference materials 
(CRMs). The CRMs are filtered seawater poisoned with HgCl(2). They are prepared 
in 500 cm3 bottles at the Scripps Institution of Oceanography (SIO) according to 
procedures given by Dickson (1990). The certified TCO2 value is obtained by 
analyzing a representative number of samples by vacuum extraction/manometry in 
the laboratory of C.D. Keeling at SIO. For this paper, the term analytical 
difference refers to the difference between the analyzed (by coulometry) and the 
certified value of the CRM (by manometry), i.e., at-sea accuracy is estimated 
from the analyzed TCO2-certified TCO2 differences.


2.6. DATA DISTRIBUTION

The complete data set has been submitted to the Carbon Dioxide Information 
Analysis Center (CDIAC) at the Oak Ridge National Laboratory (ORNL). CDIAC will 
issue a final data report which will detail the procedures for retrieving the 
data. The overall accuracy given below is considered final at this time, and the 
estimated precision is expected to remain unchanged. The CDIAC web address is 
http://cdiac.esd.ornl.gov.



3. RESULTS

During the survey, approximately 18,828 separate samples (not counting dupli-
cates) for TCO2, and 983 CRM were analyzed on the two systems (A. Kozyr, 
personal communication, November 1997).


3.1. TO-DELIVER PIPETTE VOLUME

Some 103 gravimetric determinations of the sample pipette volume were made on 28 
separate occasions during the survey (14 on each system). Four of the determina-
tions were rejected; two because they were exactly 1 cm3 too high with respect 
to the survey mean (likely due to failure to correctly note the tare weight 
determined prior to the cruise), and two because they were inexplicably 0.3% 
lower than the survey mean volumes (probably due to faulty sealing and evapora-
tion). There were no results from leg I8N because the gravimetric samples were 
collected incorrectly. Volume determinations should have been made at the start, 
middle, and at the end of each leg, or at least at the beginning and end of each 
leg. However, for a variety of reasons, this was not always the case. In order 
to consistently assign a pipette volume to each leg, a leg-specific volume 
(V(cal)) was obtained by averaging the volume determinations made closest to the 
beginning and end of the leg along with any made during that leg. Table 2 
presents the results for V(cal), and the chronological order of the pipette 
determinations used to calculate V(cal) are plotted in Fig. 4a for system I and 
Fig. 4b for system II. This averaging increases the number of determinations 
used to calculate V(cal), and ensures that V(cal) is based on at least two sets 
of determinations, separated in time, for all legs except the initial leg 
(I8SI9S) and leg I10 after the pipette was cleaned. Table 2 and Fig. 4a and b 
show the timing of events which could conceivably have affected pipette volume. 
For I8SI9S, the pipette volumes were determined in the laboratory prior to the 
cruise; however, the volume of system I had to be empirically redetermined at-
sea because its pipette was broken during transit. This was done as follows: 
after replacing the pipette, V(cal) was determined by simultaneously analyzing a 
replicate from a single seawater sample on systems I and II. Because V(cal) was 
well known for system II, the TCO2 concentration determined on system II was 
used to calculate the pipette volume of system I by rearranging Eq. 2 to solve 
for VT and letting VT be equal to V(cal) for the subsequent analyses on system I 
during leg I8SI9S. As Table 2 shows, numerous volume determinations were made 
for both systems I and II on succeeding legs.


TABLE 2: The leg-specific to-deliver pipette volume (V(cal)) and the calibration 
         temperature (T(cal)) for SOMMA-coulometer systems I and II during the 
         Indian Ocean Survey 1994-1996
______________________________________________________________________________

 Leg        n     V(cal)       S.D.     R.S.D.    T(cal)    Determinations 
                  (cm^3)      (±cm^3)    (%)       (C)      averaged (legs)
 ------    --     -------     ------    ------    -----     -----------------
 System I 
 I8SI9S     2     21.4609     0.0037     0.02     20.00     see text, 8S9S(e) 
 I9N        9     21.4543     0.0112     0.05     20.97     8S9S(e), 9N(e)
 Gas generator introduced as CG source
 I8NI5E     9     21.4443*    0.0021     0.01     20.97     9N(e), 3(m)
 I3        15     21.4471     0.0042     0.02     20.57     9N(e), 3(m), 4(s)
 Gas generator output pressure adjusted from 5 to 10  psi     
 I5WI4     10     21.4506*    0.0023     0.01     19.93     5W4(s,e) 
 I7N        8     21.4506     0.0032     0.02     20.36     7N(s,m,e) 
 I1         5     21.4462     0.0074     0.03     20.12     7N(e), 1(e) 
 Pipette dismounted,cleaned,and recalibrated     
 I10        5     21.4460     0.0110     0.05     20.08     10(e)     
 I2         8     21.4482     0.0091     0.04     20.08     10(e), 2(s,e)

 System II 
 I8SI9S    18     21.6388     0.0068     0.03     20.24     8S9S(s,e) 
 I9N        9     21.6360     0.0163     0.08     20.49     8S9S(e), 9N(e) 
 Gas generator introduced as CG source     
 I8NI5E     8     21.6239     0.0080     0.04     20.56     9N(e), 3(m)
 I3        14     21.6243     0.0068     0.03     20.31     9N(e), 3(m), 4(s)
 Gas generator output pressure adjusted from 5 to 10 psi 
 I5WI4     11     21.6293     0.0068     0.03     19.97     5W4(s,e) 
 I7N        8     21.6194*    0.0048     0.02     20.05     7N(s,m,e) 
 I1         4     21.6156     0.0035     0.02     20.00     7N(e), 1(e) 
 Pipette dismounted,cleaned,and recalibrated     
 I10        4     21.6269*    0.0017     0.01     19.95     10(e) 
 I2         9     21.6270     0.0028     0.01     19.94     10(e),2(s,e) 
______________________________________________________________________________
 The subscripts (s, m, or e) for the pipette volume determinations averaged 
 to calculate V(cal) signify determinations made at the start, middle, or end 
 of a leg, respectively. Values of V(cal) which are significantly different 
 from the V(cal) of the preceding leg are denoted by the asterisk.


Fig. 4.  The temporal record of the analytical performance of SOMMA-coulometer 
         system I (Fig. 4a) and II (Fig. 4b) during the Indian Ocean Survey 
         1994-1996. The top section of the three-part graphs shows the leg-
         specific pipette volumes, V(cal), as horizontal lines corresponding to 
         the duration of the individual legs, and the relative chronological 
         order of the means of the individual pipette determinations from which 
         V(cal) was calculated as open circles placed before, in the middle of, 
         or following the horizontal lines representing V(cal) (see text and 
         Table 2 for details). The middle section depicts the mean gas calibra-
         tion factors for each leg (horizontal lines), and the bottom section 
         shows the mean analytical differences for the CRM analyses assuming a 
         constant pipette volume (V(cal) for leg I8S) for the duration of the 
         survey (open circles) vs. the leg-specific V(cal) (closed circles). The 
         error bars through the plot symbols represent the S.D. of the determi-
         nations. Procedural changes (introduction of the gas generator, 
         pressure adjustments, and cleaning) which may have affected pipette 
         volume are indicated by the arrows.


For I10, data from the prior leg could not be used to calculate V(cal) because 
leg I10 took place after the pipettes had been dismounted for cleaning, which 
may have altered their volumes. On legs I5WI4 and I7N, replicate volume 
determinations were made at the beginning, middle, and end of the leg by the 
same measuring group so that V(cal) for these legs do not include results from 
preceding or succeeding legs. The survey mean pipette volumes and their standard 
deviations for systems I and II are 21.4502±0.0032 cm3 at 20.25°C (n=43) and 
21.6261±0.0028 cm3 at 20.14°C (n=56), respectively. The pooled standard 
deviation (sp^2) calculated according to Youden (1951) for the 28 sets of 
gravimetric determinations is ±0.0042 cm3. Individually, sp^2 for system I is 
±0.0049 cm3, and for system II sp^2 is ±0.0036 cm3, suggesting a very slightly 
higher precision for system II.

Significant differences at the 95% confidence level in V(cal) for comparisons 
between each leg with the succeeding leg were determined by two-tailed t-tests 
according to Taylor (1990), and are denoted by asterisks in Table 2. For the 
most part, leg to leg differences in V(cal) are not significant (significance in 
2 of 9 comparisons for each instrument), but it should be noted that for both 
systems, the differences between the initial leg (I8SI9S) pipette volumes and 
all leg-specific volumes after leg I9N are significant. In both systems, the to-
deliver pipette volume declines slightly with time. However, the decline is not 
consistent between instruments. In system I, significant decreases in volume 
appear earlier in the survey and may be correlated with the switch to the N2 
generator and a documented generator outlet pressure adjustment, but this is not 
the case with system II where dismounting and cleaning of the pipette late in 
the survey may have had the greatest effect.


TABLE 3: A summary of the mean analytical parameters and mean analytical 
         differences for the three batches of CRM analyzed on SOMMA-coulometer 
         systems I and II during the Indian Ocean Survey 1994-1996
_____________________________________________________________________________

 Leg       Slope     Int     Cal-    CRM   Precision,  Analytical difference
           (ec)      (ec)   factor  batch  n(±µmol/kg    const-vp/corr-vp
 -------   ------  -------  ------  -----  ----------  ---------------------
 System I 
 I8SI9S    1.0002   0.0008  1.0043   23     1.15(54)        -0.41/-0.41 
 I9N       1.0007   0.0013  1.0045   23     0.86(71)        -0.83/-0.20 

 I8NI5E    1.0007   0.0013  1.0062   23     1.36(55)        -1.71/-0.15 
 I3        1.0007   0.0013  1.0053   23     0.98(37)        -2.33/-1.31 
 I3        1.0007   0.0013  1.0053   26     0.98(20)        -2.77/-1.72 

 I5WI4     0.9998  -0.0057  1.0041   26     1.31(41)        -1.83/-0.88 
 I7N       0.9998  -0.0057  1.0043   23     1.71(6)         -1.66/-0.69 
 I7N       0.9998  -0.0057  1.0043   26     1.88(55)        -1.74/-0.78 
 I7N       0.9998  -0.0057  1.0043   27     0.88(8)         -2.91/-1.95 
 I1        0.9998  -0.0057  1.0038   27     1.10(64)        -2.82/-1.45 

 I10       0.9998  -0.0057  1.0037   27     0.72(32)        -0.58/-0.58 
 I2        0.9998  -0.0057  1.0040   27     1.11(27)        -0.57/-0.77 
 Mean                       1.0045          1.17(470)       -1.68/-0.91 
 S.D.(±)                    0.0008          0.35             0.92/ 0.58 

 System II 
 1I8SI9S   0.9996  -0.0025  1.0041   23     1.18(104)       -0.89/-0.89 
 I9N       0.9996  -0.0025  1.0039   23     0.90(70)        -1.83/-1.57 

 I8NI5E    0.9996  -0.0025  1.0041   23     1.14(59)        -1.73/-0.35 
 I3        0.9996  -0.0025  1.0045   23     0.85(35)        -2.14/-0.62 
 I3        0.9996  -0.0025  1.0045   26     0.69(13)        -2.44/-1.11 
 
 I5WI4     0.9998   0.0045  1.0050   26     0.79(41)        -2.14/-1.28 
 I7N       0.9998   0.0045  1.0051   23     0.88(5 )        -3.25/-1.47 
 I7N       0.9998   0.0045  1.0051   26     0.84(54)        -2.09/-0.32 
 I7N       0.9998   0.0045  1.0051   27     0.77(10)        -2.88/-1.10 
 I1        0.9998   0.0045  1.0041   27     1.11(70)        -3.51/-1.38 
 
 I10       0.9998   0.0045  1.0038   27     0.65(28)        -0.66/-0.66 
 I2        0.9998   0.0045  1.0035   27     1.11(24)        -1.38/-1.39 
 Mean                       1.0042          0.91(513)       -2.08/-1.01 
 S.D.(±)                    0.0005          0.18             0.87/ 0.44 
_____________________________________________________________________________
 For each CRM batch analyze d, precision is given as the standard deviation of 
 the mean of (n) analyses. Abbreviations: ec, electronic calibration; calfactor, 
 gas calibration factor; Int, intercept; const-vp, mean analytical difference 
 calculated using a constant pipette volume; corr-vp, mean analytical difference 
 calculated using the leg-specific V(cal) (Table 2). 
  (a) Gas Generator introduced as CG source. 
  (b) Gas generator output pressure adjusted from 5 to 10 psi. 
  (c) Pipette dismounted, cleaned and recalibrated.


3.2. CRM ANALYSES AND SYSTEM ACCURACY

In addition to the leg-specific pipette volumes, Fig. 4a (system I) and Fig. 4b 
(system II) show the mean analytical differences (analyzed TCO2-certified TCO2) 
and the mean gas calibration factors for each survey leg. The plots are scaled 
so that each Y-axis spans a similar range in order that the factors controlling 
system accuracy can be more readily identified. These data are also tabulated 
and summarized in Table 3. Table 3 shows that the gravimetric volume 
determinations (Table 2) have detected real changes in V(cal) during the survey. 
The mean analytical differences calculated with the corrected pipette volumes 
(corr-vp, Table 3) are -0.91 and -1.01 µmol/kg for systems I and II, 
respectively. If the pipette volumes determined at the beginning of the survey 
(const-vp) were used, the corresponding differences would be -1.61 and -2.08 
µmol/kg, showing that the routine determination of pipette volume increased 
accuracy by a factor of ~2.

Fig. 5 is a bar chart of the mean analytical difference (accuracy) for systems I 
and II as a function of cell carbon age. Both systems behave very similarly with 
the best precision and accuracy early in the cell lifetime (<10 mg C), 
increasing differences for cells of intermediate ages (>10 to <30 mg C), and 
smaller differences for carbon ages exceeding 30 mg C which are not 
significantly different from those at ages <10 mg C. No corrections based on the 
analyzed-certified TCO2 differences or cell age have been applied to the CDIAC 
data set.


Fig. 5.  A plot showing the distribution of mean analytical differences for CRM 
         analyses vs. coulometer cell age for SOMMA-coulometer systems I (open 
         bars) and II (filled bars) during the Indian Ocean Survey 1994-1996. 
         The error bars represent the 95% confidence interval for the mean 
         differences, and the numbers inside the columns are the number of 
         measurements (n) used to compute the means.


3.3. SYSTEM REPEATABILITY AND PRECISION DURING THE SURVEY

For the survey as a whole, the operating conditions and analytical performance 
of the two SOMMA systems were virtually identical. Survey-wide the mean gas 
calibration factors of the two systems were nearly identical (1.0045 for system 
I compared to 1.0042 for II). While both systems yielded slightly negative (~1.0 
µmol/kg) mean analytical differences (Table 3), the standard deviation of the 
analytical differences was slightly better on system II (±0.91 µmol/kg) than 
system I (1.17 µmol/kg). This is consistent with the gravimetric volume 
determinations where system II also exhibited a slightly higher precision 
(sp^2=±0.0036 cm^3 vs. ±0.0049 cm^3 for system I).

For the CRM analyses, the precision or pooled standard deviation (sp^2) 
calculated according to Youden (1951) is 1.19 µmol/kg (df=977). For this 
calculation, the three batches of CRM analyzed on the two systems are treated as 
six separate samples with multiple replicates. Because sp^2 includes CRM data 
measured on both systems on all legs, it applies to both systems on all legs. 
For water samples, sp^2 was calculated from duplicates analyzed on each system 
during leg I8SI9S at the start of the survey and leg I5WI4 about half way 
through the survey. The sp^2 for leg I8SI9S is ±1.26 µmol/kg (df=15), and for 
leg I5WI4, sp^2 is ±0.91 µmol/kg (df=21). These values are consistent with the 
precision of the CRM analyses given in Table 3. For the survey, the overall 
precision of the TCO2 determination is ±1.19 µmol/kg.

Fig. 6 is a plot of the analytical differences by system and CRM batch for the 
entire survey. The differences, calculated using the parameters in Table 3, 
reiterate the point that there are no significant analytical differences (bias) 
between systems or between CRM batches.


Fig. 6.  The analytical differences for the CRM analyses made on SOMMA-
         coulometer systems I and II during the Indian Ocean Survey 1994-1996 
         with separate symbols for the results from the two systems and for the 
         three batches of CRM analyzed. The beginning and end of each leg is 
         marked by vertical dashed lines. The respective salinities and 
         certified TCO2 (µmol/kg) for batches 23, 26, and 27 are 33.483 and 
         1993.10, 33.258 and 1978.34, 33.209 and 1988.10 µmol/kg.



4. DISCUSSION

The Indian Ocean CO2 Survey differed from the previous DOE CO2 Survey efforts in 
that a single ship was used for all legs, and that the measurement groups shared 
the same analytical equipment. The latter included the use of a single cache of 
coulometric reagents (two different lot numbers both of which were tested pre-
cruise with CRM), invariant sources of analytical gases, use of the same 
titration cell assemblies, standard sampling procedures, and standardized 
software. There was a pre-cruise training session, and all of the participants 
had prior experience with the sampling and measurement techniques (poisoning, 
reagent concentrations, standard calculations, glassware calibration, etc.) 
documented in the DOE Handbook of Methods (DOE, 1994). Thus, an extraordinary 
effort over several years to ensure analytical quality and uniformity culminated 
in the procedures used during the Indian Ocean Survey.

An improvement in system accuracy (Table 3) of approximately 1 part in 2000 
shows that the effort to gravimetrically determine the pipette volumes on each 
leg was worthwhile. The volumes of both systems did decrease slightly but 
significantly with time. Possible explanations include pressure changes in the 
carrier gas source (system I) or fouling of the glass pipette walls causing 
altered surface tension or displacement of small amounts of liquid (system II). 
Because the samples were poisoned with HgCl(2), it is unlikely that biological 
fouling was a problem, but the high quantity of grease used to seal the CRM 
bottles makes it possible that some of this grease found its way into the 
pipettes. After cleaning, V(cal) for leg I10 remained unchanged compared to the 
preceding leg I1 on system I and increased slightly on system II, but for both 
systems it was significantly smaller than the V(cal) determined for the initial 
leg (I8SI9S). After cleaning, the mean analytical difference for leg I10 (system 
I and II, n=2) was -0.62 µmol/kg compared to -0.40 µmol/kg on the initial leg 
I8S when the instruments were fresh from the laboratory indicating that pipettes 
were most accurate after cleaning. Whatever the cause of subsequent volume 
changes, the data confirm the importance of periodically redetermining the 
volume, and indicate that the procedure is mandatory for the highest accuracy 
over extended periods of analytical work and/or after major changes in system 
plumbing. In aggregate, both systems share a small negative analytical 
difference (-1.0 µmol/kg) for the CRM analyses throughout the survey even after 
pipette volume corrections have been applied.

The cell accuracy vs. carbon age relationship shown in Fig. 5 is typical of data 
from previous cruises (K.M. Johnson, unpublished data). The best precision and 
accuracy is found at a carbon age of 5-10 mg C, a slightly reduced accuracy 
(usually as lower recoveries of CRM carbon) is observed between 10-30 mg C, 
gradually increasing recoveries and imprecision after 30 mg C until cell death 
where cell death is defined as a positive difference ≥3.0 µmol/kg. This behavior 
underlies the recommendation that cell lifetimes be limited to a carbon age of 
≤35 mg C, i.e., to limit imprecision and because cell death normally occurs at 
carbon ages ≥35 mg C. During the survey, neither CRM or samples were run until 
the carbon age exceeded 5 mg C. This was accomplished by configuring the 
software to automatically run a test sample and three consecutive gas 
calibrations before samples were analyzed. The reasons for the observed cell 
behavior are not understood, but limiting cell lifetimes from ≥5 to ≤35 mg C 
probably helps to limit system drift which might compromise the sample analyses. 
Although the imprecision associated with cell aging is small and cell failure is 
rare at carbon ages ≤35 mg C, good analytical practice requires that samples 
should be run in random order rather than systematically in order of depth to 
avoid systematic biases which might result from any drift associated with cell 
age.

Fig. 4a and b shows no correlation between the gas calibration factors and the 
analytical differences after the CRM analyses were corrected for pipette volume 
changes (Table 3). These data do show that the overall mean gas calibration 
factor for both systems is nearly the same (1.004), but that the temporal record 
with respect to gas calibration factor variation is not. Calibration factor 
variation (R.S.D.=0.06-0.08%) is greater than the variation in V(cal) 
(R.S.D.=0.03%), and is therefore a potentially more important control on system 
accuracy. For system I, the highest mean gas calibration factor (poorest 
recovery of CO2) was observed on leg I8N, while for system II, the corresponding 
result occurred months later, on leg I7N (same measurement group, see Table 1). 
Because the system calibration factors are not correlated with the analytical 
differences, the observed variations in calibration factors are real, i.e., they 
document a change in system response shared by the calibration and sample 
analyses rather than an isolated malfunction of the gas calibration hardware 
(see Fig. 2).

The reason for gas calibration factor variation is not known. It could 
conceivably be due to procedures unique to each measurement group, e.g., 
positioning of the cathode electrode and the gas inlet tube with respect to the 
coulometer light source and photodetector (Fig. 3), plumbing differences 
resulting in leaks and small losses of CO2, or the amount of reagents used to 
dry the gas stream (Fig. 2). These procedural differences would affect sample 
determinations and gas calibration results similarly because, as Fig. 2 shows, 
the calibration gas follows the same route to the coulometer as the CO2 
extracted from samples. Table 3 suggests at least one other possible cause of 
gas calibration factor variation. The coulometers were electronically calibrated 
by the BNL group at the start of the survey (I8SI9S) and about half way through 
the survey on leg I5WI4. Between legs I8SI9S and I5WI4 the coulometer 
calibration appears to have changed by 0.08% for system I, and by 0.02% for 
system II. These calibrations were separated by many weeks so the exact 
magnitude or timing of the shift is not known. Changes in the coulometer's 
circuitry affecting the electronic slope (Slope(ec)) and intercept (Int(ec)) 
would alter the gas calibration factor but would not affect system accuracy 
because, until recalibration, the previous electronic calibration coefficients 
represent constants in Eq. 1. In both systems, the sense of the apparent change 
in electronic calibration coefficients compared to the earlier coefficients is 
qualitatively consistent with the observed short-lived variation in gas 
calibration factors, and it is possible that this variation was due to 
unexplained changes in the coulometer response.

The important point is the efficacy of the gas calibration procedure: 
corrections to data based solely on the CRM analyses which would usually be 
applied on a cruise-average basis may mask short term variation or step changes 
in system response arising from stochastic or procedural changes. The gas 
calibration procedure, in which known masses of pure CO2 are regularly analyzed, 
is an independent check of all system components except pipette volume, and it 
provides traceable documentation for the subsequent survey results.

The importance of cell assembly selection should be stressed. Investigators have 
found that the behavior of individual cell assemblies can vary significantly 
(e.g., D. Chipman, personal communication, July 1996). The factors affecting 
cell performance are still not yet completely understood. Hence, the use of 
empirical selection criteria such as those given in Section 2 is recommended. It 
is beyond the scope of the paper to go into detail, but point 'a' in Fig. 3 
illustrates one of the locations for potential problems. A faulty seal where the 
platinum electrode emerges from the glass insulator could allow infiltration and 
trapping of the cell solution in the insulator where electrochemical or chemical 
reactions could take place. Small quantities of this solution (at a pH different 
from the bulk cell solution) could randomly exchange with the bulk cell solution 
and cause titration errors. This would be difficult to detect. Assemblies which 
did not meet the empirical performance criteria in Section 2 were simply not 
used. The attention to cell assembly testing and selection is believed to a 
major reason for the success of the Indian Ocean TCO2 Survey. The survey 
assemblies were also carefully washed and dried. Drying at 55±5°C removes 
traces of the volatile and reactive cell solution from the rubber caps.


TABLE 4: Results of the crossover analysis (see text for details)
__________________________________________________________________

 Crossover  Expedition legs       Stations        TCO2 difference
 no.        Late      Early   Late       Early    ±S.D.(µmol/kg) 
 ---------  ----------------  ------------------  ---------------
     1      I1        I7N     927:931    780:784     -2.5±0.5 
     2      I1        I9N     987:990    266:270     -2.7±6.3(a) 
     3      I1        I9N     996:998    233:235     -0.9±1.7 
     4      I2        I7N     1205       728:730     -0.4±1.1 
     5      I2        I8NI5E  1137:1139  320:324      1.5±1.5 
     6      I2        I9N     1094:1096  191:193     -3.0±0.7 
     7      I2        I10     1078       1075        -1.5±1.5 
     8      I5WI4     I3      705        547:549      1.6±0.5 
     9      I3        I8NI5E  498:501    346:348     -2.6±0.7 
    10      I3        I9N     472        169          1.1±1.2 
    11      I10       I3      1039       452:454      1.1±0.3 
    12      I8NI5E    I8SI9S  404:408    9:13        -1.1±1.0 
    13      I1        I7N     861        808          1.3±0.4(b) 
   Mean                                              -0.78 
__________________________________________________________________
 The TCO difference between legs is calculated by subtracting data 
 from the earlier sampling of a crossover location from that of the 
 later sampling. The station numbers refer to the actual stations 
 used for this analysis. 
  (a) The LOESS fit diverged significantly from the data. 
  (b) Not considered reliable due to insufficient data.



5. CROSSOVER ANALYSIS

The agreement between TCO2 measurements made at similar locations, but on 
different legs of the survey, were used as a check on the internal consistency 
of the measurements. Deep measurements were used because of the lower variabil-
ity in TCO2 observed in the deep ocean. Because most motion in the ocean 
interior takes place along surfaces of constant density (isopycnals), 
comparisons were made along isopycnal surfaces rather than depth.

Our crossover analysis was performed as follows:

(1) Locations at which different cruise legs intersected were identified as 
    'crossover points.' These are identified in Table 4 and are plotted on Fig. 1.

(2) Stations located in the immediate proximity of these crossover points, for 
    which TCO2 data existed, were selected for the comparison. In general, 
    stations located within 100 km of the crossover location were selected.

(3) For water samples collected below 2500 m, smooth curves were fit through the 
    TCO2 data as a function of the density anomaly referenced to 3000 dbar 
    (sigma 3) using Cleveland's LOESS smoother (Cleveland and Devlin, 1988). A 
    separate fit was performed to the data collected from each of the two 
    intersecting legs. The tension parameter for the smoother was adjusted 
    subjectively to give a 'reasonable' fit to the data at the majority of the 
    crossover locations, and the same value for the tension parameter was used 
    for all of the crossovers. Hence, while the fits to the data may not 
    necessarily represent the best possible at each individual crossover point, 
    the smoothing function has been consistently applied to all crossovers.

(4) For each crossover, the difference between the two smooth curves was 
    evaluated at 50 evenly spaced intervals which covered the density range over 
    which the two data sets overlapped. A mean and a standard deviation of the 
    difference between the two curves was estimated based on these 50 values, 
    and these values are reported in Table 4. An illustration of a typical 
    analysis, the fitted data for crossover 4, is plotted on Fig. 7.


Fig. 7.  An example of a crossover analysis using the TCO2 vs. density fits at 
         crossover location #4. This location was first sampled on leg I7N in 
         July 1995. It was resampled during January 1996 on leg I2. The TCO2 
         data from stations within 100 km of the crossover location and 
         depths>2500 m have been plotted vs. the potential density anomaly 
         referenced to 3000 dbar (sigma 3). The solid curves represent fits to 
         the data using a LOESS smoother (see text). The difference between the 
         fits for the two separate legs was evaluated at 50 density intervals 
         spaced evenly within the overlapping density range of the two legs (see 
         Table 4). The legend shows the station numbers used for the comparison.


The results of the crossover analysis indicate that absolute leg-to-leg 
differences are always <3.0 µmol/kg (Table 4). Note that the comparisons were 
evaluated consistently such that the fit to data from the earlier leg at each 
crossover was subtracted from the fit to the later leg's data. Any uncorrected, 
long-term, monotonic drift in the calibration of the SOMMA analyzers over the 
course of the Indian Ocean expedition would therefore result in a non-zero value 
for the overall mean of these differences. The overall mean and standard 
deviation of the differences at crossovers 1-12 is -0.78 (±1.74) µmol/kg, and 
there was also no significant correlation between the individual differences 
derived from each crossover and the number of days which separated the crossover 
samplings. In general, the results of the crossover analysis are quite 
consistent with the overall precision (±1.2 µmol/kg) of the CRM analyses (see 
Section 3.3), and confirms that this precision applies to both systems 
throughout the survey. There is no suggestion in the crossover results of any 
additional significant sources of error or uncertainty.



6. CONCLUSIONS

In summary, personnel aboard the R/V Knorr have been able to use the SOMMA-
coulometer system to consistently replicate within analytical error the 
certified CRM TCO2 values. They have been able to use these systems to make, 
counting duplicates and CRM, over 20,000 determinations of TCO2 during the 14 
months of the Indian Ocean Survey without significant instrument down time. The 
measurement groups have accomplished the following.

(1) They have charted the history of the to-deliver volume of the sample 
    pipettes by gravimetric determinations, and corrected the water sample data 
    for the documented changes in the pipette volumes. The change in system 
    response due to the change in pipette volume corresponded to approximately 1 
    part in 2000 for TCO2 on both systems over the 10 months prior to recleaning 
    of the pipettes.

(2) The groups have determined that the survey precision for the TCO2 analyses, 
    irrespective of which leg or system the water samples were analyzed on, was 
    ±1.2 µmol/kg. The precision of the two instruments was nearly identical and 
    consistent throughout the 14 months of the survey.

(3) They have analyzed nearly 1000 CRM with an overall difference between the 
    analyzed and certified TCO2 of -1.0 µmol/kg (0.05%) on both systems which 
    demonstrates the equivalency of the two independent instruments, and meets 
    the survey's goal for accuracy.

(4) The measurement groups have documented the influence of factors besides 
    pipette volume which could have affected accuracy including electronic 
    calibration, gas calibration, cell age, and cell assembly selection.

For precision, the pooled standard deviation (sp^2=1.2 µmol/kg), calculated 
according to Youden (1951), is the most conservative estimate of precision 
because it includes all random analytical errors (sampling, instrumental, and 
method). The identical accuracy for the CRM analyses on both systems and the 
results of the crossover analysis (Table 4) indicate that the sp^2 statistic can 
be used to evaluate survey data sets irrespective of the leg or system the data 
originated from.

The SOMMA-coulometry systems have allowed several scientific groups in a shared 
effort to examine carbon inventories and aquatic carbon cycling. For the Indian 
Ocean Survey, the sensitivity of the TCO2 determinations defined as the ratio of 
their precision (1.2 µmol/kg) over the TCO2 dynamic range (250 µmol/kg) was 0.4% 
which approaches the 0.1% sensitivity of the salinometers used, and these 
systems were as reliable as the salinometers. If their reliability is to be 
improved, the focus should be on understanding the basic behavior of the cell 
assemblies and the chemical behavior of the cell solutions as they age, so that 
procedural corrections can be made. The accuracy and precision of the Indian 
Ocean TCO2 analyses indicates that these data will be more than adequate for 
testing applicable oceanographic models, and allow the direct measurement of the 
CO2 uptake if and when these lines are resampled.



ACKNOWLEDGEMENTS

We would like to thank the US Department of Energy's Office of Biological and 
Environmental Research for their support. The success of the Indian Ocean CO2 
Survey was due to the shared efforts of the DOE Science Team. We thank John 
Downing for his initial organization of the Science Team and assistance in 
getting the US CO2 Survey underway. We thank the chief scientists, scientific 
staff, and crew aboard the R/V Knorr for their assistance throughout. Dave 
Chipman and Taro Takahashi are acknowledged for helpful comments and advice. The 
instruments used for the survey were produced at the Equipment Development 
Laboratory (EDL) at the University of Rhode Island's Graduate School of 
Oceanography under the supervision of Dr. John King and David Butler. This 
research was performed under the auspices of the United States Department of 
Energy under Contract No. DE-AC02-98CH10886.



REFERENCES

Bates, N.R., Michaels, A.F. and Knap, A.H., 1996. Seasonal and interannual 
    variability of oceanic carbon dioxide species at the US JGOFS Bermuda 
    Atlantic time-series study (BATS) site. Deep-Sea Research II 43, pp. 347-383 

Cleveland, W.S. and Devlin, S.J., 1988. Locally-weighted regression: an approach 
    to regression analysis by local fitting. J. Am. Stat. Assoc. 83, pp. 596-610

Dickson, A.G., 1990. The oceanic carbon dioxide system: planning for quality 
    data. US JGOFS News 2:2.

DOE, 1994. Handbook of methods for the analysis of the various parameters of the 
    carbon dioxide system in sea water; version 2.0. ORNL/CDIAC-74.

Huffman Jr., E.W.D., 1977. Performance of a new automatic carbon dioxide 
    coulometer. Microchem. J. 22, pp. 567-573 

Johnson, K.M., 1995. Operator's Manual. Single-Operator Multiparameter Metabolic 
    Analyzer (SOMMA) for Total Carbon Dioxide (CT) with Coulometric Detection. 
    Version 3.0. Available from K.M. Johnson, Department of Applied Science, 
    Brookhaven National Laboratory, Upton, NY.

Johnson, K.M., Wallace, D.W.R., 1992. The single-operator multiparameter 
    metabolic analyzer for total carbon dioxide with coulometric detection. DOE 
    research summary no. 19. Carbon Dioxide Information Analysis Center, Oak 
    Ridge National Laboratory, TN.

Johnson, K.M., King, A.E. and Sieburth, J.McN., 1985. Coulometric TCO2 analyses 
    for marine studies: an introduction. Mar. Chem. 16, pp. 61-82 

Johnson, K.M., Sieburth, J.McN., Williams, P.J.leB. and Bränström, L., 1987. 
    Coulometric TCO2 analysis for marine studies: automation and calibration. 
    Mar. Chem. 21, pp. 117-133 

Johnson, K.M., Wills, K.D., Butler, D.B., Johnson, W.K. and Wong, C.S., 1993. 
    Coulometric total carbon dioxide analysis for marine studies: maximizing the 
    performance of an automated gas extraction system and coulometric detector. 
    Mar. Chem. 44, pp. 167-187 

Millero, F.J. and Poisson, A., 1981. International one-atmosphere equation of 
    state for sea water. Deep-Sea Res. 28, pp. 625-629 

Millero, F.J., Dickson, A.G., Eischeid, G., Goyet, C., Guenther, P., Johnson, 
    K.M., Lee, K., Purkerson, D., Sabine, C.L., Key, R., Schottle, R.G., 
    Wallace, D.W.R., Lewis, E.R. and Winn, C.D., 1998. Assessment of the quality 
    of the shipboard measurements of total alkalinity on the WOCE Hydrographic 
    Program Indian Ocean CO2 survey cruises 1994-1996. Mar. Chem. 63, pp. 9-20 

Sabine, C.L., Key, R.M., 1998. Surface water and atmospheric underway carbon 
    data obtained during the world ocean circulation experiment Indian Ocean 
    survey cruises (R/V Knorr, December 1994-January 1996). NDP-064, Carbon 
    Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak 
    Ridge, TN.

Taylor, J.K., 1990. Statistical techniques for data analysis. Lewis Publishers, 
    Chelsea, 200 pp.

Wilke, R.J., Wallace, D.W.R. and Johnson, K.M., 1993. A water-based, gravimetric 
    method for the determination of gas sample loop volume. Anal. Chem. 65, pp. 
    2403-2406 

Youden, W.J., 1951. Statistical Methods for Chemists. Wiley, New York, 126 pp.










                                  APPENDIX C:

                        REPRINT OF PERTINENT LITERATURE




Millero F.J., A.G. Dickson, G. Eischeid, C. Goyet, P. Guenther, K.M. Johnson, 
R.M. Key, K. Lee, D. Purkerson, C.L. Sabine, R.G. Schottle, D.W. .R. Wallace, E. 
Lewis and C.D. Winn, Assessment of the quality of the shipboard measurements of 
total alkalinity on the WOCE Hydrographic Program Indian Ocean CO2 survey 
cruises 1994-1996, Marine Chemistry 63:9-20.





Marine Chemistry 
63 1998 9 - 20


   ASSESSMENT OF THE QUALITY OF THE SHIPBOARD MEASUREMENTS OF TOTAL ALKALINITY
   ON THE WOCE HYDROGRAPHIC PROGRAM INDIAN OCEAN CO2 SURVEY CRUISES 1994-1996


        Frank J. Millero(a,*), Andrew G. Dickson(b), Greg Eischeid(c),
 Catherine Goyet(c), Peter Guenther(b), Kenneth M. Johnson(d), Robert M. Key(e),
Kitack Lee(f), Dave Purkerson(a), Christopher L. Sabine(e), Rolf G. Schottle(g), 
      Douglas W. R. Wallace(d), Ernie Lewis(d) and Christopher D. Winn(g) 

 (a) Rosenstiel School of Marine and Atmospheric Science, University of Miami, 
     Miami, FL 33149, USA
 (b) Scripps Institution of Oceanography, University of California, 
     La Jolla, San Diego, CA 92093, USA
 (c) Woods Hole Oceanographic Institute, Woods Hole, MA 02543, USA
 (d) Department of Applied Science, Brookhaven National Laboratory, 
     Upton, NY 111973, USA
 (e) Department of Geosciences, Princeton University, Princeton, NJ 08544, USA
 (f) NOAA/AOML, Miami, FL 33149, USA
 (g) Department of Oceanography, University of Hawaii, Honolulu, HI 96822, USA 

     *Corresponding author.

Received 16 January 1998; revised 31 March 1998; accepted 7 April 1998. 
Available online 8 December 1998. 





ABSTRACT

In 1995, we participated in a number of WOCE Hydrographic Program cruises in 
the Indian Ocean as part of the Joint Global Ocean Flux Study (JGOFS) CO2 
Survey sponsored by the Department of Energy (DOE). Two titration systems 
were used throughout this study to determine the pH, total alkalinity (TA) 
and total inorganic carbon dioxide (TCO2) of the samples collected during 
these cruises. The performance of these systems was monitored by making 
closed cell titration measurements on Certified Reference Materials (CRMs). A 
total of 962 titrations were made on six batches of CRMs during the cruises. 
The reproducibility calculated from these titrations was ±0.007 in pH, ±4.2 
µmol/kg-1 in TA, and ±4.1 µmol/kg-1 in TCO2. The at-sea measurements on the 
CRMs were in reasonable agreement with laboratory measurements made on the 
same batches. These results demonstrate that the CRMs can be used as a 
reference standard for TA and to monitor the performance of titration systems 
at sea. Measurements made on the various legs of the cruise agreed to within 
6 µmol/kg-1 at the 15 crossover points. The overall mean and standard 
deviation of the differences at all the crossovers are 2.1±2.1 µmol/kg-1. 
These crossover results are quite consistent with the overall reproducibility 
of the CRM analyses for TA (±4 µmol/kg-1) over the duration of the entire 
survey. The TA results for the Indian Ocean cruises provide a reliable data 
set that when combined with TCO2 data can completely characterize the 
carbonate system.



Author Keywords: alkalinity; WOCE Hydrographic Program; CO2
Index Terms:     reproductive toxicity; boron

0304-4203r98r$ - see front matter q 1998 Elsevier Science B.V. All rights 
reserved. 

PII: S0304-42039800043-7







ARTICLE OUTLINE

1. Introduction
2. Methods
2.1. Titration system
2.1.1. Electrodes
2.1.2. Standard acids
2.1.3. Volume of the cells
2.1.4. Volume of titrant
2.2. Evaluation of the carbonate parameters
3. Results and discussion
3.1. Laboratory ta measurements of CRMs
3.2. At sea measurements of TA, TCO2, and pH on CRMs
3.2.1. Accuracy of at sea measurements
3.2.2. Long term stability of a cell performance
3.3. Crossover analysis
4. Conclusion
Acknowledgements
References







1. INTRODUCTION

From 1994 to 1996, a number of cruises were made in the Indian Ocean as part 
of the World Ocean Circulation Experiment (WOCE) Hydrographic Program to 
characterize the carbon dioxide system. This survey of CO2 was an integral 
part of the Joint Global Ocean Flux Study (JGOFS). The goals of this survey 
were to: (1) Quantify the uptake of anthropogenic carbon dioxide by the 
oceans to better predict future atmospheric carbon dioxide levels; (2) 
Provide a global description of the carbon dioxide in the oceans to aid in 
the development of a 3-dimensional model of the oceanic carbon cycle; and (3) 
Characterize the transport of CO2 across the air-sea interface and the large 
scale transports of carbon dioxide within the oceans.

To satisfy these goals, it was necessary to make very precise measurements of 
at least two of the carbonate system parameters (pH; total alkalinity, TA; 
total carbon dioxide, TCO2; and the fugacity of carbon dioxide, fCO2). Within 
the United States a large part of this survey was conducted by a team of 
investigators supported by the US Department of Energy. The team selected the 
measurement of TCO2 (Johnson et al., 1998) and of TA as the parameters to be 
measured in the water column and fCO2 in the atmosphere and surface waters. 
To insure that the measurements of TCO2 and TA were as precise and accurate 
as possible Certified Reference Materials (CRMs) (Dickson, 1990a) were used 
throughout the studies. The team also developed a set of Standard Operating 
Procedures1 (DOE, 1994) and, to a large extent, shared a common approach to 
the measurement program.

For the studies in the Indian Ocean, the team shared equipment throughout the 
study. This paper presents the results of this team effort to precisely and 
accurately determine the total alkalinity during these cruises and the 
intercomparison between cruises. A companion paper (Johnson et al., 1998) 
describes the total carbon dioxide measurements.



(1) DOE, 1991. Handbook of methods for the analysis of the various parameters of 
    the carbon dioxide system in sea water, In: Dickson, A.G., Goyet, C. (Eds.), 
    Version 1.0, Unpublished manuscript



2. METHODS

The total alkalinity was determined on the JGOFS Indian Ocean cruises by the 
DOE group using systems described in detail by Millero et al. (1993). The 
total alkalinity of seawater was evaluated from the proton balance at the 
alkalinity equivalence point, pHequiv4.5, according to the exact definition mmmm
of total alkalinity (Dickson, 1981)


                     TA = [HCO¯(3)]+2[CO(3)(^2¯)]+[B(OH)¯(4)]
                         +[OH¯]+[HPO^2¯(4)]+2[PO^3¯(4)]
                          +[SiO(OH)¯(3)]+[HS¯]+[NH(3)]
                        -[H+]-[HSO¯(4)]-[HF]-[H(3)PO(4)]                   (1)


At any point in the titration, the total alkalinity of seawater can be 
calculated from the equation


                  (W(0) x TA-W x C(HCl))/(W(0)+W) = [HCO¯(3)]
                         +2[CO^2¯(3)]+[B(OH)¯(4)]+[OH¯]
                      +[HPO^2¯(4)]+2PO^3¯(4)]+SiO(OH)¯(3)]
                               +[HS¯]+[NH(3)]-[H+]
                           -[HSO¯(4)]-[HF]-[H(3)PO(4)]                     (2)



where W0 is the mass of the sample to be titrated, CHCl is the concentration 
of acid titrant, and W is the mass of acid added. In the calculations, 
volumes of the sample and of the acid were converted to mass using the 
density of seawater (Millero and Poisson, 1981) and the density of HCl in 
NaCl (Millero et al., 1977). Direct measurements made on the density of the 
acid used agreed to within 10 ppm with the equations used in the computer 
code. At the endpoint (W2) the total alkalinity is given by


                              TA = W2 x C(HCl)/W(0)                        (3)


The uncertainties in TA associated with acid concentration (0.25±0.0001 M), 
mass of acid delivered (2.5±0.0005 g), and mass of the sample (200±0.05 g) 
are ±1, ±0.5, and ±0.5 µmol/kg-1, respectively (which gives a probable error 
of about ±1 µmol/kg-1). By using the same acid, titrators, and acid 
throughout a given cruise one can obtain a precision that is comparable with 
this probable error. Discussed below are more details on the components of 
the titration systems.


2.1. TITRATION SYSTEM

The titration systems used to determine TA consist of a Metrohm 665 Dosimat 
titrator and an Orion 720A pH meter controlled by a personal computer 
(Millero et al., 1993). Both the acid titrant in a water-jacketed burette and 
the seawater sample in a water-jacketed cell were controlled to a constant 
temperature of 25±0.1°C with a Neslab constant temperature bath. The 
plexiglass water-jacketed cells used for our studies were similar to that 
used by Bradshaw and Brewer (1988) except a larger volume (about 200 cm3) was 
used to improve the precision. These cells have fill and drain valves that 
increased the reproducibility of the cell volume.

A Lab Windows C program is used to run the titration and record the volume of 
the added acid and the emf of the electrodes using RS232 interfaces. The 
titration is made by adding HCl to seawater past the alkalinity end point. A 
typical titration records the average of ten emf readings after they become 
stable (±0.09 mV) and adds enough acid to change the voltage by a pre-
assigned increment (13 mV). In contrast to the delivery of a fixed volume of 
acid, this method gives more data points in the range of a rapid increase in 
the emf near the endpoint. A full titration (25 points) takes about 20 min.


2.1.1. ELECTRODES

The electrodes used to measure the emf of the sample during a titration 
consist of a ROSS glass pH electrode and an Orion double junction Ag, AgCl 
reference electrode. A number of electrodes were screened to select those to 
be used in the titrators. Electrodes with non-Nernstian behavior (slopes more 
than 1.0 mV different from the theoretical value) were discarded. The 
reliability of the electrodes was evaluated by determining the TA, TCO2 and 
pH of Gulf Stream seawater. The titration values of TCO2 are normally higher 
than the values measured by coulometry and the values of pH are typically 
lower than the values obtained by spectrophotometric methods. These 
differences in TCO2 and pH are caused by the non-Nernstian behavior of the 
electrodes (Millero et al., 1993). We selected electrodes which gave values 
of TCO2 and pH close to the values determined by coulometric and by 
spectrophotometric methods, respectively.


2.1.2. STANDARD ACIDS

The HCl used for this study and for all of our cruises was made in the 
laboratory, standardized, and stored in 500 cm3 glass bottles. The 0.25 M HCl 
solutions were made from 1 M Mallinckrodt standard solutions in 0.45 M NaCl 
to yield an ionic strength equivalent to that of average seawater (0.7 M). 
The concentration of HCl was measured using a constant current coulometric 
technique (Taylor and Smith, 1959; Marinenko and Taylor, 1968). Coulometric 
analysis of the acids used for these cruises agreed to ±0.0001 M with the 
analyses performed independently on the same batches of acids in Dr. A. 
Dickson's laboratory at Scripps Institution of Oceanography (SIO). The mutual 
consistency of these acids was also confirmed by comparing the values of TA 
measured on Gulf Stream seawater using different batches of acids, but using 
the same titrator and electrodes. The uncertainties in TA associated with 
acid concentration (±0.0001 M) is 1 µmol/kg-1.


2.1.3. VOLUME OF THE CELLS

The volume of each of the titration cells used at sea was determined by 
comparing the values of TA obtained for Gulf Stream seawater with open and 
closed cells in the laboratory. All of the open cell laboratory TA 
measurements were made with weighed amounts of seawater in a cell with a 
small head-space. If the volume is correct, the TA from the open and closed 
cells should be the same, provided that the same acid, titrator, and 
electrodes are used. At least 10 measurements were made on each cell yielding 
an average TA that agreed with the assigned value to better than 1 µmol/kg-1. 
If the volume of a titration cell needed to be adjusted during the cruise 
(because of broken electrodes, plungers etc.), the volumes were determined 
from the daily titrations on low-nutrient surface seawater (usually collected 
before the first station) and Certified Reference Materials (CRMs) provided 
by Dr. A. Dickson (SIO). Post-cruise calibrations of the cells were made by 
comparing the values of TA for the Gulf Stream seawater and CRM with open and 
closed cells. The nominal volumes of all the cells were about 200 cm3, and 
the values were determined to ±0.05 cm3. The uncertainty in TA associated 
with this uncertainty in the volume of the cells (±0.05 cm3) is 0.5 µmol/kg-1 
obtained for the weighed samples.


2.1.4. VOLUME OF TITRANT

The volume of HCl delivered to the cell is traditionally assumed to have 
small uncertainty (Dickson, 1981) and equated to the digital output of the 
titrator. Calibrations of all the burettes of the Dosimats used were made 
with Milli-Q water at 25°C. Since the cell volumes are calibrated using 
standard solutions, errors in the accuracy of volume delivery will be 
partially canceled and included in the value of cell volumes assigned. The 
calibration of all the Dosimats used at sea and in the laboratory indicated 
that the amount of acid delivered (for a typical calculation) was uncertain 
to ±0.0005 cm3. This uncertainty in the volume delivered leads to an error in 
the TA of ±0.5 µmol/kg-1. Nevertheless, corrections to the Dosimat reading 
were made in all of our laboratory TA measurements and calibrations to insure 
that the assigned value for a different batch of CRM and Gulf Stream water 
was not affected by the use of different Dosimats. These corrections were 
also made when calculating the volume of each cell.


2.2. EVALUATION OF THE CARBONATE PARAMETERS

A FORTRAN computer program has been developed to calculate the carbonate 
parameters (pH, E*, TA, TCO2, and pK(1)) in the seawater solutions. The program 
is patterned after those developed by Dickson (1981), Johansson and Wedborg 
(1982) and Dickson (1; DOE, 1994). The fitting is performed using the STEPIT 
routine (J.P. Chandler, Oklahoma State University, Stillwater, OK 74074). The 
STEPIT software package minimizes the sum of squares of residuals by 
adjusting the parameters E*, TA, TCO2 and pK(1) of carbonic acid. The computer 
program is based on Eq. 2 and assumes that nutrients such as phosphate, 
silicate and ammonia are negligible. This assumption is strictly valid only 
for surface waters. Neglecting the concentration of nutrients in the seawater 
sample does not affect the accuracy of TA, but must be considered when 
calculating the carbonate alkalinity (CA=[HCO¯(3)]+2 [CO^2¯(3)]) from TA.

The pH and pK of the acids used in the program are on the seawater scale, 
[H+]sw[H+]+[HSO¯(4)]+[HF] (Dickson, 1984). The dissociation constants used in 
the program were taken from Dickson and Millero (1987) for carbonic acid, 
from Dickson (1990b) for boric acid, from Dickson and Riley (1979) for HF, 
from Dickson (1990c) for HSO¯(4) and from Millero (1995) for water. The program 
requires as inputs the concentration of acid, volume of the cell, salinity, 
temperature, measured emfs (E) and volumes of HCl (V). To obtain a reliable 
TA from a full titration, at least 25 data points should be collected (9 data 
points between pH 3.0 to 4.5). The precision of the fit is less than 0.4 µmol/ 
kg_1 when pK(1) is allowed to vary and 1.5 µmol/kg-1 when pK(1) is fixed. Our 
titration program has been compared to the titration programs used by others 
(Johansson and Wedborg, 1982; Bradshaw and Brewer, 1988) and the values of TA 
agree to within ±1 µmol/kg-1.



3. RESULTS AND DISCUSSION


3.1. LABORATORY TA MEASUREMENTS OF CRMS

The laboratory TA measurements made on the CRMs used throughout this study 
are summarized in Table 1. The results obtained by both laboratories 
demonstrate that no systematic differences in TA are found. With the 
exception of Batch 29, the differences in the measurements of the CRMs 
between the two laboratories are less that 2 µmol/kg-1. Since the Miami 
measurements were made with the same acid as used at sea, one cannot 
attribute the differences in Batch 29 to differences in the concentration of 
the acids (calibrated at SIO). The Miami measurements were also made using 
the same acid for all the batches of CRM within a one-week period to ensure 
the internal consistency of its results. The measurements made on the acid 
concentration in Miami and SIO by a coulometric titration were in agreement 
to ±0.0001 M, which is equivalent to an error of ±1 µmol/kg-1 in TA.


TABLE 1: Comparison of the total alkalinity of Certified Reference Materials 
         _______________________________________________________________
         
          Batch   SIO    Miami   ∆(S-M)  Cruise 
          -----  ------  ------  ------  ------------------------------
            23   2212.7  2213.7   -1.0   I8S/I9S, I9N, I8N/I5E, I3, I7N 
            24   2215.5  2215.8   -0.3   I8R    
            26   2176.6  2175.1    1.5   I3, I5W/I4, I-7N 
            27   2214.9  2214.3    0.6   I7N, I1, I10, I2 
            29   2184.8  2182.3    2.5   I8R    
            30   2201.9  2200.5    1.4   I2
         _______________________________________________________________


3.2. AT SEA MEASUREMENTS OF TA, TCO2, AND PH ON CRMS

3.2.1. Accuracy of at sea measurements

The tracts of the cruise made during the Indian Ocean studies are shown in 
Fig. 1. A total of 962 titrations were made on six batches of the CRMs during 
the cruises (Table 2). A summary of the pH, TA and TCO2 measurements made on 
CRMs (Table 3) throughout the cruise is shown in Fig. 2, Fig. 3 and Fig. 4. 
The reproducibility on the six batches of the CRMs used was ±0.007 in pH, 
±4.2 µmol/kg-1 in TA, and ±4.1 µmol/kg-1 in TCO2. The at sea TA measurements 
on the CRMs were in good agreement (2-4 µmol/kg-1) with laboratory 
measurements made on the same batches at MIAMI and SIO. These small 
differences (2-4 µmol/kg-1) are well within the overall precision of our 
measurements and can be attributed to uncertainties in the volume of cells 
assigned in the laboratory before the cruises. However, the cells used on I7 
gave significantly greater errors than the values obtained in the 
laboratories on the same batch of CRM. These large discrepancies might be 
attributed to inaccurately assigned volumes of the cells after they were 
repaired for leakage due to repositioning of a reference electrode after 
changing the inner filling solution.


Fig. 1. Cruise tracts of the Indian Ocean Studies showing crossover points.


TABLE 2: Measurements of pH, TA and TCO2 of CRM at sea
____________________________________________________________________________________

 Cruise   Start    End      Batch  Cell  N    TA     S.D.  TCO2   S.D.  pH    S.D. 
          date     date                       avg          avg          avg    
 -------  -------  -------  -----  ---- ---  ------  ---  ------  ---  -----  -----
 I8S/I9S  12/1/94  1/19/95   23    All   49  2221.5  5.1  2004.5  4.1  
                                   5     18  2223.3  4.8  2003.8  2.5  
                                   6     18  2220.8  4.0  2008.0  3.1  
                                   20    13  2220.0  6.4  2001.4  3.8  
 I9N      1/24/95  3/6/95    23    All  138  2216.2  3.3  2000.1  3.5  7.891  0.005 
                                   5     68  2215.0  3.3  1999.1  3.3  7.892  0.004 
                                   6     65  2217.5  3.3  2001.3  3.3  7.891  0.005 
                                   20     5  2214.2  3.1  1996.5  3.5  7.895  0.007 
 I8N/I5E  3/10/95  4/16/95   23    All   80  2211.6  4.9  1997.0  3.0  7.890  0.006 
                                   5     36  2213.0  5.5  1998.6  3.8  7.890  0.005 
                                   6     44  2210.1  3.6  1996.2  2.6  7.890  0.007 
 I3       4/20/95  6/7/95    23    All   65  2215.4  1.4  2002.1  1.4  7.894  0.005 
                                   2     33  2215.7  1.3  2000.7  1.4  7.898  0.006 
                                   13    35  2215.0  1.4  2003.6  1.3  7.890  0.004 
                             26    All   30  2178.0  1.2  1984.8  1.2  7.858  0.004 
                                   2     14  2178.3  1.3  1983.3  1.2  7.862  0.003 
                                   13    16  2177.7  1.2  1986.0  1.1  7.855  0.004 
 I5W/I4   6/11/95  7/11/95   26    All   79  2182.6  3.8  1990.2  3.4  
                                   2     41  2183.3  3.9  1988.0  2.4  
                                   13    38  2182.0  3.5  1992.9  2.3  
 I7N      7/15/95  8/24/95   26    All   59  2184.0  5.7  1984.7  3.4  7.862  0.009 
                                   2     33  2186.2  3.1  1984.3  2.6  7.862  0.009 
                                   13    26  2181.5  7.4  1985.2  4.0  7.858  0.006 
                             27    All    8  2221.5  3.1  1995.5  1.4  7.916  0.005 
                                   2      4  2221.4  2.4  1994.9  1.4  7.914  0.005 
                                   13     4  2221.5  4.1  1996.0  1.5  7.918  0.006 
                             23    All   10  2222.4  7.4  2002.0  4.0  7.896  0.006 
                                   2      5  2227.5  5.8  2003.2  4.1  7.897  0.005 
                                   13     5  2216.2  6.4  1999.9  3.9  7.893  0.009 
 I1       8/29/95  10/18/95  27    All  244  2219.4  3.9  1998.8  5.4  7.906  0.013 
                                   2    123  2220.1  3.2  1995.3  3.2  7.911  0.005 
                                   7     54  2219.6  3.6  1999.7  4.1  7.908  0.013 
                                   13    15  2216.2  4.7  1994.6  4.5  7.909  0.005 
                                   14    52  2217.9  4.5  2006.5  3.6  7.885  0.009 
 I10      11/6/95  11/24/95  27    All   62  2212.9  4.0  1991.3  2.9  7.912  0.006 
                                   11    30  2212.3  4.5  1989.6  2.4  7.914  0.005 
                                   16    32  2213.5  3.5  1993.1  2.0  7.910  0.006 
 I8R      9/23/95  10/24/95  29    All   36  2184.2  1.8  1914.8  2.4  8.006  0.006 
 NOAA Cruise                       4      9  2185.5  1.7  1914.5  1.9  8.006  0.005 
                                   17    17  2183.9  1.6  1914.4  2.2  8.007  0.005 
                                   18    10  2183.4  2.1  1915.7  3.1  8.004  0.009 
                             24    All   10  2216.6  2.3  1998.7  1.7  7.902  0.006 
                                   4      2  2218.5  3.8  1998.6  3.9  7.907  0.004 
                                   17     5  2215.1  0.6  1998.5  1.4  7.902  0.006 
                                   18     3  2217.3  2.6  1998.6  1.7  7.899  0.006 
 I2      11/28/95  1/19/96   27    All   67  2219.4  4.5  1994.0  2.8  7.916  0.005 
                                   11    36  2219.9  5.7  1993.1  3.3  7.918  0.005
                                   16    31  2218.9  3.2  1994.7  2.2  7.915  0.006 
                             30    All    9  2204.6  2.7  1996.8  2.1  7.879  0.004 
                                   11     4  2205.3  2.3  1995.0  2.2  7.880  0.002 
                                   16     5  2204.0  3.0  1998.4  0.8  7.879  0.006 
____________________________________________________________________________________


TABLE 3: The overall precision of at sea TA, TCO2, and pH measurements on the 
         Certified Reference Material 
                  _________________________________________

                               Precision 1σ   Number of 
                   Parameters   (µmol kg-1)   measurements 
                   ----------  -------------  ------------
                   TA              4.2            949 
                   TCO             4.1           9472 
                   pH              0.007          793(a)  
                  _________________________________________
                  (a) The numbers of the pH measurements 
                      were less than for TA and TCO2 be-
                      cause some values were not recorded.


Fig. 2.  The reproducibility of the titration pH measurements made on 
         Certified Reference Material on the Indian Ocean Study.
Fig. 3.  The reproducibility of the titration TCO2 measurements made on 
         Certified Reference Material on the Indian Ocean Study.
Fig. 4.  The reproducibility of the titration TA measurements made on 
         Certified Reference Material on the Indian Ocean Study.


3.2.2.  Long term stability of a cell performance

The at sea TA measurements on the CRMs can be used to examine the long-term 
stability of the cells used during the cruises. Overall, the TA results 
obtained using cells for a given cruise did not show any systematic trends. 
Differences in TA between laboratory and field measurements remained 
unchanged over the entire period of each cruise. However, inter-cruise 
variations in TA between laboratory and field results were observed when the 
same cells were used. For instance, cells 2 and 13 were used for four 
consecutive cruises over the period of six months. When these two cells were 
used on the first cruise (I3), the field measurements agreed to within 
±2 µmol/kg-1 with the values obtained in the laboratory. These small 
discrepancies are within the precision of our measurements. When the same 
cells were used for the later cruises, the differences in TA between 
laboratory and field measurements became significantly larger (9 µmol/kg-1). 
As mentioned in Section 3.2.1, these larger differences can be attributed to 
changes in the assigned volume of the cells due to repositioning of a 
reference electrode. These inter-cruise variations in TA can be corrected by 
normalizing the measured values obtained during the cruises using the 
corrections required to reproduce the values assigned for the CRMs by SIO 
(Table 4). This correction was applied using


                             ∆ = TA(meas,CRM)-CRM)                         (4)

                     TA(corr.) = TA(meas.) x [CRM/(CRM+∆)]                 (5)

where CRM is the SIO-certified values.


TABLE 4: Differences between TA measurements made at sea and values measured 
         in the laboratory (SIO)
______________________________________________________________________________

 Cell  I8S/I9S  I9N  I8N/I5E   I3     I5W/I4   I7N    I1   I8R   I10    I2 
 ----  -------  ---  -------  ------  ------  ------  ---  ---  ----  -------
   2                         +2.6(a)   +6.7  +9.9(a) +5.2  
   4                                                       0.7 
   5    +10.6  +2.3   +0.3   
   6    +8.1   +4.8   -2.6   
   7                                                 +4.7  
  11                                                            -2.6  +4.8(a) 
  13                         +2.1(a)   +6.0  +4.9(a) +1.3  
  14                                                 +3.0  
  16                                                            -1.4  +3.7(a) 
  17                                                      -0.9 
  18                                                      -1.4 
  20    +7.3   +1.5   
______________________________________________________________________________
(a)Based on the weighted average on different CRM. 


3.3. CROSSOVER ANALYSIS

In order to cross-check our estimates of accuracy of the TA data, which are 
derived from analyses of CRMs, we examined the agreement between TA 
measurements made at identical locations on different legs of the Indian 
Ocean expedition. All of these comparisons have been made after applying the 
corrections given in Table 4. The implicit assumption is that temporal and 
spatial gradients of TA concentrations in the deep ocean are small relative 
to measurement accuracy, so that water sampled at the same location in the 
deep ocean at two different times should have near-identical values of TA. In 
practice, vertical gradients of TA can be significant relative to measurement 
accuracy and there can also be significant vertical motions in the deep 
ocean. Hence, measurements made at the same geographical location cannot be 
compared simply on the basis of their common depth. Because most motion in 
the ocean interior takes place along surfaces of constant density 
(isopycnals), it is preferable to compare concentrations using density as the 
frame of reference rather than depth.


TABLE 5: Crossover results for the TA measurements made in the Indian Ocean 
         _________________________________________________
         
          Number  Stations      Legs               ∆TA    
          ------  ------------  ---------------  --------
            1     927,929,931,  I1-I7N            1.7±1.0    
                  780,782,784    
            2     987,990,266,  I1-I9N           -2.1±5.9    
                  268,270    
            3     996,998,      I1-I9Nb           1.2±0.8    
                  233,235    
            4     1205,728,     I2-I7N            5.6±2.4    
                  730    
            5     1137,1139,    I2-I9N/I5E        3.4±2.2    
                  320,324    
            6     1094,1096,    I2-I9N           -3.4±1.4    
                  191,193    
            7     1078,1075     I2-I10            1.8±2.4    
            8     705,547,549   I5W/I4-I3         0.7±1.7    
            9     498,499,501,  I3-I8N/I5E       -0.8±2.3    
                  346,348    
           10     472,169       I3-I9N           -0.8±0.6    
           11     1039,452,454  I10-I3           -1.0±0.7    
           12     404,406,408,  I8N/I5E-I8S/I9S  -2.7±3.8    
                  9,11,13    
           13     861,808       I1-I7N            0.3±0.6    
           14     709,707       I7N-I5W/I4        2.4±1.7    
           15     966,968,969,  I1-I8N/I5E       -4.2±4.5    
                  283,287
         _________________________________________________


Our crossover analyses were performed as follows.

(1) Locations at which different cruise legs intersected were identified as 
    crossover points. These are identified in Table 5 and Fig. 1.

(2) Stations located in the immediate proximity of these crossover points, 
    for which TA data existed, were selected for the comparison. In general, 
    stations located within 100 km of the crossover location were selected.

(3) For water samples collected below 2500 m, smooth curves were fit through 
    the TA data as a function of the density anomaly referenced to 3000 db 
    (sigma-3) using Cleveland's loess or smoother local regression (Cleveland 
    and Devlin, 1988; Cleveland and Grosse, 1991; Chambers and Hastie, 1991). A 
    separate fit was performed to the data collected from each of the two 
    intersecting legs. The tension parameter for the smoother was adjusted 
    subjectively to give a 'reasonable' fit to the data at the majority of the 
    crossover locations, and the same value for the tension parameter was used 
    for all of the crossovers. Hence, while the fits to the data may not 
    necessarily represent the best possible at each individual crossover point, 
    the smoothing function has been applied consistently. It is important to 
    note that the comparison of the data at the crossover points does not depend 
    on the fitting algorithm within the experimental error.

(4) For each crossover, the difference between the two smooth curves was 
    evaluated at 50 evenly spaced intervals that covered the density range over 
    which the two data sets overlapped. A mean and a standard deviation of the 
    difference between the two curves was estimated based on these 50 values, 
    and these values are reported in Table 5 and shown in Fig. 5. An example of 
    the crossover for cruises I3-I5W/I4 is shown in Fig. 6.


Fig. 5.  Summary of the TA reproducibility for crossover points in the Indian 
         Ocean.
Fig. 6.  Results for a typical crossover comparison (I3-I5W/I4) in the Indian 
         Ocean.


The results of the crossover analysis indicate that absolute leg-to-leg 
differences are always <6 µmol/kg-1. Note that the comparisons were evaluated 
consistently such that the fit to data from the earlier leg at each crossover 
was subtracted from the fit to the later leg's data. Any uncorrected, long-
term, monotonic drift in the calibration of the titrators over the course of 
the Indian Ocean expedition would therefore tend to result in a non-zero 
value for the overall mean of these differences. The overall mean and 
standard deviation of the differences at all the crossovers are 2.1±2.1 µmol 
kg_1. In general, the results of the crossover analysis are quite consistent 
with the overall reproducibility of the CRM analyses (±4 µmol/kg-1) over the 
duration of the entire Survey.



4. CONCLUSION

At-sea total alkalinity measurements on the several CRM batches demonstrated 
that the measurements made by various investigators were precise to about ±4 
µmol/kg-1. This level of the precision of at sea measurements was 
approximately two times worse than that in the laboratory. Differences in the 
precision between different investigators suggest that the performance of TA 
measurements was dependent upon the operators. The inter-cruise variations in 
total alkalinity between laboratory and field results clearly demonstrate 
that CRMs are an essential component to monitor the performance of titration 
systems and increase the accuracy for total alkalinity measurements in the 
field.



ACKNOWLEDGEMENTS

The authors wish to acknowledge the support of the Department of Energy for 
their support of the CO2 studies. The WOCE cruises were supported by the 
National Science Foundation, as was some of the laboratory work related to 
the preparation and standardization of Certified Reference Material.



REFERENCES

Bradshaw, A.L. and Brewer, P.G., 1988. High precision measurements of 
    alkalinity and total carbon dioxide in seawater by potentiometric titration: 
    1. Presence of unknown protolyte(s)?. Mar. Chem. 28, pp. 69-86

Chambers, J.M., Hastie, T.J., 1991. Stat. Models Sci., 309-376.

Cleveland, W.S. and Devlin, S.J., 1988. Locally-weighted regression: an 
    approach to regression analysis by local fitting. J. Am. Statist. Assoc. 83, 
    pp. 596-610

Cleveland, W.S., Grosse, E., 1991. Computational Methods for Local 
    Regression. Stat. Comput., Vol. 1.
    
Dickson, A.G., 1981. An exact definition of total alkalinity and a procedure 
    for the estimation of alkalinity and total CO2 from titration data. Deep-Sea 
    Res. 28, pp. 609-623 

Dickson, A.G., 1984. pH scales and proton-transfer reactions in saline media 
    such as seawater. Geochim. Cosmochim. Acta 48, pp. 2299-2308
    
Dickson, A.G., 1990. The oceanic carbon dioxide system: planning for quality 
    data. US JGOFS News 2 2, p. 10
    
Dickson, A.G., 1990. Thermodynamics of the dissociation of boric acid in 
    synthetic seawater from 273.15 to 318.15 K. Deep-Sea Res. 37, pp. 755-766 
    
Dickson, A.G., 1990. Standard potential of the (AgCl+1/2 H2=Ag+HCl(aq)) cell 
    and the dissociation of bisulfate ion in synthetic sea water from 273.15 to 
    318.15 K. J. Chem. Thermodyn. 22, pp. 113-127

Dickson, A.G. and Riley, J.P., 1979. The estimation of acid dissociation 
    constants in sea water media from potentiometric titrations with strong 
    base: I. The ionic production of water-KW. Mar. Chem. 78, pp. 89-99 

Dickson, A.G. and Millero, F.J., 1987. A comparison of the equilibrium 
    constants for the dissociation of carbonic acid in seawater media. Deep-Sea 
    Res. 34, pp. 1733-1743 

DOE, 1994. Handbook of methods for the analysis of the various parameters of 
    the carbon dioxide system in sea water, In: Dickson, A.G., Goyet, C. (Eds.), 
    Version 2, ORNL/CDIAC-74.

Johansson, O. and Wedborg, M., 1982. On the evaluation of potentiometric 
    titrations of seawater with hydrochloric acid. Oceanol. Acta 5, pp. 209-218
    
Johnson, K.M., Dickson, A.G., Eischeid, G., Goyet, C., Guenther, P., Key, 
    R.M., Millero, F.J., Purkerson, D., Sabine, C.L., Schottle, R.G., Wallace, 
    D.R.W., Wilke, R.J. and Winn, C.D., 1998. Coulometric total carbon dioxide 
    analysis for marine studies: Assessment of the quality of total inorganic 
    carbon measurements made during the US Indian Ocean CO2 Survey 1994-1996. 
    Mar. Chem. 63, pp. 21-37
    
Marinenko, G. and Taylor, J.K., 1968. Electrochemical equivalents of benzoic 
    and oxalic acid. Anal. Chem. 40, pp. 1645-1651
    
Millero, F.J., 1995. The thermodynamics of the carbon dioxide system in 
    oceans. Geochim. Cosmochim. Acta 59, pp. 661-677
    
Millero, F.J. and Poisson, A., 1981. International equation of state of 
    seawater. Deep-Sea Res. 28, pp. 625-629
    
Millero, F.J., Laferriere, A. and Chetirkin, P.V., 1977. The partial molal 
    volumes of electrolytes in 0.725 m sodium chloride solutions at 25°C. J. 
    Phys. Chem. 81, pp. 1737-1745

Millero, F.J., Zhang, J.Z., Lee, K. and Campbell, D.M., 1993. Titration 
    alkalinity of seawater. Mar. Chem. 44, pp. 153-160
    
Taylor, J.K. and Smith, S.W., 1959. Precise coulometric titration of acids 
    and bases. J. Res. Natl. Bur. Stds. 63A, pp. 153-159
    















                                  APPENDIX D:

                        REPRINT OF PERTINENT LITERATURE




Sabine, C. L., R. M. Key, K. M. Johnson, F. J. Millero, A. Poisson, 
J. L. Sarmiento, D. W. R. Wallace, and C. D. Winn (1999), Anthropogenic 
CO2 Inventory of the Indian Ocean, Global Biogeochem. Cycles, 13(1), 179-198.





                ANTHROPOGENIC CO2 INVENTORY OF THE INDIAN OCEAN

C.L. Sabine,(1) R.M. Key,(1) K.M. Johnson,(2) F.J. Millero,(3) A. Poisson,(4) 
J.L. Sarmiento, D.W.R. Wallace,(2,5) and C.D. Winn,(6,7)

(1) Department of Geosciences, Princeton University, Princeton, New Jersey.
(2) Oceanographic and Atmospheric Sciences Division, Brookhaven National 
    Laboratory, Upton, New York.
(3) Rosenstiel School of Marine and Atmospheric Science, University of Miami, 
    Miami, Florida.
(4) Laboratoire de Physique et Chimie Marines, Universite Pierre et Marie 
    Curie, Paris.
(5) Now at Institut für Meereskunde, Universität Kiel.
(6) Department of Oceanography, University of Hawaii at Manoa, Honolulu, 
    Hawaii.
(7) Now at Hawaii Pacific University, Kaneohe, Hawaii.



GLOBAL BIOGEOCHEMICAL CYCLES, VOL. 13, NO.1, PAGES 179-198, MARCH 1999
Copyright 1999 by the American Geophysical Union.
Paper number 1998GB900022.
0886-62361991l998GB900022$12.00




ABSTRACT. 

This study presents basin-wide anthropogenic CO2 inventory estimates for the 
Indian Ocean based on measurements from the World Ocean Circulation 
Experiment/Joint Global Ocean Flux Study global survey. These estimates 
employed slightly modified ∆C* and time series techniques originally proposed 
by Gruber et al. [1996] and Wallace [1995], respectively. Together, the two 
methods yield the total oceanic anthropogenic CO2 and the carbon increase 
over the past 2 decades. The highest concentrations and the deepest 
penetrations of anthropogenic carbon are associated with the Subtropical 
Convergence at around 30° to 40°S. With both techniques, the lowest 
anthropogenic CO2 column inventories are observed south of 50°S. The total 
anthropogenic CO2 inventory north of 35°S was 13.6±2 Pg C in 1995. The 
inventory increase since GEOSECS (Geochemical Ocean Sections Program) was 
4.1±1 Pg C for the same area. Approximately 6.7±1 Pg C are stored in the 
Indian sector of the Southern Ocean, giving a total Indian Ocean inventory of 
20.3 ±3 Pg C for 1995. These estimates are compared to anthropogenic CO2 
inventories estimated by the Princeton ocean biogeochemistry model. The model 
predicts an Indian Ocean sink north of 35°S that is only 0.61-0.68 times the 
results presented here; while the Southern Ocean sink is nearly 2.6 times 
higher than the measurement-based estimate. These results clearly identify 
areas in the models that need further examination and provide a good baseline 
for future studies of the anthropogenic inventory.



1.  INTRODUCTION

The current Intergovernmental Panel on Climate Change (IPCC) estimate for the 
oceanic sink of anthropogenic CO2 (2.0 ±D.8 Pg C yr-I) is based primarily on 
ocean models [e.g., Sarmiento et al., 1992; Sarmiento and Sundquist, 1992; 
Siegenthaler and Sarmiento, 1993; Siegenthaler and Joos, 1992; Stocker et 
al., 1994], atmospheric models [e.g., Keeling et al., 1989; Keeling and 
Shertz, 1992] or on the oceanic distribution of related species such as δ13C 
[Quay et al., 1992]. Although the basic assumptions used in these methods are 
reasonably well grounded, there will always be room for doubt with indirect 
approaches. Direct estimates of the oceanic CO2 sink, however, have been 
primarily limited by a lack of high-quality data on a global scale.

Two general approaches can be used to estimate the uptake of anthropogenic 
CO2 by the oceans. One approach, initially proposed by Tans et al. [1990], is 
to use direct measurements of the air-sea difference in CO2 partial pressure 
together with global winds and a gas exchange coefficient to estimate the net 
transfer of CO2 into the oceans. These estimates, together with an 
atmospheric transport model, predicted that the oceanic sink was only 0.3 to 
0.8 Pg C yr-', much smaller than the model predictions. The difficulty with 
the ∆C02 approach lies both in the large uncertainty in the wind speed 
dependence of the air-sea gas exchange velocity and in the ability to 
properly represent the large temporal and spatial variability of the surface 
ocean pC02 because of a lack of seasonal, global data coverage. This estimate 
has recently been revised to 0.6 to 1.34 Pg C yr-I with the addition of more 
data and a lateral advection-diffusion transport equation to help with the 
necessary temporal and spatial interpolations [Takahashi et al., 1997].

A second approach, which avoids many of the problems of temporal variability, 
is to estimate the inventory of anthropogenic CO2 stored in the oceans 
interior based on inorganic carbon measurements. Again, the problem with this 
approach in the past has been a lack of high-quality global data coverage. As 
pointed out by Broecker et al. [1979] after completion of the last global 
oceanographic survey, GEOSECS (Geochemical Ocean Sections Program), the 
precision of ocean carbon measurements at that time was two orders of 
magnitude smaller than the predicted 0.035% annual increase in surface ocean 
dissolved inorganic carbon. Nearly 20 years have passed since GEOSECS, and 
the quality of today's carbon measurements has improved significantly.

This is the first of several papers aimed at estimating the anthropogenic CO2 
inventory of the oceans based on the recent global survey of CO2 in the 
oceans. The survey was conducted as part of the JGOFS (Joint Global Ocean 
Flux Study) in close cooperation with the WOCE-HP (World Ocean Circulation 
Experiment - Hydrographic Programme). This program was a multiyear effort to 
collect high-precision inorganic carbon data with the highest possible 
spatial resolution on a global scale. This paper will focus on anthropogenic 
CO2 estimates for the Indian Ocean. Papers will soon follow with estimates 
for the other major ocean basins, with the ultimate goal of generating an 
estimate of the global oceanic anthropogenic CO2 sink based on direct carbon 
system measurements. The strength of these calculations lies not only in our 
ability to directly estimate the magnitude of the oceanic anthropogenic CO2 
sink but also in the fact that these estimates can be directly compared to 
anthropogenic CO2 inventories estimated by carbon cycle ocean general 
circulation models (GCMs). The two methods described here provide information 
over different timescales. The combined results place strong constraints on 
the uptake rate for anthropogenic CO2 and are useful for identifying 
weaknesses in the models.



2.  METHODS

Estimates of the anthropogenic CO2 inventory are determined from measured 
values using two different techniques. The first technique, referred to as 
the "time series" approach, is based on quantifying the increase in total 
carbon dioxide (TC02) since GEOSECS. The second approach quantifies the total 
anthropogenic CO2 inventory using a quasi-conservative tracer, ∆C*. Although 
the general idea for both techniques has been around for a long time, recent 
improvements in the estimation of the preserved end-member concentrations 
together with significant Improvements in the accuracy and spatial coverage 
of global carbon data give us much more confidence in these results. Given 
the difficulty of isolating the anthropogenic signal from the large TC02 
background, however, it is relevant to summarize the quality of the carbon 
data set and the techniques used to estimate the anthropogenic signal.


2.1.  DATA QUALITY

Over 20,000 water samples collected between December 1994 and July 1996 as 
part of the U.S. WOCE Indian Ocean survey were analyzed for both TC02 and 
total alkalinity (TA) using standard coulometric and potentiometric 
techniques, respectively. Figure 1 shows the locations of the 1352 stations 
occupied by U.S. WOCE as part of the Indian Ocean survey together with the 
station locations from the GEOSECS Indian Ocean Survey and the French INDIGO 
I, II, and III and CIVA-1 (WOCE designation I6S) Cruises Details of the 
WOCE/JGOFS Indian Ocean CO2 measurement program, including personnel, 
sampling procedures, measurement protocols and data quality assurance/quality 
control checks are described elsewhere [Johnson et al., 1998; Millero et al., 
1998a]. Calibrations of both the TC02 and TA systems were checked 
approximately every 12 hours by analyzing Certified Reference Material (CRM) 
samples with known concentrations of TC02 and TA [Dickson, 1990] (A.G. 
Dickson, Oceanic carbon dioxide quality control at http://www-mpl.ucsd. 
edulpeople/adicksonlC02_QC/, 1998). On the basis of these CRM analyses the 
accuracy of the TC02 and TA measurements was estimated to be ±2 and ±4 µmol 
kg-1, respectively. Primary hydrographic data from the conductivity- 
temperature-depth/Rosette were collected and analyzed following standard 
procedures [Millard, 1982]. Samples were collected for salinity on every 
bottle and analyzed with an Autosal salinometer using standard techniques 
[UNESCO, 1981]. Oxygen samples were analyzed with an automated system using a 
modified Winkler technique [Culberson et al., 1991]. Nutrients were analyzed 
on a four-channel Technicon AutoAnalyzer II following the methods of Gordon 
et al. [1992]. Chlorofluorocarbon samples were analyzed on a gas 
chromatograph using the techniques of Bullister and Weiss [1988]. Complete 
details of the analytical protocols and personnel can be obtained from the 
individual cruise reports available through the WOCE Office. 


FIGURE 1:  Station locations for WOCE Indian Ocean (circles), 
           CIVA-1/I6S (crosses), INDIGO I (stars), INDIGO 
           II (inverted triangles), INDIGO III (triangles), 
           and GEOSECS (solid squares) Indian Ocean Surveys. 
           Numbered boxes indicate location of crossovers 
           discussed in the text. Map generated using Generic 
           Mapping Tools version 3 [Wessel and Smith, 1995].


All of the data available at the time this manuscript was written have been 
included in the Indian Ocean analysis. For the primary hydrographic and 
nutrient data this means that the preliminary values available at the 
conclusion of the cruise were used. While we would prefer to use the final 
hydrographic data, typical post cruise corrections for the WOCE data sets are 
well below noise level for these calculations. Preliminary to semifinal 
chlorofluorocarbon (CFC) data were used to estimate the water age necessary 
for one of the correction terms in the ∆C* method. Although post cruise blank 
corrections can influence the final CFC concentrations, an examination of the 
existing data (except 18NI5E because data were not available at time of 
writing) indicated that the CFC-ll and CFC-12 age comparisons as well as 
comparisons of the data from one leg to the next were reasonably consistent 
with each other. The calculations were limited to waters with CFC-12 ages of 
less than 40 years where potential blank corrections are a relatively small 
fraction of the signal and mixing effects are minimized. The carbon data, 
which primarily influence the quality of the calculations, have all been 
calibrated and finalized as discussed briefly below.

Examination of Figure 1 reveals that although the WOCE survey was extensive, 
a large data gap exists in the southwestern Indian Ocean. To fill in this 
gap, data from the three French survey legs INDIGO I (February-March 1985), 
IT (April 1986) and 1lI (January-February 1987) as well as the more recent 
French cruise CIVA-1 (February-March 1993 (WOCE designation 16S» were 
included in the analysis [Poisson et al., 1988; 1989; 1990]. TC02 and TA were 
analyzed on the INDIGO cruises using standard potentiometric titration 
techniques developed by Edmond [1970]. Potentiometric titrations were also 
used to analyze the TA samples on CIVA-1, but the TC02 samples were analyzed 
using the coulometric techniques of Johnson et al. [1985]. The internal 
consistency of these cruises was examined by comparing carbon values in the 
deep waters (pressure> 2500 dbars) at the intersections of different legs. 
The stations selected for each crossover were those with carbon values which 
were closest to the intersection point. Smooth curves were fit through the 
data from each cruise as a function of sigma- 3 (density anomaly referenced 
to 3000 dbars) using Cleveland's loess function [Cleveland and Devlin, 1988; 
Cleveland et al., 1992]. The difference between the curves was evaluated at 
50 evenly spaced intervals that covered the density range over which the two 
data sets overlapped. The mean and standard deviation of the difference in TA 
and TC02 at the 35 intersections identified in Figure 1 are shown in Figure 
2. The long-term stability of the WOCE/JGOFS measurements can be estimated 
from the first 17 crossover results. The mean of the absolute values for the 
leg-to-leg differences was less than the estimated accuracy for both TC02 
(1.8 ±0.8 µmol kg-1) and TA (2.4 ±1.6 µmol kg-1). Although there is only one 
reliable crossover point between the WOCE/JGOFS cruises and the CIVA-1 (I6S) 
cruise, the differences for both parameters are within the estimated accuracy 
of the measurements. Results from the analysis of CRM samples on the CIVA-1 
cruise also support the quality of the measurements. Some of the older INDIGO 
cruises, however, did appear to have offsets relative to the WOCE/JGOFS and 
CIVA-1 data. INDIGO I and IT alkalinity values averaged 6.5 µmol kg-1 high 
and 6.8 µmol kg-1 low, respectively, while the INDIGO 1lI alkalinity values 
showed no clear offset. The INDIGO TC02 values were all consistently high 
relative to WOCE/JGOFS and CIVA-1, with differences of 10.7,9.4, and 6.4 µmol 
kg-1, respectively. These offsets are consistent with differences observed 
between at-sea values and replicate samples run at C.D. Keeling's shore-based 
TC02 facility (P. Guenther, personal communication, 1998). Since the INDIGO 
cruises were run prior to the introduction of CRMs, these offsets were 
presumed to be calibration differences, and each leg was adjusted to bring 
the values in line with the remaining cruises. The dotted boxes in Figure 2 
show the original offsets at the crossovers. The solid boxes show the final 
offsets used in the following calculations. The means of the absolute values 
for the leg-to-leg differences for all 35 crossover analyses suggest that the 
final data set is internally consistent to ± 2.2 and 3.0 µmol kg-1 for TC02 
and TA, respectively.


2.2.  "TIME SERIES" CALCULATIONS

The "time series" method for estimating the increase in the anthropogenic 
inventory uses measurements of TC02 made at a certain point in time to 
develop a predictive equation based on a multiple linear regression of the 
observed TC02 and simultaneously measured parameters such as temperature, 
salinity, oxygen, and TA (or silicate). These empirical multiparameter 
relationships have been shown to hold over large spatial scales, and their 
use drastically reduces the complicating effects of natural variability in 
determining temporal trends [Brewer et al., 1995; Wallace, 1995; Brewer et 
al., 1997]. The TC02 residuals from such predictive equations can be compared 
directly with patterns of residuals evaluated using the same predictive 
equation with TA, oxygen, and hydrographic data collected at different times 
(e.g., over decadal intervals). Since the uptake of anthropogenic CO2 will 
increase the TC02 of the waters but will not directly affect the concentrations 
of the fit parameters, systematic changes in the magnitude and distribution of 
the TC02 residuals over time provide a direct estimate of the oceanic CO2 
inventory change due to the uptake of anthropogenic CO2. The most comprehensive 
historical carbon data set for the Indian Ocean is from the GEOSECS expedition. 
By examining the WOCE data relative to that collected during the 1977-1978 
GEOSECS Indian Ocean Survey, the increase in anthropogenic inventory over the 
last 18 years can be estimated.


2.2.1.  GEOSECS FIT 

All of the GEOSECS data from the Indian Ocean (excluding Gulf of Aden and Red 
Sea regions) were fit with a single predictive equation as a function of 
potential temperature (9), salinity (S), apparent oxygen utilization (AOU), 
and TA. To minimize the influence of short-term temporal variability, only 
data from pressures greater than 200 dbars were included in the fit. Despite 
the large area covered, the GEOSECS TC02 values can be predicted from this 
equation to ± 5.2 µmol kg-1 (~ =0.992 and N = 1120). There is, however, a 
pattern in the residuals that correlates with observed hydrographic regions 
in the Indian Ocean (Figure 3).

In an attempt to improve the fit, a categorical variable based on region was 
added to the regression. The categorical variable differs from the other 
continuous variables by the fact that it is either applied or not applied 
depending on whether the sample is located within the region. The regions 
were defined as follows: I, Arabian Sea (north of l0°N and west of 78°E); 2, 
North of 10°S (excluding Arabian Sea); 3, Chemical Front (21°S to 10°S); 4, 
Central Gyre (35°S to 21°S); and 5, Southern Ocean (south of 35°S)

The addition of the regional variable resulted in a marginal improvement in 
the fit (~ =0.993 and (J =4.9 µmol kg-1) but more importantly, removed the 
regional bias in the predictive equation. The coefficients of the final fit 
are shown in Figure 4 along with a plot of the measured versus calculated 
TC02 values for all of the points used in the fit. The resulting equation was 
then used to generate TC02 values for each of the WOCE sample locations based 
on the measured temperature, salinity, oxygen, and TA values. The difference 
between the measured TC02 and the predicted TC02 reflects the CO2 increase in 
the time between the two cruises. For this work the difference is referred to 
as "excess CO2,"


FIGURE 2:  Mean difference between deep water values of (a) TA 
           and (b) TC02 for cruise intersections identified in 
           Figure 1. Bars indicate one standard deviation. Dotted 
           boxes indicate difference before adjustment (see 
           explanation in text).

FIGURE 3:  Box and whiskers plot of residuals from a multiple linear 
           regression of GEOSECS Indian Ocean data (pressure> 200 
           dbars) fit without the regional designator versus 
           oceanographic region: TC02 = 706.5 + 7.7S - 6.689 + 
           0.513TA + 0.7257AOU. Solid boxes cover the range of ±1 
           standard deviation about the mean. White lines within 
           the boxes indicate median values. The whiskers indicate 
           the range of data within the 99% confidence interval. 
           The bars outside the whiskers give the values of 
           outliers in the data set.

FIGURE 4:  Plot of measured GEOSECS TC02 versus the calculated 
           values. Solid line shows 1:1 relationship. The dashed 
           lines indicate the 99% confidence interval for the 
           fit. Text gives coefficients and related statistics. 
           The column labeled "Pr(>|T|)" gives the probability 
           that the T value in the previous column is larger 
           than the T table value in a student T test.


The residual method of estimating excess CO2 was applied to the water column 
below 200 dbars. The surface waters, however, are dominated by seasonal 
variability which can bias the residual calculations. The excess CO2 of the 
surface waters therefore was determined from the difference in the estimated 
annual mean TC02 concentrations between GEOSECS and WOCE. The annual mean 
TC02 concentration was calculated from TA and surface water ƒC02. The surface 
alkalinity was estimated from the gridded annual mean salinity and 
temperature values of Levitus et al. [1994] and Levitus and Boyer [1994] 
using a multiple linear fit of the WOCE/JGOFS surface (pressure < 60 dbars) 
TA data to the measured surface temperature and salinity. The 1978 and 1995 
surface water jC02 concentrations were estimated from the annual mean 
atmospheric concentration for the 2 years. and the annual mean ∆pC02 values 
estimated from the full correction scheme of Takahashi et al. [1997]. The 
excess TC02 values between the surface and 200 dbars were estimated with a 
linear approximation between the surface and 200 dbars values for each 1° 
grid box.


2.2.2. DATA CONSISTENCY. 

One of the major concerns with the time series technique is the necessity of 
having two data sets that are consistent with each other. This consistency 
can be well documented for both TC02 and TA today through the use of 
certified reference materials (CRMs) supplied by A. Dickson of Scripps 
Institute of Oceanography (SIO). Since CRMs were not available at the time of 
GEOSECS. the only way to infer consistency with the WOCE data set is to 
assume the deep water carbon distributions have not changed since GEOSECS. 
The most reliable way to compare the two data sets is to examine the 
difference between the predicted TC02 and the measured TC02 (excess CO2) in 
deep waters. The basic assumption with this technique is that the correlation 
between the different hydrographic parameters in the deep waters does not 
change with time. Given the long residence time of the deep and bottom waters 
in the ocean. this should be a reasonable assumption. This technique has the 
advantage that it implicitly accounts for the possibility of real variability 
in hydrographic properties between the two expeditions which would not be 
taken into account by simply comparing carbon profiles.

Examination of the excess CO2 values in waters that should be free of 
anthropogenic CO2 (pressures> 2000 dbars and containing no detectable 
chlorofluorocarbons) revealed that the GEOSECS values were 22.5 ±3 µmol kg-1 
higher than the comparable WOCE measurements. This difference is comparable 
to the correction of 18 ± 7 µmol kg-1 noted by Weiss et al. [1983] to make 
the TC02 measurements consistent with the TA and discrete CO2 partial 
pressure measurements based on the Merbach et al. [1973] dissociation 
constants. Additional support for an adjustment of the original GEOSECS data 
comes from C. D. Keeling's shore-based analysis ofTC02 samples collected on 
both the GEOSECS and the WOCE/JGOFS expeditions. Weiss et al. [1983] point 
out that the shore-based analyses of Keeling were systematically smaller than 
the at-sea measurements by 16.5 ± 5 µmol kg-1 during GEOSECS. Similar 
comparisons between the WOCE/JGOFS at-sea measurements with Keeling's shore-
based analyses indicate that the shore based samples are approximately 5 µmol 
kg-1 higher than the at sea values (P. Guenther. personal communication. 
1998). Together. the GEOSECS-Keeling-WOCE/JGOFS combination suggests an 
offset of 21.5 µmol kg-1 between GEOSECS and WOCE/JGOFS at-sea measurements. 
It is also important to note that there is no indication of a depth or 
concentration dependent correction for the GEOSECS data. The shore-based 
comparison. based only on samples collected at the surface. is within I µmol 
kg-1 of the deep comparison described above. On the basis of these results a 
constant correction of the -22.5 µmol kg-1 was applied to the GEOSECS TC02 
values to improve the consistency of the two data sets.

Ideally. the data used in the time series calculations would cover the same 
geographic region with as much of a time difference as possible. The trade-
off. however. is that the quality and spatial coverage of the older data sets 
is generally very limited. Given the relatively small area of overlap between 
the WOCE/JGOFS and INDIGO data sets and the shorter time difference between 
cruises (9 years versus 18 years for WOCE - GEOSECS). the time series 
analysis was limited to a comparison between WOCE/JGOFS and GEOSECS in the 
main Indian Ocean basin.


2.2.3.  EVALUATION OF ERRORS 

An estimate of the random errors associated with the excess CO2 calculation 
can be made with a simple propagation of errors based on the fit to the 
GEOSECS data and the estimated precision of the WOCE/JGOFS data. With a 
standard deviation of 4.9 µmol kg-1 for the GEOSECS fit and an estimated 
long-term precision of ±2 µmol kg-1 in the WOCE/JGOFS TC02 values the excess 
CO2 error is estimated to be approximately ±5 µmol kg-1. This value compares 
well with the standard deviation of 3.5 µmol kg-1 for the excess CO2 below 
the maximum anthropogenic CO2 penetration depth (pressure> 1500 dbars).

Systematic errors with this technique are very difficult to evaluate. The 
largest potential systematic error is probably associated with the surface 
water estimates. Because the same ∆pC02 value is used to estimate the TC02 
for both years. the excess CO2 (1995 TC02 - 1978 TC02) is not very sensitive 
to potential errors associated with the actual ∆pC02 values used. The surface 
estimate is sensitive. however. to the assumption that the ∆pC02 has not 
changed over time (i.e.. that the surface ocean increase has kept pace with 
the atmospheric increase). It is not likely that the surface ocean has 
increased at a faster rate than the atmosphere. but it is conceivable that 
the rate is slower. The current assumption results in a total inventory of 
0.8 Pg C in the surface layer. If the surface ocean were increasing at half 
the rate of the atmosphere. the systematic bias in the final inventory would 
be around 0.4 Pg C. Below the surface layer the most likely systematic error 
would result from the uncertainty in fitting the GEOSECS data. Systematic 
errors associated with calibration differences between cruises are 
potentially quite large. but the analysis and subsequent correction given in 
section 2.2.2 should remove these biases. The estimated uncertainty for the 
GEOSECS adjustment was ±3 µmol kg-1. If this value is integrated for the area 
north of 35°S between 200 m and the average penetration depth of the excess 
CO2 (~800 m). the potential error would be ±0.9 Pg C. Propagating the errors 
for the surface and deeper layers gives an estimated error of approximately 
±1 Pg C in the total excess CO2 inventory. Clearly, there are other ways of 
estimating the potential errors in these calculations. but we feel that this 
is a reasonable estimate based on the available data.


2.3.  ∆C* CALCULATIONS

Gruber et al. [1996] developed a method to estimate the total anthropogenic 
CO2 inventory which has accumulated in the water column since pre-industrial 
times. Although the details of the calculation are thoroughly discussed by 
Gruber et al. the basic concept of the calculation can be expressed in terms 
of the following equation:

                C(anth)(µmol/kg) = C(m)-∆C(bio)-C(280)-∆C(dis)            (1)

where
     C(anth) anthropogenic carbon concentration;
     C(m)    measured total carbon concentration;
     ∆C(bio) change in TC02 as a result of biological activity;
     C(280)  TC02 of waters in equilibrium with an atmospheric CO2
             concentration of 280 µatm;
     ∆C(dis) air-sea difference in CO2 concentration expressed in µmol
             kg-1 of TC02.

The Gruber et al. technique employs a new quasi-conservative tracer ∆C*, 
which is defined as the difference between the measured TC02 concentration, 
corrected for biology, and the concentration these waters would have at the 
surface in equilibrium with a pre-industrial atmosphere (i.e., ∆C* = C(m) - 
∆C(bio) - C(280)). Rearranging (I) shows that ∆C* reflects both the 
anthropogenic signal and the air-sea CO2 difference (i.e., ∆C* = C(anth) + 
∆C(dis)). The airsea disequilibrium component can then be discriminated from 
the anthropogenic signal using either information about the water age (e.g., 
from transient tracers such as CFCs or 3H-3He) or the distribution of ∆C* in 
regions not affected by the anthropogenic transient. The details of this 
technique will not be covered here except as necessary to explain small 
modifications that were necessary for use with the WOCE Indian Ocean data 
set.


2.3.1.  PREFORMED ALKALINITY EQUATION 

The first modification to the Gruber et al. [1996] technique involved a 
recalculation of the preformed alkalinity equation. The preformed alkalinity 
(Alk^0) of a subsurface water parcel is an estimate of the alkalinity that 
the water had when it was last at the surface. This value is necessary to 
determine the equilibrium concentration (C(280)) of the waters. Gruber et al. 
generated a single global equation for estimating Alk^0 from salinity and the 
conservative tracer "PO" (PO = 02+l70*P) [Broecker, 1974] based on the data 
collected during GEOSECS, South Atlantic Ventilation Experiment, Transient 
Tracers in the Ocean/North Atlantic Study and Transient Tracers in the Ocean/ 
Tropical Atlantic Study. Given the limited representation of the Indian Ocean 
in these data and the improved quality of today's measurements, the Gruber et 
al. fit was examined for a possible bias with respect to the WOCE/JGOFS 
results. Alk^0 values calculated from the Gruber et al. equation were found 
to be, on average, 7 ±12 µmol kg-1 lower than the WOCE/JGOFS measured surface 
alkalinity values. Rather than making assumptions about which parameters 
would provide the best fit to the surface alkalinity data, several possible 
parameters were tested based on previously noted correlations. Although 
salinity has been shown to generally correlate very strongly with surface 
alkalinity [Brewer et al., 1986; Millero et al., 1998b], some areas, such as 
the high-latitude regions, require additional parameters to fit regional 
changes in alkalinity. Some investigators have used temperature as an 
additional variable [e.g., Chen and Pytkowicz, 1979; Chen, 1990; Millero et 
al., 1998b]. Others, such as Gruber et al. [1996], have used other 
conservative tracers to compensate for the regional differences. The best fit 
for the WOCE/JGOFS, INDIGO, and CIVA Indian Ocean data, with pressures less 
than 60 dbars, is given by (2):

                 Alk^0 = 378.1+55.22 x S+0.0716 x PO-1.236 x θ            (2)

Alk^0 has units of µmol kg-1 when salinity (S) is on the practical salinity 
scale, PO is in µmol kg-1, and potential temperature (θ) is in degrees 
Celsius. The standard error in the new Alk^0 estimate is ±8.0 µmol kg-1 based 
on 2250 data points. A standard ANOVA analysis of the fit shows that all four 
terms are highly significant (Table 1). Reevaluating the Alk^0 equation not 
only removed the 7 µmol kg-1 offset of Gruber's equation but also resulted in 
a 35% reduction in the uncertainty. 


TABLE 1: Results From ANOVA Analysis of Alk^0 Fit.
         ____________________________________________________

                   Coefficient  Standard  T Value   Pr(>|T|)
                                 Error
                   -----------  --------  --------  --------
          Intercept  378.1       8.9       42.2715  0.0000
          Salinity    55.22      0.23     235.0369  0.0000
          PO           0.0716    0.0041    17.4693  0.0000
          Theta       -1.236     0.061    -20.3697  0.0000
         ____________________________________________________
          The column labeled "Pr(>rr1)" gives the probability 
          that the T value in the previous column is larger 
          than the T table value in a student T test. Alk^0 
          is preformed alkalinity. an estimate of the alka-
          linity of a parcel of subsurface water when it was 
          last at the surface.


2.3.2.  DENITRIFICATION CORRECTION 

A second modification to the original ∆C* technique was necessary to properly 
account for the anoxic regions in the northern Indian Ocean. The C(bio) term 
in (1) assumes that the remineralization of carbon in the interior of the 
ocean occurs in proportion to the oxygen uptake based on a standard Redfield 
type stoichiometry. The ratios used for these calculations were based on the 
global estimates of Anderson and Sarmiento [1994]. Gruber et al. [1996] 
demonstrated that the errors in the ∆C* calculation due to uncertainties in 
the C:O stoichiometric ratio only become significant for AOU values greater 
than 80 µmol kg-1. Given that most of the anthropogenic CO2 is found in 
relatively shallow waters with low AOU, this error, on average, is small. For 
some regions of the Arabian Sea, however, oxygen depletion can be quite 
extensive at relatively shallow depths [Sen Gupta et al., 1976; Deuser et 
al., 1978; Naqvi and Sen Gupta, 1985]. In areas where the waters become 
anoxic, denitrification can significantly alter the dissolved carbon to 
oxygen ratio [Naqvi and Sen Gupta, 1985; Anderson and Dyrssen, 1994; Gruber 
and Sarmiento, 1997]. The dissolved carbon generated by denitrification shows 
up as high ∆C* values as demonstrated at the northern end of the section in 
Figure 5a. The distribution of ∆C* values along the density surface σθ = 26.9-
27.0 shows maximum values at both the northern and southern ends of the 
section. One would expect the uptake of anthropogenic CO2 to generate the 
highest values close to the outcrop region in the south, but this surface 
does not outcrop in the north. Following the methods of Gruber and Sarmiento 
[1997], the denitrification signal can be estimated using the N* tracer. N* 
is a quasi-conservative tracer which can be used to identify nitrogen (N) 
excess or deficits relative to phosphorus (P). Using the global equation of 
Gruber and Sarmiento [1997], N* is defined as

                       N*(µmol/kg) = 0.87(N - 16P + 2.90)                 (3)

Figure 5b shows the magnitude of the denitrification signal along the σθ = 
26.9-27.0 surface. The N* values were converted from nitrogen units to µmol C 
kg-1 based on a denitrification carbon to nitrogen ratio of 106:-104 [Gruber 
and Sarmiento, 1997]. Negative values reflect nitrogen fixation, while 
positive values indicate denitrification. As expected, the values of N* are 
essentially zero in the main Indian Ocean basin but show a strong 
denitrification signal at middepths in the Arabian Sea. The low N* values at 
the north end of this surface (Figure 5b) are from the Bay of Bengal and show 
little or no denitrification in this region. Subtracting a denitrification 
correction term from the original ∆C* equation lowers the high ∆C* values at 
the northern end of the section leaving the expected maximum near the outcrop 
region (Figure 5c).

The final definition for ∆C* as used in this work is given by (4)

                  ∆C* = TCO2^meas - TCO2^(S,T,Alk^0,280)
                      -(117/-170)(O2-O2^(sat))
                      -(1/2)(TA-ALK^0+(16/-170))O2-O2^(sat))
                      -(I06/-104)N*                                       (4)

where TC02(meas), TA, and O2 are the measured concentrations for a given 
water sample in µmol kg-1. Alk^0 is the preformed alkalinity value as 
described in section 2.3.1. 02^(sat) is the calculated oxygen saturation 
value that the waters would have if they were adiabatically raised to the 
surface. TCO2^(S,T,Alk^0,280) is the TC02 value the waters would have at the 
surface with a TA value equal to Alk^0 and ƒC02 value of 280 µatm.


2.3.3.  ESTIMATION OF AIR-SEA DISEQUILIBRIUM 

To isolate the anthropogenic CO2 component from ∆C*, the air-sea 
disequilibrium values (∆C(dis)) must be determined. Gruber et al. [1996] 
described two techniques for estimating these values on density surfaces. For 
deeper density surfaces one can assume that the waters far away from the 
outcrop region are free from anthropogenic CO2. The mean ∆C* values in these 
regions therefore reflect only the disequilibrium value. For shallower 
surfaces the air-sea disequilibrium can be inferred from the ∆C*I tracer.

∆C*(t) is the difference between C* and the concentration the waters would 
have in equilibrium with the atmosphere at the time they were last at the 
surface. The time since the waters were in contact with the surface is 
estimated from CFC-12 age (t) and the atmospheric CO2 concentration history 
as a function of time (ƒC02{t(sample)-t}). The atmospheric CO2 time history 
from 1750 through 1996 was determined from a spline fit to ice core and 
measured atmospheric values [Neftel et al., 1994; Keeling and Whorf, 1996]. 
The CFC-12-based ages were determined following the technique described by 
Warner et al. [1996]. The apparent age of the water is determined by matching 
the CFC-12 partial pressure (pCFC-12) of the waters with the atmospheric CFC-
12 concentration history (procedures and atmospheric time history provided by 
J. Bullister). Although CFCs do not give a perfect representation of the true 
calendar age of the waters, Doney et al. [1997] have shown that the CFC-12 
and 3H-3He ages in the North Atlantic agree within 1.7 years for ages less 
than 30 years. Gruber [1998] successfully used both CFC and 3H-3He ages for 
his disequilibrium calculations in the Atlantic and has thoroughly discussed 
the assumptions and caveats associated with these techniques. The disequilibrium 
values on shallow density surfaces presented here were calculated using CFC-12 
ages modified from the ∆C*(t) equation of Gruber [1998] to include the 
denitrification correction: 

       ∆C*(t) = TCO2^meas - TCO2^(S,T,Alk^0,ƒC02{t(sample)-t})
              -(117/-170)(O2-O2^(sat))
              -(1/2)(TA-ALK^0+(16/-170))O2-O2^(sat))
              -(I06/-104)N*                                               (5)

where TCO2^(S,T,Alk^0,ƒC02{t(sample)-t}) is the TC02 the waters would have at 
the surface with a TA value of Alk^0 and ƒC02 value in equilibrium with the 
atmospheric CO2 concentration at the time the waters were last at the surface 
(date of sample collection minus CFC age).

The CFC age method was used for waters with densities less than σθ = 27.25 and 
CFC-12 ages less than 40 years. The anthropogenic CO2 of the waters with 
pressures less than 150 dbars or densities less than σθ = 25.95 was determined 
by subtracting the ∆C*(t) value estimated at each sample location from the 
corresponding ∆C* value. Given that the Indian Ocean does not extend into the 
high northern latitudes, the major outcrop region for Indian Ocean waters 
below the mixed layer is toward the south. Although other tracers might be 
used to identify multiple end-members, the CFC-12 ages on each density 
surface get steadily older toward the north, and the ∆C*(t) values are 
reasonably constant (see diamonds in Figure 6). This suggests that most of 
the water in the Indian Ocean is derived from the south or, at least in terms 
of the air-sea disequilibria, cannot be distinguished from other sources. The 
∆C(dis) term for the main Indian Ocean basin therefore was determined from a 
mean ∆C*(t) value on each surface. The mean ∆C(dis) terms were then 
subtracted from the individual ∆C* values to determine the anthropogenic 
component. Table 2 summarizes the ∆C(dis) values for the density surfaces 
estimated exclusively from the ∆C*(t) method.

One major exception to the southern source waters is observed in the Arabian 
Sea. Although none of the surfaces with σθ values greater than 26.0 outcrop in 
the Arabian Sea, a number of higher density surfaces do outcrop in the Red 
Sea and Persian Gulf. These outcrops could provide pathways for the 
introduction of CFCs and anthropogenic CO2 into the northern Arabian Sea and 
could reset the disequilibria term. Wyrtki [1973] noted that the Red Sea and 
Persian Gulf waters mix in the Arabian Sea to form the high-salinity North 
Indian Intermediate Water (NIIW). The ∆C*(T) values in the Arabian Sea do 
vary significantly and generally have a strong correlation with salinity. The 
CFC-12 ages also begin to get younger toward the northern end of the Arabian 
Sea. These high salinity waters appear to have a higher disequilibria term 
than the lower-salinity waters that make up the majority of the Indian Ocean 
intermediate waters.

To account for this phenomenon, the Arabian Sea waters (north of 5°N and west 
of 78°E) were isolated, and the ∆C*(t) values were fit against salinity with 
a linear regression. Thus this region was treated as a two-end-member mixing 
scenario between the high salinity NIIW and the lower-salinity waters of the 
main Indian Ocean basin. The ∆C(dis) values in this region were determined 
based on the relative contributions of the two end-members using salinity as 
a conservative tracer. The coefficients for the Arabian Sea fits are given in 
Table 2. The difference between the high salinity and lower-salinity 
disequilibria generally decreased as densities increased (note decreasing 
slope values in Table 2) to the point where the Arabian Sea disequilibria 
values were no longer distinguishable from the main Indian Ocean basin 
values. The additional terms were dropped for surfaces where the two end 
member mixing terms resulted in values within the error of the basin-wide 
mean (Table 2).

As stated previously, the disequilibria term for the deeper, CFC free 
surfaces was determined directly from the mean ∆C* value of each density 
interval. Careful examination of the extent of CFC penetration along the 
density surface was used to limit data used in determining the ∆C(dis) term. 
Only regions where CFC concentrations were below a reasonable blank (0.005 
pmol kg-1) were considered. The ∆C(dis) values determined using this method 
are summarized in the lower half of Table 3 (σθ > 27.5).


FIGURE 5:  ∆C* values for data on the 26.9 - 27.0 σθ surface: 
           (a) calculated without denitrification, (b) denitri-
           fication signal put in terms of ∆C*, (c) with 
           denitrification correction (i.e., data in Figure 
           5a minus the data in Figure 5b).


Determination of the ∆C(dis) values for either shallow or deep surfaces is 
relatively straightforward using the techniques mentioned above. It is not 
straightforward, however, to estimate the ∆C(dis) values for intermediate 
levels where the CFC ages are relatively old and may be significantly 
influenced by mixing and yet the waters could have enough anthropogenic CO2 
to influence the estimates based on ∆C*. The effect of using the ∆C* 
technique in waters that actually have anthropogenic CO2 would be to 
overestimate the ∆C(dis) term and thus underestimate the anthropogenic CO2, 
The effect of mixing on the CFC ages, however, generally results in an 
underestimation of the CFC age which would lead to an underestimation of the 
∆C(dis) term and an overestimation of the anthropogenic CO2, The CFC age 
technique has additional problems in waters with σθ values greater than 27.25, 
because the waters with the younger ages are all found in the very high 
latitudes of the Southern Ocean and generally are not directly ventilated in 
these regions. Therefore the basic assumption that the ∆C(dis) term can be 
determined by following the density level to its outcrop and examining the 
younger waters there is not valid.


TABLE 2: Values of ∆C(dis) Determined on Potential Density (σθ) Intervals
_______________________________________________________________________

  Potential    Main Basin     Main       Arabian    Arabian  Arabian
   Density     Mean (SDM)    Basin #    Intercept    Slope    # of 
    Range                   of Points     (SD)       (SD)    Points
 -----------  ------------  ---------  ----------  --------  -------
 25.95-26.05   -1.3(±0.88)     56      -740(±92)   21.3(±3)  12
 26.05-26.15   -0.7(±1.21)     42      -745(±130)  21.4(±4)  12
 26.15-26.25   -3.4(±0.65)     63      -699(±76)   20.0(±2)  11
 26.25-26.35   -4.8(±0.62)     61      -516(±90)   14.8(±3)  12
 26.35-26.45   -5.6(±0.48)     83      -316(±84)    9.1(±2)  20
 26.45-26.55   -7.1(±0.34)    103      -558(±87)   15.9(±2)  21
 26.55-26.65   -7.2(±0.32)    123      -512(±53)   14.5(±I)  28
 26.65-26.75   -8.9(±0.27)    152      -397(±52)   11.2(±I)  34
 26.75-26.85   -9.1(±0.23)    254      -428(±66)   12.0(±2)  28
 26.85-26.95  -11.2(±0.31)    209      -285(±115)   7.9(±3)   6
 26.95-27.00  -12.2(±0.35)    104
 27.00-27.05  -13.8(±0.48)     92
 27.05-27.10  -15.2(±0.4O)     90
 27.10-27.15  -16.3(±0.47)     84
 27.15-27.20  -17.1(±0.51)     89
 27.20-27.25  -19.5(±0.56)     74
_______________________________________________________________________
 Standard deviations (SD) are given for the slope and intercept terms 
 for the Arabian Sea data. Standard deviation of the mean (SDM, i.e., 
 standard deviation weighted by the number of individual determinations) 
 is given for each main basin estimate.  Values of ∆C(dis) are given in 
 µmol kg-1. Dashes indicate value not determined.


TABLE 3: Values of ∆C(dis) Determined on Potential Density (σθ) Intervals
_________________________________________________________________________

  Potential     Mean ∆C*      # of     Mean ∆C*l     # of    Final Mean
   Density        (SDM)      Points      (SDM)      Points   ∆Cdi5(SDM)
    Range        
 -----------  ------------   ------   ------------  ------  ------------
 27.25-27.30   -2.3(±0.45)     42     -19.7(±0.98)    22    -8.3(±1.l3)
 27.30-27.35   -4.0(±0.49)     45     -21.0(±0.84)    19    -9.1(±1.06)
 27.35-27.40   -5.3(±0.44)     72     -22.5(±1.25)     7    -6.8(±0.69)
 27.40-27.45   -7.1(±0.26)     92     -23.5(±0.83)    10    -8.7(±0.54)
 27.45-27.50   -7.9(±0.30)     98     -25.0(±1.65)     7    -9.0(±0.51)
 27.50-27.55   -9.3(±0.28)     93                            -9.3(±0.28)
 27.55-27.60  -10.7(±0.28)     92                           -10.7(±0.28)
 27.60-27.65  -11.3(±0.34)    125                           -11.3(±0.34)
 27.65-27.70  -13.0(±0.36)    127                           -13.0(±0.36)
 27.70-27.75  -14.8(±0.30)    184                           -14.8(±0.30)
 27.75-27.80  -15.3(±0.24)    349                           -15.3(±0.24)
    >27.80    -18.6(±0.15)    629                           -18.6(±0.15)
_________________________________________________________________________
 Standard deviation of the mean given in brackets (SDM, i.e., standard 
 deviation weighted by the number of individual determinations). Values 
 of ∆C(dis) are given in µmol kg-l. Dashes indicate value not determined.


As a general rule, the errors associated with the CFC age technique increase 
at higher density levels, and the errors associated with the ∆C* technique 
decrease at higher density levels. To minimize the errors in the final 
∆C(dis) determination, waters with σθ values between 27.25 and 27.5 were 
evaluated using a combination of the two methods mentioned above. The 27.25 
cut in the CFC age technique was chosen because this density corresponds with 
the core of the Antarctic Intermediate water and also generally the highest-
density water that outcrops in this region [Wirtki, 1973; Levitus and Boyer, 
1994; Levitus et al., 1994]. To help ensure that the ∆C(dis) values were 
determined on waters moving into the main Indian Ocean basin, mean ∆C*(t) 
values were only estimated from samples north of 35°S with CFC-12 ages less 
than 40 years. Mean ∆C* values were also determined on the same density 
surfaces for samples where CFCs were measured, but concentrations were below 
0.005 pmol kg-1. The final mean value used for the ∆C(dis) correction on each 
surface was determined from the mean of the combined individual estimates 
from each method (Table 3).

Examination of the individual and combined means in Table 3 indicates that 
there is a sizeable spread in the estimates from the two techniques in the 
overlap region. This difference is maximized since these density levels are 
pushing the limits of both techniques, and the errors in both estimates serve 
to increase this difference. Since the number of points available from the 
CFC age technique generally decreased at greater density levels and the 
number of points from the ∆C* technique generally increased at greater 
density levels, the mean becomes progressively more heavily weighted toward 
the ∆C* technique as the density levels increased. Although this is not the 
ideal solution, we believe that this minimizes the potential errors as much 
as possible. The technique used to estimate final ∆C(dis) values in this 
region could systematically bias the anthropogenic CO2 inventory estimates. 
The magnitude of this potential error on the final inventory was estimated to 
be approximately ±1.8 Pg C by integrating the difference between the two 
methods over the effected water volume. This estimate represents a maximum 
potential error since the known limitations of each method work to increase 
the differences in ∆C(dis).


2.3.4.  TIME ADJUSTMENT FOR INDIGO DATA 

One difficulty with combining data from different cruises for a time-
dependent calculation like the anthropogenic CO2 inventory is the issue of 
getting the data sets referenced to a common time. One of the advantages of 
the WOCE/JGOFS Indian Ocean survey data is the fact that all of the samples 
were collected in a little over a year's time. In terms of the CO2 inventory 
this is essentially a synoptic data set. The couple of years between the 
CIVA-1 cruise and the WOCE/JGOFS data are also not distinguishable in terms 
of the anthropogenic increase. The INDIGO data, however, were collected 8-10 
years before the WOCE/JGOFS data set and must be adjusted to reflect the 
anthropogenic uptake during that time. Unfortunately, any correction of this 
sort can have large errors and potentially bias the results. This problem 
must be weighed against the errors of ignoring the time difference between 
cruises or omitting these data entirely. The decision to correct the INDIGO 
data was based on two factors. First, analysis of the change in anthropogenic 
inventory between GEOSECS and WOCE (discussed below) indicated that a 
significant fraction of the total anthropogenic uptake has occurred in the 
past 2 decades. Second, careful examination of objective maps of 
anthropogenic CO2 made prior to the INDIGO correction showed obvious, 
anomalously low concentrations in the regions strongly dependent on the 
INDIGO data. Two different adjustment functions were made depending on 
whether the stations were located in the main Indian Ocean basin or in the 
Southern Ocean.

North of 30°S, where portions of the INDIGO data were located relatively near 
WOCE stations, a crossover comparison of the INDIGO anthropogenic CO2 
concentrations as a function of density was made with the WOCE/JGOFS data in 
that region. The difference between the two data sets was evaluated at σθ 
intervals of 0.05 from the surface to σθ = 27.5 and added to the INDIGO data. 
This correction ranged from approximately 12 µmol kg-1 at the surface down to 
zero at 27.5.

South of 30°S, there were very few WOCE or CIVA-1 stations close enough for a 
proper crossover comparison. It was clear from the northern data, however, 
that some correction was necessary. Given that the isolines for most 
properties in the Southern Ocean run east-west, we decided to correct the 
southern INDIGO data based on a crossover comparison with all results from 
CIVA-1 and WOCE cruises in that region. The average adjustment for the 
southern stations was approximately 11 µmol kg-l over the same density range. 
The magnitude of the corrections in both regions is consistent with the 
expected increase over the time period between cruises.


2.3.5.  EVALUATION OF ERRORS 

Error evaluation is much more difficult for the ∆C* method than for the time 
series approach because of potential systematic errors associated with some 
of the parameters (Le., the biological correction). The random errors 
associated with the anthropogenic CO2 can be determined by propagating 
through the precision of the various measurements required for the 
calculation of (4). 

{σ(C(anth))}^2 = {σ(C)}^2 + {σ(C(eq))}^2

               + {(-R(CO) - 0.5R(NO))σ(O(2))}^2

               + {(R(CO) + 0.5R(NO))σ(O(2[sat]))}^2

                                     ∂C(eq)
               + {-0.5σ(TA)}^2 + {(- ------ + 0.5)σ(Alk^0)}^2
                                      ∂TA

               + {0.8667σ(N)}^2 + +{13.867σ(P)}^2

                              N-16P+2.9
               + {0.8667(-P - --------- )σ(R(N:P[nitr]))}^2
                                 120

               + {-0.00111(N - 16P + 2.9)σ(R(N:P[denitr]))}^2

               - {σ(∆C(dis))}^2                                           (6)

where 
      σ(C)         = 2 µmol kg-1 ;
      σ(C(eq))     = 4 µmol kg-1 ;
      σ(O(2))      = 1 µmol kg-1 ;
      σ(O(2[sat])) = 4 µmol kg-1 ;
      σ(TA)        = 4 µmol kg-1 ;

      ∂C(eq)
      -------      = 0.842 ;
      ∂TA

      σ(Alk^0)     = 7.8 µmol kg-1 ;
      σ(N)         = 0.2 µmol kg-1 ;
      σ(P)         = 0.02 µmol kg-1 ;
      σ(R(N:P[nitr])) = 0.25 ;
      σ(R(N:P[denitr])) = 15


The equation for the random error analysis is adapted from Gruber et al. 
[1996] (excluding those terms that involve the C:O Redfield error) with 
additional terms for the error propagation of the N* correction [Gruber and 
Sarmiento, 1997]. The terms involving the C:O are evaluated separately below 
because the random errors cannot be isolated from potential systematic 
errors. The sigma values used in (6) were either taken from the appropriate 
WOCE cruise reports or from previously determined estimates of Gruber et al. 
[1996] and Gruber and

Sarmiento [1997]. The error in the ∆C(dis) term is taken from the average 
value for the standard deviation of the mean for the examined surfaces 
(σ(∆C(dis)) = 0.5 µmol kg-1). The formulation given in (6) results in an 
estimated error of 6.1 µmol kg-1. This estimate is larger than the standard 
deviation of the ∆C* values below the deepest anthropogenic CO2 penetration 
depth (±2.8 µmol kg-1 for pressure > 2000 dbars) suggesting that the 
propagated errors may be a maximum estimate of the random variability.

The potential systematic errors associated with the anthropogenic CO2 
calculation are much more difficult to evaluate. The random error estimate 
above includes all terms except those associated with the C:O biological 
correction. Although other terms involving N:O and N:P corrections 
potentially have systematic offsets associated with errors in the ratio 
estimates, the only potentially significant errors involve the C:O 
corrections [Gruber et al., 1996; Gruber, 1998].

There is evidence, however, that the Anderson and Sarmiento [1994] 
stoichiometric ratios must be reasonably close to the actual remineralization 
ratios observed in the Indian Ocean. Figure 6 is a plot of ∆C*t based on CFC-
12 ages for the density interval from σθ = 27.1 to σθ = 27.15. The diamonds 
are the values calculated from (5). These values represent the preserved air-
sea disequilibrium value for the past 40 years and should be constant if the 
air-sea disequilibrium has not changed over time (Le., that the surface ocean 
CO2 is increasing at the same rate at the atmosphere). A linear regression of 
the diamonds in Figure 6 yields a slope that is not significantly different 
from zero. The circles and pluses are the ∆C*t values one would get by using 
a C:O ratio of -0.60 and -0.78 in (5), respectively. These C:O values 
represent one standard deviation from the Anderson and Sarmiento [1994] mean 
value of 0.69. The -0.60 ratio results in values with a significant positive 
slope. This slope would imply that the surface ocean CO2 is increasing much 
slower than the atmospheric increase. While this is possible, the -0.60 ratio 
is much larger than historical Redfield estimates and would be very difficult 
to justify. The -0.78 ratio is more typical of historical estimates but 
results in a significant negative slope in the ∆C*t values with time. A 
negative slope would imply that carbon is accumulating in the ocean faster 
than the atmosphere. Neither of these scenarios seems very likely. The fact 
that none of the ∆C*t values on the examined surfaces exhibit a statistically 
significant slope suggests that the C:O value of -0.69 does accurately 
represent the remineralization ratio for these waters and supports the 
methodology of taking a mean value of ∆C*t on these density surfaces.


FIGURE 6:  Plot of ∆C*t based on CFC-12 ages for the density 
           interval from σθ =27.1 to 27.15 versus CFC-12 age. 
           The diamonds were calculated using the Anderson 
           and Sarmiento [1994] c:o (-0.69). The circles and 
           pluses were calculated from C:O of -0.60 and -0.78, 
           respectively. Lines and text give results from a 
           linear regression of the three sets of data.


A sensitivity study was also used to evaluate the potential error associated 
with an incorrect C:O value. Two additional estimates of anthropogenic CO2 
were determined using the -0.60 and -0.78 C:O values. Since the C:O 
correction applies to both ∆C* and the ∆C*(t) terms, the disequilibrium 
values were reevaluated in the same manner as described above. The range of 
anthropogenic values from these three estimates varied as a function of 
apparent oxygen utilization (AOU) from 0.0 to 22 with an average difference 
of only 4.2 µmol kg-1. Because the C:O correction affects both the ∆C* and 
∆C*(t) terms together, much of the systematic error in the final 
anthropogenic estimate (∆C*-∆C*(t)) cancels out.


2.4. INVENTORY ESTIMATES

Basin-wide anthropogenic and excess CO2 concentrations (WOCE/JGOFS - GEOSECS) 
were evaluated on a 1° grid at 100 m intervals between the surface and 2600 m 
using the objective mapping techniques of Sarmiento et al. [1982]. Total 
anthropogenic CO2 was mapped over an area from 20° to 120°E and 70°S to 30°N 
(excluding areas of land, the Red Sea, the Persian Gulf, and the South China 
Sea). Because the WOCE/JGOFS data set did not cover much of the Southern 
Ocean, the excess CO2 maps were limited to the area north of 35°S. The values 
at each level were multiplied by the volume of water in the 100 m slab and 
summed to generate the total anthropogenic or excess CO2 inventory. The 
method of integrating mapped surfaces compared very well with the technique 
of vertically integrating each station and mapping the station integrals.

It is extremely difficult to evaluate a reasonable estimate of the potential 
errors associated with the inventory estimates. A simple propagation of 
errors implies that the random errors associated with any individual 
anthropogenic estimate is approximately ±6.1 µmol kg-1, but these errors 
should essentially cancel out for an integrated inventory based on nearly 
25,000 individual estimates. Systematic errors have by far the largest impact 
on the inventory estimates. Potential errors as large as ±1.8 Pg C have been 
estimated for the ∆C(dis) term. Sensitivity studies with the C:O variations 
give a range of total inventory estimates of ±2.5 Pg C. Other systematic 
errors could also be generated from the denitrification term, the terms 
involving N:0, the time correction for the INDIGO data, and the mapping 
routines used in the inventory estimates. The magnitude of these errors is 
believed to be much smaller than the uncertainty in either the C:O correction 
or the ∆C(dis) determination. Propagation of the two estimated uncertainties 
gives an overall error of approximately ±3 Pg C for the total inventory. An 
error of roughly 15% is comparable to previous error estimates using this 
technique [Gruber et al., 1996; Gruber, 1998]. Errors for regional 
inventories are assumed to scale to the total.



3. RESULTS AND DISCUSSION

The excess CO2 concentrations for the Indian Ocean range from 0 to 25 µmol 
kg-1. The most prominent feature in the excess CO2distribution, as shown with 
representative sections in the eastern and western Indian Ocean (Figure 7), 
is the maximum in concentrations at midlatitudes (~40°S). This maximum is 
coincident with the relatively strong gradient in surface density associated 
with the Subtropical Convergence and the transition from the high salinity 
subtropical gyre waters to the low-salinity Antarctic waters. The outcropping 
of these density surfaces and subsequent sinking of surface waters provides a 
pathway for excess CO2 to enter the interior of the ocean. Relatively high 
excess CO2 concentrations can also be observed at the very northern end of 
the western section (Figure 7a). Although not readily evident from this 
section, the distribution of concentration gradients indicates that excess 
CO2 is entering the northern Indian Ocean from the Persian Gulf and Red Sea 
regions. This is likely to result from the outcropping of density surfaces in 
these areas which are not ventilated in the main Indian Ocean basin. The 
implied Red Sea and Persian Gulf sources of CO2 are consistent with uptake 
estimates of anthropogenic CO2 in these areas as observed by Papaud and 
Poisson [1986]. The third major feature observed in the excess CO2 
distribution is a dramatic shoaling of the excess CO2 isolines south of 
approximately 40°S. Poisson and Chen [1987] attributed the low anthropogenic 
CO2 concentrations in Antarctic Bottom Water to a combination of the pack sea 
ice blocking air-sea gas exchange and the upwelling of old Weddell Deep 
Water. This explanation is consistent with the observed excess CO2 
distributions in this study.

The general features observed with excess CO2 are also observed in the 
anthropogenic CO2 distribution (Figure 8). The range of values, however, 
extends up to 55 µmol kg-1. The maximum depth of the 5 µmol kg-1 contour is 
approximately 1300 m at around 40°S, only 200 m deeper than the maximum depth 
of the 5 µmol kg-1 contour of excess CO2, The similarity in maximum 
penetration depth between the 200 year and the 18 year anthropogenic CO2 
accumulation, together with the wide range of depths covered by the 5 µmol 
kg-1 isoline, indicates that the primary pathway for CO2 to enter the ocean's 
interior is from movement along isopycnals, not simple diffusion or cross 
isopycnal mixing from the surface. The 1300 m penetration results from the 
downwarping of the isopycnals in the region of the Subtropical Convergence. 
Likewise, the low anthropogenic CO2 concentrations in the high-latitude 
Southern Ocean result from the compression and shoaling of isopycnal surfaces 
in that region. Although the complex oceanography of the Southern Ocean may 
call into question some of the assumptions regarding mixing and nutrient 
uptake ratios with these techniques, both the time series excess CO2 and the 
∆C* anthropogenic CO2calculations clearly indicate that the anthropogenic CO2 
concentrations south of approximately 50°S are relatively small.

The distribution of anthropogenic CO2 determined in this study is similar to 
the distribution presented by Chen and Chen [1989] based on GEOSECS and 
INDIGO data. Although the penetration depth at 40°S was slightly deeper than 
observed with this study (1400-1600 m for the 5 µmol kg-1 isoline), Chen and 
Chen also observed a significant shoaling of the anthropogenic CO2 isolines 
toward the south. They suggest that anthropogenic CO2 has only penetrated a 
few hundred meters into the high-latitude (>50°S) Southern Ocean.

There has been debate in the literature over recent years as to the 
importance of the Southern Ocean as a sink for anthropogenic CO2 [e.g., 
Sarmiento and Sundquist, 1992; Keeling et al., 1989; Tans et al., 1990]. Most 
of the recent data-based estimates, however, indicate a relatively small 
Southern Ocean sink [Poisson and Chen, 1987; Murphy et al., 1991; Gruber, 
1998; this study]. The lack of observed anthropogenic CO2 in the Southern 
Ocean is also qualitatively consistent with ∆14C estimates which show no 
measurable storage of bomb 14C in the Southern Ocean since GEOSECS [Leboucher 
et al., 1998; R. Key, unpublished data, 1998]. Recent studies by Bullister et 
al. [1998], which show evidence of deep CFC penetration in the Southern 
Ocean, may appear to contradict these low anthropogenic CO2 estimates, but we 
believe it is further evidence that one must be careful when inferring 
anthropogenic carbon distributions from other tracers. One possible 
explanation of this apparent discrepancy may be the CFC equilibration rate of 
days which is significantly faster than the CO2 equilibration time of months 
[e.g., England, 1995; Warner and Weiss, 1985; Tans et al., 1990]. This can 
become an issue in the Southern Ocean where upwelling and convection may 
allow the CFCs to equilibrate to a greater extent than the CO2, Again, we 
acknowledge the limitations of the methods used in the Southern Ocean, and it 
is possible that the apparent discrepancy in the CFC penetration versus the 
CO2 penetration may also be an issue of detection limits. With a detection 
limit that is approximately 6 µmol kg-1, it is not possible to say with this 
technique that the concentration of anthropogenic CO2 below 500 m at 60°S is 
zero. However, we can say with some confidence that the concentration is not 
10 µmol kg-1 or greater. Since there is no natural oceanic source of CFCs and 
these compounds are not biologically utilized, the ability to detect them is 
much greater. If mixing has diluted the anthropogenic signal to 
concentrations just below detection limits, it is possible that carbon 
measurement based techniques would underestimate the Southern Ocean sink.


FIGURE 7:  Sections of excess CO2 along (a) -57°E and (b) -92°E. 
           Dots indicate sample locations used in sections.  Note 
           that I6S data along 30°E were brought into the line of 
           section for contours south of 40°S in Figure 7a.


The total anthropogenic CO2 inventory for the main Indian Ocean basin (north 
of 35°S) was 13.6±2 Pg C in 1995. The increase in CO2 inventory since GEOSECS 
was 4.1±1 Pg C for the same area. This represents a nearly 30% increase in 
the past 18 years relative to the total accumulation since pre-industrial 
times. The relative oceanic increase is very similar to the 31% increase 
observed in atmospheric concentrations over the same time period [Keeling and 
Whorf, 1996]. This similarity suggests that the oceans, at least for now, are 
keeping pace with the rise in atmospheric CO2, Approximately 6.7±1 Pg C are 
stored in the Indian sector of the Southern Ocean giving a total Indian Ocean 
inventory (between 20° and 120°E) of 20.3±3 Pg C in 1995.

To put these results in a global perspective, the total inventory for the 
Indian Ocean is only half that of the Atlantic (40±6 Pg C [Gruber, 1998]), 
but it contains an ocean volume that is nearly 80% of the Atlantic. The main 
difference between the two oceans, of course, is that the Indian Ocean does 
not have the high northern latitude sink that the Atlantic has. The big 
unknown at this point is the anthropogenic inventory of the Pacific. With 
nearly 50% of the total ocean volume the Pacific has the potential to be the 
largest oceanic reservoir for anthropogenic CO2.



4. COMPARISON WITH PRINCETON OCEAN BIOGEOCHEMISTRY MODEL

Current IPCC anthropogenic estimates are primarily based on global carbon 
models. Ultimately, these models are necessary to predict the oceanic 
response to future climate scenarios. It is important, however, to validate 
these models. One way to compare results is to examine profiles of the 
average anthropogenic concentrations such as those shown in Figure 9. The 
model presented here is the Princeton Ocean Biogeochemistry Model (OBM). The 
Princeton OBM is based on the circulation of Toggweiler et al. [1989] with 
explicit parameterization for the biological and solubility carbon pumps 
[Sarmiento et al., 1995; Murnane et al., 1998]. On this scale the model-based 
concentrations for both the total anthropogenic CO2 and the increase since 
GEOSECS appear to be reasonably consistent with the data. The primary 
difference is slightly higher values at middepths in the data-based 
estimates. A more detailed examination, however, indicates that the regional 
distribution of the model-based estimates is significantly different than the 
data-based distribution. Figure 10 presents maps of the vertically integrated 
excess CO2 normalized to a unit area. The model shows a consistent decrease 
in column inventory toward the north. The lowest inventories in the data-
based map are in a narrow band just south of the equator. The highest values 
are found in the southeastern Indian Ocean. Relatively high values are also 
observed in the Arabian Sea in the regions near the Red Sea and the Persian 
Gulf. The small patch of lower values immediately outside the Gulf of Aden 
does not result from low concentrations but rather results from the shallow 
water depth associated with the mid-ocean ridge in that area. The low values 
east of there, however, do result from lower concentrations near the southern 
tip of India. The total model-based inventory for the region north of 35°S is 
approximately 0.61 times the data-based inventory (Table 4). 


FIGURE 8:  Sections of anthropogenic CO2 along (a) -57°E and 
           (b) -92°E. Dots indicate sample locations used in sections.
FIGURE 9:  Profile of area weighted mean anthropogenic CO2 
           concentrations for model (solid symbols) and 
           data-based (open symbols) estimates for main 
           Indian Ocean basin (north of 35°S). Circles 
           show increase since GEOSECS (1978-1995). Triangles 
           show total increase since pre-industrial times.


Figure II shows maps of total anthropogenic CO2 column inventory. As with the 
excess CO2, the model predicts decreasing anthropogenic concentrations north 
of 35°S. The data-based distribution pattern is similar to the data-based 
excess C02 pattern with a minimum inventory band south of the equator and 
higher values toward the north and south. Similar to the findings with excess 
CO2, the model-based anthropogenic inventory north of 35°S is approximately 
0.68 times the data-based inventory (Table 4). The largest difference between 
the data-based results and the model is evident, however, in the Southern 
Ocean (south of 35°S). In this region the model anthropogenic inventory is 
nearly 2.6 times the data-based inventory (Table 4). The primary reason for 
this difference is the presence of a large convective cell in the model at 
approximately 55°S and 90°E in the Southern Ocean. This is a region of 
intense, unrealistic convection which pumps relatively high concentrations of 
anthropogenic CO2 down in excess of 4000 m. This problem is a known 
shortcoming with the mixing scheme used in several GCMs [e.g., England, 1995] 
but has never before been quantified in terms of its direct effect on 
anthropogenic CO2 storage by the models. It is beyond the scope of this paper 
to examine the details of the model physics; however, this same general trend 
of getting too much anthropogenic CO2 into the Southern Ocean has been 
observed in comparisons with three other global carbon models with a range of 
mixing and advective schemes [C. Sabine, unpublished results, 1998]. This 
cursory comparison with the Princeton OBM clearly demonstrates the diagnostic 
usefulness of comparing the data distributions with models.



5. CONCLUSIONS

Although the general techniques proposed by Gruber et al. [1996] and Wallace 
[1995] can be important tools for estimating global anthropogenic CO2, 
careful consideration must be used when applying these techniques to new 
regions. Complicating factors such as those found in the Arabian Sea can 
influence the quality of the estimates if not properly addressed. An 
additional term had to be added to the basic ∆C* calculation to account for 
denitrification in the Arabian basin. For the excess CO2 calculations a 
categorical variable was used to remove regional biases in the GEOSECS fit.

With the above mentioned modifications the anthropogenic inventory of the 
Indian Ocean was shown to be relatively small, approximately half of that 
found in the Atlantic. This study provides an important baseline for future 
studies of the Indian Ocean. The calculations presented here suggest that the 
oceanic increase in carbon storage (30%) has roughly kept pace with the 
atmosspheric increase (31%) over the past 18 years. Models predict that this 
trend is likely to change as atmospheric CO2 concentrations continue to rise 
in the future [Sarmiento et al., 1995]. As more CO2 enters the ocean, the 
carbonate ion concentration will become depleted. This will decrease the 
buffering capacity of the ocean and its ability to continue carbon uptake at 
the current rate. Comparison of future survey cruises in the Indian Ocean 
with the anthropogenic and total carbon values from this study will allow us 
to document future changes in ocean chemistry and better understand the 
oceanic response to global change. 


FIGURE 10: Maps of vertically integrated excess CO2 based on 
           (a) data and (b) model estimates. Contours are in mol 
           m^-2. Solid regions indicate land mask used for inventory 
           estimates. Thin lines in Figure 10b indicate land regions 
           used in Figure 10a.


TABLE 4: Summary of Data Based and Model Based Inventory Estimates
         ______________________________________________________________

                   Total    Southern  Main Basin  Main Basin  Increase
                   Anthro-  Ocean     Anthro-     Excess      since
                   pogenic  Anthro-   pogenic     CO2(χ)      GEOSECS
                   CO2(α)   pogenic   CO2(χ)      PgC         %
                   PgC      CO2(β)    PgC
                            PgC
                   -------  --------  ----------  ----------  --------
          Data     20.3±3     6.7±1     13.6±2      4.1±1       29.9
          based
          
          Model     26.7      17.4       9.3         2.5        26.7
          based
         ______________________________________________________________
          (α) Area between 20°-120°E.
          (β) Latitude is < 35°S.
          (χ) Latitude is > 35°S.


Finally, comparison of the spatial distribution of the anthropogenic
carbon can be a powerful tool for understanding the carbon
uptake of the models. The methods presented here provide a two point
calibration for examining the response of the models to
observed atmospheric CO2 increases. The anthropogenic CO2 data
can also be subtracted from the TC02 measurements to provide an
estimate of the pre-industrial TC02 distribution. Comparing these
estimates with the steady state model distributions can provide
insight into whether differences in the model and data-based
anthropogenic inventories result from problems with the uptake
parameterization or the basic physics and initialization parameters
of the model. This paper is just the first step in the interpretation of
the WOCE/JGOFS data set. Subsequent papers will analyze additional
cruise data as they become available. Together. these analyses
will significantly improve our understanding of the global
carbon cycle.


FIGURE 11: Maps of vertically integrated anthropogenic CO2 
           based on (a) data and (b) model estimates. Contours 
           are in mol m-2. Solid regions indicate land mask 
           used for inventory estimates. Thin lines in Figure 
           11b indicate land regions used in Figure 1la.



ACKNOWLEDGMENTS. 

This work was accomplished with the cooperative efforts of the DOE CO2 
Science Team. We thank B. Warren for organizing the WOCE Indian Ocean 
expedition, the captain and crew of the R/V Knorr, and the WOCE-HP personnel 
at sea. We. thank the chief scientists (M. McCartney, A. Gordon, L. Talley, 
W. Nowlin, J. Toole, D. Olson, J. Morrison, N. Bray, and G Johnson) and the 
CFC PIs (1. Bullister, R Fine, M. Warner, and R Weiss) for giving us access 
to their preliminary data for use in this publication. We also thank N. 
Metzl, G Eischeid, and C. Goyet for providing carbon data and T. Takahashi 
for providing S4I data and ∆pC02 maps. We thank R Murnane and T. Hughes for 
providing model results. Strong collaboration, cooperation, and input from N. 
Gruber and investigators in the NOAA Ocean Atmosphere Carbon Exchange Study 
(R Wanninkhof, R. Feely, J. Bullister, and T.-H. Peng) is also acknowledged 
along with the helpful comments of two anonymous reviewers. This work was 
primarily funded by DOE grant DE-FG02-93ER61540 with additional support by 
NSF/NOAA grant OCE-9120306. 



REFERENCES

Anderson, L., and D. Dyrssen, Alkalinity and total carbonate in the Arabian
    Sea: Carbonate depletion in the Red Sea and Persian Gulf, Mar. Chem., 
    47, 195-202, 1994.

Anderson, L.A., and J.L. Sarmiento, Redfield ratios of remineralization 
    determined by nutrient data analysis, Global Biogeochem. Cycles, 8, 6580, 
    1994.

Brewer, P.G, A.L. Bradshaw, and RT. Williams, Measurement of total carbon 
    dioxide and alkalinity in the North Atlantic Ocean in 1981, in The 
    Changing Carbon Cycle: A Global Analysis, edited by J.R. Trabalka and 
    D.E. Reichle, pp. 348-370, Springer-Verlag, New York, 1986.

Brewer, P.G, D.M. Glover, C. Goyet, and D.K. Shafer, The pH of the North 
    Atlantic Ocean: Improvements to the global model for sound absorption in 
    seawater, J. Geophys. Res., 100, 8761-8776, 1995.

Brewer, P.G, C. Goyet, and G Freiderich, Direct observation of the oceanic 
    CO2 increase revisited, Proc. Natl. Acad. Sci. U.S.A., 94, 8308-8313, 
    1997.

Broecker, W.S., 'NO', a conservative water-mass tracer, Earth Planet. Sci. 
    Lett., 23, 100-107, 1974.

Broecker. W.S., T. Takahashi, H.J. Simpson, and T.-H. Peng, Fate of fossil 
    fuel carbon dioxide and the global carbon budget, Science, 206, 
    409-418, 1979.

Bullister, J.L., and R.E. Weiss, Determination of CCl3F and CCl2F2 in 
    seawater and air, Deep Sea Res., Part A, 35, 839-853, 1988.

Bullister, J.L., D.P. Wisegrave, W.M. Smethie, and M.J. Warner, The 
    appearance of CFCs and carbon tetrachloride in the abyssal waters of the 
    Samoa Passage, Eos, Trans. AGU, 79, 1998.

Chen, C.-T., Rates of calcium carbonate dissolution and organic carbon 
    decomposition in the North Pacific Ocean, J. Oceanogr. Soc. Jpn., 46, 
    201-210, 1990.

Chen; C.-T., and R.M. Pytkowicz, On the total CO2-titration alkalinity-oxygen 
    system in the Pacific Ocean, Nature, 281, 362-365, 1979.

Chen, D.W., and C.-T. Chen, The anthropogenic CO2 signals in the Indian 
    Ocean, J. Environ. Prot. Soc. Repub. China, 12 (2), 46-65, 1989.

Cleveland, W.S., and S.J. Devlin, Locally-weighted regression: An approach to 
    regression analysis by local fitting, JASA J. Am. Stat. Assoc., 83,596-
    610, 1988.

Cleveland, W.S., E. Grosse, and W.M. Shyu, Local regression models, in 
    Statistical Models in S, edited by J.M. Chambers and T.J. Hastie, pp. 
    309-376, Wadsworth and Brooks, Pacific Grove, California, 1992.

Culberson, C.H., et al., A comparison of methods for the determination of 
    dissolved oxygen in seawater, WHPO Rep. 91-2, World Ocean Circ. Exp. 
    Hydrogr. Programme Off., Woods Hole, Massachusetts, 1991.

Dickson, A.G., The ocean carbon dioxide system: Planning for quality data, US 
    JGOFS News, 2(2), 2, 1990.

Doney, S.C., W.J. Jenkins, and J. Bullister, A comparison of ocean tracer 
    dating techniques on a meridional section in the eastern North Atlantic, 
    Deep-Sea Res., Part 1, 44, 603-626, 1997.

Dueser, W.G., E.H. Ross, and Z.J. Mlodzinska, Evidence for and rate of 
    denitrification in the Arabian Sea, Deep Sea Res., 25, 431-445, 1978.

Edmond, J.M., High precision determination of titration alkalinity and total 
    carbon dioxide content of seawater by potentiometric titration, Deep Sea 
    Res., 17, 737-750, 1970.

England, M.H., Using chlorofluorocarbons to assess ocean climate models, 
    Geophys. Res. Lett., 22, 3051-3054 ,1995.

Gordon, L.I., J.C. Jennings Jr., A.A. Ross, and J.M. Krest, A suggested 
    protocol for continuous flow automated analysis of seawater nutrients in 
    the WOCE Hydrographic Programme and the Joint Global Ocean Fluxes Study, 
    Grp. Tech. Rep. 92-1, Coll. of Oceanogr., Oregon State Univ., Corvallis, 
    1992.

Gruber, N., Anthropogenic CO2 in the Atlantic Ocean, Global Biogeochem. 
    Cycles, 11, 165-191, 1998.

Gruber, N., and J.L. Sarmiento, Global patterns of marine nitrogen fixation 
    and denitrification, Global Biogeochem. Cycles, 11, 235-266, 1997.

Gruber, N., J.L. Sarmiento, and T.F. Stocker, An improved method for 
    detecting anthropogenic CO2 in the oceans, Global Biogeochem. Cycles, 
    10, 809-837, 1996.

Johnson, K.M., A.E. King, and J.M. Sieburth, Coulometric TC02 analyses for 
    marine studies: An introduction, Mar. Chem., 16, 61-82, 1985.

Johnson, K.M., et al., Coulometric total carbon dioxide analysis for marine 
    studies: Assessment of the quality of total inorganic carbon measurements 
    made during the US Indian Ocean CO2 Survey 1994-1996, Mar. Chem., 63, 
    21-37, 1998.

Keeling, C.D., and T.P. Whorf, Atmospheric CO2 records from sites in the SIO 
    air sampling network, in Trends: A Compendium of Data on Global Change, 
    Carbon Dioxide Inf. Anal. cent., Oak Ridge Nat. Lab., Oak Ridge, Tenn., 
    1996.

Keeling, C.D., S.C. Piper, and M. Heimann, A three-dimensional model of 
    atmospheric CO2 transport based on observed winds, 4, Mean annual 
    gradients and interannual variations, in Aspects of Climate Variability 
    in the Pacific and the Western Americas, Geophys. Monagr. Ser., Vol. 55, 
    edited by D.H. Peterson, pp. 305-363, AGU, Washington D.C., 1989.

Keeling, R.F., and S.R. Shertz, Seasonal and interannual variations in 
    atmospheric oxygen and implications for the global carbon cycle, Nature, 
    358, 723-727, 1992.

Leboucher, V., 1. Orr, P. Jean-Baptiste, M. Arnold, P. Monfray, N. 
    Tisnerat-Laborde, A. Poisson, and J. Duplessy, Oceanic radiocarbon   
    between Antarctica and South Africa along WOCE section 16 at 30°E, 
    Radiocarbon, in press, 1998.

Levitus, S., and T.P. Boyer, Temperature, in NOAA Atlas NESDIS 3: World Ocean 
    Atlas 1994, NOAA Tech. Rep. 4, Natl. Environ. Satell. Data and Inf. 
    Serv., Silver Spring, MD, 1994.

Levitus, S., R. Burgett, and T.P. Boyer, Salinity, in NOAA Atlas NESDIS 3: 
    World Ocean Atlas 1994, NOAA Tech. Rep. 3. Natl. Environ. Satell. Data 
    and Inf. Serv., Silver Spring, MD, 1994.

Merbach, C., C.H. Culberson, J.E. Hawley, and R.M. Pytkowicz, Measurements of 
    the apparent dissociation constants of carbonic acid in seawater at 
    atmospheric pressure, Limnol. Oceanogr.,18, 897-907, 1973.

Millard, R.C., Jr., CTD calibration and data processing techniques at WHOI 
    using the practical salinity scale, paper presented at International SID 
    Conference and Workshop, Mar. Tech. Soc., La Jolla, Calif., 1982.

Millero, F.J., et al., Total alkalinity measurements in the Indian Ocean 
    during the WOCE Hydrographic Program CO2 survey cruises 1994-1996, Mar. 
    Chem., 63, 9-20, 1998a.

Millero, F.J., K. Lee, and M. Roche, Alkalinity as a major variable in the 
    marine carbonate system, Mar. Chem., 60, 111-130, 1998b.

Murnane, R., 1.L. Sarmiento, and C. LeQuere, The spatial distribution of air-
    sea CO2 fluxes and the interhemispheric transport of carbon by the 
    oceans, Global Biogeochem. Cycles, in press, 1999.

Murphy, P.P., R.A. Feely, R.H. Gammon, K.C. Kelly, and L.S. Waterman, Autumn 
    air-sea disequilibrium of CO2 in the South Pacific Ocean, Mar. Chem., 35, 
    77-84, 1991.

Naqvi, S.W.A., and R Sen Gupta, "NO" a useful tool for the estimation of 
    nitrate deficits in the Arabian Sea, Deep Sea Res., Part A, 32, 665-674, 
    1985.

Neftel, A., H. Friedli, E. Moor, H. Lotscher, H. Oeschger, U. Siegenthaler, 
    and B. Stauffer, Historical CO2 record from the Siple station ice core, 
    in Trends '93: A Compendium of Data on Global Change, edited by T. Boden 
    et al., Rep. ORNL/CDIAC-65, pp. 11-14, Carbon Dioxide Inf. Anal. Cent., 
    Oak Ridge Nat. Lab., Oak Ridge, Tenn., 1994.

Papaud, A, and A. Poisson, Distribution of dissolved CO2 in the Red Sea and 
    correlation with other geochemical tracers, J. Mar. Res., 44, 
    385-402, 1986. 

Poisson, A., and C.-T.A Chen, Why is there little anthropogenic CO2 in the 
    Antarctic Bottom Water?, Deep Sea Res., Part A, 34, 1255-1275, 1987.

Poisson, A., B. Schauer, and C. Brunet, Les Rapports des campagnes a la mer, 
    MD43/INDIGO I, in Les publications de fa mission de recherche des Terres 
    Australes et Antarctiques Fancaises, Rep. 85-06, 267 pp., Univ. Pierre et 
    Marie Curie, Paris, France, 1988.

Poisson, A., B. Schauer and C. Brunet, Les Rapports des campagnes a la mer, 
    MD49/IND1GO 2, in Les publications de fa mission de recherche des Terres 
    Australes et Antarctiques Fancaises, Rep. 86-02, 234 pp., Univ. Pierre et 
    Marie Curie, Paris, France, 1989.

Poisson, A., B. Schauer and C. Brunet, Les Rapports des campagnes a la mer. 
    MD53/INDIGO 3, in Les publications de la mission de recherche des Terres 
    Australes et Antarctiques Fancaises, Rep. 87-02, 269 pp., Univ. Pierre et 
    Marie Curie, Paris, France, 1990.

Quay, P.D., B. Tilbrook, and C.S. Wong, Oceanic uptake of fossil fuel CO2: 
    Carbon-13 evidence, Science, 256, 74-79,1992.

Sarmiento, J.L., and E.T. Sundquist, Revised budget for the oceanic uptake of 
    anthropogenic carbon dioxide, Nature, 356, 589-593, 1992.

Sarmiento, J.L., J. Willebrand, and S. Hellerman, Objective analysis of 
    Tritium observations in the Atlantic Ocean during 1971-74, OTL Tech Rep. 
    I, 19 pp., Ocean Tracers Lab., Princeton Univ., Princeton, NJ, 1982.

Sarmiento, J.L., J.C. Orr, and U. Siegenthaler, A perturbation simulation of 
    CO2 uptake in an ocean general circulation model, J. Geophys. Res., 
    97, 3621-3645, 1992.

Sarmiento, J.L., R. Murnane, and C. LeQuere, Air-sea C02 transfer and the 
    carbon budget of the North Atlantic, Philos. Trans. R. Soc. London, Ser. 
    B, 348, 211-219, 1995.

Sen Gupta, R., M.D. Rajagopal, and S.Z. Qasim, Relationship between dissolved 
    oxygen and nutrients in the northwestern Indian Ocean, Indian J. Mar. 
    Sci., 5, 201-211, 1976.

Siegenthaler, U., and F. Joos, Use of a simple model for studying oceanic 
    tracer distributions and the global carbon cycle, Tellus, Ser. B, 44, 
    186-207, 1992. 

Siegenthaler, U., and J.L. Sarmiento, Atmospheric carbon dioxide and the 
    ocean, Nature, 365, 119-125, 1993.

Stocker, T.F., W.S. Broecker, and D.G. Wright, Carbon uptake experiments with 
    a zonally-averaged global circulation model, Tellus, Ser. B, 46, 103-122, 
    1994.

Takahashi, T.T., R.A. Feely, R.F. Weiss, R.H. Wanninkhof, D.W. Chipman, S.C. 
    Sutherland, and T.T. Takahashi, Global air-sea flux of CO2: An estimate 
    based on measurements of sea-air pC02 difference, Proc. Natl. Acad. Sci. 
    U.S.A., 94, 8292-8299, 1997.

Tans, P.P., I.Y. Fung, and T. Takahashi, Observational constraints on the 
    global atmospheric CO2 budget, Science, 247, 1431-1438, 1990.

Toggweiler, J.R., K. Dickson, and K. Bryan, Simulations of radiocarbon in a 
    coarse-resolution world ocean model, 1. Steady state pre-bomb 
    distributions, J. Geophys. Res., 94, 8217-8242, 1989.

UNESCO, Background papers and supporting data on the Practical Salinity 
    Scale, 1978 UNESCO Tech. Pap. in Mar. Sci., 37, 1981.

Wallace, D.W.R, Monitoring global ocean inventories, OOSDP Background Rep. 5, 
    54 pp., Ocean Observ. Syst. Dev. Panel, Texas A&M Univ., College Station, 
    TX, 1995.

Warner, M.J. and R.F. Weiss, Solubilities of chlorofluorocarbons 11 and 12 in 
    water and seawater, Deep-Sea Res., 32, 1485-1497, 1985.

Warner, M.J., J.L. Bullister, D.P. Wisegarver, R.H. Gammon, and R.F. Weiss, 
    Basin-wide distributions of chlorofluorocarbons CFC-II and CFC-12 in the 
    North Pacific: 1985-1989, J. Geophys. Res., 101, 20525-20542, 1996.

Weiss, RE, W.S. Broecker, H. Craig, and D. Spencer, GEOSECS Indian Ocean 
    Expedition Vol. 5, Hydrographic Data 1977-1978, 48 pp., Natl. Sci. 
    Found., U.S. Gov. Print. Off., Washington D.C., 1983.

Wessel, P., and W.H.F. Smith, New version of generic mapping tools released, 
    Eos Trans. AGU, 76, 329, 1995.

Wyrtki, K., Physical oceanography of the Indian Ocean, in Ecological Studies: 
    Analysis and Synthesis, vol. 3, pp. 18-36, edited by B. Zeitzschel, 
    Springer-Verlag, New York, 1973. 

________________


   K.M. Johnson, Oceanographic and Atmospheric Sciences Division,
Brookhaven National Laboratory, Upton, NY 11973.

   RM. Key, C.L. Sabine, and J.L. Sarmiento, Department of Geosciences,
Princeton University, Princeton, NJ 08544. (key@geo.princeton.edu,
sabine@geo.princeton.edu, andjls@splash.princeton.edu)

   F.J. Millero, Rosenstiel School of Marine and Atmospheric Sciences,
University of Miami, 4600 Rickenbacker Cswy., Miami, FL 33149.
(fmillero@rsmas.miami.edu)

   A. Poisson, Laboratoire de Physique et Chimie Marines, Universite
Pierre et Marie Curie, 4 Place Jussieu, Tour 24-25, 75720 Paris Cedex 05
France. (apoisson@ccr.jussieu.fr)

   D.W.R. Wallace, Abteilung Meereschemie, Institut fur Meereskunde an
der Universität Kiel, Duesternbrooker Weg 20, D-24105 Kiel, Germany.
(dwallace@ifm.uni-kieI.de)

   C.D. Winn, Marine Science Program, Hawaii Pacific University, 45045
Kamehameha Highway, Kaneohe, HI 96744-5297.
(cwinn@soest.hawaii.edu)



(Received May 11, 1998; revised November 24,1998; accepted November 24,1998.)









                                  APPENDIX E:

                        REPRINT OF PERTINENT LITERATURE




Key R. M., and P. D. Quay. 2002. U.S. WOCE Indian Ocean Survey: Final Report 
for Radiocarbon.  Technical Report. Princeton University, Princeton, N.J.





                          U.S. Woce Indian Ocean Survey:
                           Final Report for Radiocarbon 

                          Robert M. Key and Paul D. Quay

                  Ocean Tracer Laboratory; Technical Report 02-1
                                  July 12, 2002



1.0  General Information 


The U.S. WOCE Indian Ocean Survey consisted of 9 cruises covering the period 
December 1, 1994 to January 22, 1996.  All of the cruises used the R/V Knorr 
operated by the Woods Hole Oceanographic Institute.  A total of 1244
hydrographic stations were occupied with radiocarbon sampling on 366 stations. 
The radiocarbon stations are shown as black dots in Figure 1.  To give an 
indication of the total radiocarbon coverage for the Indian Ocean, the figure 
includes radiocarbon stations from WOCE sections S4I (Key, 1999; red dots) and 
I6S (Leboucher, et al., 1999; white dots) and from the earlier GEOSECS (Stuiver 
and Ostlund, 1983; brown dots) and INDIGO (Bard, et al., 1988; yellow dots) 
expeditions.  Specific summary information on the 9 WOCE survey cruises is given 
in Table 1. 


TABLE 1: Summary for Survey Sections 
         __________________________________________________________________

                      Chief                               ∆14C     ∆14C
          Cruise    Scientist     Start       End       Stations  Samples  
          ------  ------------  ----------  ----------  --------  --------
          I8SI9S  M. McCartney  12/01/94    01/19/95       26       662  
                  T. Whitworth  Fremantle   Fremantle 
                                Australia   Australia      
          I9N     A. Gordon     01/24/95    03/05/95       22       364  
                  D. Olson      Fremantle   Colombo 
                                Australia   Sri Lanka      
          I8NI5E  L. Talley     03/10/95    04/15/95       20       414
                  M. Baringer   Colombo     Fremantle 
                                Sri Lanka   Australia       
          I3      W. Nowlin     04/20/95    06/07/95       20       462
                  B. Warren     Fremantle   Port Louis
                                Australia   Mauritius       
          I5WI4   J. Toole      06/11/95    07/11/95       15       361 
                                Port Louis  Port Louis
                                Mauritius   Mauritius      
          I7N     D. Olson      07/15/95    08/24/95       22       373 
                  S. Doney      Port Louis  Muscat 
                  D. Musgrave   Mauritius   Oman   
          I1      J. Morrison   08/29/95    10/16/95       24       426 
                  H. Bryden     Muscat      Singapore      
                                Oman        China       
          I10     N. Bray       11/11/95    11/28/95        6       127  
                  J. Toole      Dampier     Singapore      
                                Australia   China      
          I2      G. Johnson    12/02/95    01/22/96       28       651 
                  B. Warren     Singapore   Mombasa 
                                China       Kenya   
         __________________________________________________________________



2.0 Personnel 

∆14C sampling for cruise I8SI9S was carried out by Melinda Brockington 
(University of Washington). Personnel for the remainder of the cruises came from 
the Ocean Tracer Lab (OTL Princeton University) and included G. McDonald, A. 
Doerty, R. Key, T. Key, and R. Rotter. ∆14C (and accompanying δ13C) analyses 
were performed at the National Ocean Sciences AMS Facility (NOSAMS) at Woods 
Hole Oceanographic Institution. R. Key collected the data from NOSAMS, merged 
the files with hydrographic data, assigned quality control flags to the ∆14C 
and submitted the results to the WOCE office (4/02).  R. Key is P.I. for the 14C 
data.  P. Quay (U.W.) and A. McNichol (WHOI/NOSAMS) are P.I.s for the 13C data. 
In addition to collecting samples the shipboard 14C person was also responsible 
for operation of the underway pCO2 system provided by the OTL (Sabine and Key, 
1997; Sabine, et al., 2000). 


3.0 Results 


This ∆14C data set and any changes or additions supersedes any prior release. 


3.1 Hydrography 

Hydrographic data from these cruises were submitted to the WOCE office by the 
chief scientists and are described in various reports which are available from 
the web site (http://whpo.ucsd.edu/data/tables/onetime/1tim_ind.htm). 

3.2 

∆14C The ∆14C values described here were originally distributed in the NOSAMS 
data reports listed in Table 2 and given in full in the References.  Those 
reports included results which had not been through the WOCE quality control 
procedures.  For WOCE applications, this report supersedes the NOSAMS reports. 


TABLE 2: NOSAMS Data Report Summary 
                               _________________

                                Cruise   Report
                                -------  ------
                                I8SI9S   99-089  
                                I7NI9N   99-144  
                                I1       99-199  
                                I8N      00-218  
                                I3I5WI4  01-013  
                                I2       02-001  
                               _________________

All of the AMS samples from these cruise have been measured using the AMS 
methods outlined in Key et al., 1996 and citations therein (especially Mcnichol 
et al., 1994; Osborne et al. 1994; and Scheideret al. 1995).  Table 3 summarizes 
the number of samples analyzed and the quality control flags assigned for each 
cruise.  Approximately 98% of the samples collected were deemed to be "good" 
(flagged 2 or 6).  Quality flag values were assigned to all ∆14C measurements 
using the code defined in Table 0.2 of WHP Office Report WHPO 91-1 Rev. 2 
section 4.5.2. (Joyce, et al., 1994).  No measured values have been removed from 
this data set. 


TABLE 3: Sample Analysis and QC Summary 
                   _______________________________________

                            Samples     QC Flag Totals  
                    Cruise  Analyzed  2    3   4   5   6  
                    ------  -------- ---  --- --- --- ---
                    I8SI9S    662    636   6   8   0  12
                    I9N       368    354   4   3   4   3  
                    I8NI5E    416    401   6   0   2   7  
                    I3        463    448   5   0   1   9  
                    I5WI4     366    342   3   1   5  15  
                    I7N       383    370   3   0  10   0  
                    I1        430    421   2   2   4   1  
                    I10       127    127   0   0   0   0  
                    I2        655    636  13   2   4   0  
                    Total    3870   3735  42  16  30  47  
                   _______________________________________



4.0 Data Summary 

Figures 2-6 summarize the ∆14C data collected during the Indian Ocean survey. 
Only ∆14C measurements with a quality flag value of 2 ("good") or 6 
("replicate") are included in the figures.  Figure 2 shows the ∆14C values with 
2σ error bars plotted as a function of pressure.  The mid depth ∆14C minimum 
which occurs around 2500 meters in the Pacific is not apparent in these data.
In fact, there is very little variation in the deep and bottom water other than 
the previously reported decrease in ∆14C from south to north. All of the 
samples collected at a depth greater than 1000 meters have a mean ∆14C = 
-165.±25‰ (standard error = 0.5‰ with n=2086).  A substantial fraction of this 
variability is due to the difference between the Southern Ocean and main basin 
deep waters. 

Figure 3 shows the deep (>1000m) ∆14C values plotted against silicate.  The 
black and red points are from north and south of 35°S, respectively.  The 
straight line shown in the figure is the least squares regression relationship 
derived by Broecker et al. (1995) based on the GEOSECS global data set. 
According to their analysis, this line (∆14C = -70 - Si) represents the 
relationship between naturally occurring radiocarbon and silicate for most of 
the ocean.  They noted that the technique could not be simply applied at high 
latitudes as confirmed by this data set. 

Figure 4 shows all of the radiocarbon values plotted against potential 
alkalinity (defined as [alkalinity + nitrate]*35/salinity).  The straight line 
is the regression fit (14C = -59 -0.962(PALK-2320) derived by Rubin and Key 
(2002) using GEOSECS measurements assumed to have no bomb-produced ∆14C.  The 
value 2320 is the estimated surface ocean mean potential alkalinity.  As with 
Figure 3 the black and red points in Figure 4 indicate measurements taken north 
and south of 35°S, respectively. Unlike the silicate plot (Figure 3), there is 
no apparent difference in the relationship for Southern Ocean vs Indian Ocean 
deep waters.  The distance a point falls above the regression line is an 
estimate of the bomb radiocarbon contamination for the sample. 

Figures 5-9 show gridded sections of the ∆14C data. In each figure the water 
column is divided into upper (0-1000m) and lower (1000-bottom) portions.  The 
data were gridded using the loess method (Chambers et al., 1983; Chambers and 
Hastie, 1991; Cleveland,1979; Cleveland and Devlin, 1988).  The span for the 
fits was adjusted to be minimum and yet capture the large scale features.  The 
contour interval is 10‰ for the upper water column and 20‰ for intermediate and deep water. 

Figure 5 and Figure 6 show the meridional ∆14C distribution in the eastern and 
western Indian Ocean.  In both figures the distribution pattern is very similar 
to that seen in the Pacific Ocean WOCE samples.  In the Pacific the maximum 
∆14C values were frequently found in shallow water, but beneath the surface.  
In the Indian Ocean data a subsurface maximum is not so common.  Both sections 
show intrusion of Circumpolar Deep Water from the south along the bottom and 
return flow of deep water at 2000-3000m.  As with the Pacifiic the middepth 
waters have the lowest ∆14C values, however the middepth Indian Ocean waters 
have significantly higher values that corresponding Pacific waters.  This 
pattern is consistent with a mean ageing of waters from the Atlantic to Indian 
to Pacific. 

Figure 7, Figure 8 and Figure 9 show zonal ∆14C sections along the WOCE lines 
I1 (~10°N), I2(~8°S) and I3(~20°S).  Except for the western ends, the ∆14C 
contours in the upper kilometer are relatively flat.  In each section the deep 
waters of the western basins have somewhat higher ∆14C than at the same depth 
in the eastern basins.  The strength of this signal decreases from south to 
north and is almost certainly due to the western basins having a higher fraction 
of North Atlantic Deep Water. 

Figure 10 shows the meridional distribution of bomb produced ∆14C (via Rubin 
and Key, 2002) in the eastern and western Indian Ocean.  The eastern section 
used all WOCE samples collected at depths less than 1000m and east of 85°E.  The 
western section uses the same depth range, but samples from west of 75°E.  Both 
sections are contoured and colored in potential density space rather against 
depth.  One might expect a priori that the distributions would differ north of 
the equator due to the geography and difference in chemistry between the Bay of 
Bengal and Arabian Sea.  Perhaps unexpected is the fact that the distributions 
differ significantly as far as 40°S.  In the eastern section the maximum bomb 
∆14C values are found between 40°S and 20°S and more or less uniformly from the 
surface down to the level where σθ~26.5.  The western section has a maximum in 
the same latitude range but in this case the maximum occurs as a subsurface 
lens. 


Figure  1: AMS 14C station map for WOCE S4I. 
Figure  2: ∆14C results shown with 2σ error bars. 
Figure  3: ∆14C as a function of silicate for samples collected deeper than 
           1000m.  The black points are from north of 35°S and the red points 
           south of 35°.  The straight line shows the relationship proposed by 
           Broecker, et al., 1995 (∆14C = -70 - Si with radiocarbon in ‰ and 
           silicate in µmol/kg). 
Figure  4: Based on the potential alkalinity method (Rubin and Key, 2002), the 
           samples which plot above the line and have potential alkalinity 
           values less than about 2400 µmole/kg are contaminated with bomb-
           produced 14C. 
Figure  5: ∆14C, along I8S and I9N in the eastern Indian Ocean. 
Figure  6: ∆14C along I7 in the western Indian Ocean. 
Figure  7: ∆14C along I1 in the northern Indian Ocean. 
Figure  8: ∆14C along I2 in the southern tropical Indian Ocean. 
Figure  9: ∆14C along I3 in the southern subtropical Indian Ocean at 
           approximately 20°S. . 
Figure 10: Mean bomb-produced ∆14C sections in the eastern (A) and western 
           (B) Indian Ocean, shown in potential density space for samples from 
           the upper 1000m. 
Figure 11: (A) ∆14C and (B) bomb-produced ∆14C for the surface Indian Ocean 
           from WOCE measurements. 
Figure 12: (A) ∆14C and (B) bomb-produced ∆14C on σθ=24.0. 
Figure 13: (A) ∆14C and (B) bomb-produced ∆14C on σθ=25.0 
Figure 14: (A) ∆14C and (B) bomb-produced ∆14C on σθ=26.0 
Figure 15: (A) ∆14C and (B) bomb-produced ∆14C on σθ=26.5 
Figure 16: (A) ∆14C and (B) bomb-produced ∆14C on σθ=26.8 
Figure 17: (A) ∆14C and bomb-produced (B) ∆14C on σθ=27.1 



5.0 REFERENCES AND SUPPORTING DOCUMENTATION 


Bard, E., M. Arnold, H.G. Ostlund, P. Maurice, P. Monfray and J.-C. Duplessy, 
    Penetration of bomb radiocarbon in the tropical Indian Ocean measured by 
    means of accelerator mass spectrometry, Earth Planet. Sci. Lett., 87, 379-
    389, 1988. 

Broecker, W.S., S. Sutherland and W. Smethie, Oceanic radiocarbon: Separation 
    of the natural and bomb components, Global Biogeochemical Cycles, 9(2), 
    263-288, 1995. 

Chambers, J.M. and Hastie, T.J., 1991, Statistical Models in S, Wadsworth & 
    Brooks, Cole Computer Science Series, Pacific Grove, CA, 608pp. 

Chambers, J.M., Cleveland, W.S., Kleiner, B., and Tukey, P.A., 1983, Graphical 
    Methods for Data Analysis, Wadsworth, Belmont, CA. 

Cleveland, W.S., 1979, Robust locally weighted regression and smoothing 
    scatterplots, J. Amer. Statistical Assoc., 74, 829-836. 

Cleveland, W.S. and S.J. Devlin, 1988, Locally-weighted regression: An approach 
    to regression analysis by local fitting, J. Am. Statist. Assoc, 83:596-610. 

Elder, K.L., A.P. McNichol and A.R. Gagnon, Reproducibility of seawater, 
    inorganic and organic carbon 14C results at NOSAMS, Radiocarbon, 40(1), 
    223-230, 1998 

Joyce, T., and Corry, C., eds., Corry, C., Dessier, A., Dickson, A., Joyce, T., 
    Kenny, M., Key, R., Legler, D., Millard, R., Onken, R., Saunders, P., 
    Stalcup, M., contrib., Requirements for WOCE Hydrographic Programme Data 
    Reporting, WHPO Pub. 90-1 Rev. 2, 145pp., 1994. 

Key, R.M., WOCE Pacific Ocean radiocarbon program, Radiocarbon, 38(3), 415-423, 
    1996. 

Key, R.M., P.D. Quay, G.A. Jones, A.P. McNichol, K.F. Von Reden and R.J. 
    Schneider, WOCE AMS Radiocarbon I: Pacific Ocean results; P6, P16 & P17, 
    Radiocarbon, 38(3), 425-518, 1996. 

Key, R.M. and P. Schlosser, S4P: Final report for AMS 14C samples, Ocean Tracer 
    Lab Technical Report 99-1, January, 1999, 11pp. 

Leboucher, V., J. Orr, P. Jean-Babtiste, M. Arnold, P. Monfrey, N. Tisnerat-
    Laborde, A. Poisson and J.C. Duplessey, Oceanic radiocarbon between 
    Antarctica and South Africa along WOCE section I6 at 30°E, Radiocarbon, 41, 
    51-73, 1999. 

McNichol, A.P., G.A. Jones, D.L. Hutton, A.R. Gagnon, and R.M. Key, Rapid 
    analysis of seawater samples at the National Ocean Sciences Accelerator 
    Mass Spectrometry Facility, Woods Hole, MA, Radiocarbon, 36 (2):237-246, 
    1994. 

NOSAMS, National Ocean Sciences AMS Facility Data Report #99-043, Woods Hole 
    Oceanographic Institution, Woods Hole, MA, 02543, 2/16/1999. 

Osborne, E.A., A.P. McNichol, A.R. Gagnon, D.L. Hutton and G.A. Jones, Internal 
    and external checks in the NOSAMS sample preparation laboratory for target 
    quality and homogeneity, Nucl. Instr. and Methods in Phys. Res., B92, 158-
    161, 1994. 

Rubin, S. and R.M. Key, Separating natural and bomb-produced radiocarbon in the 
    ocean: The potential alkalinity method, Global Biogeochem. Cycles, in 
    press, 2002. 

Sabine, C.L. and R.M. Key, Surface Water and Atmospheric Underway Carbon Data 
    Obtained During the World Ocean Circulation Experiment Indian Ocean Survey 
    Cruises (R/V Knorr, December 1994-January 1996), ORNL/CDIAC-103, NDP-064, 
    Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, 
    Oak Ridge TN, 89 pp., 1997. 

Sabine, C.L., R. Wanninkhof, R.M. Key, C. Goyet, R. Millero, Seasonal CO2 
    fluxes in the tropical Indian Ocean, Mar. Chem., 72, 33-53, 2000. 

Schneider, R.J., A.P. McNichol, M.J. Nadeau and K.F. von Reden, Measurements of 
    the oxalic acid I/oxalic acid II ratio as a quility control parameter at 
    NOSAMS, In Proceedings of the 15th International 14C Conference, 
    Radiocarbon, 37(2), 693-696, 1995. 

Stuiver, M. and H.G. Ostlund, GEOSECS Indian Ocean and Mediterranean 
    radiocarbon, Radiocarbon, 25(1), 1-29, 1983. 






_____________________________________________________________________________________________________________
_____________________________________________________________________________________________________________





CCHDO DATA PROCESSING NOTES

Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  ---------------------
10/03/97  Zimmermann   CTD              Submitted          ready for DQE
          I've just sent you the processed CTD data for the WOCE legs I8S + I9S,
          ready for the DQE step. I've put the files into the incoming directory
          of whpo.ucsd.edu.   
            The files sent are:
            I8SI9S.RP1, the CTD report
            I8SI9S.SUM
            I8S.SEA
            I9S.SEA
            KA45D001.CTD through KA45D147.CTD
          I am leaving for a cruise Oct 5 and will be back from sea Nov 20.  If 
          there are any questions I will not be able to answer them until then.

10/06/97  McCartney    CTD/BTL          Data are Public    Not Yet DQE'd
          The I8S and I9S data should be made available to the community 
          with the caveat as you mentioned, that the data have not been DQE'd.
          
02/23/98  Kozyr        TCARBN/ALKALI    Website UpDate     Take Data Offline
          I have recently looked at the PUBLIC data files for the WOCE 
          I8S/I9S Sections that are currently posted through WHPO WEB site. 
          I discovered that the TCO2 and Alkalinity are completely deferent 
          from those I have from BNL PIs Ken Johnson and Doug Wallace. I 
          thing the TCO2 and TALK data you have are from the Chief Scientist 
          and are the row data from the cruise records. These data have to 
          be removed from the final data set on the WEB. I am currently pre-
          paring WOCE formatted CO2 data files for this and other Indian Ocean 
          cruises, and will send them to you as soon as I finish.
          
02/26/98  Diggs        BTL              Data NonPublic     by PI request
          WOCE Indian Ocean bottle data from Mike McCartney (I08S/I09S) 
          has been encrypted as requested by the PI until further notice.
          
03/09/98  Kozyr        ALKALI/TCARBN    Submitted          Data are Final
          I have put the final CO2-related data for I8S/I9S Indian Ocean Line 
          to the WHPO ftp INCOMING area.
          
03/27/98  Whitworth    CTD/BTL          Data are Public    Includes i08 i03 s04i icm03 pcm09
          Steve Rutz put the ICM3 data on the WHPO FTP site in January, along 
          with the I3 and S4I data, with the provision that it not be made 
          publicly available yet.  We see from the web site that I3 and s4(I) 
          are there, and have resent the ICM3 data to the FTP site.

          The PCM9 deployment and recovery cruise data (Rapuhia and Monowai, respectively) 
          were submitted to NODC in July, 1995.  We will also place these on the FTP site.
          These data are available to the public.
          
09/09/98  Talley       SUM              Data Update:       deleted xtra header lines,  I8 changed to i08s
          Steve - there were 2 extra header lines in i08ssu.txt. I removed them, 
          and also change the section names from I8 and I9 to I08S and I09S. I 
          placed the new file in whpo.ucsd.edu INCOMING.
          
09/30/98  Talley       BTL              Data Update:       corrected expocode
          I made a small change to the first header line of the i08shy.txt and 
          i09shy.txt files - they are from the same cruise and neither of them 
          had the righ expocode. Expocode was changed to 316N145_5 in both 
          files.

10/13/98  Muus         CTD/SUM          Update needed      reformatted files not online
          I08S CTD data has been reformatted and put on imani anonymous ftp 
          pub/INCOMING/i08sCTD+SUM.tar.gz. The tar file also includes a comment 
          file and a corrected copy of the summary file. The original summary 
          file is still on the web but Sarilee reformatted it Feb 6, 1998 (web 
          sample summary file) and I reformatted it Aug 12th. I corrected the 
          Station 102 date on my reformatted file and included it in the tar 
          file but it still has the /s in the EXPOCODEs as per instuctions last 
          August.

12/01/98  Diggs        BTL              Website Updated:   cfcs, carbon data Public
          CFCs removed (masked) from bottle files and decrypted for public 
          consumption per McCartney's instructions.  Also removed ALKALI and 
          TCARBN as well as replacing the string FC02 (with a zero) with the 
          string FCO2 (with an 'o') in both the i09s and i08s bottle files.

12/01/98  McCartney    BTL              Update Needed      Change status to Public
          Someone pointed out to me that the bottle files for I08S and I09S are 
          still encrypted and in non public status. I do not recall there being 
          some reason for this but as far as I am concerned, they should be 
          realeased for public use.  - Mike McCartney
          
          

Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  ---------------------
12/01/98  McCartney    CFCs/TCARBN/ALK  Data UpDate        Data encrypted by PI request
          CFCs removed (masked) from bottle files and decrypted for public 
          consumption per McCartney's instructions.  Also removed ALKALI and 
          TCARBN as well as replacing the string FC02 (with a zero) with the 
          string FCO2 (with an 'o') in both the i09s and i08s bottle files.
          
06/16/99  Diggs        CTD/BTL          Website UpDate     corrected units in BTL file, reformatted CTD 
          You are correct, the values were in ml/l and the CTD files were in 
          a non-WOCE format.  I have rectified this situation by replacing  
          both the CTD zipfile and the hydro file with newer versions that 
          are in WOCE formt (CTD) and a newer hydro file wih the correct 
          units for Oxygen.    -sd

          Stephen - I downloaded the data for I08S and I09S today, 26 May. I 
          compared the water sample data to data I had retrieved in April 
          1995 from the Indian Ocean preliminary data site at WHOI available 
          to Indian Ocean PIs (I work for Arnold Gordon).  The data from 
          your WHPO site has less resolution than the data from 1995.  The 
          oxygens in the hydro files have a resolution of only one decimal 
          place, compared to three in 1995.  Phosphate has two compared to 
          three.  The difference seems to be more than a rounding error, as 
          the 1995 data rounded to one decimal place does not result in the 
          value I retrieved.  I suppose if the data were updated and then 
          rounded, this could account for the difference. Also, I see in the 
          data description that the CTD data was reformatted by WHPO.  The 
          data downloaded is still in the original WHOI format, dated Aug 
          1995.  Is there a final version?   - Phil Mele
          
08/17/99  Anderson     SUM/HYD          Data Update:       No errors detected
          I have checked the .sum and .sea/.hyd files for lines A08, A12, 
          I08S/I09S, and P14S.  The files on the web page for A08 and 
          I08S/I09S adhere to the WHP format specifications, and I have run 
          them over the programs wocecvt and sumchk without any errors. 
          
09/29/99  Falkner      BA               Update Needed:     Data quallity does not meet WOCE standards
          The quality of the Ba data from most WOCE legs in the Indian Ocean 
          turned out to be quite  poor; far worse than attainable analytical 
          precision (+/-20% as opposed to 2%). We recorded many vials which 
          came back with loose caps and evaporation associated with that 
          seems to be the primary problem.  The only hope I have of 
          producing a decent data set is to run both Ba and a conservative 
          element simultaneously and then relating that to the original 
          salinity of the sample.  We will be taking delivery on a high 
          resolution ICPMS here at OSU sometime this winter which would make 
          the project analytically feasible and economical.  I do not 
          presently have the funds in hand to do this and so have archived 
          the samples for the time being. I don't think the WHPO would 
          derive any benefit from the present data set.  -- KKF
          
12/16/99  Bullister    BTLNBR           Update Needed:     stations missing or replicated
          Stations 1-3 are absent (they were present iin an earlier versioni 
          of the file)
          sta 39 samp 1 is repeated 19 times
          sta 129 samp 1 is repeated 4 times
          
12/16/99  Bullister    CFCs             Submitted          Data are Final & Public
          Post these revised files at the web site, with the CFC data 
          'public' for these cruises.
          
02/08/00  Newton       CFCs/CO2         Website UpDate     Date Merged into hyd file
          Notes on merging of CFC's, TCARBN, ALKALI   316N145_5   I08S/09S  
          In i08shy.txt. Removed 18 duplicate lines at Station 39 cast 1 
          sampno 1.
            Removed 4 duplicate lines at station 129 cast 1 sampno 1.
            Merged in CFC11 and CFC12.  Merged in TCARBN and ALKALI. Source for
            TCARBN and ALKALI was f8.2, Rounded to f8.1 for merged .hyd file.
            Stations 1,2,3 were not in .hyd file, but were in cfc file.
          David Newton   09Feb2000
          
02/09/00  Diggs        CFCs/CO2         Website UpDate     CFCs Public, CO2 masked
          David Newton and I have done some work on I08S/I09S bottle data.  
          The CFCs have been updated with values from J. Bullister's 12/1999 
          data submission and Alex Kozyr's carbon values.  The carbon values 
          on-line have been masked out pending public release from Alex.
          All tables and files have been updated accordingly.
          
02/14/00  Kozyr        TCARBN/ALKALI    Submitted          Data Final, DQE Complete
          I've just put a total of 13 files [carbon data measured in Indian 
          (6 files) and Atlantic (7 files) oceans] to the WHPO ftp area. 
          Please let me know if you get data okay.
          
03/01/00  Whitworth    CTD              Update Needed      Incorrect Oxy units
          Bob Key tells me he's notified you of the nutrients units problem 
          on the I8S and I9S bottle files.  The CTD files have the same 
          problem with oxygen - e.g. values in ml/l interpreted as umol/kg.
          
          
Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  -------------------          
03/24/00  Schlosser    He/Tr            Data are Public    Not final
          As mentioned in my recent message, we will release our data with a 
          flag that indicates that they are not yet final. We started the 
          process of transferring the data and we will continue with the 
          transfer during the next weeks. I had listed the expected order of 
          delivery in my last message.
          
04/25/00  Anderson     NUTs             Data Update:       Units changed from UMOL/L to UMOL/KG
          Nutrients were labeled UMOL/KG but were really UMOL/L. Converted 
          mislabeled nutrients from UMOL/L to UMOL/KG. Subtracted NITRIT 
          from NO2+NO3 to get NITRAT.  
          
04/25/00  Anderson     NUTs/CTDOXY      Update Needed      Correct NUTs Units not yet online
          In March of 1998 I reformatted (this was before our accepted 
          format was in place) I08S, I09S.  At that time I noted that the O2 
          was in ML/L and the nuts had the wrong unit headings, which I 
          changed from UMOL/KG to UMOL/L.  Perhaps that file was never put 
          up on the web site, but the file there now has the O2 in the 
          correct units UMOL/KG, and the nuts are as stated by Orsi in 
          UMOL/L but say UNOL/KG.  Also there is NO2+NO3 and NITRIT.  All of 
          the above I can correct in a short period of time.  Should I go 
          ahead and do this?   
          I note that the ctd files for this line still have O2 in  ml/l.  
          
04/27/00  Diggs        NUTs/CTDOXY      Website Update:    Reformatted NUTs/CTDOXY online
          I have replaced the older I08S/I09S files with the ones that 
          Sarilee recently sent.  All tables and meta files have been 
          updated.
          
05/05/00  Quay         DELC14           Data are Public    I08S DEL14C data are public
          You can make the 14C data from I8S open to the public.
          
06/23/00  Schlosser    HELIUM/NEON      Submitted          also NEON
          2000.10.27 KJU
          Moved files from ftp-incoming.2000.10.23/ Files contain 
          documentation and bottle data. Could not determine who sent the 
          files. No relevant email was found. They were received on June 23, 
          2000 along with other cruises that had the same format. Path is 
          i08/i08s/original/2000.06.23_I8S_DOC_SEA.
          
08/04/00  Warren       NUTs             Update Needed      Units & DQE status unclear
          Was I right that the I2 nutrients were in per liter rather than 
          per kilogram, and that I8N and I9N were in per kilogram?  Also, 
          did Joe Jennings and Lou Gordon ever review the I1 nutrients?  All 
          I have seen is the shipboard data, and the silicic acid values 
          there for Stations 973, 974, 975, and 996 appear high at all 
          depths, suggestive of a standardization problem.
          
08/31/00  Kozyr        OXYGEN           Update Needed      Units are ml/l, should be umol/kg
          in the I9S/I8S and I1 .hy files oxygen is given in ml/l instead 
          umol/kg as it is in the rest of section and in WHPO manual 
          suggested.
          
09/26/00  Schlosser    TRITUM           No Data Submitted  Data not yet calibrated
          Tritium data will be submitted later (after intercalibration). 
          We hold tritium data for a subset of our He lines only. 
          WHP lines with tritium: 
          S4P> S4I (East)> I8S> I9S> P9
          
09/27/00  Kappa        Cruise Report    Data UpDate        New PDF & TXT files completed
          
09/29/00  Huynh        Cruise Report    Website Updated:   pdf, txt versions online
          
02/07/01  Mantyla      NUTs/S/O         DQE Begun          Agreed to DQE Indian BTL data
          I would be glad to look over the Indian Ocean data for you. 
          Sarilee has started plotting up I01 for me to start on. - Arnold
          
02/26/01  Schlosser    HELIUM/DELHE3    Data are Public    minor corrections may be needed
          Following up on Bill Jenkins's message, I would like to ask you to 
          make public all ldeo woce tritium/he data that have been submitted 
          to you.  Because the tritium/he community has not yet finished the 
          final calibration of the data, I might have to apply minor 
          corrections to these data once the intercal. Effort has been 
          completed.  Our acce work was funded over a 5-year period that 
          ended in 2000.  Consequently, this data set is further behind in 
          quality control before submission, but i expect that we will get 
          these data ready soon.
          
06/21/01  Uribe        CTD/BTL          Website Updated:   Exchange file online
          CTD and bottle exchange files were put online.
          
          
Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  ---------------------          
09/18/01  Wisegarver   CFCs             Submitted          Data final & public
          This is information regarding line:  I08S
          ExpoCode:  316N145_5
          Cruise Date:  1994/12/01 - 1995/01/19
          From:  WISEGARVER, DAVID
          Email address:  WISE@PMEL.NOAA.GOV
          Institution:  NOAA
          Country:  USA
          The directory this information has been stored in is:  20010918.171618_WISEGARVER_I08S
          The format type is:   ASCII   
          The data type is:  BottleFile 
          The Bottle File has the following parameters:  CFC-11, CFC.12
          The Bottle File contains:  CastNumber StationNumber BottleNumber SampleNumber
          WISEGARVER, DAVID would like the data PUBLIC.
          And would like the following done to the data:   MERGE FINAL DQE CFC DATA
          Any additional notes are:  SUBMITTED FOR D.WYLLIE. CFCS ON SIO98 SCALE.

12/24/01  Uribe        CTD              Website Updated:   Exchange file online
          CTD has been converted to exchange using the new code and put online.
          
01/03/02  Hajrasuliha  CTD              WHPO QC done       .ps & *check.txt files created
          created .ps files for this cruise. created *check.txt file for this cruise.
          
02/01/02  Anderson     TCARBN/ALKALI    Website Updated:  Data merged into online file, new CSV file added
          Merged TCARBN and ALKALI into bottle file and made new exchange file. 
          Put both new files online.
          
03/04/02  Bartolacci   CFC's            Submitted          Data are Final, DQE'd
          I have placed the DQEd CFC data sent by D. Wisegarver in the appropriate 
          I08S original directory.  Included in the directory are website submission 
          README file and data file containing CFC11/12 and quality flags. Data are 
          in need of merging at this time.
          
04/01/02  Gerlach      DELC13           Submitted          Data are Public, with Q flags
          The data disposition is:   Public 
          The file format is:   Plain Text (ASCII) 
          The archive type is:   NONE - Individual File 
          The data type(s) is:  Other:  flagged 13C data
          The file contains these water sample identifiers:
          • Cast Number (CASTNO)
          • Station Number (STATNO)
          • Bottle Number (BTLNBR)
          GERLACH, DANA would like the following action(s) taken on the data:
          • Merge Data
          • Place Data Online
          Any additional notes are:
          If there are questions, concerns, or problems, please contact:
          • Dana Gerlach (dgerlach@whoi.edu) or
          • Ann McNichol (amcnichol@whoi.edu)  
          The data disposition is: Public  
          The file format is:      Plain Text (ASCII) 
          The archive type is:     NONE - Individual File 
          The data type(s) is:     Documentation
          • Other:  flagged 13C replicate data
          The file contains these water sample 
          • Cast Number    (CASTNO)
          • Station Number (STATNO)
          • Bottle Number  (BTLNBR)
          GERLACH, DANA would like the following action(s) taken on the data:
          • Other: provide as reference
          Any additional notes are:
          • This description file lists the individual flags for the replicate values.  
            It is a detailed listing of those stations which have c13f = 6. 
            DELC13 replicate samples (from 20020401.104111_GERLACH_I08S & 
            20020401.104233_GERLACH_I09S)
          WHPID:  I08S
          expocode:  316N145/5
          depth_corr  station cast    niskin  del_c13 c13f    average num_reps
          57.483       9       1       22      1.608   2       1.611   2
          57.483       9       1       22      1.614   2
          57.354       21      1       22      1.457   2       1.469   2
          57.354       21      1       22      1.480   2
          56.247       27      1       22      1.499   2       1.514   2
          56.247       27      1       22      1.530   2
          57.804       41      1       34      1.656   2       1.671   2
          57.804       41      1       34      1.686   2
          57.179       56      1       34      1.621   2       1.627   2
          57.179       56      1       34      1.633   2
          52.786       75      1       32     -0.323   2      -0.323   1
          52.786       75      1       32     -4.722   4
          54.059       85      1       34      0.679   2       0.726   2
          54.059       85      1       34      0.773   2
          54.582      100      1       20      0.728   4       1.517   2
          54.582      100      1       20      1.517   2
          54.908      110      1       34      1.513   2       1.484   2
          54.908      110      1       34      1.455   2
          33.123      122      1       35      1.329   2       1.462   2
          33.123      122      1       35      1.595   2
          57.748      134      1       34      1.625   2       1.616   2
          57.748      134      1       34      1.607   2
          494.872     143      1        2      1.260   2       1.247   6
          494.872     143      1        2      1.216   2
          494.872     143      1        2      1.284   2
          494.872     143      1        2      1.288   2
          494.872     143      1        2      1.199   2
          494.872     143      1        2      1.237   2
          
08/09/02  Anderson     ALKALI           Website Updated:   TCARBN/ALKALI/C13/C14/CFCs Online
          Merged the DELC14 and C14ERR from Key, the DELC13 from Gerlach, and 
          the TCARBN and ALKALI from Kozyr. Created QUALT2 flags by copying 
          the QUALT1 flags. Merged the CFCs from Wisegarver. Made new exchange file. 
          Notes for i08s/i09s:
          Merged the DELC14 and C14ERR from file I8SI9S.C14 found in
          /usr/export/html-public/data/onetime/indian/i08/i08s/original/20020410_KEY_I8SI9S_C14 
          into online file 20020201WHPOSIOSA.
          Remerged the TCARBN and ALKALI from file 
          2000.02.14_CO2_KOZYR_i8si9sdat.txt found in /usr/export/html-     
          public/data/onetime/indian/i08/i08s/original/moved _from_ftp-incoming.2000.02.14 
          into online file.  This file contains moreup-to-date data.
          Merged the DELC13 (i08s only) from file 20020401.102044_GERLACH_I08S_whpo_i08s.txt found in 
          /usr/export/html-public/data/onetime/indian/i08/i08s/original/20020401.102044_GERLACH_I08S 
          into online file. Merged the DELC13 (i09s only) from file 20020401.102306_GERLACH_I09S_whpo_i09s.txt 
          found in
          /usr/export/html-public/data/onetime/indian/i09/i09s/original/20020401.102306_GERLACH_I09S 
          into online file.
          Created QUALT2 flags by copying the QUATL1 flags.
          Merged CFC11 and CFC12 from file 20010918.171618_WISEGARVER_I08S_i08s_CFC_DQE.dat found in 
          /usr/export/html-public/data/onetime/indian/i08/i08s/original/2001.09.18_I08S_CFC_DQE_WISEGARVER 
          into online file.
          
          
Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  ---------------------          
08/14/02  Anderson     HELIUM/NEON      Website Updated:   HELIUM/DELHE3/NEON Online
          Merged the DELHE3, DELHER, HELIUM, HELIEER, NEON, and NEONER into 
          online file. Made new exchange file.
          Merge notes for i08s:
          DELHER, HELIUM, HELIER, NEON, and NEONER from file i8SHeNe.SEA found i
          /usr/export/html-public/data/onetime/indian/ i08/i08s/original/2000.06.23_I8S_DOC_SEA
          into online file 20020809WHPOSIOSA.
          Merged DELHE3, DELHER, HELIUM, HELIER, NEON, and NEONER from file i9SHeNe.SEA found in 
          /usr/export/html-public/data/onetime/indina/i09/i09s/original/2000.06.23_I9S_DOC_SEA 
          into online file.
          
09/19/02  Anderson     CTDSAL/CTDOXY    Update needed      flag problems
          Alex Kozyr noticed that the online bottle file had 1 flags for almost 
          all the CTDSAL and CTDOXY. In looking at the file I noticed that there 
          are many other flag problems. These need to be investigated and corrected 
          when time allows.
          
09/30/02  Kozyr        NUTs             Update Needed      stn 138, btl 1 should all be flagged 4
          In I08SI09S files i08shy.txt and i08s_hy1.csv, station 138, bottle 1 
          (last line for this station) all nutrients are obvious outliers and should 
          be flagged 4 (bottle has flag 3).
          
06/20/03  Anderson     TRITUM           Website Updated:   Data Reformatted/OnLine
          Merged TRITIUM and TRITER sent by Bob Newton April 29, 2002 into online 
          file. Made new exchange file.
          June 20, 2003
          Merged TRITIUM and TRITER into online file.  Tritium file sent by Bob 
          Newton April 29, 2002. There were 7 stations in the tritium file that 
          had duplicate tritium values.  The merge program uses the first value.  
          Below are the duplicate values.
            sect_id  stnnbr  castno  sampno  depth    Tritum     flags     TrEr
              I08S      4       1      21      800  1.021137124   22    0.013095889
              I08S     35       1      28      159  1.047400814   44    0.033042912
              I08S     35       1      27      259  0.744451287   44    0.012347122
              I08S     35       1      25      457  0.306869291   22    0.007097154
              I08S     59       1       6     3885  0.146283093   44    0.006416019
              I09S     97       1      11     2402  0.123630035   44    0.004821189
              I09S    114       1      32      107  0.406992072   22    0.007447354
         The original file from Newton only had one quality flag.  I copied that into the QUALT2 field.

01/10/05  Key          DELC14           Report Submitted   covers 9 Indian Ocean cruises, 1/94-1/96
          The U.S. WOCE Indian Ocean Survey consisted of 9 cruises covering the 
          period December 1,1994 to January 22,1996.All of the cruises used the 
          R/V Knorr operated by the Woods Hole Oceanographic Institute. A total 
          of 1244 hydrographic stations were occupied with radiocarbon sampling 
          on 366 stations.
          
05/18/05  Anderson     CTD              Website Updated:   changed number of records to 1573
          Changed the NO. RECORDS in file i08s0037.WCT from 1572 to the correct 
          value1573. Rezipped i08sct.zip and put new file online. 

11/08/07  Swift        S/O2/NUTs        update needed      data & Q flag problems detailed  
          Noted following problems w/ I8S/I9S Exchange bottle data file:
          Bottle quality flag 9, but there are values for some bottle data 
          parameters (station/cast/sample):
          27   1  23
          47   1  32
          37   1  33
          28   1  19
          I found no bottle quality code 1 flags, though I thought Lynne mentioned some.
          
          Bottle quality flag 2 but no bottle data:
          52   1  29
          92   1  36
          104  1  21
          143  1  2
          
          
Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  ---------------------
11/08/07  Swift        S/O2/NUTs        update needed      data & Q flag problems detailed (continued)
          Bottle quality flag 3, but missing or bad bottle data:
          36   1  11
          123  1  34
          32   1  30
          
          Bottle quality flag 3, but bad salts and oxygens (some may have nuts coded 2?):
          15   1  11  63   1  11
          145  1  6   138  1  1
          10   1  2   66   1  11
          129  1  3   119  1  25
          
          similar to previous, but may have code 3 for salt or oxygen:
          18  1  7
          26  1  13
          44  1  32
          
          Bottle code 2 but no bottle data (at least for S, O2, and, I think, nuts):
          52  1  29  104  1  21
          92  1  36  143  1  2
          
          Bottle coded 3, but with good oxygens and, in all but three cases, good 
          salts (should these be code 2 bottles??):
          15   1  24  62   1  20  21   1  1   12   1  16
          55   1  19  128  1  10  7    1  15  11   1  16
          109  1  17  66   1  24  39   1  4   138  1  24
          124  1  23  116  1  10  20   1  1   144  1  22
          122  1  35  63   1  30  80   1  12  141  1  28
          4    1  1   124  1  10  17   1  1   135  1  28
          133  1  14  39   1  10  39   1  1   126  1  23
          138  1  13  96   1  17  103  1  11  132  1  32
          72   1  26  122  1  10  19   1  1   18   1  23
          100  1  11  113  1  13  39   1  2   120  1  28
          98   1  26  44   1  10  80   1  10  27   1  20
          109  1  24  89   1  24  86   1  24  130  1  24
          74   1  24  85   1  29  93   1  11  48   1  19
          111  1  24  89   1  23  126  1  21  128  1  28
          62   1  25  39   1  8   86   1  17  14   1  23
          96   1  28  113  1  10  121  1  22  42   1  29
          39   1  12  21   1  13  141  1  22  43   1  34
          120  1  12  34   1  4   46   1  2   114  1  36
          34   1  11  36   1  20  114  1  24  48   1  23
          115  1  13  115  1  6   22   1  15  97   1  31
          28   1  11  74   1  13  42   1  22  113  1  36
          68   1  24  111  1  11  65   1  10  108  1  36
          94   1  30  80   1  17  89   1  9   108  1  34
          122  1  12  64   1  16  103  1  7   62   1  34
          62   1  22  39   1  5   125  1  22  64   1  36
          134  1  11  92   1  15  140  1  22  94   1  34
          91   1  28  118  1  3   85   1  5   
          
Date      Contact      Data Type        Action             Summary
--------  -----------  ---------------  -----------------  --------------------------------------------------------------------------------------
11/09/07  Muus         S/O2/NUTs        Update Needed:     will correct errors noted by J.Swift
          The oldest I08S, I09S 1994 bottle files I can find are i8s.sea and 
          i9s.sea dated Oct 3, 1997, Expocode 31ka45, which appear to me to 
          be the original WHOI files.  They contain the same quality flag 
          problems that have been carried  through to the present.
          I cannot find any record of a DQE. Jerry's hard copy book has a 
          message from you, dated Feb 6, 2001, to Arnold Mantyla requesting 
          he DQE all Indian Ocean WOCE cruises together with Arnold's response 
          saying he would be glad to do it and would start with I01. But the 
          book has no further reference to a DQE for I08S/I09S.
          I will correct the problems you found and then recheck for any other problems.

11/21/07  Jennings     NITRAT           Update Needed:     flag stn 60, 3358.4 dbar "-9"
          There was a bubble in the nitrite flow cell which caused the high 
          absorbance reading (station 60, 3358.4 dbar value of 1.08). Since it 
          is an obvious problem, I'd replace the bad value with a -9.
          
03/24/08  Muus         DELC13/CFCs      Website Updated    Qual flag correctioins
          Notes on changes to I08S-I09S_1994 20051213 bottle data files:   
          EXPOCODE 316N145_5
          
          1. No PCO2 data so PCO2 column deleted.
          2. CTDSAL & CTDOXY quality flag "1"s(Stations 4-59) and "3"s(Stations 60-147) 
             that first appeared in the 19980616 bottle files were changed back to the 
             original quality flag "2"s.
          3. Inconsistent quality flags were changed to more logical values based on plots of data values:
          
          STNNBR  CASTNO  SAMPNO  BTLNBR  CTDPRS      20051213 QUALT1           NEW QUALT2                  
             4       1       8  SIH029  1725.7 31129222299559995999 22229222299559995999   Btl 2 vs 3
             4       1       1  SIH036  1937.0 31122222222422625222 22222222222422625222   Btl 2 vs 3
             7       1      15  SIH021   810.6 31122222299259225999 22222222299259225999   Btl 2 vs 3
            10       1       2  SIH001  4373.7 31144222299559995999 32244444499559995999   Nuts 4 vs 2, Bottle appears to have leaked.
            11       1      16  SIH022   608.7 31122222222559225999 22222222222559225999   Btl 2 vs 3
            12       1      16  SIH022   609.6 31122222222559995999 22222222222559995999   Btl 2 vs 3 
            14       1      23  SIH033    30.6 31122222299559995999 22222222299559995999   Btl 2 vs 3
            15       1      24  SIH037    12.6 31192222299559995999 22292222299559995999   Btl 2 vs 3
            15       1      11  SIH015  1516.4 31144444299559335999 32244444499559335999   PO4 4 vs 2
            17       1       1  SIH036  4192.5 31122222233559225999 22222222233559225999   Btl 2 vs 3  
            18       1      23  SIH033    33.4 31122222299229995922 22222222299229995922   Btl 2 vs 3
            18       1       7  SIH009  2734.4 31143444299439995933 32243444499439995933   PO4 4 vs 2
            19       1       1  SIH036  4160.5 31122222299559625999 22222222299559625999   Btl 2 vs 3
            20       1       1  SIH036  3896.1 31122222222559995999 22222222222559995999   Btl 2 vs 3
            21       1      13  SIH018  1202.3 31122222222552225299 22222222222552225299   Btl 2 vs 3
            22       1      15  SIH021   812.2 31122222299559995999 22222222299559995999   Btl 2 vs 3
            26       1      13  SIH018  1009.8 31143222299559995999 22243222299559995999   Btl 2 vs 3
            27       1      23  SIH033    32.0 93399222299559995999 92399444499559995999   Nuts 4 vs 2. All are exact dupe of sample 23 so 
                                                                                                        probably no water sample obtained.
            27       1      20  SIH028   208.3 31122222222552295299 22222222222552295299   Btl 2 vs 3.
            28       1      19  SIH027   307.5 93399222299559995999 22399222299559995999   Btl 2 vs 9, Nuts only water samples given but look OK.
            28       1      11  SIH015  1415.0 31122222299559995999 22222222299559995999   Btl 2 vs 3.
            32       1      30  SIH029   206.3 31149444299559225999 32249444499559445999   PO4 4 vs 2. TCARBN 14 low, ALKALI 10 hi @ 272db qf2 4 
                                                                                                       vs 2 [ctds max, ctdo min, btls-ctds=-.08]
            34       1      11  SIH015  1829.9 31122222299559995999 22222222299559995999   Btl 2 vs 3
            34       1       4  SIH004  2943.7 31122222299559995999 22222222299559995999   Btl 2 vs 3
            36       1      20  SIH028   916.6 31122222299559995999 22222222299559995999   Btl 2 vs 3
            37       1      33  SIH026    82.9 93399444299559995999 93399444499559995999   PO4 4 vs 2
            37       1      31  SIH024   157.5 22221222222559295999 22229222222559295999   O2  9 vs 1
            37       1      18  SIH003  1115.2 41144444222559395999 42244444444559495999   PO4,f11,f12 4 vs 2; TCARBN 4 vs 3
            39       1      12  SIH016  1721.5 31122222222559225999 22222222222559225999   Btl 2 vs 3
            39       1      10  SIH013  2135.7 31122222222559225999 22222222222559225999   Btl 2 vs 3
            39       1       8  SIH010  2565.7 31122222222559225999 22222222222559225999   Btl 2 vs 3
            39       1       5  SIH006  3158.2 31122222222559235999 22222222222559235999   Btl 2 vs 3
            39       1       4  SIH004  3360.3 31122222222559225999 22222222222559225999   Btl 2 vs 3
            39       1       2  SIH001  3698.6 31122222222559225999 22222222222559225999   Btl 2 vs 3
            39       1       1  SIH036  3743.6 31122222222559695999 22222222222559695999   Btl 2 vs 3
            41       1      11  SIH015  1823.8 31129222299559225999 22229222299559225999   Btl 2 vs 3
            42       1      34  SIH029    56.3 93399999999549995924 72399999999549995944   Dhe3 4 vs 2
                                                                                           Btl 7 vs 9 DELHE3,HELIUM,NEON submitted i8SHeNe.SEA
            42       1      29  SIH014   256.1 31122222299559995999 22322222299559995999   Btl 2 vs 3
            42       1      22  SIH031   706.1 31122222299559995999 22222222299559995999   Btl 2 vs 3
            43       1      34  SIH029    56.7 31122222299559225999 22222222299559225999   Btl 2 vs 3
            44       1      32  SIH023   105.2 31133222266559995999 22233222266559995999   Btl 2 vs 3
            44       1      10  SIH013  2320.4 31122222222559995999 22222222222559995999   Btl 2 vs 3
            46       1       2  SIH001  3810.3 31122222299559225999 22322222299559225999   Btl 2 vs 3
            47       1      32  SIH023    81.8 93399222299559995999 92399444499559995999   Nuts 4 vs 2
            48       1      23  WHF017    26.7 31122222222552995299 22222222222552995299   Btl 2 vs 3
            48       1      19  WHF014   297.8 31122222299552995299 22222222299552995299   Btl 2 vs 3
            52       1      29  SIH014   253.6 23399999999559995999 73399999999559995999   Btl 7 vs 2  no water sample data
            55       1      19  SIH027  1026.2 31142333322559995999 32243333333559995999   Oxy 3 vs 2; f11&f12 3 vs 2
            62       1      34  SIH029    57.9 33322222299552225299 22222222299552225299   Btl 2 vs 3
            62       1      25  SIH019   464.6 33322222222552225299 22222222222552225299   Btl 2 vs 3
            62       1      22  SIH031   721.8 33322222299559925999 22222222299559925999   Btl 2 vs 3
            62       1      20  SIH028   913.3 33322222299552995299 22222222299552995299   Btl 2 vs 3
            63       1      30  SIH017   207.4 33322222299559995999 22222222299559995999   Btl 2 vs 3
            63       1      11  SIH015  2438.8 33344222299559995999 32244444499559995999   Nuts 4 vs 2, data indicates leak.
            64       1      36  SIH035    10.8 33322222222559625999 22222222222559625999   Btl 2 vs 3
            64       1      16  SIH022  1414.9 33322222299559225999 22222222299559225999   Btl 2 vs 3
            65       1      10  SIH013  2736.4 33322222999559995999 22222222299559995999   Btl 2 vs 3
            66       1      24  SIH037   503.2 33322222299559225999 22222222299559225999   Btl 2 vs 3
            66       1      11  SIH015  2431.5 33344222299559225999 32244444499559445999   Nuts 4 vs 2, ALKALI & TCARBN 4 vs 2, data indicate leak
            67       1       3  SIH025  2941.1 43344222299559995999 42244444499559995999   Nuts 4 vs 2, data indicates mistrip or leak.
            68       1      24  SIH037   508.2 33322222222429995922 22222222222429995922   Btl 2 vs 3
            72       1      26  SIH005   104.4 33322222299559225999 22222222299559225999   Btl 2 vs 3
            74       1      24  SIH037   307.1 33322222299559995999 22222222299559995999   Btl 2 vs 3
            74       1      13  SIH018  1214.6 33322222299559995999 22222222299559995999   Btl 2 vs 3
            80       1      17  SIH024  1214.8 33322222299559995999 22222222299559995999   Btl 2 vs 3
            80       1      12  SIH016  1721.2 33322222299559995999 22222222299559995999   Btl 2 vs 3
            80       1      10  SIH013  1924.6 33322222299559995999 22222222299559995999   Btl 2 vs 3
            85       1      29  SIH014   254.8 33322222222552295299 22222222222552295299   Btl 2 vs 3
            85       1       5  SIH006  2330.4 33322222222552225299 22222222222552225299   Btl 2 vs 3
            85       1      24  SIH037  2908.2 93329999999559995999 92299999999559995999   SALNTY 9 vs 2 value = -9
            86       1      24  SIH037   506.2 33322222299229225922 22222222299229225922   Btl 2 vs 3
            86       1      17  SIH024  1217.7 33322222299249225924 22222222299249225924   Btl 2 vs 3
            89       1      30  SIH017   206.2 43333222299459995999 42233333399459995999   Nuts 3 vs 2, nut data look ok but do not know reason 
                                                                                                        for btl qf=4 and no other values 2.
            89       1      24  SIH037   509.2 33322222222559995999 22222222222559995999   Btl 2 vs 3
            89       1      23  SIH033   609.7 33322222299559995999 22222222299559995999   Btl 2 vs 3
            89       1       9  SIH012  2235.9 33322222222429995922 22222222222429995922   Btl 2 vs 3
            91       1      28  SIH011   306.4 33322222222559995999 22222222222559995999   Btl 2 vs 3
            92       1      36  SIH035    10.9 23399999999559995999 73399999999559995999   Btl 7 vs 2, No water sample data.
                                                                                                       Original CTDS&O qf=3(i9s_sea.txt)
            92       1      15  SIH021  1629.7 33322222222529225922 22222222222529225922   Btl 2 vs 3
            93       1      11  SIH015  2436.5 33322222299559995999 22222222299559995999   Btl 2 vs 3
            94       1      34  SIH029    57.3 33322222299559995999 22222222299559995999   Btl 2 vs 3
            94       1      30  SIH017   202.4 33322222299559225999 22222222299559225999   Btl 2 vs 3
            96       1      28  SIH011   306.1 33322222299559225999 22222222299559225999   Btl 2 vs 3
            96       1      17  SIH024  1217.0 33322222299559225999 22222222299559225999   Btl 2 vs 3
            97       1      31  SIH020   156.1 33322222299559995999 22222222299559995999   Btl 2 vs 3
            98       1      26  SIH005   409.0 33322222299559995999 22222222299559995999   Btl 2 vs 3
           100       1      11  SIH015   456.5 33322222222552235299 22222222222552235299   Btl 2 vs 3
           103       1      11  SIH015  2435.7 33322222222229995922 22222222222229995922   Btl 2 vs 3 
           103       1       7  SIH009  3664.6 33322222222229995922 22222222222229995922   Btl 2 vs 3
           104       1      31  SIH020   157.5 43344222299554995399 42244444499554995399   Nuts 4 vs 2 Data indicates mistrip
           104       1      21  SIH030   812.6 23399999999559995999 73399999999559995999   Btl 7 vs 2, No water sample data.
                                                                                                       Original CTDS&O qf=3(i9s_sea.txt)
           108       1      36  SIH035    13.3 33322222299559225999 22222222299559225999   Btl 2 vs 3
           108       1      34  SIH029    56.9 33322222299559225999 22222222299559225999   Btl 2 vs 3
           109       1      24  SIH037   507.1 33322222222559995999 22222222222559995999   Btl 2 vs 3
           109       1      17  SIH024  1213.2 33342222299559995999 22242222299559995999   Btl 2 vs 3
           111       1      24  SIH037   510.5 33322222299559995999 22222222299559995999   Btl 2 vs 3
           111       1      11  SIH015  1915.5 33322222299559995999 22222222299559995999   Btl 2 vs 3
           113       1      36  SIH035    11.3 33322222299559995999 22222222299559995999   Btl 2 vs 3
           113       1      13  SIH018  1613.6 33322222299559995999 22222222299559995999   Btl 2 vs 3
           113       1      10  SIH013  1920.7 33322222299559995999 22222222299559995999   Btl 2 vs 3
           114       1      36  SIH035    10.9 33322222222559625999 22222222222559625999   Btl 2 vs 3
           114       1      24  SIH037   507.0 33322222299559995999 22222222299559995999   Btl 2 vs 3
           115       1      13  SIH018  1625.0 33322222299559995999 22222222299559995999   Btl 2 vs 3
           115       1       6  SIH038  2743.7 33322222299559995999 22222222299559995999   Btl 2 vs 3
           116       1      10  SIH013  2026.2 33322222299559225999 22222222299559225999   Btl 2 vs 3
           117       1      14  SIH002  1516.5 43344222299559995999 42244444499559995999   Nuts 4 vs 2, data indicate mistrip
           118       1       3  SIH025  3457.1 33322222299459225999 22222222299459225999   Btl 2 vs 3
           119       1      25  SIH019   457.1 33344222299559995999 32244444499559995999   Nuts 4 vs 2, data indicate problem
           120       1      28  SIH011   308.4 33322222299559225999 22222222299559225999   Btl 2 vs 3
           120       1      12  SIH016  1931.9 33322222222559225999 22222222222559225999   Btl 2 vs 3
           121       1      22  SIH031   710.9 33322222299559995999 22222222299559995999   Btl 2 vs 3
           122       1      35  SIH032    33.8 33332222222556225699 22232222222556225699   Btl 2 vs 3
           122       1      12  SIH016  2114.2 33322222222422225222 22222222222422225222   Btl 2 vs 3
           122       1      10  SIH013  2544.1 33322222299552225299 22222222299552225299   Btl 2 vs 3
           124       1      23  SIH033   608.7 33332222222559225999 22232222222559225999   Btl 2 vs 3
           124       1      10  SIH013  2640.9 33322222299559225999 22222222299559225999   Btl 2 vs 3
           125       1      22  SIH031   710.8 33322222222559995999 22222222222559995999   Btl 2 vs 3
           126       1      23  SIH033   608.6 33322222222559225999 22222222222559225999   Btl 2 vs 3
           126       1      21  SIH030   811.3 33322222222559225999 22222222222559225999   Btl 2 vs 3
           128       1      28  SIH011   309.9 33322222299553225499 22222222299553225499   Btl 2 vs 3
           128       1      10  SIH013  2733.5 33322222299559225999 22222222299559225999   Btl 2 vs 3
           130       1      29  SIH014   252.5 23344222299559225999 42244444499559445999   Nuts, Alk & TCARBN 4 vs 2, btl 4 vs 2, data indicate 
                                                                                                       btl closed early
           130       1      24  SIH037   503.4 33322222299559295999 22222222299559295999   Btl 2 vs 3
           132       1      32  SIH023   109.1 33322222299559225999 22222222299559225999   Btl 2 vs 3
           133       1      14  SIH002  1821.4 33322222299559995999 22222222299559995999   Btl 2 vs 3
           134       1      11  SIH015  2743.5 33322222299552225299 22222222299552225299   Btl 2 vs 3
           135       1      28  SIH011   309.8 33322222299559995999 22222222299559995999   Btl 2 vs 3
           138       1      24  SIH037   612.4 33322222299552225299 22222222299552225299   Btl 2 vs 3
           138       1      13  SIH018  2322.5 33322222299559225999 22222222299559225999   Btl 2 vs 3
           138       1       1  SIH036  5351.6 33344444499559635999 32244444499559335999   TCO2 3 vs 6, TCARBN ok but most other water smpls 
                                                                                                        indicate problem. TCARBN values same @ 
                                                                                                        1800db as this level.
           140       1      22  SIH031   711.9 33322222299559225999 22222222299559225999   Btl 2 vs 3
           141       1      28  SIH011   311.1 33322222299559995999 22222222299559995999   Btl 2 vs 3
           141       1      22  SIH031   712.9 33322222299559995999 22222222299559995999   Btl 2 vs 3
           143       1       2  SIH001   499.6 23399999999556995699 23399999999556995699   C14 & c13 qf = 6, no other water samples, left as is.
           144       1      22  SIH031   450.2 33322222299559295999 22222222299559295999   Btl 2 vs 3
           145       1       6  SIH038  1111.4 33344222299559935999 32244444499559935999   Nuts 4 vs 2, data indicate leak
          
         20080324/dm

4/22/08  Kappa         Cruise Report Expanded              Added 3 reports, expanded Data Processing Notes
         Updated & expanded these Data Processing notes.
         Added 3 reports to pdf and text versions of cruise report:
           1) Assessment of the quality of total inorganic carbon measurements (Appendix B)
           2) Assessment of the quality of the shipboard measurements of total alkalinity (Appendix C)
           3) Anthropogenic CO2 Inventory of the Indian Ocean (Appendix D)
          

I08S_1994 • I09S_1995
CCHDO Data Processing Notes