circulation. In addition there are the tasks associated with Goal 2 that have been discussed
in Section 7.4.1.
7.4.3.2 Remote sensing
Remote sensing was an important component of
the strategy, particularly for surface wind vectors
and surface topography. As it transpired, the
approach for accurate altimetric measurements for
surface topography proved to be extremely effective and clearly stands out as one of the key successes of WOCE. However, as we have witnessed
recently, altimetric measurements have utility over
and above that for determining global ocean circulation. The 1997–98 El Niño was among the
largest ever recorded and certainly was the best
observed. Altimetric measurements from TOPEX/
POSEIDON greatly assisted in the monitoring and
description of the event (McPhaden, 1999) and
have been shown useful for predictions (Ji et al.,
2000).
While the utility of altimeteric data for studying
seasonal-to-interannual variability was anticipated, it is only in recent times that this use has
been demonstrated for real-time monitoring and
prediction (e.g. Le Traon and Dibarboure, 1999).
The altimeter data have also been extremely effective for placing interannual variability in the context of global variability. Nerem et al. (1999)
show that the 1997–98 event had a first-order
effect on global mean sea level as estimated from
TOPEX/POSEIDON data. This result also suggests that interannual effects will almost certainly
affect the representativeness of the WOCE data.
The strategy for surface wind vector measurements relied heavily on the ability of Numerical
Weather Prediction Centers to provide accurate
surface fields and on the availability of highquality remotely sensed winds (WCRP, 1989b).
The failure of NSCAT and the less-than-expected
improvement in numerical weather prediction
model estimates limited the quality of WOCE
products (Taylor et al., 2001). The data from
TAO remain critical for good estimates of the
tropical Pacific wind fields (McPhaden et al.,
1998) and the lack of such data in other tropical
regions remains a key issue (e.g. in the Atlantic;
Servain et al., 1998). The key message has two
aspects. First, we retain a heavy reliance on highquality direct surface wind measurements for climate prediction and, second, for the required
spatial resolution, sampling density and accuracy,
we cannot rely on numerical weather prediction
model estimates alone and must continue the quest
for further remote measurements.
7.4.3.3 Direct measurements
WOCE recognized the need for high-quality direct
measurements to complement the data collected remotely and as a ‘ground reference’ for
fields derived from numerical weather prediction.
Some improvements were made to the Volunteer
Observing Ship (VOS) programme, though as yet
there have been few examples of effective action
(Taylor et al., 2000). The surface drifter programme was perhaps the most successful WOCE
‘increment’ to the surface marine network (Niiler,
Chapter 4.1), providing badly needed extra surface
pressure measurements in remote regions in addition to the direct samples of the surface circulation
that were required by WOCE.
In terms of climate prediction, the most obvious
synergies between WOCE and other observational
efforts occurred in XBT measurements from the
Ship-of-Opportunity Programme (SOOP) and sealevel measurements. A single programme was
developed to address jointly the needs of WOCE
and TOGA, combining broad-scale and frequently
repeated sampling, which are most apt for tropical
sampling, with high-density basin sections at low
frequency (TWXXPC, 1993). As with the VOS,
the most serious deficiency of this strategy was the
inability to deliver truly global coverage. This is
less of an issue with ENSO prediction since the
focus is regional, but the lack of temporal sampling becomes an issue. Perhaps the most important attribute was that SOOP did at least provide
some global context for interannual variability and
a baseline for extending interest beyond the tropical Pacific.
Smith et al. (1999b) reviews the application of
XBT data, particularly within the WOCE programme. The high-density XBT lines provided a
unique capability within WOCE to measure the
upper ocean transports of heat, mass and fresh
water on interannual time scales, overcoming the
sampling and representativeness issues for at least
some regions of the oceans. Roemmich and Gilson
(2000) have analysed data from the Pacific lines
P37/10/44 together with TOPEX/POSEIDON altimetric data to determine the scale-dependent correlation of subsurface temperature and dynamic
SECTION 7 INSIGHTS FOR THE FUTURE
590
in Section 7.4.1.
7.4.3.2 Remote sensing
Remote sensing was an important component of
the strategy, particularly for surface wind vectors
and surface topography. As it transpired, the
approach for accurate altimetric measurements for
surface topography proved to be extremely effective and clearly stands out as one of the key successes of WOCE. However, as we have witnessed
recently, altimetric measurements have utility over
and above that for determining global ocean circulation. The 1997–98 El Niño was among the
largest ever recorded and certainly was the best
observed. Altimetric measurements from TOPEX/
POSEIDON greatly assisted in the monitoring and
description of the event (McPhaden, 1999) and
have been shown useful for predictions (Ji et al.,
2000).
While the utility of altimeteric data for studying
seasonal-to-interannual variability was anticipated, it is only in recent times that this use has
been demonstrated for real-time monitoring and
prediction (e.g. Le Traon and Dibarboure, 1999).
The altimeter data have also been extremely effective for placing interannual variability in the context of global variability. Nerem et al. (1999)
show that the 1997–98 event had a first-order
effect on global mean sea level as estimated from
TOPEX/POSEIDON data. This result also suggests that interannual effects will almost certainly
affect the representativeness of the WOCE data.
The strategy for surface wind vector measurements relied heavily on the ability of Numerical
Weather Prediction Centers to provide accurate
surface fields and on the availability of highquality remotely sensed winds (WCRP, 1989b).
The failure of NSCAT and the less-than-expected
improvement in numerical weather prediction
model estimates limited the quality of WOCE
products (Taylor et al., 2001). The data from
TAO remain critical for good estimates of the
tropical Pacific wind fields (McPhaden et al.,
1998) and the lack of such data in other tropical
regions remains a key issue (e.g. in the Atlantic;
Servain et al., 1998). The key message has two
aspects. First, we retain a heavy reliance on highquality direct surface wind measurements for climate prediction and, second, for the required
spatial resolution, sampling density and accuracy,
we cannot rely on numerical weather prediction
model estimates alone and must continue the quest
for further remote measurements.
7.4.3.3 Direct measurements
WOCE recognized the need for high-quality direct
measurements to complement the data collected remotely and as a ‘ground reference’ for
fields derived from numerical weather prediction.
Some improvements were made to the Volunteer
Observing Ship (VOS) programme, though as yet
there have been few examples of effective action
(Taylor et al., 2000). The surface drifter programme was perhaps the most successful WOCE
‘increment’ to the surface marine network (Niiler,
Chapter 4.1), providing badly needed extra surface
pressure measurements in remote regions in addition to the direct samples of the surface circulation
that were required by WOCE.
In terms of climate prediction, the most obvious
synergies between WOCE and other observational
efforts occurred in XBT measurements from the
Ship-of-Opportunity Programme (SOOP) and sealevel measurements. A single programme was
developed to address jointly the needs of WOCE
and TOGA, combining broad-scale and frequently
repeated sampling, which are most apt for tropical
sampling, with high-density basin sections at low
frequency (TWXXPC, 1993). As with the VOS,
the most serious deficiency of this strategy was the
inability to deliver truly global coverage. This is
less of an issue with ENSO prediction since the
focus is regional, but the lack of temporal sampling becomes an issue. Perhaps the most important attribute was that SOOP did at least provide
some global context for interannual variability and
a baseline for extending interest beyond the tropical Pacific.
Smith et al. (1999b) reviews the application of
XBT data, particularly within the WOCE programme. The high-density XBT lines provided a
unique capability within WOCE to measure the
upper ocean transports of heat, mass and fresh
water on interannual time scales, overcoming the
sampling and representativeness issues for at least
some regions of the oceans. Roemmich and Gilson
(2000) have analysed data from the Pacific lines
P37/10/44 together with TOPEX/POSEIDON altimetric data to determine the scale-dependent correlation of subsurface temperature and dynamic
SECTION 7 INSIGHTS FOR THE FUTURE
590