vii
Abstract
__________________________________
The central Pacific Ocean equatorial circulation system is
comprised of surface currents, mainly the South Equatorial Current,
the Equatorial Undercurrent, the North Equatorial Countercurrent,
and the North Equatorial Current and underlying water massesSurface Layer Water and Intermediate Water, specifically North Pacific
Intermediate Water and Antarctic Intermediate Water. Annual
variation of these waters is highly influenced by short-term climactic
changes in atmospheric circulation, especially the El Nino Southern
Oscillation cycle.
Using data from CTD deployments, two stages in the decline of
the 2002-2003 El Niño were geostrophically calculated. These
observations showed that by April of 2003 the effects of this El Niño
were minimal with only slight temperature and salinity anomalies
remaining. The second cruise, from late March through early May,
54
Appendix a
__________________________________
Hydrostatic equilibrium
In order for an equilibrium to be maintained it is necessary to have a
balancing of forces. It is often stated that the Coriolis force and the pressure
force must balance due to this, but why is this the case?
The vector form of the equation of motion is,
dV
= −α∇p − 2Ω × V + g + F ,
dt
where V is the total vector velocity, and the RHS reads from left to right, the
pressure term, the Coriolis term, gravity and all other forces (per unit mass).
The x-component (horizontal) of this is (Pond and Pickard, 1983),
du
∂p
= −α
+ 2Ω sin φ v − 2Ω cos φ w + Fx .
∂x
dt
There is no gravity term in the pure horizontal equation, but rather there are
two Coriolis terms, one involving the y-component of velocity (v) and another
using the z-component of velocity (w).
55
When frictional forces and eddy viscosities are then taken into
account, this equation becomes (Pond and Pickard, 1983),
∂u
∂u
∂u
∂u
∂p
∂ 2u
∂ 2u
∂ 2u
+u
+v +w
= −α
+ fv − 2Ω cos φ wz + Ax 2 + Ay 2 + Az 2 .
∂t
∂x
∂y
∂z
∂z
∂x
∂y
∂z
However, when the magnitudes of these terms are investigated, most of them
are negligible. The magnitude of the pressure term is unknown. All other
terms are at least 10-2 times smaller than the first Coriolis term (Pond and
Pickard, 1983). Therefore, to an order of accuracy of 1%
0 = −α
∂p
+ fv .
∂x
This holds true for all three directions. In the z-direction the pressure
term is balanced by gravity. This means that for any column of water, the
weight must be equal, even if the volume is not.
56
Bibliography
__________________________________
Bjerknes J. (1969) Atmospheric Teleconnections from the Equatorial
Pacific. Monthly Weather Review. 97(3): 163-172.
Bradshaw AL and KE Schleicher. (1980) Electrical Conductivity of
Seawater. IEEE Journal of Oceanographic Engineering. OE-5(1): 50-62.
Chelliah, M. (1990) The Global Climate for June-August, 1989: A
season of near normal conditions in the tropical Pacific. Journal of
Climate. 3:138-160.
D’Aleo Joseph S. with Pamela Grube. The Oryx Resource Guide to El
Nino and La Nina. Oryx Press: Westport, CT, 2002.
Dauphinee TM, J Ancsin, HP Klein and MJ Phillips. (1980) The Effect
of Concentration and Temperature on the Conductivity Ratio of
Potassium Chloride Solutions to Standard Seawater of Salinity 35‰
(Cl 19.3740‰). IEEE Journal of Oceanographic Research. OE-5(1): 1728.
Dauphinee TM, J Ancsin, HP Klein and MJ Phillips. (1980) The
Electrical Conductivity of Weight Diluted and Concentrated
Standard Seawater as a Function of Salinity and Temperature. IEEE
Journal of Oceanographic Engineering. OE-5(1): 28-50.
57
Donguy J and Meyers G. (1996) Mean annual variation of transport
of major currents in the tropical Pacific Ocean. Deep Sea Research I,
41(7): 1105-1122.
Eldin, G. (1983) Eastward flows of the south equatorial central pacific.
Journal of Physical Oceanography. 13(): 1461-1467.
Emery WJ and Pickard GL. Descriptive Physical Oceanography: An
Introduction. Pergamon Press: Oxford, 1990.
Fofonoff, NP. (1985) Physical Properties of Seawater: A New Salinity
Scale and Equation of State for Seawater. Journal of Geophysical
Research. 90(C2): 3332-3342.
Johnson GC, Sloyan BM, Kessler WS, and McTaggart CE. (2002) Direct
Measurements of upper ocean currents and water properties
across the tropical Pacific during the 1990’s. Progress in
Oceanography. 52(1): 31-61.
Katz Richard W. (2002) Sir Gilbert Walker and a Connection between
El Niño and Statistics. Statistical Science. 17(1): 97-112.
Kling J and Rosenberg K. (2002) Pacific Equatorial Current Structure
with respect to early onset of an El Nino Southern Oscillation
Event. SSV Robert C. Seamans, Sea Education Association: Woods
Hole, MA, Class 179.
Lewis, EL. (1980) The Practical Salinity Scale 1978 and Its Antecedents.
IEEE Journal of Oceanographic Engineering. OE-5(1): 3-8
Lu P, McCreary JP, and Klinger BA. (1998) Meridional Circulation
Cells and the Source Waters of the Pacific Equatorial
Undercurrent. Journal of Physical Oceanography 28(1): 62-84.
McPhaden MJ, AJ Busalacchi, R Cheney, J Donguy, K Gage, D
Halpern, M Ji, P Julian, G Meyers, G Mitchum, PP Niiler, J Picaut,
RW Reynolds, N Smith, K Takeuchi. (1998) The Tropical OceanGlobal Atmosphere observing system: A decade of progress.
Journal of Geophysical Research. 103(C7): 14,169-14,240.
58
Perkin RG and EL Lewis. (1980) The Practical Salinity Scale 1978:
Fitting the Data. IEEE Journal of Oceanographic Engineering. OE-5(1):
9-16.
Picaut J and R Tournier. (1991) Monitoring the 1979-1985 Equatorial
Pacific current transports with expendable bathythermography
data. Journal of Geophysical Research. 96(supplement): 3263-3277.
Pond, Stephen and George L Pickard. Introduction to Dynamical
Oceanography; 2nd Edition. Butterworth-Heinemann Ltd: Woburn,
MA, 1983.
Schmitz WJ. On the World Ocean Circulation (Volume II): The Pacific
and Indian Oceans/A Global Update. Woods Hole Oceanographic
Institution: Woods Hole, Massachusetts, 1996.
Taylor AC, Richardson P (Peter the Exalted), Hunter M, Crotty EA,
Gross SH and Scherr MS. (2003) Central Pacific Ocean Equatorial
Water Circulation. SSV Robert C. Seamans, Sea Education
Association: Woods Hole, MA, Class 186.
Tomczak M and JS Godfrey. (2003) Regional Oceanography: an
Introduction 2nd edn
Walker GW. (1928) World Weather. Monthly Weather Review. 56(5):
167-170.
Wyrtki K and Kilonsky B. (1984) Mean Water and Current Structure
during the Hawaii-to-Tahiti Shuttle Experiment. Journal of Physical
Oceanography. (14): 242-254.
53
Conclusion
__________________________________
This work has shown that the methods used to calculate
geostrophic currents do show the large scale shifts in current structure
as seen in the phases of the El Niño Southern Oscillation. In particular,
the weakening of the 2002-2003 El Niño is observed. However, the
error incurred by this method of approximation is extreme. In order to
even decrease the error by half, the increased number of
measurements necessary would be logistically prohibitive for a vessel
with multiple research objectives.
36
discussion
__________________________________
El niño and the currents
The same mechanisms that cause the weakening of westerly flowing
currents during the onset of El Niño also result in restoring those
currents during El Niño's decline. The primary mechanism is the
change in the strength of easterly trade winds. As these easterly trade
winds increase in strength, the surface waters should be blown into
areas of greater potential height difference. With the increase in the
difference in geostrophic height, stronger currents will theoretically
return. This can be seen in the change of maximum velocities of the
South and North Equatorial Currents between cruises 185 and 186
(Figure 5.1). The maximum velocity of the SEC changed from 0.363
ms-1 to 0.875 ms-1 in the six weeks between passes. The NEC's velocity
changed from 0.195 ms-1 to 0.315 ms-1. But how much of this is due to
where CTDs were taken rather than the current structure? For both
37
Core Current Velocities
1
Velocity (m/s)
0.9
0.8
0.7
0.6
Cruise 185
0.5
Cruise 186
0.4
0.3
0.2
0.1
0
SEC(SS)
SEC(S)
SEC(N)
SSCC
EUC
NEC
NECC
NSCC
Current
Figure 5.1. Comparison of maximum current velocities on cruise 185 (blue) and
cruise 186 (purple).
cruises, there were between three and four CTD drops in the region of
the NEC, but those for cruise 185 were more evenly distributed from
10°N to 17°N while those for cruise 186 were more densely distributed,
primarily from 9°N to 12°N. Due to this closer distribution on cruise
186, the area of greatest velocities may have been observed more
thoroughly than the rest of the current, leading to a greater average
velocity. The geopotentials are within the same range, and could
possibly include the slope seen in cruise 186, but this is unknown. For
the SEC, there are much better data for cruise 185 than for 186. Cruise
185 has seven data points fairly evenly spaced, with six falling between
14.5°S and 2.9°S, whereas there is a large gap in data from cruise 186
between 12.3°S and 5.2°S. Despite this lack of data, it appears highly
38
Geopotential Distance (Nm)
Geopotential Sea Surface
21
SEC
NECC NEC
19
17
15
13
-20
-10
0
Latitude
185 surface
10
20
186 surface
Figure 5.2. Geopotential sea surface of cruise 185 and cruise 186 measured as
a geopotential distance and graphed as a function of latitude. The currents
which flow at the surface are labeled along the top of the graph.
unlikely that the sea surface of 186, even with more points added,
would have a different slope more similar to that of cruise 185 (Figure
5.2).
When considering the EUC in particular, there are two ways to
conceptualize the effects of El Niño. One would be that the westerly
flowing currents are weaker than normal, creating less friction on the
EUC, so it can therefore flow faster back towards the west. Another
possibility is that the weakening of winds could cause less water to be
piled up in the western side of the basin. This could cause a
weakening of the EUC due to a decrease in dynamic height on the
western side of the basin and a resulting decrease in the pressure
gradient.
39
Realistically, it is more likely that El Niño is a combination of
these two effects. Although waters are not being pushed west with as
much force as usual, there is still a great deal of water there that is not
being kept piled up by strong easterly winds. So when the winds
slacken, there is plenty of piled up water to flow back east, and less
friction from the weakened westerly currents to slow it down.
Therefore, during El Niño there should be a significant strengthening
of the EUC, as well as the NECC—but why? If this current is driven
geostrophically and the westerly currents weaken, why would the
easterly currents increase in velocity and transport as well? When
looking at the geopotential sea surface (Figure 5.2), it is difficult to
even see that the NECC is stronger. In fact, it seems to have a slightly
shallower slope that the NECC during cruise 186. This is a reminder of
the complex nature of the coupled ocean-atmosphere system.
For cruise 185, the core velocities for the EUC and NECC were
0.705 and 0.447 ms-1, respectively. This top speed for the NECC is at a
depth of 67 m, which could explain some of the confusion about the
lack of difference in the sea surface slope. The geopotential anomaly
must have a greater slope at this depth, but not at the surface. This
depth corresponds to the start of both the thermocline and pycnocline,
40
signaling quick changes in both temperature and density. During
cruise 186, the core velocities found for the EUC and NECC were 0.178
and 0.315 ms-1 respectively.
As to whether these differences are due to the actual currents, or
whether they are anomalous to the models, there is resounding
evidence that currents cause the observed differences, especially in the
case of the NECC. For the EUC however, there are very few data
points, especially in cruise 185. If the sea surface heights in the near
equator region are compared between cruises 185 and 186, it appears
that the points that do exist are at the maximum and minimum
attributed to the EUC. So although the slope is somewhat less during
185, the bounds of the geopotential are reasonable. Although the slope
seems to be less at the surface, there are greater differences at depth,
creating the large difference in velocity and transport that is seen
between cruise 185 and 186.
Comparison with tao/triton
To link this research back to reality, the findings were compared
to the ‘normal’ of Wyrtki and Kilonski (1984). Since this is an El Niño,
41
albeit a weak one, it is expected that the westerly currents (the SEC
and NEC) will be weaker than normal for both cruises. The easterly
flowing currents are expected to be stronger during the El Niño. All
currents should be approaching their normal conditions as El Niño
wanes.
This expectation held true south of the equator in the South
Equatorial Current and the South Equatorial Counter Current. The
Equatorial Undercurrent (using Johnson’s method, 2002) seems to
oscillate around its expected value, and the currents of the northern
hemisphere are diverging from normal rather than the expected
opposite (Figure 5.3). However, as can be seen in the SOI for this time
Average Velocity (m/s)
Comparison of Current Velocities
0.5
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
185: Wyrtki and
Kilonski (1984)
186: Wyrtki and
Kilonski (1984)
Normal
186: Johnson
(2002)
SEC(SS)
SEC(S)
SEC(N)
SSCC
EUC
NEC
NECC
NSCC
185: Johnson
(2002)
Current
Figure 5.3. Comparison of average current velocities using both the Wyrtki and
Kilonski (1984) and Johnson (2002) methods of defining currents. Ideally, each
current should have a chevron of bars, with each current either decreasing or increasing
in velocity toward its normal, for both methods of defining the currents.
42
period (Figure 1.1), the 2002-2003 El Niño did not wane smoothly,
but rather weakened and restrengthened during the period of
measurement.
Comparison of Cruises 185 and 186 with Normal conditions
SEC(SS) SEC(S) SEC(N) SSCC EUC NEC NECC NSCC
AREA [km^2]
Normal
274
193
50
90
73 285
92
133
185: Wyrtki and Kilonski
228
172
24
32 292 247
58
81
(1984)
185: Johnson (2002)
126
37.5
91 173
59
186: Wyrtki and Kilonski
207
201
43
37 102 244
72
23
(1984)
186: Johnson (2002)
138
65
57 124
68
0.5
CORE VELOCITY
Cruise 185
0.127
0.363 0.285 0.043 0.705 0.195 0.447 0.045
(depth)
(7)
(17)
(7) (262) (182) (7) (67) (312)
Cruise 186
0.117
0.497 0.875 0.046 0.178 0.367 0.315 0.061
(depth)
(127)
(17)
(12) (232) (132) (42) (57) (272)
AVG VELOCITY [m/s]
Normal
0.049
0.137 0.302 0.048 0.31 0.082 0.214 0.067
185: Wyrtki and Kilonski
0.052
0.135 0.097 0.035 0.43 0.056 0.156 0.04
(1984)
185: Johnson (2002)
0.151 0.094
0.45 0.057 0.155
186: Wyrtki and Kilonski
0.076
0.15 0.396 0.035 0.083 0.073 0.113 0.039
(1984)
186: Johnson (2002)
0.189 0.286
0.077 0.1 0.13 0.01
TRANSPORT [Sv]
Normal
13.4
26.5
15.1
4.3 22.8 23.3 19.8
8.9
185: Wyrtki and Kilonski
13.4
22.5
2.1
0.1 125.7 17.8
8.7
3.2
(1984)
185: Johnson (2002)
21.7
2.5
40.6 11.8
8.6
186: Wyrtki and Kilonski
16
29.2
1.7
0.7 8.5 13.5
5.3
2
(1984)
186: Johnson (2002)
25.3
17.7
4.4 6.4
4.4 0.006
Table 5.1. Comparison of ‘Normal’ conditions as published by Wyrtki and Kilonski (1984) to
calculated values for area, velocity and volume transport for both Wyrtki and Kilonski (1984)
and Johnson (2002) definitions.
43
Questions of methodology
This area of physical oceanography is fraught with many
questions of methodology. Mixed units are a common practice, and
there is a conspicuous lack of clear definitions. There is no one
definition of the currents, one best reference depth, or one set of clear
terminology.
The two papers that are being used as bases for comparison,
Wyrtki and Kilonski (1984) and Johnson (2002), have different methods
of defining the boundaries of currents. What are the reasons for each,
and is one better than the other? Since the currents of the equatorial
Pacific are highly changeable in their velocities and depths depending
on the state of the ocean-atmosphere system, it makes some sense to
have a fluid boundary of the lower limits of these currents. This seems
to be the thought process supported by Johnson (2002). However, this
is facilitated by looking at a limited number of currents, only the
NECC, NEC and EUC. Johnson (2002) defined the NEC and NECC as
having a density less than 26.0 kgm-3 line (pink line; figure 5.4), and the
EUC with density less than 26.5 kgm-3. This makes sense because the
44
Cruise 185
NECC
SEC
NEC
EUC
NSCC
Latitude (degrees)
Figure 5.4. Cross section of cruise 185 current velocity contours every 0.1 ms-1
and at ±0.02 ms-1 with respect to latitude and depth. Shaded region shows area
that the model takes as eastward flow. The pink line shows the depth with
latitude of the 26.0 kgm-3 density line used by Johnson (2002) as a lower
boundary for surface currents. The yellow line is the 26.5 kgm-3 used by
Johnson (2002) as the lower boundary for the EUC.
currents being directly studied are surface currents that are supposed
to reside above the thermocline. The pycnocline tends to start at ~26.0
kgm-3, which corresponds to the upper limit of the thermocline. Of the
two cruises, there is only one CTD along the equator, during cruise
186.
45
Figure 5.5. Density with depth of 186 CTD_067 which shows the small
pycnocline at 26.5 kgm-3 possibly related to Johnson’s (2002) choice of
minimum density of EUC.
The density plot of this one CTD seems to reveal a second smaller
pycnocline at this density (Figure 5.5).
46
Cruise 186
NECC NEC
SEC
EUC
SSCC
NSCC
Latitude (degrees)
Figure 5.6. Cross section of cruise 186 current velocity contours every 0.1 ms-1
and at ±0.02 ms-1 with respect to latitude and depth. Shaded region shows area
that the model takes as eastward flow. The pink line shows the depth with
latitude of the 26.0 kgm-3 density line used by Johnson (2002) as the lower
boundary for surface currents. The yellow line is the 26.5 kgm-3 used by
Johnson (2002) as the lower boundary for the EUC.
It is possible that this is a coincidence, or an anomaly of that one
CTD drop. However, it is also possible that this miniature pycnocline
is tied into the EUC, since this line cuts directly below the EUC as it
appears in Figure 5.6.
47
Johnson’s (2002) method of defining the currents seems much
more appropriate that the artificial distinction of one depth that cuts
surface waters from intermediate waters. However, the distinction by
Wyrtki and Kilonski (1984) that is based on currents with velocities
greater than 0.02 ms-1 seems like a wise choice. There are large
portions of the equatorial cross-section with velocities that are very
near zero and have minimal transport. These areas are shown in white
on Figure 5.7. This area also depends on the chosen reference depth.
With a deeper reference depth there will be steeper gradients and
higher velocities.
The TAO/TRITON data have a reference depth of 500 m,
whereas Wyrtki and Kilonski (1984) and Johnson (2002) use a reference
depth of 1000 m. Does this reference depth make a difference? In
order for there to be the same difference in geopotential between
stations, there would have to be a constant volume between 500 and
1000 m across the entire latitudinal system. If the geopotential were
calculated using a reference depth of 1000 m, the geopotential
calculated using 500 m was subtracted off, and these two reference
48
(a)
SEC
NECC NEC
EUC
NSCC
SSCC
Latitude (degrees)
(b)
NECC
NEC
SEC
EUC
SSCC
NSCC
Latitude (degrees)
Figure 5.7. Velocity profile of cruise 185 (a) and cruise 186 (b) as a function of
latitude and depth. Contours are at ±0.02 ms-1 and at every 0.1 ms-1. The solid lines
represent flow in the eastward direction. The dashed lines are westward flow. Flow
is also shown in a color scale from 0.8 ms-1 to –0.7 ms-1 where negative velocities
represent westward flow.
49
depths were equivalent, then the result would be a constant. When
this was done, a constant was not the result.
In order to see the differences between these two reference
depths, the results of this calculation and the distribution between
them are analyzed. For cruise 185, the average difference was 4.845
m2s-2, with a standard deviation of 0.105; For cruise 186 the average
difference is 4.796 m2s-2 with a standard deviation of 0.180. However,
in cruise 186, there is an outlier just north of the equator (0.13°N) that
when removed gives an average difference of 4.838 m2s-2 and a
standard deviation of 0.086. With this point removed, the average
difference between the two cruises is almost the same, which gives
confidence to the comparison of these two systems.
Given that any data compared with TAO/TRITON were
reprocessed with a 500 m reference depth may make this seem like a
moot point. However, this allows the question to be asked, is it
necessary to look as far down the water column as 1000 m, or is 500 m
sufficient? These data suggest that because there is little variation
within the difference between the surface found using 500 m and 1000
m, the 500 m reference depth is sufficient. However, if the outlier is
taken into account, the 1000 m reference depth is better. Given the
50
choice, the 1000 m reference depth is also attractive because there is
consistent westward flow below the EUC, sometimes referred to as the
Equatorial Intermediate Current (EIC), that often has a boundary
below 400 m. In this case, the 500 m reference depth would not be
sufficient to model this currents. Even in the case that the EIC is not
studied, it causes outliers such as the one seen in this data. Therefore,
if possible, a lower basis depth should be used.
Errors
By looking at only the North-South gradient of the currents, a
flow directly East-West was assumed. This is fairly reasonable given
the structure of the current system, but allowing no meanders may be
unrealistic. Also, the fact that there are only small changes in salinity
and temperature lends great importance to the specificity of these
measurements. Although these measurements were precise, the errors
incurred were extremely large. Despite summation over the top
1000 m, the error of the specific volume anomaly of each station was
small compared to that of the geopotential anomaly. This is due to the
fact that the difference between specific volume anomaly of stations is
51
used to calculate the geopotential. In order to decrease the error by a
factor of two, the number of measurements would need to increase
four-fold. This would require a CTD deployment approximately every
5 hrs of the voyage. The lack of error analysis in papers in this area of
research may have to do with the logistical difficulties that would be
involved in attempting to lessen this error combined with
acknowledgement of the large error.
As previously mentioned, a significant source of error is the use
of two CTDs to create velocity measurements. This in and of itself is
not a problem, however, the velocities that this method finds are then
applied at their given depth to the entire latitudinal range between the
two CTDs which were used originally. This creates block-like swaths
of velocities that are not realistic (Figure 5.6). This is of particular
concern in the case of the equatorial currents of cruise 185. Because the
scientific deployments of this cruise were not designed with this
experiment in mind, they were not strategically placed to observe the
transitions between currents. A particularly telling example is the gap
in cruise 185 between the two CTDs on either side of the equator: 2°58'
S to 3°48' N. Also, scientific cruises are rarely planned with one
primary experiment in mind, but rather a collection of all possible
52
data. In the case of the SSV Robert C. Seamans, multiple
experiments were performed to look at the plant and animal
ecosystems surrounding islands, seamounts and atolls as well as the
hydrographic structure caused by these land masses. As a result of
this, cruise 185 had irregular CTDs that could not be used for the
geostrophic balance due to the influence of the surrounding land. For
cruise 186, the CTDs taken in these areas were not of sufficient depth
to be used for geostrophic calculations, and therefore, do not create
confusion.
30
Results
__________________________________
CTD data locate the boundaries of the surface and subsurface
currents according to both the definitions of Wyrtki and Kilonski
(1984) and Johnson (2002) within the vicinity of the mean boundaries.
Geopotential Height (Nm)
The South Equatorial Counter Current was not detected.
21
19
17
15
-20
-15
-10
-5
0
5
Latitude (degrees)
points affected by landmass
10
15
points used for measurements
Figure 4.1. Geopotential height at surface for all cruise 185 CTDs. The
points affected by landmass (blue) were not included in calculations of
flow (pink).
20
31
Using Wyrtki’s definitions, currents are generally larger, and
therefore have a greater volume transport (Table 4.1 and Table 4.2).
This is particularly noticeable in the EUC and NEC.
The sea surface calculated for cruise 185 using all CTDs of over
Cruise 185 (Southbound)
SEC(SS) SEC(S) SEC(N) SSCC
BOUNDARIES [m]
Wyrtki and Kilonski (1984)
(latitude)
Johnson (2002)
(latitude)
400
(-11.26)
315
(-9)
267
(-8)
228
172
0.127
±2.619
(7)
400
400
(4) (-4.24)
160
(4.4)
EUC
NEC
400
400
(all)
(10)
240
230
(all) (15.82)
NECC
NSCC
145
(6.29)
150
(5.68)
400
(12.24)
247
58
81
126
0.363
±7.589
(17)
37.5
91
173
0.285 0.043 0.705 0.195
±28.58 ±6.420 ±26.14 ±8.890
(7) (262) (182)
(7)
59
0.447
±12.96
(67)
0.045
±1.718
(312)
0.135
±1.173
0.151
±1.244
0.097 0.035
0.43 0.056
±1.508 ±0.999 ±1.478 ±0.370
0.094
0.45 0.057
±2.504
±3.581 ±0.484
0.156
±1.093
0.155
±0.478
AREA [km2]
Wyrtki and Kilonski (1984)
Johnson (2002)
CORE VELOCITY
(depth)
AVG VELOCITY [m/s]
Wyrtki and Kilonski (1984)
0.052
±0.217
Johnson (2002)
AVG SALINITY [psu]
Wyrtki and Kilonski (1984)
Johnson (2002)
AVG TEMP. [C]
Wyrtki and Kilonski (1984)
Johnson (2002)
MAX TRANSPORT [Sv]
TRANSPORT [Sv]
Wyrtki and Kilonski (1984)
Johnson (2002)
24
32
292
0.04
±0.269
35.81
35.59
34.92 34.83 34.84
34.5
34.45
34.53
±6.3E-4 ±6.9E-4 ±1.1E-3 ±1.8E-3 ±1.0E-3 ±6.6E-4 ±1.1E-3 ±1.7E-3
36.65
34.94
31.91 30.26
34.43
±7.2E-4 ±1.5E-3
±1.5E-3 ±9.1E-4 ±4.6E-4
21.42
23.87
23.27 11.46 14.88 17.22
23.57
10.18
±1.7E-4 ±2.1E-4 ±2.8E-4 ±4.3E-4 ±2.5E-4 ±1.7E-4 ±3.2E-4 ±4.0E-4
24.53
26.55
17.34 17.92
24.19
±2.2E-4 ±4.2E-4
±4.1E-4 ±2.5E-4 ±3.1E-4
0.172
1.058
1.058 0.065 2.616 0.179
0.301
0.084
±3.56
±11.5
±106 ±9.75
±97 ±5.06
±8.72
±3.22
13.4
±698
22.5
±1535
21.7
±1359
2.1
±125
2.5
±341
0.1
±197
125.7
17.8
±6774 ±1132
40.6
11.8
±2460 ±403
8.7
±486
8.6
±501
3.2
±142
Table 4.1. Database of boundaries, area, velocity, salinity, temperature and transport of
equatorial currents for Cruise 185 and associated errors.
32
1000 m produces a surface with noticeable anomalies (Figure 4.1).
However, upon closer inspection, if the data points that correspond to
locales downstream of atolls, islands and seamounts are removed, a
more reasonable picture forms.
Cruise 186 (Northbound)
SEC(SS) SEC(S) SEC(N) SSCC
BOUNDARIES
Wyrtki and Kilonski (1984)
(latitude)
Johnson (2002)
(latitude)
400
325
(-13.15) (-8.77)
245
(-8)
135
340
(3.69) (-4.13)
140
(3.69)
EUC
NEC
335
400
(3.69) (8.92)
240
190
(1.55) (13.43)
NECC NSCC
290
400
(4.99) (11.05)
240
165
(1.55) (1.55)
AREA [km2]
Wyrtki and Kilonski (1984)
Johnson (2002)
CORE VELOCITY
(depth)
AVG VELOCITY [ms-1]
Wyrtki and Kilonski (1984)
Johnson (2002)
AVG SALINITY
Wyrtki and Kilonski (1984)
Johnson (2002)
AVG TEMP. [C]
Wyrtki and Kilonski (1984)
Johnson (2002)
MAX TRANSPORT [Sv]
TRANSPORT [Sv]
Wyrtki and Kilonski (1984)
Johnson (2002)
207
201
43
37
102
244
72
23
.117
±1.45
(127)
138
0.497
±10.8
(17)
65
0.875
±21.0
(12)
0.046
±9.33
(232)
57
0.178
±19.4
(132)
124
0.367
±6.01
(42)
68
0.315
±9.64
(57)
0.5
0.061
±7.22
(272)
0.076
0.15
±0.208 ±1.025
0.189
±1.120
0.396 0.035 0.083 0.073 0.113 0.039
±2.531 ±2.404 ±2.706 ±0.395 ±1.052 ±1.064
0.286
0.077
0.1
0.13
0.01
±2.087
±2.664 ±0.572 ±1.310 ±8.178
35.67 35.53
34.9 34.86 34.96 34.57 34.58 34.66
±1.0E-3 ±1.1E-3 ±1.2E-3 ±3.8E-3 ±1.8E-3 ±7.8E-4 ±1.1E-3 ±1.7E-3
35.63
34.87
35.04 34.48 34.63 34.47
±1.1E-3 ±1.0E-3
±1.6E-3 ±1.1E-3 ±1.0E-3 ±1.1E-2
21.63 22.25
25.64 11.86 14.18
16.2 21.13 10.53
±2.0E-4 ±2.4E-4 ±3.4E-4 ±6.9E-4 ±2.9E-4 ±1.8E-4 ±2.7E-4 ±3.9E-4
24.38
25.28
15.27 22.01 21.48 13.37
±2.7E-4 ±3.0E-4
±3.6E-4 ±2.6E-4 ±2.7E-4 ±2.6E-3
0.462
0.63
1.38 0.055
0.28
0.34
0.28 0.032
±5.73 ±13.0
±32.3 ±11.2 ±30.7 ±5.57 ±7.92 ±3.73
16
±487
29.2
±1061
25.3
±980
1.7
±782
17.7
±2419
0.7
8.5
±335 ±1827
4.4
±1101
13.5
±1074
6.4
±425
5.3
±743
4.4
±782
2.0
±160
0.006
±4.23
Table 4.2. Database of boundaries, area, velocity, salinity, temperature and transport
of equatorial currents for Cruise 186 and associated errors.
33
The ‘boundaries’ presented are the lowest depth at which a
given current is found. Wyrtki and Kilonski (1984) only look at the
currents above 400 m. This is seen in all currents but the NECC and
SEC(S) having their maxima there. During cruise 186, the maxima is
not at the 400 m extremity, but is still below 300 m .
The boundaries for currents defined using Johnson’s (2002)
method are most often the point at which density is equal to
26.0 kgm-3. The only exception to this is the EUC, which has a
maximum depth where the density is equal to 26.5 kgm-3. The net area
of the current is in km2 and is used to calculate the average velocity,
salinity and temperature of the currents.
The data points taken on the ship were not on the boundaries
defined by Wyrtki and Kilonski (1984) or Johnson (2002). Therefore, in
order to find the area covered by currents with these definitive
boundaries, only a percentage of the total area covered by the CTD
was used. The core velocity is the maximum velocity of the current in
ms-1, while the maximum transport is the transport at this point as is
given in Sverdrups, which are x106 m3s-1.
34
Dynamic Height (N-cm)
160
120
80
40
-10
-5
0
5
10
Latitude
S185 TAO
S186 TAO
S185 geostrophic
S186 geostrophic
Figure 4.2. Comparison of ship based Dynamic Height measurement for cruises 185
and 186 and the data from the TAO/TRITON buoy array.
In order to ground the models in reality, data were compared to
TAO/TRITON, which uses a reference depth of 500 m. The resulting
surfaces (Figure 4.2) do not agree with the TAO/TRITON data exactly.
The greatest differences between the geostrophic and TAO/TRITON
data are at the ITCZ, which also corresponds with the area of largest
error. The average difference in dynamic height between the TAO
data point and the same latitude point of cruise 185 data is 0.038 m2s-2
and 0.042 m2s-2 for cruise 186. These differences are on the order of the
change in dynamic height between stations.
In previous studies, the errors on volume transport and velocity
calculated by geostrophic methods were not addressed. This is
35
peculiar given the extremely large size of the error. The error in the
velocity and volume transport measurements for individual CTDs
were typically under 100 times the size of the measurement, but in one
case was almost 700 times greater. The errors for all measurements are
given in Tables 4.1 and 4.2.
24
methods
__________________________________
Data for this project were collected aboard the SSV Robert C.
Seamans, cruises 185 and 186, from February 10, 2003 to May 2, 2003
(Figure 3.1). Cruise 185 collected data along the track from Honolulu,
Hawaii (20° N, 156°
W) to Papeete,
Tahiti (17° 30’ S,
150° W) to Nuku
Hiva, Marquesas (9°
S, 140° W), whereas
cruise 186 sailed
back from Tahiti to
Hawaii. The object
Figure 3.1. Cruise track 186 of the Robert C. Seamans, from
Papeete, Tahiti to Honolulu, Hawaii.
of both cruises was
25
to examine the Central Pacific Equatorial circulation system. Data
were collected and analyzed using the following instruments: Seabird
SBE 45 Thermosalinograph, and the Seabird SEACAT Profiler Model
SBE 19 Conductivity-Temperature-Depth (CTD) units. Data from the
TAO/TRITON archive were also used. The thermosalinograph
provided temperature and salinity readings taken at the surface every
minute. The CTD was calibrated at 4 m for every deployment and
measured the conductivity, temperature and depth of water flowing
through the device four times per second.
Salinity and temperature were measured at depth using the
CTD. Sampling using the CTD was conducted at stations along the
cruise tracks (Figure 3.2) and to varying depths. CTD sensors were
Cruise 185
Cruise 186
Latitude
10
5
0
-5
-10
Latitude
25
20
15
0
-5
-10
-15
-20
-25
-15
-20
-25
-170
-160
-150
Longitude
-140
-130
25
20
15
10
5
-170
-160
-150
-140
Longitude
-130
Figure 3.2. CTD deployments (pink) along cruises 185, from Honolulu, Hawaii south to
Papeete, Tahiti between February and March of 2003, and 186, from Papeete, Tahiti north
to Honolulu, Hawaii between March and May of 2003. Hawaii can be seen on this map at
approximately 20ºN, 155ºW. The black dots represent land masses.
26
calibrated at 4 m depth for 2 minutes. Water masses were identified
from CTD data. Dynamic height and geostrophic flow were analyzed
using data from CTDs. Theoretical surface currents’ velocities and
transport were also calculated using CTD data to calculate volumetric
anomalies and resulting flows. For a more detailed explanation of
why this works, please refer to Appendix A.
CTD conductivity, temperature (t) and depth (z) data were
processed by SeatermAF for salinity (s), density and specific volume
anomaly ( δ ), averaged every 5 m. The specific volume anomaly
represents the difference between the volume of 1 kg of the sampled
seawater and the volume (V) of a standard (salinity (s), temperature (t),
and pressure (p)) determined by depth (Pond and Pickard, 1983).
δ =108(V(s,t,p)-V(35,0,p)) [10-8 m3/kg]
The volume of sampled seawater is calculated using the International
Equation of State for Seawater, which is dependent on the density and
pressure of a body of water (Fofonoff, 1985). This volumetric anomaly
is then converted to ∆Φ , a ‘height’ anomaly. The ∆Φ is calculated by
multiplying the average δ of two depths by the change in pressure
between them.
27
Dynamic height is measured in dynamic meters [Nm] using
the change in geopotential (∆Φ) between two locations (Pond and
Pickard, 1983). The change in geopotential is related to dynamic
height (D) by a factor of ten (Pond and Pickard, 1983).
∆Φ = δ ∆p
D=
∆Φ
10
[m 2s-2 ]
[m 2s -2 ]
By setting the geopotential equal to zero at a reference depth, and then
summing the change over the depth, a dynamic sea surface height is
derived. Geopotential distance [m2s-2] can only be converted to
geometric meters if the reference depth is chosen where the velocity
equals zero (Pond and Pickard, 1983). In the region of the equator
there is no one depth that consistently has a zero velocity. Therefore a
1000 m reference depth was chosen to facilitate comparison with
Wyrtki and Kilonski (1984), but the measurements of geopotential
distance were left in units of dynamic meters. When comparisons
were made with TAO/TRITON data, a reference depth of 500 m was
used, and geopotential distance was converted to dynamic height.
Based upon the difference in geopotential between two stations,
the velocity and volume transport were calculated (Pond and Pickard,
1983).
v=
28
∆Φ A − ∆Φ B
Lf
V = w zv
[ms -1 ]
[m 3 s -1 ]
The measurement of velocity (v) takes into account the Coriolis
factor ( f ) that drives the direction of flow. The Coriolis factor is the
magnitude of the Coriolis force:
ˆ × kˆ
FC = 2Ω sin φ V
H
f = 2Ω sin φ
The Coriolis factor is directly related to the angular momentum ( Ω ) of
the Earth, which is 7.29x10-5 rad s-1, and the latitude ( φ ) of the flow. In
this model it is assumed that all of the flows are directly along a northsouth meridian, so the direction of the Coriolis force is the cross
product of this with the radial vector and is directly east-west. Volume
transport (V) is a function of depth (z), velocity (v), and the standard
width (w). For calculations that required two stations’ data, the data
were plotted using that average longitude, and the average latitude.
The maximum possible error for the initial conductivity and
temperature measurements were then taken into account. The error
was then propagated using the International Equation of State to find
the error involved in the specific volume anomaly, and thereby the
29
velocity and volume transport. For each equation the variables were
defined, and then the error was calculated as a sum of the absolute
value of the partial derivatives multiplied by the error in each variable.
For example, for a function F
∆F ( x, y, z ) =
∂F
∂F
∂F
∆x +
∆y +
∆z ,
∂x
∂y
∂z
where delta signifies the error of a function or variable.
8
Background
__________________________________
Introduction
Pacific equatorial circulation is controlled by a set of complex
relationships between wind driven geostrophic surface currents and
density driven water masses. The waters in the Pacific Ocean may be
divided into water masses according to their respective densities:
Surface Layer low density Water (SLW), Intermediate Water (IW), and
Deep/Bottom high density Water (DW/BW). Within the SLW are the
wind-driven surface currents. In the equatorial Pacific, the five main
surface currents are the South Equatorial Current (SEC), the Equatorial
Undercurrent (EUC), the North and South Equatorial Countercurrents
(NECC, SECC), and the North Equatorial Current (NEC). Below the
SLW are density-driven Intermediate Waters; at the Equator, there are
the North Pacific Intermediate Water (NPIW) and the Antarctic
Intermediate Water (AAIW). The surface currents and the salinity-
9
and temperature-driven currents beneath them are responsible for the
equatorial ocean heat budget.
El niño southern oscillation
Variations in wind patterns and even slight changes of
temperature or distribution in central Pacific Ocean waters may lead
to, or be an effect of, regional and global climate effects. The El Niño
Southern Oscillation (ENSO) cycle in the equatorial Pacific reflects
these kinds of fluctuations. ENSO is an inter-annual cycle of
atmospheric variation, affecting the trade winds above the Pacific
Ocean and, by extension, the speed, depth and distribution of
Equatorial currents (D’Aleo, 2002). The ENSO cycle is associated with
movement of a low pressure system across the Pacific Ocean.
Normally the low is centered over Darwin, Australia, and along the
western Pacific boundary there is more precipitation than in the
middle and eastern Pacific. One phase of this oscillation is referred to
as La Niña, in which precipitation on the western portion of the
Pacific, correlated with drought on the eastern portion, is extreme
(D’Aleo, 2002). Predominate trade winds increase in strength and
10
cause the westbound surface
currents to strengthen. As a
result, there is a build up of
water in the west, deepening
the thermocline there and
promoting regionally
increased downwelling. In
the east, there is increased
upwelling resulting in a
shallower thermocline and
Figure 2.1. Sea surface height anomaly during
the 1999 La Niña. The red and white areas
indicate above normal conditions, whereas the
blue and purple indicate below normal
conditions. (NASA-JPL)
decreased surface temperatures (Figure 2.1).
El Niño is the other extreme of this oscillation, most notably
indicated by a decrease in strength in easterly trade winds (D’Aleo,
2002). The low-pressure zone focused near Darwin, Australia moves
eastward to Tahiti, bringing wetter weather with it and leaving
Australia drier. This change in the pressure system over the Pacific
alters the surface currents. The westbound currents slacken, and
upwelling in the east decreases as does downwelling in the west. In
consequence, the thermocline levels out allowing decreased upwelling
and an increase in sea surface temperature (Figure 2.2).
11
El Niño and La Niña
are only the extremes of the
ENSO cycle. Each cycle lasts
approximately three to seven
years. The Southern
Oscillation Index (SOI)
compares surface air pressure
in Darwin, Australia to that in
Tahiti in French Polynesia to
Figure 2.2. Sea Surface height anomaly during
the 1997 El Niño. (NASA-JPL)
derive an index of the strength of a certain cycle in any given year.
The SOI is calculated by subtracting the month average of sea level
pressure in Darwin (D) from that in Papeete (T) and dividing out by
the average difference for all months between 1951-1980 (Chelliah,
1990). During an El Niño event, the pressure in Darwin is greater than
that in Papeete, yielding a negative SOI. The opposite holds true
during La Niña. Data are also available for comparison
SOI =
T(month, yr ) − D(month, yr )
S
from the TAO/TRITON arrays, buoys placed at specific locations
across the Equatorial Pacific that measure physical properties of the
12
water at depth. TAO data provide a means for comparing current
data with recent past conditions.
Steady state
The Pacific equatorial surface currents are set in motion by the
north- and south- easterly trade winds. In a world without continents,
the winds would converge at the equator. However, there is a larger
proportion of land mass in the northern hemisphere that results in
warmer, lower-pressure air. The southeasterly trade winds therefore
cross the Equator and converge with the northeasterly trade winds
slightly north of the Equator in a region known as the Intertropical
Convergence Zone (ITCZ), located at ~7° N (Emery and Pickard, 1990).
The trade winds cause a deflection of surface waters known as
'Ekman wind drift' (Apel JR, 1987). This wind drift is typically
between 10° and 45° to the right in the northern hemisphere and to the
left in the Southern Hemisphere. This veering effect continues
downward through the water column while decaying exponentially
(Figure 2.3). This effect dies off within 10-20 m of the surface, and
13
when integrated has a net transport 90° from the wind direction,
known as Ekman transport (Apel, 1987).
Due to the trade winds and Ekman transport there is a unique
pattern of dynamic sea surface height in the equatorial Pacific.
Between the Equator and 30°, there are predominantly easterly trade
Figure 2.3. Rotation and magnitude of near-surface velocities
through the Ekman layer of the ocean. Wind direction is indicated
by the topmost vane. (From Ekman, 1905)
winds in the Hadley cells, whereas the Ferrel cells (between 30°and
60°) have predominantly westerly winds. Where these two opposing
14
winds meet, at approximately 30° N and S latitudes, there are
subtropical ocean convergent zones that have high dynamic sea height
(Figure 2.4). Another area of convergence with high dynamic sea
height is centered at ~3° N, below the ITCZ (Emery and Pickard, 1990).
This area of convergence is due to the shift of the trade winds across
the Equator and the resulting movement of water due to Ekman
transport. In actuality, the area of high dynamic height at 3°N
combines with the ITCZ forming one dynamic high of approximately
CURRENTS
30° N
20° N
10° N
20° S
WINDS
Westerlies
convergence
NE
TradeWinds
North
Equatorial
Current
Equatorial
CounterCurrent
Equator
10° S
TRANSPORTS
South
Equatorial
Current
divergence
convergence
Equatorial
Divergence
Doldrums
SE
TradeWinds
convergence
30° S
Westerlies
Figure 2.4. Equatorial Pacific Water Circulation, surface diagram of currents,
transports and winds. (Emery and Pickard, 1990)
15
1.75 Nm at 4.5°N (Wyrtki, 1984). Between these areas of high
dynamic sea height, there are lows at approximately 10°N (between
the North Pacific Subtropical gyre and the ITCZ) and at the equator
(between the ITCZ and the South Pacific Subtropical Gyre). The North
and South Pacific Subtropical Gyres are major surface circulations
(Pickard and Emery, 1975).
Primary geostrophic currents
South Equatorial Current
Geostrophic currents (Figure 2.5) are associated with the pattern
of dynamic sea height. The SEC and the NEC (Figure 2.4) are both
westward flowing currents; they are the low-latitude limbs of the
South and North
Pacific subtropical
gyre. The SEC
covers a large
range of latitudes,
~10° S to 4° N. It
Figure 2.5. Areas occupied by main zonal currents between
Hawaii and Tahiti. Blue areas are westward flow, red are
eastward flow. (Based on Wytki and Kilonski, 1984)
16
has two branches as a result of the trade winds crossing the Equator:
in the southern hemisphere, SEC(S) is located along the northern edge
of the South Subtropical Gyre; and in the northern hemisphere, SEC(N)
flows between the Equator and the 3° N convergence (Pickard and
Emery, 1992).
Wyrtki (1984) defines the SEC in three parts- all with velocities
over 0.02 ms-1 and above 400 m in depth: an SEC(N), which is between
the Equator and 4° N; SEC(S), between the Equator and 9° S; and a
third branch which describes any flow south of 9° S, here referred to as
SEC(SS). Johnson (2002) defines the SEC as any westward flow with a
density less than 26.0 kgm-3 and splits it into two parts: SEC(N), which
is between the equator and the NECC, and the SEC(S) from the
Equator to 8° S. In a normal year, there is a weaker section of the SEC
that flows directly over the equator.
The SEC is located above the thermocline with a mean
temperature of ~27° C and a salinity range from 34.8 to 36.0 ppt (Picaut
and Tournier, 1991). The mean depth of the SEC is ~150 m, but it has
been detected as deep as 400 m; thus, it may affect the underlying
intermediate water masses. The mean transport of the SEC is
~55x106 m3s-1 with a normal maximum velocity of about 0.6 ms-1
17
(Johnson et al., 2002). A shallow thermocline intensifies its velocity,
so as the SEC moves westward where the thermocline deepens, its
velocity decreases. During El Niño conditions, decreased wind speeds
result in diminished geostrophic flow of the equatorial currents, most
noticeably in the SEC. Areas of high precipitation associated with the
eastward movement of low-pressure systems during El Niño bring
fresh water to the equator, decreasing salinity and density of SEC
water.
North Equatorial Current
The westward-flowing North Equatorial Current is the southern
limb of the North Subtropical Gyre, located between 8° and 20° N. The
NEC flows above the thermocline with a mean depth of 300 m and
increases in velocity towards the west. For analysis Wyrtki (1984)
defines the NEC as any westward flow with a velocity greater than
0.02 ms-1 above 400 m whereas Johnson (2002) defines it as any
westward flow above 26.0 kgm-3 north of the NECC. It has an average
temperature of 15°-25°C and a salinity profile having little variabilitya range of 34.8-35.0 ppt. The speed of the NEC increases as it flows
westward. Mean transport of this current is 22.1x106 m3s-1 (Picaut and
Tournier, 1991). The NEC acts similarly to the SEC during an El Niño
18
year; that is, it increases in temperature and decreases in velocity due
to slackening trade winds and leveling of the thermocline. Changes in
salinity due to El Niño are unknown.
Countercurrents
There are five countercurrents that flow eastward, against
predominate westward-flowing winds and surface currents. The
South and North Equatorial Countercurrents (SECC & NECC,
respectively) are eastward flowing surface currents that are generally
rather shallow (less than 200 m deep) and slow moving (maximum of
0.4 ms-1) compared to westward flowing currents. The SECC
fluctuates between 7° S and 14° S and typically has two branches, one
at 8°-10° S and the other from 11°-13° S, with a combined mean
transport of 0.6x106 m3s-1 (Wyrtki and Kilonski, 1984). However, these
data may be affected by mixing with the SEC; when sections are
analyzed individually, the mean transport is 3x106 m3s-1 (Eldin, 1983).
Wyrtki (1984) defines this current as eastward flow above 0.02 ms-1 in
surface waters south of the Equator. Johnson only looks at the SEC,
SECC and EUC specifically. However, in order to have a full second
set of currents, not defined by a lower depth, Johnson’s criterion of a
19
26.0
kgm-3
lower boundary line was extrapolated to the NEC and
NECC. Salinity of the SECC ranges from 35.6-36.2 ppt.
The North Equatorial Counter Current (NECC) is a geostrophic
current that flows east between the northern edge of the area of high
dynamic height at ~3º N and the ITCZ at 7º N, although it occasionally
lies as far south as 2.5º N. It borders the SEC(N) to the south,
occasionally merging with the EUC at depths. The mixing that occurs
along the boundaries of the NECC leads to some uncertainty over the
volume of flow, but it is generally greater than 12x106 m3s-1 (Wyrtki
and Kilonsky, 1984). The parameters used by Wyrtki and Kilonski
(1984) to define this current were any eastward flow above 0.02 ms-1 to
a depth of 400 m. Johnson (2002) defines this current as eastward flow
north of 2° N with densities below 26 kgm-3. The NECC is at its
weakest in May in the central region of the Pacific carrying ~5 x 106
m3s-1 (Donguy and Meyers, 1996).
Subsurface countercurrents, weaker than the Equatorial
Countercurrents, also flow eastward in the equatorial regions, but
generally flow below their surface counterparts. The South Subsurface
Countercurrent (SSCC) found at 4° S is identified by low oxygen and
high nutrient levels. The SSCC has a weaker second branch between
20
6°S and 7°S at a slightly greater depth. The transport of the two
branches combined is 4.3x106 m3s-1 (Wyrtki and Kilonski, 1984). There
is also a North Subsurface Countercurrent (NSCC), located between
200 m and 400 m in depth and 2°N and 6°N, with a maximum velocity
of 0.1 ms-1. This current, related to the SSCC, has a mean transport of
8x106 m3s-1 (Wyrtki and Kilonski, 1984).
Effects of El Niño
Previous research has not addressed El Niño's effect on the
countercurrents. There is a possibility that the velocity of these
countercurrents increases in an El Niño event due to the slackening of
the westward winds whose friction normally would impede their flow.
Countercurrents could also increase due to the necessity of water
flowing back to the east, attempting to flatten the thermocline.
However, if the countercurrents are propelled by the downwelling
caused by the pile-up of other currents, then the decrease in the
transport of westward flowing currents would result in less transport
to the east, causing these currents to weaken.
Equatorial Undercurrent
The most prominent of the eastward flowing currents is the
Equatorial Undercurrent (EUC) that transports the downwelled water
21
from the western side of the basin back east. The EUC is important
to the equatorial circulation and transport, and it is directly affected by
ENSO cycles. It is a unique current because there is no deflection as a
result of Ekman transport and the Coriolis force directs any northward
and southward meanders back toward the equator. The result is that
the EUC rarely strays beyond 2°N and 2°S. Johnson (2002) defines this
current as eastward flow having density between 23 and 26.5 kgm-3
from 2° S to 2° N; Wyrtki and Kilonski (1984) classify it as eastward
flow between 4° S and 4° N and above 400 m.
The EUC flows eastward under the equator within the
thermocline at a depth range of 50 to 300 m, with a mean depth of
100 m. The EUC is thin in comparison with other currents, ~81 km2 in
area, and diminishes as it flows eastward. It surfaces in the eastern
basin, has a mean core speed of 0.31 ms-1 (Emery,1990), and a
maximum velocity of 0.5 ms-1 (Pond and Pickard, 1984). The
temperature range of the EUC is 15°-25°C (Lu et al., 1998) and the
salinity ranges from 34.6-35.2 ppt. The EUC is warmest and most
saline at the dateline. Most mixing occurs above the thermocline
(hence above the EUC) as eddies and turbulence move surface waters
22
into lower waters. This mixing with the surface waters has minimal
effect on the temperature and salinity of the EUC.
Previous work using ADCP data suggest that the EUC
originated in waters from the New Guinea Coastal Undercurrent
(Emery,1990). During normal ENSO periods, there is high dynamic
sea surface height coupled with a deep thermocline along the western
boundary. The steep thermocline creates a zonal pressure gradient
that drives the eastward movement of the EUC (Johnson et al., 2002).
According to data collected by Wyrtki and Kilonsky (1984), the mean
volume of transport for the EUC is ~30.5x106 m3s-1 during a normal
year.
The EUC is subject to high variability during the ENSO cycle.
Johnson et al. (2002) found that the SEC and the EUC are both weaker
during an El Niño event. At 155° W the velocity of the SEC dropped
from 0.70 ms-1 to 30 ms-1 between La Niña (SOI=+1) and El Niño
(SOI=-1). Under similar conditions, the velocity of the EUC decreased
from 1.10 ms-1 to 0.70 ms-1. In some extreme El Niño periods, the EUC
has disappeared entirely. During an El Niño event, the thermocline
becomes shallower and the zonal pressure gradient that drives the
EUC shoals more parallel to the surface. Johnson also found that the
23
highest EUC speeds occur when easterly trade winds are weakest.
This seems contradictory since the easterlies are at their weakest
during an El Niño, but the EUC also weakens. This indicates that the
velocity of the EUC is influenced by both surface winds and friction,
yet also by the slope of the zonal pressure gradient created in the
thermocline.
conclusion
To observe the abnormal, the normal must be understood. This
holds true in general, and specifically in the case of the Pacific
Equatorial system. An analysis of salinity, temperature, and surface
current velocity provides insight into the state of the ocean with
regards to the ENSO cycle. During an El Niño the thermocline flattens,
wind strength decreases, and the SEC and EUC weaken as do the
countercurrents. During a La Niña the thermocline steepens, the
winds strengthen, and the surface currents increase in velocity and
volume transport. Therefore, the position of the ocean-atmosphere
system within the ENSO cycle can be observing through the current
state of the ocean.
1
introduction
__________________________________
Variation in global weather patterns occurs on widely varying
time-scales, from minutes to centuries. Phenomena such as
abnormally cold winters, warm summers, short term droughts or
periods of flooding can occur anywhere on the globe. Sometimes,
widely spaced events can be correlated (e.g. a warm, wet period in
Ecuador and drought conditions in northern Australia).
In the past few decades, scientists have realized that global
circulation patterns can produce complicated and varied affects. One
of these global circulation patterns is the El Niño Southern Oscillation
(ENSO).
“The El Niño Southern Oscillation (ENSO) is an
interannual coupled oscillation of the atmosphere and ocean
of the tropical Pacific. In the atmosphere, the east-west seesaw of surface pressure and the related patterns of clouds,
winds, temperatures, and precipitation. In the ocean, the
2
east-west flip-flop of the location and depth of warm and cool
pools of water.”
D’Aleo, 2002, p.1
The term, ENSO, comes from the combination of two older
terms, “El Niño” and “the Southern Oscillation”. The Southern
Oscillation is “an interannual see-saw of sea surface pressure across
the tropical Pacific” first described by Sir Gilbert Walker in the mid1920’s (D’Aleo, 2002). The Southern Oscillation is currently measured
by the Southern Oscillation Index (SOI), which chronicles the
difference in sea level pressure between Darwin, Australia, and Tahiti,
taking into account normal variation.
The second term, El Niño, refers to the warming of sea waters
off the coast of Ecuador and Peru near Christmas time. El Niño means
“little boy” in Spanish and refers to the Christ child. It was named by
Peruvian fisherman in the 1800s who noticed the warm waters off the
coast and the associated decline in fish populations.
Although warm waters may be associated with plentiful
tropical fish populations, this is often not the case. Off the coast of
South America, deep, cold waters are drawn up to the surface when
the warmer surface waters are blown away from the coast. With these
3
cold waters come nutrients that support a productive fishery. In turn,
these fish population are the backbone of the coastal economy.
history
The two major contributors to the formulation of the ENSO
cycle are Sir Gilbert Walker and Jacob Bjerknes (D’Aleo, 2002; Glantz,
2001). Walker was the first to describe the Southern Oscillation, and
Bjerknes was the first to combine this with what was known of El Niño
to formulate the El Niño Southern Oscillation.
Even before Walker, other meteorologists had noticed a
correlation between pressure in the North Atlantic and weather in
Europe (Walker, 1928). However, Hildebrandsson was the first to
realize its global consequences by looking at 68 weather stations using
purely graphical techniques (Walker, 1928).
Walker was appointed director-general of observatories in India
in 1904, directly after the 1899-1900 El Niño. This El Niño drastically
changed the normal climactic patterns of India, causing a major
drought during the regular monsoon season, and thereby became Sir
Walker’s primary research. The limited statistical work that had been
4
done with meteorology by a contemporary, however, was not
convincing to Walker. Beginning in 1886, official monsoon forecasts
had been published that were devoid of any rigorous statistical or
meteorological basis. These predictions did lead to the identification
of possible variables for drought conditions (e.g. Himalayan snow
cover and distant atmospheric pressure), which Walker believed could
be connected to form quasi-periodic behavior (Walker, 1925). Quasiperiodic behavior can be described as a function of discrete frequencies
that are related by an irrational ratio (Fisher, 1995).
The British statistician, George Udny Yule, had meanwhile
devised a description of second order quasi-periodic behavior (Katz,
2002) that Walker extended for the more complex system he was
studying. Walker looked at the correlations between Port Darwin,
Australia, Zanzibar and Samoa. These differences in pressure were the
beginnings of the Southern Oscillation Index, which is now defined as
the difference in pressure between Port Darwin, Australia and Tahiti.
Walker began to publish in 1910 on the “Correlation of Seasonal
Variations in Weather” and wrote in 1923 that,
“there is a swaying of pressure on a big scale backwards
and forwards between the Pacific Ocean and the Indian
Ocean, there are swayings, on a much smaller scale, between
5
the Azores and Iceland, and between the areas of high and
low pressure in the N. Pacific.”
Katz, 2002, p.101
In these simple five lines, Sir Walker describes not only the Southern
Oscillation but also the North Atlantic Oscillation and the North
Pacific Oscillation (two other important global circulation patterns).
Unfortunately, it was not until Bjerknes made the connection
between El Niño and the Southern Oscillation that Walker’s work on
the Southern Oscillation was widely regarded as more than a “climate
curiosity” (Rasmusson, 1991). In March of 1969, Bjerknes wrote,
“The maxima of the sea temperature in the eastern and
central equatorial Pacific occur as a result of anomalous
weakening of the trade winds of the Southern Hemisphere
with inherent weakening of the equatorial upwelling. These
anomalies are shown to be closely tied to the ‘Southern
Oscillation’ of Sir Gilbert Walker.”
Bjerknes, 1969, p.169
Bjerknes thus showed that El Niño was not a local phenomena, but
instead basin-wide (Glantz, 2001). Bjerknes’ conclusions effectively
ended the lull in Southern Oscillation and El Niño research that had
been in place since the 1940s, and the first scientific workshop
specifically on El Niño took place in December 1974, paving the way
6
for further research into the timing and causes of this important
climate/weather phenomena.
Research SUMMARY
This research focuses on the effects of El Niño on the Pacific
equatorial ocean basin around 143°W. This thesis examines the effects
of El Niño as witnessed in two cruises of the SSV Robert C. Seamans
between Papeete, Tahiti and Honolulu, Hawaii between February and
May of 2003. These effects are compared with those seen by Wyrtki
and Kilonski (1984) and Johnson (2002).
Southern Oscillation Index 2002-2004
Figure 1.1. The Southern Oscillation Index from January 2002 through
December 2003 showing the standardized sea level pressure over time.
Negative (blue) values indicate El Niño conditions and positive (red) values
indicate La Niña conditions.
7
This thesis demonstrates that data collected while on the
Robert C. Seamans reflect the waning of an El Niño cycle. This can be
seen both in the Southern Oscillation Index (Figure 1.1) and in the
trends of the currents that were modeled using geostrophic techniques.
Geostrophic techniques use the balancing of the Coriolis force with the
pressure gradient of the sea surface slope to approximate the velocity,
direction and volume of current transport. Presently, there are many
more accurate ways of measuring currents than by geostrophic
methods, but these methods are often more expensive and logistically
prohibitive.
viii
shows a noticeable decrease in the countercurrent velocities and an
increase in the SEC and NEC westward velocities.
In particular, the
South Subsurface Countercurrent, a very strong and shallow current,
similar to the EUC, and with higher temperatures and lower salinities
than found by Wyrtki and Kilonski (1984), was observed during the
southern portion of cruise 186 from Papeete, Tahiti to Nuku Hiva,
Marquesas and on towards Honolulu, Hawaii (from ~20 S to 20 N and
along ~143 W) in April 2003.