SECTION 4 THE GLOBAL FLOW FIELD
236
salty west Indian Ocean water to exchange with
fresh east Indian water.
Several climatologies of Indian Ocean currents
and transports are available (Reverdin, 1987;
Molinari et al., 1990a; Rao et al., 1989; Hastenrath and Greischar, 1991). The first two papers
examined the equatorial region in detail. Reverdin
(1987) used ship-drift observations in 1°S–1°N to
generate a mean seasonal cycle of surface zonal
flow along the equator (Fig. 4.3.10). The WJs are
clearly visible. Molinari et al. (1990a) found that a
climatology derived from satellite-tracked buoys
was very comparable to the earlier ones from ship
drifts, within the error limits of both. The WJs are
moderately well represented in typical OGCMs,
e.g. Fig. 4.3.10, from Visbeck and Schott (1992).
One major source of discrepancy between models
and observation seems to lie in uncertainties in
equatorial winds, e.g. Anderson and Carrington
(1993). Han et al. (1999) investigated the semiannual WJs in some detail, in a numerical model;
they found that wind forcing is the primary cause
of the WJs. However, Rossby waves, resonance,
and mixed-layer shears are all necessary to produce jets with realistic strength. Fresh water input
in summer tends to make the fall WJ stronger but
shallower than the spring WJ; however, dissipation around the Maldive Islands slows the jets and
tends to make them more equal.
Molinari et al. (1990a) compared annual mean
currents from ship drift and buoys; Figure 4.3.10
shows the buoy version. Strong eastward equatorial currents are seen in most longitudes, except
near the western boundary. This is one region
where ship drift and buoy climatologies do not
agree; marked convergence onto the equator is
apparent in Fig. 4.3.10 in the western boundary,
but it does not appear in the ship-drift climatology
(not shown). This suggests that ships may not
resolve an eddy-like feature centred on (0°, 50°E),
which shows up frequently in individual buoy
tracks. It is known as the ‘Southern Gyre’, and it
may play a role in transporting heat and other
properties eastwards and/or across the equator.
However, the strong surface convergence evident
in Fig. 4.3.10c at the equatorial western boundary
suggests that there may be annual mean sinking
here. This will be discussed further, in connection
with the western boundary current system.
Compared with other oceans, data on subsurface Indian Ocean flows is very sparse. Reverdin
(1987) provided a qualitative indication of the
EUC, by showing the equatorial ‘bulging’ of the
layer between the mixed layer base and the 20°C
isotherm depth, relative to 2.5°N, 2.5°S. He found
that near 55°E the EUC is confined to the period
January–June, and is strongest in March. It
reaches the eastern boundary, but is weak west of
80°E. Donguy and Meyers (1995) examined XBT
data from lines that crossed the equator; they
found that subsurface temperature–salinity relations were sufficiently well defined for use in estimating transports along their lines. They estimated
seasonal cycles of geostrophic transport relative to
400 db between 2.5°S and 0.5°S, and between
0.5°N and 2.5°N, along three lines that cross the
equator near 55°E, 65°E and 80°E. These show
major longitudinal differences in seasonal cycle. At
55°E, at the eastern end of the Southern Gyre,
the seasonal cycle shows even larger differences
between 2.5°S and 0.5°S and 0.5°N and 2.5°N.
Interannual variability of zonal currents is also
a strong feature of the equatorial Indian Ocean.
Reppin et al. (1999) examined data from six
moorings along 80°30ЈE, between July 1993 and
September 1994. Their Acoustic Doppler Current
Profiler (ADCP) observations at the equator did
not agree with earlier climatologies; the autumn
1993 WJ (35 Sv) was much stronger than the
spring 1994 one (5 Sv). Similarly, the EUC was
present in January to June 1994 as expected, but it
reappeared strongly in August 1994. Reppin et al.
related these results to strong interannual wind
variations in 1994. They first compared historic
ship-drift data with zonal winds and with the
Southern Oscillation Index (SOI:Tahiti–Darwin
pressure), showing that all three were quite
strongly correlated on 6-month time scales. When
they correlated their own 25-m currents with local
winds and the SOI on monthly time scales, they
found remarkably high correlations. The unusual
features in summer 1994 were due to unseasonal
winds at that time, resulting in upwelling and cool
SST in the equatorial eastern Indian Ocean. However, Saji et al. (1999) and Webster et al. (1999)
suggest an alternative explanation for the 1994
event, involving ENSO-like coupled air–sea interactions within the Indian Ocean, which they refer
to as the ‘Indian Ocean Dipole’.
As noted earlier, the annual mean flow along the
equator may contribute to zonal water exchange
from the salty Arabian Sea to the fresh Bay of
236
salty west Indian Ocean water to exchange with
fresh east Indian water.
Several climatologies of Indian Ocean currents
and transports are available (Reverdin, 1987;
Molinari et al., 1990a; Rao et al., 1989; Hastenrath and Greischar, 1991). The first two papers
examined the equatorial region in detail. Reverdin
(1987) used ship-drift observations in 1°S–1°N to
generate a mean seasonal cycle of surface zonal
flow along the equator (Fig. 4.3.10). The WJs are
clearly visible. Molinari et al. (1990a) found that a
climatology derived from satellite-tracked buoys
was very comparable to the earlier ones from ship
drifts, within the error limits of both. The WJs are
moderately well represented in typical OGCMs,
e.g. Fig. 4.3.10, from Visbeck and Schott (1992).
One major source of discrepancy between models
and observation seems to lie in uncertainties in
equatorial winds, e.g. Anderson and Carrington
(1993). Han et al. (1999) investigated the semiannual WJs in some detail, in a numerical model;
they found that wind forcing is the primary cause
of the WJs. However, Rossby waves, resonance,
and mixed-layer shears are all necessary to produce jets with realistic strength. Fresh water input
in summer tends to make the fall WJ stronger but
shallower than the spring WJ; however, dissipation around the Maldive Islands slows the jets and
tends to make them more equal.
Molinari et al. (1990a) compared annual mean
currents from ship drift and buoys; Figure 4.3.10
shows the buoy version. Strong eastward equatorial currents are seen in most longitudes, except
near the western boundary. This is one region
where ship drift and buoy climatologies do not
agree; marked convergence onto the equator is
apparent in Fig. 4.3.10 in the western boundary,
but it does not appear in the ship-drift climatology
(not shown). This suggests that ships may not
resolve an eddy-like feature centred on (0°, 50°E),
which shows up frequently in individual buoy
tracks. It is known as the ‘Southern Gyre’, and it
may play a role in transporting heat and other
properties eastwards and/or across the equator.
However, the strong surface convergence evident
in Fig. 4.3.10c at the equatorial western boundary
suggests that there may be annual mean sinking
here. This will be discussed further, in connection
with the western boundary current system.
Compared with other oceans, data on subsurface Indian Ocean flows is very sparse. Reverdin
(1987) provided a qualitative indication of the
EUC, by showing the equatorial ‘bulging’ of the
layer between the mixed layer base and the 20°C
isotherm depth, relative to 2.5°N, 2.5°S. He found
that near 55°E the EUC is confined to the period
January–June, and is strongest in March. It
reaches the eastern boundary, but is weak west of
80°E. Donguy and Meyers (1995) examined XBT
data from lines that crossed the equator; they
found that subsurface temperature–salinity relations were sufficiently well defined for use in estimating transports along their lines. They estimated
seasonal cycles of geostrophic transport relative to
400 db between 2.5°S and 0.5°S, and between
0.5°N and 2.5°N, along three lines that cross the
equator near 55°E, 65°E and 80°E. These show
major longitudinal differences in seasonal cycle. At
55°E, at the eastern end of the Southern Gyre,
the seasonal cycle shows even larger differences
between 2.5°S and 0.5°S and 0.5°N and 2.5°N.
Interannual variability of zonal currents is also
a strong feature of the equatorial Indian Ocean.
Reppin et al. (1999) examined data from six
moorings along 80°30ЈE, between July 1993 and
September 1994. Their Acoustic Doppler Current
Profiler (ADCP) observations at the equator did
not agree with earlier climatologies; the autumn
1993 WJ (35 Sv) was much stronger than the
spring 1994 one (5 Sv). Similarly, the EUC was
present in January to June 1994 as expected, but it
reappeared strongly in August 1994. Reppin et al.
related these results to strong interannual wind
variations in 1994. They first compared historic
ship-drift data with zonal winds and with the
Southern Oscillation Index (SOI:Tahiti–Darwin
pressure), showing that all three were quite
strongly correlated on 6-month time scales. When
they correlated their own 25-m currents with local
winds and the SOI on monthly time scales, they
found remarkably high correlations. The unusual
features in summer 1994 were due to unseasonal
winds at that time, resulting in upwelling and cool
SST in the equatorial eastern Indian Ocean. However, Saji et al. (1999) and Webster et al. (1999)
suggest an alternative explanation for the 1994
event, involving ENSO-like coupled air–sea interactions within the Indian Ocean, which they refer
to as the ‘Indian Ocean Dipole’.
As noted earlier, the annual mean flow along the
equator may contribute to zonal water exchange
from the salty Arabian Sea to the fresh Bay of