(1998) (referred to below as TOGAObs). The
output of Ocean General Circulation Models
(OGCMs) driven with observed flux products,
with and without the assimilation of observed
ocean data, will also provide a valuable analysis
tool.
The complexity does not stop here. In the tropics, much water mass formation occurs due to
Kelvin–Helmholtz instability and mixing in layers
with low Richardson numbers. Mixing is thus a
highly non-linear function of the mean circulation,
so it may vary considerably with seasonal and
non-seasonal circulation changes. Tidal mixing,
mixing due to flow over sills and horizontal mixing are similarly non-linear processes, and research
on all of them is still at the early stages of development. All of them are probably confined to locations that are quite small compared with that of a
typical WOCE ‘box’. In the following sections, we
mainly describe our present understanding of flow
patterns, but we also discuss water mass formation
processes where appropriate. As will be seen, the
variety of processes is very rich in the equatorial
region; we will emphasize those we feel particularly need to be better parameterized, to create a
more effective ocean component of a coupled climate model. This implies that we will discuss some
near-surface processes that result in little transport
of heat, but are nevertheless of importance for
changing SST.
Sections 4.3.2, 4.3.3 and 4.3.4 each deal with
observations and some modelling results, in the
Pacific, Atlantic and Indian Oceans, respectively,
relating to the main flows near the equator and
their water mass sources. The presentation varies
for each ocean, since the outstanding issues and
available data vary among the three basins. We
use the fact that the dynamics of the equatorial
Atlantic and Pacific basins are similar to shorten
the discussion of the former. However, the Indian
Ocean is qualitatively different from the other
two, so rather more discussion is devoted to it
than to the Atlantic. Some conclusions are given in
Section 4.3.5. We deal quite cursorily with the
ENSO phenomenon and shorter-time-scale variability, referring readers wherever possible to the
June 1998 issue of the Journal of Geophysical
Research – a special issue devoted to reviews of
TOGA. Similarly, we will not devote much attention to the Indonesian Throughflow, since this was
the subject of another special issue of the Journal
of Geophysical Research in May 1996. Nor have
we tried to deal at all comprehensively with crossequatorial sections undertaken in WOCE, which
are described (for the Pacific) in the special issue
of Journal of Geophysical Research, June 1998.
A comparable special issue for the Atlantic is
Deep-Sea Research, January 1999; and Geophysical Review Letters devoted a special section to the
Indian Ocean WOCE Expedition in 1997.
4.3.2 Equatorial phenomena in the
Pacific Ocean
4.3.2.1 Flows within the Pacific equatorial
thermocline
A schematic diagram of equatorial flows within
the thermocline in the Pacific (adapted from
Philander, 1990) is seen in Fig. 4.3.1, while a midPacific annual mean cross-section, due to Wyrtki
and Kilonsky (1984), is shown in Fig. 4.3.2. Mean
easterly winds along the equator drive the South
Equatorial Current (SEC) westward at the surface,
piling warm water into the western Pacific. An
along-equatorial pressure gradient therefore develops over the top 250 m to roughly balance the
wind stress; this drives the Equatorial UnderCurrent
(EUC) eastward. The EUC shoals and upwells (as
can be seen in the equatorial tongue of east-flowing
water reaching towards the surface in Fig. 4.3.2),
supplying the bulk of the surface water that
diverges from the equatorial East Pacific in meridional Ekman transports. As indicated in Fig. 4.3.1,
the EUC is in turn supplied by the meridional
geostrophic inflow that compensates the Ekman
transports, including inflows at the western boundary. The geostrophic inflow is indicated by tongues
of high-salinity water in Fig. 4.3.2 flowing
equatorward from both hemispheres.
The SEC is a broad shallow current, extending
from the subtropical South Pacific to 2–5°N; it
mainly occurs in the top 200 m (Fig. 4.3.2). Its
breadth is set by patterns of wind stress curl. Wind
curls also generate the North Equatorial CounterCurrent (NECC), found north of 2°N in the west,
and 5°N in the east Pacific. The temperature field
seen in Fig. 4.3.2 provides the main geostrophic
balance for the SEC, NECC and EUC. These
flows, together with the surface Ekman flows and
the equatorial western boundary currents including the Indonesian Throughflow (IT), all vary
seasonally and through the irregular ENSO cycle.
SECTION 4 THE GLOBAL FLOW FIELD
216
output of Ocean General Circulation Models
(OGCMs) driven with observed flux products,
with and without the assimilation of observed
ocean data, will also provide a valuable analysis
tool.
The complexity does not stop here. In the tropics, much water mass formation occurs due to
Kelvin–Helmholtz instability and mixing in layers
with low Richardson numbers. Mixing is thus a
highly non-linear function of the mean circulation,
so it may vary considerably with seasonal and
non-seasonal circulation changes. Tidal mixing,
mixing due to flow over sills and horizontal mixing are similarly non-linear processes, and research
on all of them is still at the early stages of development. All of them are probably confined to locations that are quite small compared with that of a
typical WOCE ‘box’. In the following sections, we
mainly describe our present understanding of flow
patterns, but we also discuss water mass formation
processes where appropriate. As will be seen, the
variety of processes is very rich in the equatorial
region; we will emphasize those we feel particularly need to be better parameterized, to create a
more effective ocean component of a coupled climate model. This implies that we will discuss some
near-surface processes that result in little transport
of heat, but are nevertheless of importance for
changing SST.
Sections 4.3.2, 4.3.3 and 4.3.4 each deal with
observations and some modelling results, in the
Pacific, Atlantic and Indian Oceans, respectively,
relating to the main flows near the equator and
their water mass sources. The presentation varies
for each ocean, since the outstanding issues and
available data vary among the three basins. We
use the fact that the dynamics of the equatorial
Atlantic and Pacific basins are similar to shorten
the discussion of the former. However, the Indian
Ocean is qualitatively different from the other
two, so rather more discussion is devoted to it
than to the Atlantic. Some conclusions are given in
Section 4.3.5. We deal quite cursorily with the
ENSO phenomenon and shorter-time-scale variability, referring readers wherever possible to the
June 1998 issue of the Journal of Geophysical
Research – a special issue devoted to reviews of
TOGA. Similarly, we will not devote much attention to the Indonesian Throughflow, since this was
the subject of another special issue of the Journal
of Geophysical Research in May 1996. Nor have
we tried to deal at all comprehensively with crossequatorial sections undertaken in WOCE, which
are described (for the Pacific) in the special issue
of Journal of Geophysical Research, June 1998.
A comparable special issue for the Atlantic is
Deep-Sea Research, January 1999; and Geophysical Review Letters devoted a special section to the
Indian Ocean WOCE Expedition in 1997.
4.3.2 Equatorial phenomena in the
Pacific Ocean
4.3.2.1 Flows within the Pacific equatorial
thermocline
A schematic diagram of equatorial flows within
the thermocline in the Pacific (adapted from
Philander, 1990) is seen in Fig. 4.3.1, while a midPacific annual mean cross-section, due to Wyrtki
and Kilonsky (1984), is shown in Fig. 4.3.2. Mean
easterly winds along the equator drive the South
Equatorial Current (SEC) westward at the surface,
piling warm water into the western Pacific. An
along-equatorial pressure gradient therefore develops over the top 250 m to roughly balance the
wind stress; this drives the Equatorial UnderCurrent
(EUC) eastward. The EUC shoals and upwells (as
can be seen in the equatorial tongue of east-flowing
water reaching towards the surface in Fig. 4.3.2),
supplying the bulk of the surface water that
diverges from the equatorial East Pacific in meridional Ekman transports. As indicated in Fig. 4.3.1,
the EUC is in turn supplied by the meridional
geostrophic inflow that compensates the Ekman
transports, including inflows at the western boundary. The geostrophic inflow is indicated by tongues
of high-salinity water in Fig. 4.3.2 flowing
equatorward from both hemispheres.
The SEC is a broad shallow current, extending
from the subtropical South Pacific to 2–5°N; it
mainly occurs in the top 200 m (Fig. 4.3.2). Its
breadth is set by patterns of wind stress curl. Wind
curls also generate the North Equatorial CounterCurrent (NECC), found north of 2°N in the west,
and 5°N in the east Pacific. The temperature field
seen in Fig. 4.3.2 provides the main geostrophic
balance for the SEC, NECC and EUC. These
flows, together with the surface Ekman flows and
the equatorial western boundary currents including the Indonesian Throughflow (IT), all vary
seasonally and through the irregular ENSO cycle.
SECTION 4 THE GLOBAL FLOW FIELD
216