interannual time scales (also see Leuliette and
Wahr, 1999). The figure indicates that the ocean’s
heat storage increased during the 1997–98 El
Niño. Nerem and Mitchum (2000) also computed
the global mean total water vapour content using
the radiometer measurement from T/P. They
found that the water vapour content was highly
correlated with the mean sea level (correlation
coefficient of 0.9 with zero lag). This finding, coupled with the observation of decreased outgoing
long-wave radiation during El Niño (Wong et al.,
1998), provides supporting evidence for the extra
heat storage of the ocean during El Niño. Nerem
et al. (1999) also showed the simulation of meansea-level change during the 1997–98 El Niño by
an ocean general circulation model forced by wind
and sea surface temperature. Their results indicated that most of the extra heat was stored in the
upper 200 m of the ocean.
Using T/P and ERS-1 data, Minster et al. (1999)
studied the effects of the earth’s hydrological cycle
on the annual variation of the global mean sea
level. They removed steric effects estimated from
climatology and found that the remaining signal
had an amplitude close to 1 cm with a maximum
in mid-September. This signal is due to the annual
change of the mass of the ocean. This finding is
consistent with the annual variation of water mass
in the atmosphere and the land. On seasonal time
scales, the hydrological cycle actually has a larger
effect on global mean sea level than the steric
effect, which is responsible for an amplitude of
only 0.5 cm with a maximum in early March.
3.3.4 Currents and eddies
The sampling capability of a single altimeter is a
tradeoff between spatial and temporal resolutions.
Although the instrument makes high-resolution
(6–7 km) sampling along its ground tracks, the
spacing between adjacent ground tracks is hundreds of kilometres, depending on the orbit repeat
period. The two-dimensional structures of currents
and eddies are thus often not well mapped by satellite altimetry (Greenslade et al., 1997; Le Traon
and Dibarboure, 1999). However, the systematic
along-track sampling of the global mesoscale variability by satellite altimetry has made a profound
impact on the understanding of the dynamics of
currents and eddies. The first demonstration of the
potential of satellite altimetry was due to Cheney
et al. (1983). The qualitative features of the distribution of global mesoscale variability was revealed
by using only 24 days’ worth of SEASAT data.
The cross-track resolution of SEASAT was very
coarse (longitudinal cross-track spacing of 930 km
at the equator) due to its short 3-day repeat
period. With a longer repeat period (17 days)
GEOSAT produced much denser tracks (164 km
equatorial cross-track spacing) at the expense of
temporal sampling. More detailed maps of the
global mesoscale variability were produced (Zlotnicki et al., 1989; Shum et al., 1990). The 35-day
repeat orbit of ERS-1 and -2 has undersampled
the temporal variability but provided very good
spatial coverage (78 km equatorial cross-track
spacing). The sampling of the 10-day repeat orbit
of T/P is a compromise, leading to fairly good
coverage in both space and time at high latitudes.
The spatial coverage at low latitudes (315 km
equatorial cross-track spacing) is certainly too
coarse for mapping eddies. The combination of
observations from T/P and ERS, however, has
led to improved mapping of mesoscale eddies
(Le Traon et al., 1998; Le Traon and Ogor, 1998).
The study of currents and eddies is the most
productive application of satellite altimetry. First
of all, eddies and currents represent the most energetic component of ocean circulation. Their large
signals stood out first in the background of the
substantial measurement errors in the early altimetric measurements. Second, the scales of eddies
and currents are sufficiently smaller than those of
orbit errors, which have to be filtered out from
the poorly determined orbits of early altimetry
missions. Therefore, eddies and currents are not
significantly affected by the filtering. A large body
of literature has been produced on the subject. It is
beyond the scope of this chapter to present a comprehensive review (see Le Traon and Morrow,
2000). In this section I will focus on topics that are
illustrative of the utility of altimetry of interests
to WOCE.
3.3.4.1 Global sea surface height variability
The geographic distribution of the mesoscale variability derived from satellite altimetry provides a
first-order test of the validity of eddy-resolving
ocean general circulation models. Although these
models are able to simulate the basic features of
the distribution of mesoscale energy, there have
been significant discrepancies in both energy level
SECTION 3 NEW WAYS OF OBSERVING THE OCEAN
162
Wahr, 1999). The figure indicates that the ocean’s
heat storage increased during the 1997–98 El
Niño. Nerem and Mitchum (2000) also computed
the global mean total water vapour content using
the radiometer measurement from T/P. They
found that the water vapour content was highly
correlated with the mean sea level (correlation
coefficient of 0.9 with zero lag). This finding, coupled with the observation of decreased outgoing
long-wave radiation during El Niño (Wong et al.,
1998), provides supporting evidence for the extra
heat storage of the ocean during El Niño. Nerem
et al. (1999) also showed the simulation of meansea-level change during the 1997–98 El Niño by
an ocean general circulation model forced by wind
and sea surface temperature. Their results indicated that most of the extra heat was stored in the
upper 200 m of the ocean.
Using T/P and ERS-1 data, Minster et al. (1999)
studied the effects of the earth’s hydrological cycle
on the annual variation of the global mean sea
level. They removed steric effects estimated from
climatology and found that the remaining signal
had an amplitude close to 1 cm with a maximum
in mid-September. This signal is due to the annual
change of the mass of the ocean. This finding is
consistent with the annual variation of water mass
in the atmosphere and the land. On seasonal time
scales, the hydrological cycle actually has a larger
effect on global mean sea level than the steric
effect, which is responsible for an amplitude of
only 0.5 cm with a maximum in early March.
3.3.4 Currents and eddies
The sampling capability of a single altimeter is a
tradeoff between spatial and temporal resolutions.
Although the instrument makes high-resolution
(6–7 km) sampling along its ground tracks, the
spacing between adjacent ground tracks is hundreds of kilometres, depending on the orbit repeat
period. The two-dimensional structures of currents
and eddies are thus often not well mapped by satellite altimetry (Greenslade et al., 1997; Le Traon
and Dibarboure, 1999). However, the systematic
along-track sampling of the global mesoscale variability by satellite altimetry has made a profound
impact on the understanding of the dynamics of
currents and eddies. The first demonstration of the
potential of satellite altimetry was due to Cheney
et al. (1983). The qualitative features of the distribution of global mesoscale variability was revealed
by using only 24 days’ worth of SEASAT data.
The cross-track resolution of SEASAT was very
coarse (longitudinal cross-track spacing of 930 km
at the equator) due to its short 3-day repeat
period. With a longer repeat period (17 days)
GEOSAT produced much denser tracks (164 km
equatorial cross-track spacing) at the expense of
temporal sampling. More detailed maps of the
global mesoscale variability were produced (Zlotnicki et al., 1989; Shum et al., 1990). The 35-day
repeat orbit of ERS-1 and -2 has undersampled
the temporal variability but provided very good
spatial coverage (78 km equatorial cross-track
spacing). The sampling of the 10-day repeat orbit
of T/P is a compromise, leading to fairly good
coverage in both space and time at high latitudes.
The spatial coverage at low latitudes (315 km
equatorial cross-track spacing) is certainly too
coarse for mapping eddies. The combination of
observations from T/P and ERS, however, has
led to improved mapping of mesoscale eddies
(Le Traon et al., 1998; Le Traon and Ogor, 1998).
The study of currents and eddies is the most
productive application of satellite altimetry. First
of all, eddies and currents represent the most energetic component of ocean circulation. Their large
signals stood out first in the background of the
substantial measurement errors in the early altimetric measurements. Second, the scales of eddies
and currents are sufficiently smaller than those of
orbit errors, which have to be filtered out from
the poorly determined orbits of early altimetry
missions. Therefore, eddies and currents are not
significantly affected by the filtering. A large body
of literature has been produced on the subject. It is
beyond the scope of this chapter to present a comprehensive review (see Le Traon and Morrow,
2000). In this section I will focus on topics that are
illustrative of the utility of altimetry of interests
to WOCE.
3.3.4.1 Global sea surface height variability
The geographic distribution of the mesoscale variability derived from satellite altimetry provides a
first-order test of the validity of eddy-resolving
ocean general circulation models. Although these
models are able to simulate the basic features of
the distribution of mesoscale energy, there have
been significant discrepancies in both energy level
SECTION 3 NEW WAYS OF OBSERVING THE OCEAN
162