4.3 The Tropical Ocean Circulation
233
Godfrey, Johnson, McPhaden, Reverdin and Wijffels
and interannual variability. This was also clearly
illustrated in numerical model simulations of the
equatorial Atlantic seasonal cycle (Böning and
Herrmann, 1994). The difficult next step will
be to combine optimally the individual sections
and other observations in order to separate
properly the mean transports from their variability, at subseasonal, seasonal and interannual time
scales.
4.3.4 Near-equatorial circulation in the
Indian Ocean
4.3.4.1 Introduction
The tropical Indian Ocean has a wind system that
is grossly different from those in the Atlantic and
Pacific. Wind-driven upwelling, with its associate
water mass conversion and cool SSTs, occurs
mainly in the seasonally reversing western boundary currents rather than in the eastern equatorial
region. Thus a completely different set of mechanisms control heat and freshwater absorption. The
sparsity of Indian Ocean data also implies a need
to use information that has been superseded by
better products in the other two oceans. For these
reasons the discussion here takes a somewhat different form.
The question of greatest interest for WOCE may
be: how does the heat absorbed into the northern
Indian Ocean (north, say, of 7°S, the southernmost
latitude for which the Indian Ocean can be
regarded as a closed basin) get advected southward? We now have plausible qualitative ideas on
the three-dimensional pathways by which thermocline water enters across 7°S; moves (with much
vacillation due to seasonal and intraseasonal variability) to upwelling sites; and is removed across
7°S, near the surface. The seasonally reversing
winds play a very important role here (Fig. 4.3.9);
to first approximation, their zonal component
changes sign across the equator, in summer, winter
and on annual mean. Consequently there is meridional Ekman transport with the same sign on both
sides of the equator. It is southward on annual
mean, which is believed to be a major driver of the
heat absorption in the northern Indian Ocean. The
small annual mean westerly component of wind
along the equator (Fig. 4.3.9c) may maintain the
observed annual mean pressure gradient along the
equator, holding temperatures low in the western
Indian Ocean as observed.
In this section we review our understanding of
heat transport, water mass formation, surface heat
flux and SST, in the tropical Indian Ocean. We
first review the present understanding of the mean
seasonal cycle of flow along and across the equator, in and above the thermocline and away from
the western boundary; and also the open-ocean
surface heat budget. Next we discuss the heat
transport properties of the highly non-linear
western boundary currents, consisting of the East
African Coast Current and Somali Current (SC),
plus the various recycling flows that occur a few
hundred kilometres offshore; these are referred to
below as the ‘Somali Current System’. It is then
possible to describe two mechanisms for crossequatorial heat transport in the thermocline,
which involve links between the western boundary
and equatorial flows. Then we consider the relation of equatorial flows to higher-latitude flows,
outside the western boundary current. Finally,
we consider deeper flow patterns, including the
Deep Reversing Jets found on the equator, and the
deep western boundary currents; this raises questions regarding how heat penetrates through the
thermocline.
4.3.4.2 Flow above the thermocline
Zonal flows along the equator
Early work by Wyrtki (1973b) showed that strong
east-flowing semiannual ‘Wyrtki Jets’ (WJs) occur
along the equator between each monsoon. They
are driven by westerly winds at the monsoon
transitions in the central equatorial Indian Ocean.
They are associated with shoaling of the 20°C
isotherm in the western Indian Ocean, and deepening in the eastern. Another striking feature of
Indian Ocean zonal flows is the spasmodic occurrence of eastward undercurrents, beneath a westward surface current. Knox (1976) and McPhaden
(1982a,b) showed, from weekly temperature and
current data from Gan (0°41ЈS, 73°10ЈE), that
at this location an undercurrent usually occurred
towards the end of the winter monsoon in
February–March, after a period of easterly equatorial winds. Zonal flow during the summer monsoon
is relatively weak. Annual mean flow (Fig. 4.3.10c),
like the annual mean wind stress, was eastward at
all depths in 0–200 m at Gan, and was 0.3 m s
91 at
the surface. This result may be of particular significance for understanding Indian Ocean properties,
since it suggests a well-defined steady pathway for
233
Godfrey, Johnson, McPhaden, Reverdin and Wijffels
and interannual variability. This was also clearly
illustrated in numerical model simulations of the
equatorial Atlantic seasonal cycle (Böning and
Herrmann, 1994). The difficult next step will
be to combine optimally the individual sections
and other observations in order to separate
properly the mean transports from their variability, at subseasonal, seasonal and interannual time
scales.
4.3.4 Near-equatorial circulation in the
Indian Ocean
4.3.4.1 Introduction
The tropical Indian Ocean has a wind system that
is grossly different from those in the Atlantic and
Pacific. Wind-driven upwelling, with its associate
water mass conversion and cool SSTs, occurs
mainly in the seasonally reversing western boundary currents rather than in the eastern equatorial
region. Thus a completely different set of mechanisms control heat and freshwater absorption. The
sparsity of Indian Ocean data also implies a need
to use information that has been superseded by
better products in the other two oceans. For these
reasons the discussion here takes a somewhat different form.
The question of greatest interest for WOCE may
be: how does the heat absorbed into the northern
Indian Ocean (north, say, of 7°S, the southernmost
latitude for which the Indian Ocean can be
regarded as a closed basin) get advected southward? We now have plausible qualitative ideas on
the three-dimensional pathways by which thermocline water enters across 7°S; moves (with much
vacillation due to seasonal and intraseasonal variability) to upwelling sites; and is removed across
7°S, near the surface. The seasonally reversing
winds play a very important role here (Fig. 4.3.9);
to first approximation, their zonal component
changes sign across the equator, in summer, winter
and on annual mean. Consequently there is meridional Ekman transport with the same sign on both
sides of the equator. It is southward on annual
mean, which is believed to be a major driver of the
heat absorption in the northern Indian Ocean. The
small annual mean westerly component of wind
along the equator (Fig. 4.3.9c) may maintain the
observed annual mean pressure gradient along the
equator, holding temperatures low in the western
Indian Ocean as observed.
In this section we review our understanding of
heat transport, water mass formation, surface heat
flux and SST, in the tropical Indian Ocean. We
first review the present understanding of the mean
seasonal cycle of flow along and across the equator, in and above the thermocline and away from
the western boundary; and also the open-ocean
surface heat budget. Next we discuss the heat
transport properties of the highly non-linear
western boundary currents, consisting of the East
African Coast Current and Somali Current (SC),
plus the various recycling flows that occur a few
hundred kilometres offshore; these are referred to
below as the ‘Somali Current System’. It is then
possible to describe two mechanisms for crossequatorial heat transport in the thermocline,
which involve links between the western boundary
and equatorial flows. Then we consider the relation of equatorial flows to higher-latitude flows,
outside the western boundary current. Finally,
we consider deeper flow patterns, including the
Deep Reversing Jets found on the equator, and the
deep western boundary currents; this raises questions regarding how heat penetrates through the
thermocline.
4.3.4.2 Flow above the thermocline
Zonal flows along the equator
Early work by Wyrtki (1973b) showed that strong
east-flowing semiannual ‘Wyrtki Jets’ (WJs) occur
along the equator between each monsoon. They
are driven by westerly winds at the monsoon
transitions in the central equatorial Indian Ocean.
They are associated with shoaling of the 20°C
isotherm in the western Indian Ocean, and deepening in the eastern. Another striking feature of
Indian Ocean zonal flows is the spasmodic occurrence of eastward undercurrents, beneath a westward surface current. Knox (1976) and McPhaden
(1982a,b) showed, from weekly temperature and
current data from Gan (0°41ЈS, 73°10ЈE), that
at this location an undercurrent usually occurred
towards the end of the winter monsoon in
February–March, after a period of easterly equatorial winds. Zonal flow during the summer monsoon
is relatively weak. Annual mean flow (Fig. 4.3.10c),
like the annual mean wind stress, was eastward at
all depths in 0–200 m at Gan, and was 0.3 m s
91 at
the surface. This result may be of particular significance for understanding Indian Ocean properties,
since it suggests a well-defined steady pathway for