pCO 2 levels. Hence the contemporary local net
transfer of CO 2 across the air–sea interface can in
principle be smaller than that which took place in
1750, despite the fact that atmospheric pCO 2
levels are now much higher. It is quite possible
that certain regions of the North Atlantic may
indeed be behaving in this way. This behaviour is,
essentially, unique to CO 2 : it does not apply for
example to non-reactive gases such as the CFCs.
In reality within the North Atlantic, CO 2 does
not completely equilibrate between the atmosphere
and surface ocean during this cooling. Nevertheless, the buffer capacity change acts to decrease the
driving force for gas exchange (i.e. the air–sea
⌬pCO 2 ) and hence tends to decrease the net flux
from the atmosphere to the ocean.
6.3.5 Conclusions and outlook
The JGOFS/WOCE CO 2 Survey made use of the
sampling opportunities provided by the WOCE
Hydrographic Programme to collect an ocean carbon data set of unprecedented extent and quality.
As such the Survey represented an excellent example of cooperation at the interface between two
large global change research programmes. With
respect to implementation, an early focus on measurement techniques development and standardization, quality control, and a team-oriented
approach to project coordination paid off with a
significant improvement in data quality compared
with prior efforts.
The data resulting from this large field effort
are now being used in the early stages of synthesis.
Initial efforts are focusing on the distribution of
the Excess CO 2 inventory for comparison with
GCM-based predictions. A global picture of the
ocean’s Excess CO 2 inventory will likely be produced within the next 3 years. Present results are
revealing important areas of agreement and disagreement with models, the causes of which need
to be investigated in detail. The possibility of
detecting the temporal increase of Excess CO 2 in
the oceans has been demonstrated, implying that
future measurement campaigns can now be
designed to estimate, independently of models, the
regional and global Excess CO 2 uptake. Further
development, refinement and testing of methods to
reliably extract the Excess CO 2 signal from ocean
data are required. However, the critical point
is that the Survey has provided an accurately
determined baseline against which future inventory increases can be reliably gauged.
Efforts are also underway to make use of ocean
carbon data collected along WOCE heat-flux lines
to investigate meridional carbon transport. A
major driving force for this type of analysis is to
estimate the preindustrial ocean transport, in order
to clarify which natural factors may have led to a
preindustrial interhemispheric gradient of atmospheric CO 2 . To date most studies of ocean carbon
transport have focused on the Atlantic Ocean,
however such research is also underway in the
Pacific (P. Robbins, personal communication). In
the Atlantic it appears that there may be significant disagreements between GCM-based transports and observed transports that need to be
further investigated. Advection plays a potentially
very important role for the storage of Excess CO 2
in the North Atlantic and it is therefore critical
that carbon cycle models accurately parameterize
advection (and heat transport), together with mixing and deep-water formation, in order to assess
accurately the sensitivity of Excess CO 2 uptake to
future change.
Last but not least, one of the major goals of the
Survey was to provide a global description of the
distribution of carbon in the oceans in order to aid
the further development of models. To this end,
various new methods of interpolating the stillsparse carbon-related data set in space and time are
being developed. For example, Goyet et al. (1997,
2000) have applied interpolation and mapping
tools to existing global climatologies of correlated
hydrographic properties (e.g. the Levitus climatology) in order to make an initial global climatology
for CO 2 -related parameters. It is expected that
efforts along these lines will continue: a goal is to
provide global-scale data products from the CO 2
Survey that will be of considerable use to carboncycle investigations of the future.
Acknowledgements
The Global CO 2 Survey was unarguably the brainchild of Peter Brewer. His scientific inspiration and
imagination, as well as his hard work in the long,
convoluted corridors of the US Department of
Energy’s headquarters, are the main reasons that
the Survey took place. Support as well as imaginative programme management by John Downing
and Mike Riches made the Survey a reality. I
SECTION 6 LARGE-SCALE OCEAN TRANSPORTS
520
transfer of CO 2 across the air–sea interface can in
principle be smaller than that which took place in
1750, despite the fact that atmospheric pCO 2
levels are now much higher. It is quite possible
that certain regions of the North Atlantic may
indeed be behaving in this way. This behaviour is,
essentially, unique to CO 2 : it does not apply for
example to non-reactive gases such as the CFCs.
In reality within the North Atlantic, CO 2 does
not completely equilibrate between the atmosphere
and surface ocean during this cooling. Nevertheless, the buffer capacity change acts to decrease the
driving force for gas exchange (i.e. the air–sea
⌬pCO 2 ) and hence tends to decrease the net flux
from the atmosphere to the ocean.
6.3.5 Conclusions and outlook
The JGOFS/WOCE CO 2 Survey made use of the
sampling opportunities provided by the WOCE
Hydrographic Programme to collect an ocean carbon data set of unprecedented extent and quality.
As such the Survey represented an excellent example of cooperation at the interface between two
large global change research programmes. With
respect to implementation, an early focus on measurement techniques development and standardization, quality control, and a team-oriented
approach to project coordination paid off with a
significant improvement in data quality compared
with prior efforts.
The data resulting from this large field effort
are now being used in the early stages of synthesis.
Initial efforts are focusing on the distribution of
the Excess CO 2 inventory for comparison with
GCM-based predictions. A global picture of the
ocean’s Excess CO 2 inventory will likely be produced within the next 3 years. Present results are
revealing important areas of agreement and disagreement with models, the causes of which need
to be investigated in detail. The possibility of
detecting the temporal increase of Excess CO 2 in
the oceans has been demonstrated, implying that
future measurement campaigns can now be
designed to estimate, independently of models, the
regional and global Excess CO 2 uptake. Further
development, refinement and testing of methods to
reliably extract the Excess CO 2 signal from ocean
data are required. However, the critical point
is that the Survey has provided an accurately
determined baseline against which future inventory increases can be reliably gauged.
Efforts are also underway to make use of ocean
carbon data collected along WOCE heat-flux lines
to investigate meridional carbon transport. A
major driving force for this type of analysis is to
estimate the preindustrial ocean transport, in order
to clarify which natural factors may have led to a
preindustrial interhemispheric gradient of atmospheric CO 2 . To date most studies of ocean carbon
transport have focused on the Atlantic Ocean,
however such research is also underway in the
Pacific (P. Robbins, personal communication). In
the Atlantic it appears that there may be significant disagreements between GCM-based transports and observed transports that need to be
further investigated. Advection plays a potentially
very important role for the storage of Excess CO 2
in the North Atlantic and it is therefore critical
that carbon cycle models accurately parameterize
advection (and heat transport), together with mixing and deep-water formation, in order to assess
accurately the sensitivity of Excess CO 2 uptake to
future change.
Last but not least, one of the major goals of the
Survey was to provide a global description of the
distribution of carbon in the oceans in order to aid
the further development of models. To this end,
various new methods of interpolating the stillsparse carbon-related data set in space and time are
being developed. For example, Goyet et al. (1997,
2000) have applied interpolation and mapping
tools to existing global climatologies of correlated
hydrographic properties (e.g. the Levitus climatology) in order to make an initial global climatology
for CO 2 -related parameters. It is expected that
efforts along these lines will continue: a goal is to
provide global-scale data products from the CO 2
Survey that will be of considerable use to carboncycle investigations of the future.
Acknowledgements
The Global CO 2 Survey was unarguably the brainchild of Peter Brewer. His scientific inspiration and
imagination, as well as his hard work in the long,
convoluted corridors of the US Department of
Energy’s headquarters, are the main reasons that
the Survey took place. Support as well as imaginative programme management by John Downing
and Mike Riches made the Survey a reality. I
SECTION 6 LARGE-SCALE OCEAN TRANSPORTS
520