the Pacific west of 165°E and between 10°N and
14°S show a divergence of 6.4 Sv (Wijffels, 1993) –
comparable to the Indonesian Throughflow, whose
mean value in the thermocline appears to lie in the
5–10 Sv range (e.g. Meyers et al., 1995; Gordon
et al., 1999a). But this water cannot be upwelled
from the depth range of the EUC, which is in good
mass balance in the west Pacific; it appears to come
from further down. However, attempts to close the
western Pacific volume budget are difficult due to
the strong intraseasonal and interannual variability
in the flows there. Using long-term averaged XBT
lines, Morris et al. (1996) was able to balance
these flows over a much larger region encompassing the whole western half of the Pacific Ocean
and the Indonesian Seas.
JMcP find that geostrophic convergence towards
the equator, below the surface mixed layer,
increases to about 65 Sv when the limits are taken
at 3°S and on the trough between the SEC and the
NECC. The excess of 25 Sv over what is needed to
supply the EUC is presumably warmed in the westward-flowing SEC, subducted in the east or central
Pacific, and then upwelled again to flow out in the
surface mixed layer further west. (This phenomenon is referred to as the ‘Tropical Cell’, as opposed
to the ‘Subtropical Cell’ that supplies subtropical
water to the equator, e.g. Lu et al., 1998.) It should
be noted that the picture of a meridionally symmetric system near the equator is not quite correct;
mean meridional flow across the equator is not
zero, but is given quite well by the Sverdrup relation (e.g. Joyce, 1988).
Budget studies
One means by which our ideas on equatorial
dynamics can be refined is by use of budget studies,
and several of these have been undertaken in the
equatorial Pacific. Bryden and Brady (1985) used
repeat hydrographic sections along 150°W and
110°W, from 5°N to 5°S, to determine annual
mean geostrophic flows through the boundaries of
this rectangle, above 500 db. This region includes
the climatological mean position of the EUC maximum strength (about 140°W). After adding Ekman
transports, vertical velocities (assumed to be mainly
confined to 0.75°S–0.75°N) were determined by
mass conservation. They find that about half the
57 W m
92 absorbed in the region is used to heat the
zonal SEC as it traverses the region, and half heats
the meridional outflows. Zonal flow in the EUC is
basically along the zonally sloping isotherms, so
that upwelling through isotherms is typically only
1/3 of the net upwelling. Stress divergence can be
inferred from the mean momentum budget; it
implies that eddy mixing of momentum extends
well below the core of the Undercurrent, consistent with Toole et al.’s (1987) finding that low
Richardson numbers occurred below the EUC as
well as above.
Wijffels (1993) examined top-to-bottom sections along 10°N and 14°S in the Pacific, and
repeat meridional sections at 165°E. She used
these to form mass and heat budgets for the
region. Her resulting description of flow patterns
at each depth is basically compatible with that
given earlier. A major uncertainty is the size of the
Ekman fluxes across the zonal sections. Based on
an assumption of small vertical mixing through
barrier layers, Godfrey and Lindstrom (1989) concluded that the heat flux into the far western equatorial Pacific is only of order 0–10 W m
92
. When
Wijffels used this as an average over the whole
western part of her domain, it implied that the
Indonesian Throughflow must be smaller than 7 Sv
if heat is to be conserved. The reason is that subthermocline inflow from south of Australia must
warm enough to enter the thermocline, a temperature change of about 10°C. This in turn would
imply that Gordon’s (1986; also Gordon, Chapter
4.7) warm water pathway for the ‘Conveyor Belt’
circulation is weak.
The TOGA Coupled Ocean–Atmosphere
Response Experiment of 1992–93 (COARE; see
Godfrey et al., 1998, for a review) provided some
answers to the questions Wijffels raised. Fluxes
were measured over 4 months from a moored
buoy (Weller and Anderson, 1996), and detailed
near-surface budgets were undertaken using these
fluxes in order to test closure, and thus reduce
some of the uncertainties in surface heat and fresh
water flux measurements. Feng et al. (1998b)
demonstrated that the heat budgets closed to better
than 10 W m
92
, on each of three periods of a few
weeks. Both Feng et al. and Cronin and McPhaden
(1997) found that the heat budget was primarily
one-dimensional, with periods near wind bursts
when advection was important. The 4-month
average heat flux into the surface in COARE was
nearly 25 W m
92 (Weller and Anderson, 1996).
Climatologies show small seasonal variations of
heat flux at this location, so this may be close to an
4.3 The Tropical Ocean Circulation
219
Godfrey, Johnson, McPhaden, Reverdin and Wijffels
14°S show a divergence of 6.4 Sv (Wijffels, 1993) –
comparable to the Indonesian Throughflow, whose
mean value in the thermocline appears to lie in the
5–10 Sv range (e.g. Meyers et al., 1995; Gordon
et al., 1999a). But this water cannot be upwelled
from the depth range of the EUC, which is in good
mass balance in the west Pacific; it appears to come
from further down. However, attempts to close the
western Pacific volume budget are difficult due to
the strong intraseasonal and interannual variability
in the flows there. Using long-term averaged XBT
lines, Morris et al. (1996) was able to balance
these flows over a much larger region encompassing the whole western half of the Pacific Ocean
and the Indonesian Seas.
JMcP find that geostrophic convergence towards
the equator, below the surface mixed layer,
increases to about 65 Sv when the limits are taken
at 3°S and on the trough between the SEC and the
NECC. The excess of 25 Sv over what is needed to
supply the EUC is presumably warmed in the westward-flowing SEC, subducted in the east or central
Pacific, and then upwelled again to flow out in the
surface mixed layer further west. (This phenomenon is referred to as the ‘Tropical Cell’, as opposed
to the ‘Subtropical Cell’ that supplies subtropical
water to the equator, e.g. Lu et al., 1998.) It should
be noted that the picture of a meridionally symmetric system near the equator is not quite correct;
mean meridional flow across the equator is not
zero, but is given quite well by the Sverdrup relation (e.g. Joyce, 1988).
Budget studies
One means by which our ideas on equatorial
dynamics can be refined is by use of budget studies,
and several of these have been undertaken in the
equatorial Pacific. Bryden and Brady (1985) used
repeat hydrographic sections along 150°W and
110°W, from 5°N to 5°S, to determine annual
mean geostrophic flows through the boundaries of
this rectangle, above 500 db. This region includes
the climatological mean position of the EUC maximum strength (about 140°W). After adding Ekman
transports, vertical velocities (assumed to be mainly
confined to 0.75°S–0.75°N) were determined by
mass conservation. They find that about half the
57 W m
92 absorbed in the region is used to heat the
zonal SEC as it traverses the region, and half heats
the meridional outflows. Zonal flow in the EUC is
basically along the zonally sloping isotherms, so
that upwelling through isotherms is typically only
1/3 of the net upwelling. Stress divergence can be
inferred from the mean momentum budget; it
implies that eddy mixing of momentum extends
well below the core of the Undercurrent, consistent with Toole et al.’s (1987) finding that low
Richardson numbers occurred below the EUC as
well as above.
Wijffels (1993) examined top-to-bottom sections along 10°N and 14°S in the Pacific, and
repeat meridional sections at 165°E. She used
these to form mass and heat budgets for the
region. Her resulting description of flow patterns
at each depth is basically compatible with that
given earlier. A major uncertainty is the size of the
Ekman fluxes across the zonal sections. Based on
an assumption of small vertical mixing through
barrier layers, Godfrey and Lindstrom (1989) concluded that the heat flux into the far western equatorial Pacific is only of order 0–10 W m
92
. When
Wijffels used this as an average over the whole
western part of her domain, it implied that the
Indonesian Throughflow must be smaller than 7 Sv
if heat is to be conserved. The reason is that subthermocline inflow from south of Australia must
warm enough to enter the thermocline, a temperature change of about 10°C. This in turn would
imply that Gordon’s (1986; also Gordon, Chapter
4.7) warm water pathway for the ‘Conveyor Belt’
circulation is weak.
The TOGA Coupled Ocean–Atmosphere
Response Experiment of 1992–93 (COARE; see
Godfrey et al., 1998, for a review) provided some
answers to the questions Wijffels raised. Fluxes
were measured over 4 months from a moored
buoy (Weller and Anderson, 1996), and detailed
near-surface budgets were undertaken using these
fluxes in order to test closure, and thus reduce
some of the uncertainties in surface heat and fresh
water flux measurements. Feng et al. (1998b)
demonstrated that the heat budgets closed to better
than 10 W m
92
, on each of three periods of a few
weeks. Both Feng et al. and Cronin and McPhaden
(1997) found that the heat budget was primarily
one-dimensional, with periods near wind bursts
when advection was important. The 4-month
average heat flux into the surface in COARE was
nearly 25 W m
92 (Weller and Anderson, 1996).
Climatologies show small seasonal variations of
heat flux at this location, so this may be close to an
4.3 The Tropical Ocean Circulation
219
Godfrey, Johnson, McPhaden, Reverdin and Wijffels