WOCE established a group of 13 Data Assembly Centres (DACs), each of which would be
responsible for one or more of the WOCE data
types: hydrographic data, meteorology, current
meter data, and so on. Once the bulk of the early
planning for the WHP had been completed, the
main function of the WHP Office was to be the
DAC for CTD and bottle data.
For some measurements, the requirements for
accuracy/reproducibility demanded revision of seagoing methodology in order to meet the high standards. Experts in each of the WHP measurements
were invited to set down in detail the techniques
they used and which were known to work. The
expert knowledge thus became available to anyone
wishing to participate in the programme, and was
also translated into Spanish. The Spanish version
was printed and distributed by the US WOCE
Office at Texas A&M University. This material
(WHPO, 1994a), often called the WHP ‘cookbook’,
was produced in order to document methods
that had previously not been widely disseminated.
Unfortunately some techniques remained underdescribed, in that they did not allow an experienced
technician, who followed the technique but without any specific prior knowledge, to consistently
produce the desired result. However, the exercise
did help propagate certain types of technical
knowledge into the community. Also, the fact that
standards for accuracy had been published and
techniques and equipment recommended provided
PIs with important leverage to help secure investment in equipment and staff.
Temperature has been measured well enough
for many purposes related to inferring ocean circulation for nearly a century. Carefully used, deepsea reversing thermometers had the potential to be
accurate to 0.01°C. While platinum resistance
thermometers provided the advantage of continuous measurement, there are few regions where the
improvement in accuracy to order 0.002°C is crucial. However, the calibration stability of platinum
thermometers means that with a well-maintained
instrument the acquisition of high-quality fineresolution temperature data can be routine.
Salinity measurements of equivalent accuracy (i.e.
contributing equivalent uncertainty to density in the
equation of state) were not feasible in a routine way
until the development of a laboratory salinometer.
With the improvement in routine determinations of
sample conductivity, the equation of state itself
becomes a limiting factor, because of differences
in ionic composition in the Atlantic and Pacific.
While CTDs provide continuous measurement of
conductivity, it is still the case at the end of
WOCE that shipboard calibration of the data is
required. Conductivity cells drift, and the required
accuracy of 0.002 in salinity can only be recovered
by adjustment of the data to agree with salinometer analyses of bottle samples. The availability of
bottle samples during WOCE meant that there
was little requirement to develop improved conductivity cells for the WHP. But in the post-WOCE
era, increasing use of autonomous platforms,
where samples for analysis are not available, has
made this an urgent problem. Several manufacturers are bringing out new products with improved
stability. There is now a requirement for stability
of unattended salinity measurements at the level of
0.01 for up to 5 years.
At the onset of WOCE, dissolved oxygen was a
difficult parameter to measure with the required
accuracy. However, many laboratories took advantage of developments in equipment and techniques,
notably with the widespread introduction of automatic endpoint determination of titrations. With
suitable equipment an experienced analyst could
make repeat measurements with a precision of
0.01 mol kg
91 . Nutrient measurements did not
see such a significant improvement: the techniques
are essentially the same as those in use during
GEOSECS. The lack of standard reference materials for both nutrient and oxygen measurements
remains a significant limitation.
Data quality assurance
A system of Data Quality Experts (DQEs), also
overseen by the WHP Office, was conceived to
ensure that the final WOCE data set was of a uniformly high quality and consistency. The idea was
that all data submitted to the WHP Office would
be sent to a scientist familiar with the oceanography of the region, and experienced in making and
interpreting the observations in question. The
DQE would examine the data for internal consistency and reproducibility. In addition, the DQE
would be provided with relevant historical data
from the region, and might receive data from several adjacent or intersecting WOCE cruises at once
so that discrepancies could be identified and suggestions made for their resolution. Difficulties or
queries identified by the DQE would be referred
3.1 Shipboard Observations during WOCE
107
King, Firing and Joyce
responsible for one or more of the WOCE data
types: hydrographic data, meteorology, current
meter data, and so on. Once the bulk of the early
planning for the WHP had been completed, the
main function of the WHP Office was to be the
DAC for CTD and bottle data.
For some measurements, the requirements for
accuracy/reproducibility demanded revision of seagoing methodology in order to meet the high standards. Experts in each of the WHP measurements
were invited to set down in detail the techniques
they used and which were known to work. The
expert knowledge thus became available to anyone
wishing to participate in the programme, and was
also translated into Spanish. The Spanish version
was printed and distributed by the US WOCE
Office at Texas A&M University. This material
(WHPO, 1994a), often called the WHP ‘cookbook’,
was produced in order to document methods
that had previously not been widely disseminated.
Unfortunately some techniques remained underdescribed, in that they did not allow an experienced
technician, who followed the technique but without any specific prior knowledge, to consistently
produce the desired result. However, the exercise
did help propagate certain types of technical
knowledge into the community. Also, the fact that
standards for accuracy had been published and
techniques and equipment recommended provided
PIs with important leverage to help secure investment in equipment and staff.
Temperature has been measured well enough
for many purposes related to inferring ocean circulation for nearly a century. Carefully used, deepsea reversing thermometers had the potential to be
accurate to 0.01°C. While platinum resistance
thermometers provided the advantage of continuous measurement, there are few regions where the
improvement in accuracy to order 0.002°C is crucial. However, the calibration stability of platinum
thermometers means that with a well-maintained
instrument the acquisition of high-quality fineresolution temperature data can be routine.
Salinity measurements of equivalent accuracy (i.e.
contributing equivalent uncertainty to density in the
equation of state) were not feasible in a routine way
until the development of a laboratory salinometer.
With the improvement in routine determinations of
sample conductivity, the equation of state itself
becomes a limiting factor, because of differences
in ionic composition in the Atlantic and Pacific.
While CTDs provide continuous measurement of
conductivity, it is still the case at the end of
WOCE that shipboard calibration of the data is
required. Conductivity cells drift, and the required
accuracy of 0.002 in salinity can only be recovered
by adjustment of the data to agree with salinometer analyses of bottle samples. The availability of
bottle samples during WOCE meant that there
was little requirement to develop improved conductivity cells for the WHP. But in the post-WOCE
era, increasing use of autonomous platforms,
where samples for analysis are not available, has
made this an urgent problem. Several manufacturers are bringing out new products with improved
stability. There is now a requirement for stability
of unattended salinity measurements at the level of
0.01 for up to 5 years.
At the onset of WOCE, dissolved oxygen was a
difficult parameter to measure with the required
accuracy. However, many laboratories took advantage of developments in equipment and techniques,
notably with the widespread introduction of automatic endpoint determination of titrations. With
suitable equipment an experienced analyst could
make repeat measurements with a precision of
0.01 mol kg
91 . Nutrient measurements did not
see such a significant improvement: the techniques
are essentially the same as those in use during
GEOSECS. The lack of standard reference materials for both nutrient and oxygen measurements
remains a significant limitation.
Data quality assurance
A system of Data Quality Experts (DQEs), also
overseen by the WHP Office, was conceived to
ensure that the final WOCE data set was of a uniformly high quality and consistency. The idea was
that all data submitted to the WHP Office would
be sent to a scientist familiar with the oceanography of the region, and experienced in making and
interpreting the observations in question. The
DQE would examine the data for internal consistency and reproducibility. In addition, the DQE
would be provided with relevant historical data
from the region, and might receive data from several adjacent or intersecting WOCE cruises at once
so that discrepancies could be identified and suggestions made for their resolution. Difficulties or
queries identified by the DQE would be referred
3.1 Shipboard Observations during WOCE
107
King, Firing and Joyce