of high
3
H/
3 He ratio, emitted from hydrothermal
vents at the East Pacific Rise near 110°W; the
plumes drift westward near 2500 m. Stommel
(1982) and Hautala and Riser (1993) developed a
model in which this heat source, or the heat source
in combination with broader Stommel–Arons
(Stommel and Arons, 1960a,b) circulation, might
explain the westward flow. To test this idea further, TJ obtained a map of potential temperature
on the 36.96 2 surface (near 2700 m), using data
from several WOCE deep hydrographic sections
and earlier sources. Broad westward tongues of
warm water are seen near 5°N and 10°S – but a
narrower tongue of cold water suggests eastward
flow along 2°S, coinciding closely with the ‘3000-m
current’ seen in Fig. 4.3.3a. Similar features are
also found in the Atlantic near 2000 m. In that
ocean, Böning and Schott (1993) find a Kelvin
wave-like eastward jet along the equator, in a
numerical model. They suggest that eddy-diffusive
mechanisms, associated with resolved flow and
salinity components in the model, can generate the
observed salinity tongue slightly south of the main
flow (near 2°S in the Atlantic).
Johnson and Talley (1997) (referred to below as
JT) explored the currents between about 2000 and
3000 m in more detail. They took temperature (T)
and salinity (S) data from deep ConductivityTemperature-Depth (CTD) sections at 10°N and
15°S, plus three WOCE meridional sections at
increasing distance east of the East Pacific Rise.
They first defined a single ‘background’ –S relation for all sections, against which to define anomalies. In their zonal sections of salinity anomaly
(Fig. 4.3.4; see Plate 4.3.4, p. 300), plumes of salty
water ‘reminiscent of smoke from a chimney’ are
seen west of the East Pacific Rise, appearing to
emanate off the top of the Rise. The three meridional sections also show such plumes, with greatest
strengths at 2700 m near 10°S and 8°N. A cold,
fresh tongue lies between them at 2°S. The two
salty plumes are about 600 km across meridionally, 800 m vertically, and decay to half-strength in
2800 km westward. Consistent with the idea that
these plumes are formed by geothermal heat
release from the East Pacific Rise, JT also find that
buoyancy frequency N
2 shows similar plumes –
minimum buoyancy occurs within the salinity
plume, with an underlying plume of maximum
buoyancy near 3200 m. A vertical derivative of the
thermal wind relation yields fѨ
2
u/Ѩz
2
:9ѨN
2
/Ѩy,
so a rough estimate of velocity and transport is
obtained by vertical integration using the above
dimensions. The result is the prediction of two
flow cells overlying one another – a cyclonic cell
(eastward flow nearest the equator, i.e. between
about 4° and 8° from the equator) near 3200 m,
and an anticyclonic cell near 2700 m. The lower
one was predicted earlier, but JT provide the first
observational evidence for it. The order of magnitude and sign of the predicted flows on the equatorial side of these cells are comparable to the steady
flows seen near 3000 m and 2000 m, respectively,
near 2°S in Fig. 4.3.3a. JT describe three different
dynamical scenarios to account for their inferred
flows; the most realistic is also the most complex,
and the dynamical relationship between the salty
plumes and the eastward 3000 m jet in Fig. 4.3.3a
remains unclear. However, JT note that the (previously unremarked) westward flow seen near 2°S,
2000 m in Fig. 4.3.3a may be related to the flow at
3000 m in a vertical couplet sense, as in the flow
generated by the plumes.
Reversing jets
In the 17-month average on the equator
(Fig. 4.3.3a), four wavelengths of an apparent wave
can be seen between 1000 and 2200 m depth. Such
reversing jets have been observed in the other
oceans (e.g. Luyten and Swallow, 1976; Eriksen,
1981; Ponte and Luyten, 1989, 1990). However,
the lengths of these other records are inadequate
to test whether the jets have a substantial component with period of more than a year, as Fig.
4.3.3a shows to be the case in the Pacific. It seems
plausible that this phenomenon may be related to
enhanced vertical diffusion and upwelling along
the equator. Several authors (e.g. Bryan, 1987;
Weaver and Sarachik, 1990; Suginohara and Aoki,
1991; Wang, 1995) obtain reversing equatorial
jets in their models, associated with an enhanced
overturning cell on the equator. However, Wang
(1995) finds that the density stratification in these
models is much too low, and that the cells disappear if a more realistic stratification is introduced.
On the other hand, Semtner and Chervin (1992)
also obtained reversing jets in the equatorial
Pacific, in a global model where deep temperature
and salinity – and therefore stratification – was
restored to observed values. A distinctive feature of their run was the use of the Pacanowski–
Philander (Pacanowski and Philander, 1981)
SECTION 4 THE GLOBAL FLOW FIELD
224
3
H/
3 He ratio, emitted from hydrothermal
vents at the East Pacific Rise near 110°W; the
plumes drift westward near 2500 m. Stommel
(1982) and Hautala and Riser (1993) developed a
model in which this heat source, or the heat source
in combination with broader Stommel–Arons
(Stommel and Arons, 1960a,b) circulation, might
explain the westward flow. To test this idea further, TJ obtained a map of potential temperature
on the 36.96 2 surface (near 2700 m), using data
from several WOCE deep hydrographic sections
and earlier sources. Broad westward tongues of
warm water are seen near 5°N and 10°S – but a
narrower tongue of cold water suggests eastward
flow along 2°S, coinciding closely with the ‘3000-m
current’ seen in Fig. 4.3.3a. Similar features are
also found in the Atlantic near 2000 m. In that
ocean, Böning and Schott (1993) find a Kelvin
wave-like eastward jet along the equator, in a
numerical model. They suggest that eddy-diffusive
mechanisms, associated with resolved flow and
salinity components in the model, can generate the
observed salinity tongue slightly south of the main
flow (near 2°S in the Atlantic).
Johnson and Talley (1997) (referred to below as
JT) explored the currents between about 2000 and
3000 m in more detail. They took temperature (T)
and salinity (S) data from deep ConductivityTemperature-Depth (CTD) sections at 10°N and
15°S, plus three WOCE meridional sections at
increasing distance east of the East Pacific Rise.
They first defined a single ‘background’ –S relation for all sections, against which to define anomalies. In their zonal sections of salinity anomaly
(Fig. 4.3.4; see Plate 4.3.4, p. 300), plumes of salty
water ‘reminiscent of smoke from a chimney’ are
seen west of the East Pacific Rise, appearing to
emanate off the top of the Rise. The three meridional sections also show such plumes, with greatest
strengths at 2700 m near 10°S and 8°N. A cold,
fresh tongue lies between them at 2°S. The two
salty plumes are about 600 km across meridionally, 800 m vertically, and decay to half-strength in
2800 km westward. Consistent with the idea that
these plumes are formed by geothermal heat
release from the East Pacific Rise, JT also find that
buoyancy frequency N
2 shows similar plumes –
minimum buoyancy occurs within the salinity
plume, with an underlying plume of maximum
buoyancy near 3200 m. A vertical derivative of the
thermal wind relation yields fѨ
2
u/Ѩz
2
:9ѨN
2
/Ѩy,
so a rough estimate of velocity and transport is
obtained by vertical integration using the above
dimensions. The result is the prediction of two
flow cells overlying one another – a cyclonic cell
(eastward flow nearest the equator, i.e. between
about 4° and 8° from the equator) near 3200 m,
and an anticyclonic cell near 2700 m. The lower
one was predicted earlier, but JT provide the first
observational evidence for it. The order of magnitude and sign of the predicted flows on the equatorial side of these cells are comparable to the steady
flows seen near 3000 m and 2000 m, respectively,
near 2°S in Fig. 4.3.3a. JT describe three different
dynamical scenarios to account for their inferred
flows; the most realistic is also the most complex,
and the dynamical relationship between the salty
plumes and the eastward 3000 m jet in Fig. 4.3.3a
remains unclear. However, JT note that the (previously unremarked) westward flow seen near 2°S,
2000 m in Fig. 4.3.3a may be related to the flow at
3000 m in a vertical couplet sense, as in the flow
generated by the plumes.
Reversing jets
In the 17-month average on the equator
(Fig. 4.3.3a), four wavelengths of an apparent wave
can be seen between 1000 and 2200 m depth. Such
reversing jets have been observed in the other
oceans (e.g. Luyten and Swallow, 1976; Eriksen,
1981; Ponte and Luyten, 1989, 1990). However,
the lengths of these other records are inadequate
to test whether the jets have a substantial component with period of more than a year, as Fig.
4.3.3a shows to be the case in the Pacific. It seems
plausible that this phenomenon may be related to
enhanced vertical diffusion and upwelling along
the equator. Several authors (e.g. Bryan, 1987;
Weaver and Sarachik, 1990; Suginohara and Aoki,
1991; Wang, 1995) obtain reversing equatorial
jets in their models, associated with an enhanced
overturning cell on the equator. However, Wang
(1995) finds that the density stratification in these
models is much too low, and that the cells disappear if a more realistic stratification is introduced.
On the other hand, Semtner and Chervin (1992)
also obtained reversing jets in the equatorial
Pacific, in a global model where deep temperature
and salinity – and therefore stratification – was
restored to observed values. A distinctive feature of their run was the use of the Pacanowski–
Philander (Pacanowski and Philander, 1981)
SECTION 4 THE GLOBAL FLOW FIELD
224