length-scales of tens to hundreds of metres. Direct
estimates of dissipation rates within the unstable
zones (Ri:0.33) are as large as 10
97 W kg
91
, but
the depth average is about 7.510
910 W kg
91
.
This dissipation rate, and the shear and strain
spectral densities, are 3–5 times larger than comparable values between about 3200 m and 1000 m
depth at 140°W in the Pacific (Gregg et al., 1995).
Muench and Kunze (1999) find that the replenishment of short wavelength internal waves can be
strongly altered in the presence of jet-like structures like those observed, which may result in an
energy source adequate to maintain the jets against
dissipation.
Modelling issues associated with deep overturning circulation in the Indian Ocean are discussed
by Warren (1994) and Lee and Marotzke (1998).
4.3.4.4 Summary: Implications for heat and
fresh water transport, water mass changes,
heat fluxes and SST
As stated in the introduction, we are concerned
here with using WOCE and other data to help us
understand the physical mechanisms relating heat
and fresh water transport, the associated water
mass changes, surface heat and fresh water fluxes,
and SST. We need to understand these relations on
all time scales (not just on long-term mean), if we
are to understand and simulate the Indian Ocean
SST changes that are so critical to climate variability. Section 4.3.4 has presented a qualitative
picture of two possible modes of heat transport
in the upper equatorial Indian Ocean. One mode
involves the directly wind-driven overturning cell.
A second involves an ‘eddy transport’ internal to
the Somali Current System, which appears to draw
deep water in from the East African Coast Current, warm it in the Somali Current, export it just
across the equator in the western boundary, and
thence feed annual mean zonal flow along the
equator. The latter exits in the South Equatorial
Countercurrent. We also briefly considered connections to deep overturning circulations.
The question of the relative sizes of the direct
wind-driven and annual ‘flushing’ mechanisms is
important for purposes of understanding interannual variability, because the first is now fairly
well understood physically, but the second is not.
Unfortunately it will be hard to tease the two mechanisms apart, since the main effect of the annual
flushing mechanism is to change the temperature
distribution of the geostrophic inflows and outflows to the basin. However, we do not know
what physically sets this distribution, either in the
presence or the absence of the eddy transport
mechanism.
Hopefully, traditional analyses of WOCE Indian
Ocean sections may settle the question of how
much heat is mixed down through the thermocline, to warm incoming Bottom Water and drive
the deep overturning cell. To do so, they will need
to allow for possible aliasing of ‘snapshot’ data by
short time scale effects, such as the large variations
in transport associated with the Intraseasonal
Oscillations (Fig. 4.3.11). It will then be necessary
to find the sources of the deep mixing – and of the
mixing of heat through the thermocline. Equatorial mixing such as that associated with the Deep
Reversing Jets is a strong candidate; but mixing
within the Somali Current System is another one.
Unfortunately, many issues associated with the
SCS are difficult to test observationally, because
they require work within the Exclusive Economic
Zones of several countries.
4.3.5 Overall conclusions
The comparison between the three equatorial
oceans seems to confirm that the qualitative
resemblance between current systems in the Pacific
and Atlantic covers a wide variety of phenomena.
Both have an SEC, EUC and NECC with an
annual cycle of similar phase, though the EUC
display some marked quantitative differences.
Furthermore, the Atlantic NEUC and SEUC seem
to be somewhat similar to the NSCC and SSCC
(Tsuchiya Jets) in the Pacific. If this is confirmed,
the terminology should be rationalized so that
similar phenomena have the same name in each
ocean. Observational work has tended to take a
different approach in the two oceans, resulting in
complementary insights. For example, the inventory of currents feeding the Pacific EUC could usefully be examined for the Atlantic. The apparent
analogies may extend to the Equatorial Intermediate Currents and to the Deep Reversing Jets. Conversely, the links of Atlantic current variability to
ring formation may have Pacific analogues.
One major difference between the Atlantic and
the other two oceans is, of course, that the entire
equatorial Atlantic circulation from the surface to
the North Atlantic Deep Water outflow is part of
4.3 The Tropical Ocean Circulation
245
Godfrey, Johnson, McPhaden, Reverdin and Wijffels
estimates of dissipation rates within the unstable
zones (Ri:0.33) are as large as 10
97 W kg
91
, but
the depth average is about 7.510
910 W kg
91
.
This dissipation rate, and the shear and strain
spectral densities, are 3–5 times larger than comparable values between about 3200 m and 1000 m
depth at 140°W in the Pacific (Gregg et al., 1995).
Muench and Kunze (1999) find that the replenishment of short wavelength internal waves can be
strongly altered in the presence of jet-like structures like those observed, which may result in an
energy source adequate to maintain the jets against
dissipation.
Modelling issues associated with deep overturning circulation in the Indian Ocean are discussed
by Warren (1994) and Lee and Marotzke (1998).
4.3.4.4 Summary: Implications for heat and
fresh water transport, water mass changes,
heat fluxes and SST
As stated in the introduction, we are concerned
here with using WOCE and other data to help us
understand the physical mechanisms relating heat
and fresh water transport, the associated water
mass changes, surface heat and fresh water fluxes,
and SST. We need to understand these relations on
all time scales (not just on long-term mean), if we
are to understand and simulate the Indian Ocean
SST changes that are so critical to climate variability. Section 4.3.4 has presented a qualitative
picture of two possible modes of heat transport
in the upper equatorial Indian Ocean. One mode
involves the directly wind-driven overturning cell.
A second involves an ‘eddy transport’ internal to
the Somali Current System, which appears to draw
deep water in from the East African Coast Current, warm it in the Somali Current, export it just
across the equator in the western boundary, and
thence feed annual mean zonal flow along the
equator. The latter exits in the South Equatorial
Countercurrent. We also briefly considered connections to deep overturning circulations.
The question of the relative sizes of the direct
wind-driven and annual ‘flushing’ mechanisms is
important for purposes of understanding interannual variability, because the first is now fairly
well understood physically, but the second is not.
Unfortunately it will be hard to tease the two mechanisms apart, since the main effect of the annual
flushing mechanism is to change the temperature
distribution of the geostrophic inflows and outflows to the basin. However, we do not know
what physically sets this distribution, either in the
presence or the absence of the eddy transport
mechanism.
Hopefully, traditional analyses of WOCE Indian
Ocean sections may settle the question of how
much heat is mixed down through the thermocline, to warm incoming Bottom Water and drive
the deep overturning cell. To do so, they will need
to allow for possible aliasing of ‘snapshot’ data by
short time scale effects, such as the large variations
in transport associated with the Intraseasonal
Oscillations (Fig. 4.3.11). It will then be necessary
to find the sources of the deep mixing – and of the
mixing of heat through the thermocline. Equatorial mixing such as that associated with the Deep
Reversing Jets is a strong candidate; but mixing
within the Somali Current System is another one.
Unfortunately, many issues associated with the
SCS are difficult to test observationally, because
they require work within the Exclusive Economic
Zones of several countries.
4.3.5 Overall conclusions
The comparison between the three equatorial
oceans seems to confirm that the qualitative
resemblance between current systems in the Pacific
and Atlantic covers a wide variety of phenomena.
Both have an SEC, EUC and NECC with an
annual cycle of similar phase, though the EUC
display some marked quantitative differences.
Furthermore, the Atlantic NEUC and SEUC seem
to be somewhat similar to the NSCC and SSCC
(Tsuchiya Jets) in the Pacific. If this is confirmed,
the terminology should be rationalized so that
similar phenomena have the same name in each
ocean. Observational work has tended to take a
different approach in the two oceans, resulting in
complementary insights. For example, the inventory of currents feeding the Pacific EUC could usefully be examined for the Atlantic. The apparent
analogies may extend to the Equatorial Intermediate Currents and to the Deep Reversing Jets. Conversely, the links of Atlantic current variability to
ring formation may have Pacific analogues.
One major difference between the Atlantic and
the other two oceans is, of course, that the entire
equatorial Atlantic circulation from the surface to
the North Atlantic Deep Water outflow is part of
4.3 The Tropical Ocean Circulation
245
Godfrey, Johnson, McPhaden, Reverdin and Wijffels