quality, the US DOE CO 2 Survey Science Team
developed and distributed standardized instrumentation for TCO 2 analysis (Johnson and Wallace,
1992). Several training sessions were organized, as
well as an international workshop on coulometric
TCO 2 analysis.
CO 2 Survey Measurement groups were expected
to include at least one other independent measure
of the carbonate system on their cruise legs: no
restriction was imposed on the choice of this second parameter. Some groups measured total alkalinity, and others chose to make measurements of
pCO 2 or pH on discrete water samples.
Over the course of the CO 2 Survey significant
methodological advances were made concerning
the measurement of TCO 2 (e.g. Johnson et al.,
1993), total alkalinity (e.g. Millero et al., 1993;
Mintrop et al., 2000), continuous-flow (e.g.
Körtzinger et al., 2000) and discrete-sample
(Wanninkhof and Thoning, 1993; Neill et al.,
1997) pCO 2 measurement, and pH (e.g. McElligott et al., 1998). Most importantly, standardized
methods for sampling, analysis and data workup
were documented in a Handbook of Methods
(DOE, 1994). This publication greatly assisted the
rapid dissemination of accurate measurement techniques, worldwide, during the Survey.
Certified Reference Materials
A critical advance of the Global Survey, once
again implemented by the US DOE CO 2 Survey
Science Team, and particularly by A. Dickson of
Scripps Institution of Oceanography, was the
development and distribution of Certified Reference Materials (CRMs) for TCO 2 and, later, for
total alkalinity (Dickson and Anderson, 2001).
The use of reference materials had long been advocated (SCOR, 1985; Poisson et al., 1990). However, it required the prospect of long-term support
by the DOE programme, as well as community
acceptance of the reference materials, to bring
such long-standing recommendations to reality.
The CRMs are samples of surface seawater that
were initially collected in the Pacific Ocean and
then UV-sterilized, poisoned, equilibrated with the
atmospheric pCO 2 and then sealed within carefully-stoppered 500 ml glass bottles. A random
subset of the bottles from each batch are analysed
by Certified methods for TCO 2 (with a long-term
reproducibility of the order of 1 mol kg
91 by vacuum extraction/manometry in the laboraratory of
C. D. Keeling) and total alkalinity (by potentiometric titration in A. Dickson’s laboratory with a
long-term reproducibility of the order of 0.8 mol
kg
91
), as well as for salinity and nutrient content.
Tabulated values of Certified concentrations for
the 44 separate batches of CRMs produced since
January 1990 are listed at http://www.mpl.ucsd.
edu/people/adickson/CO2_QC/index.html. (See also
Dickson and Anderson, 2001.) The CRMs have
been extensively tested and proven to be stable
with respect to their Certified concentrations for
periods of at least 2 years after bottling.
The CRM bottles are then shipped to the various laboratories making inorganic carbon measurements in seawater. Within these laboratories,
regular analyses of CRMs are interspersed with
normal sample analysis. Between 1990 and 1998,
the Dickson laboratory produced and distributed
about 20 000 CRM bottles. The majority (65%)
were used within the USA, and the remainder were
distributed to 18 other countries. At least 50%
of these were used to support CO 2 Survey field
measurements (Dickson and Anderson, 2001). On
board WOCE cruises, CO 2 Survey investigators
analysed CRMs at least daily in order to check the
performance of their seagoing analytical systems
and to provide quality assurance of the results
obtained (see Section 6.3.3.4).
Shore-based analyses: a link to the past?
At the outset of the Survey the US DOE CO 2 Survey
Science Team established an additional data assessment mechanism, as a means to relate the WOCEera data (made with the benefit of CRMs) to
pre-WOCE ocean carbon data sets (collected when
CRMs were not available). The laboratory of
C. D. Keeling at Scripps Institution of Oceanography
was supported to make replicate analyses of TCO 2
on a limited number of samples collected from each
cruise on which a US-DOE supported CO 2 measurement team had participated. These samples
were collected, sealed, and shipped back to California (DOE, 1994). The samples were subsequently
analysed using a gas extraction/manometry technique identical to that used to assign the Certified
TCO 2 value for the CRMs. The concept behind the
shore-based analyses was two-fold:
1 Quality assessment. The shore-based programme
potentially allowed ‘closure of the circle’ with
6.3 Storage and Transport of Excess CO 2 in the Oceans
499
Wallace
developed and distributed standardized instrumentation for TCO 2 analysis (Johnson and Wallace,
1992). Several training sessions were organized, as
well as an international workshop on coulometric
TCO 2 analysis.
CO 2 Survey Measurement groups were expected
to include at least one other independent measure
of the carbonate system on their cruise legs: no
restriction was imposed on the choice of this second parameter. Some groups measured total alkalinity, and others chose to make measurements of
pCO 2 or pH on discrete water samples.
Over the course of the CO 2 Survey significant
methodological advances were made concerning
the measurement of TCO 2 (e.g. Johnson et al.,
1993), total alkalinity (e.g. Millero et al., 1993;
Mintrop et al., 2000), continuous-flow (e.g.
Körtzinger et al., 2000) and discrete-sample
(Wanninkhof and Thoning, 1993; Neill et al.,
1997) pCO 2 measurement, and pH (e.g. McElligott et al., 1998). Most importantly, standardized
methods for sampling, analysis and data workup
were documented in a Handbook of Methods
(DOE, 1994). This publication greatly assisted the
rapid dissemination of accurate measurement techniques, worldwide, during the Survey.
Certified Reference Materials
A critical advance of the Global Survey, once
again implemented by the US DOE CO 2 Survey
Science Team, and particularly by A. Dickson of
Scripps Institution of Oceanography, was the
development and distribution of Certified Reference Materials (CRMs) for TCO 2 and, later, for
total alkalinity (Dickson and Anderson, 2001).
The use of reference materials had long been advocated (SCOR, 1985; Poisson et al., 1990). However, it required the prospect of long-term support
by the DOE programme, as well as community
acceptance of the reference materials, to bring
such long-standing recommendations to reality.
The CRMs are samples of surface seawater that
were initially collected in the Pacific Ocean and
then UV-sterilized, poisoned, equilibrated with the
atmospheric pCO 2 and then sealed within carefully-stoppered 500 ml glass bottles. A random
subset of the bottles from each batch are analysed
by Certified methods for TCO 2 (with a long-term
reproducibility of the order of 1 mol kg
91 by vacuum extraction/manometry in the laboraratory of
C. D. Keeling) and total alkalinity (by potentiometric titration in A. Dickson’s laboratory with a
long-term reproducibility of the order of 0.8 mol
kg
91
), as well as for salinity and nutrient content.
Tabulated values of Certified concentrations for
the 44 separate batches of CRMs produced since
January 1990 are listed at http://www.mpl.ucsd.
edu/people/adickson/CO2_QC/index.html. (See also
Dickson and Anderson, 2001.) The CRMs have
been extensively tested and proven to be stable
with respect to their Certified concentrations for
periods of at least 2 years after bottling.
The CRM bottles are then shipped to the various laboratories making inorganic carbon measurements in seawater. Within these laboratories,
regular analyses of CRMs are interspersed with
normal sample analysis. Between 1990 and 1998,
the Dickson laboratory produced and distributed
about 20 000 CRM bottles. The majority (65%)
were used within the USA, and the remainder were
distributed to 18 other countries. At least 50%
of these were used to support CO 2 Survey field
measurements (Dickson and Anderson, 2001). On
board WOCE cruises, CO 2 Survey investigators
analysed CRMs at least daily in order to check the
performance of their seagoing analytical systems
and to provide quality assurance of the results
obtained (see Section 6.3.3.4).
Shore-based analyses: a link to the past?
At the outset of the Survey the US DOE CO 2 Survey
Science Team established an additional data assessment mechanism, as a means to relate the WOCEera data (made with the benefit of CRMs) to
pre-WOCE ocean carbon data sets (collected when
CRMs were not available). The laboratory of
C. D. Keeling at Scripps Institution of Oceanography
was supported to make replicate analyses of TCO 2
on a limited number of samples collected from each
cruise on which a US-DOE supported CO 2 measurement team had participated. These samples
were collected, sealed, and shipped back to California (DOE, 1994). The samples were subsequently
analysed using a gas extraction/manometry technique identical to that used to assign the Certified
TCO 2 value for the CRMs. The concept behind the
shore-based analyses was two-fold:
1 Quality assessment. The shore-based programme
potentially allowed ‘closure of the circle’ with
6.3 Storage and Transport of Excess CO 2 in the Oceans
499
Wallace