The Twelfth OMISAR Workshop on Ocean Models

Predicting Movement of the Warm Pool, the

Salinity Front, and the Convergence Zone in the
Western and Central Part of Equatorial Pacific
Using a Coupled Hydrodynamical-Ecological
Model
Supangat1, A., T. Rameyo Adi1, Widodo S. Pranowo1, and N.S. Ningsih2.
1
Agency for Marine Affairs and Fisheries, Indonesia
2
Study Program of Oceanography, Bandung Institute of Technology, Indonesia

ABSTRACT
A coupled hydrodynamical-ecological model for regional and shelf seas
called ‘COHERENS’ and developed by Luyten et al. (1999) was used to
predict the water dynamics, distribution of temperature and salinity in the
western and central part of equatorial Pacific. The model results have
demonstrated well equatorial processes that exist in the ocean, such as the
major currents, the western equatorial Pacific warm pool, equatorial upwelling
in the central Pacific, and the permanent convergence of surface-layer water
masses on the eastern edge of the warm pool. Lehodey et al. (1997)
published that one of the most successful fishing grounds is located in the
vicinity of the convergence zone between the warm (> 28 – 29 oC) low-salinity
water of the warm pool and the cold saline water of the central Pacific
equatorial upwelling. This zone of convergence, identified by a well-marked
salinity front and approximated by the 28 – 29 oC sea surface temperature
(SST) isotherm, has been simulated well by the model. Therefore, seasonal
variations of movement of the 28 – 29 oC isotherm, the salinity front, and the
convergence zone are important to be predicted. We would expect that the
model results might be able to be used for the prediction of good fishing
grounds.

Key Words: coupled-models, major currents, warm pool, equatorial

upwelling, convergence zone, salinity front, Pacific, fishing grounds.

INTRODUCTION
The basic feature and the existence of equatorial processes in the Pacific Ocean,
such as the major currents, the western equatorial Pacific warm pool (WEPWP, defined
by sea surface temperatures ≥ 28 oC), equatorial upwelling in the central Pacific, and

Printed on recycled paper 11 - 1

The Twelfth OMISAR Workshop on Ocean Models

the permanent convergence of surface-layer water masses on the eastern edge of the
warm pool, have been documented by previous scientists (e.g., Tomczak and Godfrey,
1994; Wyrtky, 1989; Stewart, 2002). Understanding the spatial and temporal evolution
of the equatorial processes is an important factor for atmospheric circulation and
fisheries.

Stewart (2002) summarized characteristics of the equatorial processes in the

Pacific Ocean, as follows: (i) there are permanent major currents in the equatorial
Pacific (e.g., the wind-driven North and South Equatorial Currents, North and South
Equatorial Counter Current, Equatorial Under Current, and Equatorial Intermediate
Current), (ii) surface waters are warmest in the WEPWP, (iii) the divergence of the
Ekman transport produces upwelling of cold water on the equator, and (iv) there is a
permanent convergence of surface-layer water masses on the eastern edge of the
warm pool that is close to cold tongue of water around the equator.

It has been recognized that the major currents in the equatorial Pacific redistribute
heat, which is then released by the rain. The released heat, variability of currents and
temperatures in the equatorial region plays an important role in the atmospheric
circulation (Stewart, 2002). In addition, movement of the WEPWP has been the subject
of intense researches recently because it is related to the development of ENSO (El
Niño Southern Oscillation) events and the earth’s climate in general (Philander, 1990;
Wyrtky, 1989; Yan et al., 1997). Meanwhile, Lehodey et al. (1997) published that nearly
70 % of the world’s annual tuna harvest comes from the Pacific Ocean, in which
catches are highest in the WEPWP. In addition, they also reported that one of the most
successful fishing grounds is located in the vicinity of the convergence zone between
the warm (> 28 – 29 oC) low-salinity water of the warm pool and the cold saline water of
the central Pacific equatorial upwelling. This zone of convergence is identified by a
well-marked salinity front and approximated by the 28 – 29 oC sea surface temperature
(SST) isotherm.

The equatorial processes respond quickly to variations in the wind fields, mainly
the Trade winds (the northeastern and southeastern Trades), which are highly
persistent winds converging near the equator in a narrow Intertropical Convergence
Zone. Previous scientists have documented that although the Trades are remarkably
steady, they do vary from month to month and can change dramatically from year to
year (e.g., Tomczak and Godfrey, 1994; Wyrtky, 1989; Stewart, 2002). Therefore, the
strength and position of the equatorial currents and upwelling vary both seasonally
associated with meridional variations of sunlight and annually associated with the
ENSO phenomenon. In addition, the warm pool, the salinity front, and the convergence
zone also move both seasonally and annually.

11 – 2 Printed on recycled paper

 The Twelfth OMISAR Workshop on Ocean Models

In the present study, we address to predict seasonal variations of the equatorial

processes, such as strength and position equatorial currents and upwelling, and
movement of the warm pool, the salinity front, and the convergence zone in the
western and central part of equatorial Pacific by using numerical model. A coupled
hydrodynamical-ecological model called ‘COHERENS’ and developed by Luyten et al.
(1999) was used to predict the water dynamics, distribution of temperature and salinity
in the western and central part of equatorial Pacific. For sake of simplicity and as a
preliminary study, we neglected interannual variability of the equatorial processes
although it has been documented by previous scientists (e.g., Tomczak and Godfrey,
1994; Wyrtky, 1989; Stewart, 2002).

MODEL CONFIGURATION
A three-dimensional prognostic calculation of the COHERENS was used in this
study. Location of the study area is from of 20 oS - 20 oN, and 120 oE -150 oW. The
domain is covered by 181 x 81 grid cells at approximately 0.5o resolution. The vertical
grid consists of 19 σ-levels concentrated over the upper and intermediate depths in
order to adequately resolve the respective wind-influenced layers. The bathymetry was
interpolated from the GTOPO30 data set (http://topex.ucsd.edu/cgi-bin/get_data.cgi).
The external and internal mode time steps used in the simulation are 60 s and 1200 s,
respectively. The main forcing for the model is the 5 daily-wind fields derived from
SSM/I (Special-Sensor Microwave/Imager), PODAAC (Physical Oceanography
Distributed Active Archive Center). The initial temperature and salinity data are taken
from the Levitus data (1982). In addition to the wind field data, the following surface
data were supplied to the model, namely monthly mean values of air temperature,
atmospheric pressure, relative humidity, cloud coverage, evaporation and precipitation
rates, which are derived from DASILVA
(http://Ingrid.ldeo.columbia.edu/SOURCE/.DASILVA). The simulations covered a full
yearly cycle for the year 1993 providing a prediction of monthly averaged fields of the
equatorial processes.

Results and Discussion

MAJOR CURRENTS IN THE EQUATORIAL PACIFIC

The model results reproduce well the existence of permanent major currents in the
equatorial Pacific. Monthly averaged surface distributions of currents and temperature
during 1993 are shown in Fig. 1. The figure shows the major westward components of
the equatorial current system, namely the North and South Equatorial Current (NEC
and SEC, respectively). It can be seen that the NEC and SEC vary seasonally in

Printed on recycled paper 11- 3

The Twelfth OMISAR Workshop on Ocean Models

strength. The NEC flows steadily during the whole year between 8 oN and 15 oN with
speeds of 0.5 m s-1 or less. In the northern winter, the NEC seems to be stronger, for
example in January and December (Fig. 1). During the month of April 1993 (weak
event of El Niño), the NEC remained strong and it reached speeds of 0.5 m s-1.
Meanwhile, the strength of NEC was reduced in the northern summer (i.e., in July and
August) and during the months of September and October with speeds of 0.2 m s-1 or
less. On the other hand, the SEC was very weak in January and increased afterwards
gradually until it is most strongly developed in April with speeds of 1.0 m s-1 or less. In
addition, in the month of April, the SEC extends so far to the northern hemisphere
reaching 3 oN. Meanwhile, during the month of July and October, the SEC was still
developed. However, its strength decreased and its position moved further south
around 6 - 13 oS.

The simulations also shows the most important eastward flow in the equatorial
current system, namely the North Equatorial Counter Current (NECC) and the
Equatorial Under Current (EUC), as shown in Fig. 1 and 2, respectively. The NECC is
strongly seasonal, in which in May it starts to be fed from both hemispheres, flows
between 2 - 7 oN, and reaches its greatest strength during July – December with
speeds of 0.7 m s-1 or less (Fig. 1). Meanwhile, during January – April the NECC is
weak and flows around 2 - 5 oN. According to Tomczak and Godfrey (1994), during the
months of January – April, the Northwest Monsoon prevents the SEC from feeding the
NECC, so that the strength of NECC was reduced (Fig. 1).

Figure 2 shows simulation results of monthly averaged zonal velocity at 170 oW

for January, April, July, and October 1993. The existence of the EUC below the surface
on the equator can be seen in the figure. The EUC flows strongest during February –
July but it is much weaker in August – January. In the month of January, the EUC is
very weak at speeds of 0.05 m s-1 or less. This may be due to a weakening of the SEC
in January. Tomczak and Godfrey (1994) concluded that the EUC belongs entirely to
the southern circulation system and its source waters originate mainly from the
southern hemisphere. In addition, they reported that the EUC is fed from the SEC.
Therefore, the EUC is associated with variable strength of the SEC. In this study, the
core of EUC is found at a depth of from 300 to 600 m. Meanwhile, according to
Tomczak and Godfrey (1994), its existence is found at 200 m depth. An explanation for
the difference of results obtained from both studies is still lacking. A good explanation
might be obtained by extending the simulation periods, which represent periods of
marked interannual variations in the equatorial Pacific (i.e., associated with the El Niño
and La Niña phenomena). Existence of the Equatorial Intermediate Current (EIC),
flowing from East to West, is also found by the model (Fig. 2). The current core is found
at a depth of from 900 to 1900 m.

11 – 4 Printed on recycled paper

 The Twelfth OMISAR Workshop on Ocean Models

WESTERN EQUATORIAL PACIFIC WARM POOL

Figure 1 also shows monthly averaged spatial distribution of the western
equatorial Pacific warm pool (WEPWP), which is defined by sea surface temperatures
≥ 28 oC during 1993. The spatial and temporal movement of the warm pool can be
seen in the figure. During the months of January – April, the WEPWP mainly existed in
the southern hemisphere and then moved mostly northward associated with the
meridional movement of the sun. Meanwhile, around the western equator the SST
decreased from ≥ 28o to 27- 28 oC in June – October, and to 26- 27 oC in November –
December. It is not clear whether this decrease in temperature is due to the 1993 El
Niño event or the zonal transport of cold water originating in the central Pacific
equatorial upwelling by the westward-flowing south equatorial current.

EQUATORIAL UPWELLING IN THE CENTRAL PACIFIC

The existence of equatorial upwelling in the central Pacific is well indicated in Fig.
1 and identified by a cold tongue of water around the equator. During the months of
January – March, a cold tongue of 26 – 27 oC reached around 167 oW, and then
extended further westward in April – September with a maximum distance of 160 oE in
July. The westward displacement is induced by the strong south equatorial current
(SEC). In April, a 25 – 26 oC cold tongue started to present reaching around 173 oW.
The cold tongue also extended further westward during the months of May – December
and reaching a maximum of 175 oE in December (Fig. 1).

In addition, the presence of equatorial upwelling can be also seen in Fig. 3. The
figure showed simulation results of monthly averaged temperature for January, April,
July, and October 1993 along a vertical transect at 170 oW. In January, the upwelling
strength was weak indicated by the 27 – 28 oC isotherms, which are upwelled to the
surface on the equator. In the other months, the upwelling strength increased and it
was indicated by the isotherm of 26 – 27 oC in April, and of 25 – 26 oC in July and
October.

CONVERGENCE ZONE AND SALINITY FRONT

Figure 4 and 5 show monthly averaged distribution of surface salinity for the
months of January – June and July – December 1993, respectively. The thin solid lines
indicating the 26, 27, 28, and 28.5 oC isotherms and the thin dash line pointing to the
34.6 isohaline are superimposed in the figures. Overlaying the cold tongue of water
around the equator on the spatial distribution of surface salinity for the same month
indicates that the warm low-salinity water of the warm pool would converge with the
cold saline water of the central Pacific equatorial upwelling. This pattern appears to be

Printed on recycled paper 11- 5

The Twelfth OMISAR Workshop on Ocean Models

permanent, as shown in Fig. 4 and 5. Therefore, the major feature on this figure is the
presence of a permanent convergence of surface-layer water masses on the eastern
edge of the warm pool that is close to cold tongue of water around the equator. The
convergence zone is identified by a well-marked salinity front and approximated by the
26 – 29 oC isotherm. Here, the salinity front is approximated by the 34.6 isohaline. The
model results show the spatial and temporal evolution of the convergence zone, which
moved both zonally and meridionally. Meanwhile, we also superimposed lines that
represent the convergence zone in Fig.4 and 5, namely the thick solid line with arrows
on the both corners (close to the 28.5 oC isotherm), the dash solid line with arrows on
the both corners (close to the 28 oC isotherm), the square dot line with arrows on the
both corners (close to the 27 oC isotherm), and the long dash line (close to the 26 oC
isotherm).

During the months of January, February, and March, the convergence zone is
identified by a well-marked salinity front and approximated by the 28 – 29 oC isotherm,
as delineated on Fig. 4 by the thick solid line with arrows on the both corners. In the
periods, the convergence zone appeared around 163 oE – 150 oW and 13 oS - 6 oN. In
April, the convergence zone, marked by the thick and dash solid line with arrows on the
both corners (close to the 28.5 and 28 oC isotherms, respectively), was found around
165 oE – 158 oW and 8 oS - 5 oN. In May, it moved further westward reaching around
156 oE and was marked by the dash solid line with arrows on the both corners (close to
the 28 oC isotherm). In June – August (Fig. 4 and 5), the convergence zone existed
around 162 oE – 162 oW and 13 oS - 4 oN and was marked by the dash-solid and the
square-dot lines with arrows on the both corners, which close to the 28 oC and 27 oC
isotherms, respectively. Meanwhile, during the months of September – December (Fig.
5), it appeared mostly in the southern hemisphere (163 oE – 164 oW and 13 oS - 3 oN)
and was delineated by the square-dot lines with arrows on the both corners and by the
long dash line, which close to the 27 oC and 26 oC isotherms, respectively. Meanwhile,
around the western equator the SST decreased from ≥ 28 to 27- 28 oC in June –
October, and to 26- 27 oC in November – December. As mentioned previously, during
the months of June – December there was a decrease in temperature of WEPWP, so
that we used the >26 and >27 oC isotherms to indicate the convergence zone.

As reported by Lehodey et al. (1997), one of the most successful fishing grounds
is located in the vicinity of the convergence zone. For model verification in the present
study, we used the location of US purse-seiners catch per unit effort (CPUE) data >10
mt d-1 (metric tonnes per fishing day on a 1 week time resolution) of skipjack tuna,
which is published by Lehodey et al. (1998) for the month of June 1994. The location is
approximated by the ellipse shapes in Fig. 4 (for the month of June). In general, the
present simulated convergence zone is in agreement with the location of high skipjack

11 – 6 Printed on recycled paper

 The Twelfth OMISAR Workshop on Ocean Models

tuna CPUE although there is a shift in space appears between them. The space shift
could be due to the difference of the periods of simulation (1993) and the CPUE data
(1994) and also because the surface data supplied to the model (i.e., monthly mean
values of air temperature, atmospheric pressure, relative humidity, cloud coverage,
evaporation and precipitation rates) are not representative. We would expect that the
model results might be able to be improved significantly if we used a shorter interval of
the surface data, namely at least 12-hourly data in order to represent the daytime and
nighttime data.

Concluding Remarks
Equatorial processes that exist in the western and central part of equatorial Pacific Ocean (e.g.,
the major currents, the western equatorial Pacific warm pool, equatorial upwelling in the central
Pacific, and the permanent convergence of surface-layer water masses on the eastern edge of the
warm pool) have been predicted by using the three-dimensional prognostic calculation of
COHERENS developed by Luyten et al. (1999). The model results have shown the spatial and
temporal evolution of the warm pool, the salinity front, and the convergence zone. In general,
during the period of simulation (1993), the convergence zone, as the one of the most successful
fishing grounds, was found around 156 oE – 150 oW and 13 oS - 6 oN. The present simulated
convergence zone is generally encouraging. However, to obtain a significant improvement of
the predicting movement of the warm pool, salinity front, and convergence zone, it is important
to extend the simulation periods in order to acquire some insight on the interannual variability
of equatorial processes. In addition, to improve the model results, it is also necessary to use a
shorter interval of the surface data (at least 12-hourly data), which are supplied to the model,
such as values of air temperature, atmospheric pressure, relative humidity, cloud coverage,
evaporation and precipitation rates. By extending the simulation periods and using the shorter
interval of the surface data as well as the availability of good and complete data of skipjack tuna
catch and fishing effort, we would expect to establish the link of behaviour, migration pattern,
and seasonal distribution of the fish to the physical properties and variability of the water. Once
the link of them are established, knowledge of climate variability and the ocean’s role in the
convergence zone can then be used for the prediction of good fishing grounds, namely
increasing the catch, while keeping the sustainable stocks.

REFERENCES
Lehodey, P., Bertignac, M., Hampton, J., Lewis, A. and Picaut, J. (1997) El Niño
Southern Oscillation and Tuna in the Western Pacific. Nature 389: 715-718.
Lehodey, P., Andre, J-M., Bertignac, M., Hampton, J., Stoens, A., Menkes, C., Memery,
L. and Grima, N. (1998) Predicting Skipjack Tuna Forage Distributions in the
Equatorial Pacific using A Coupled Dynamical Bio-geochemical Model. Fish
Oceanogr. 7(3/4): pp. 1-9.
Levitus, S. (1982) Climatological Atlas of the World Ocean. U.S. Department of

Printed on recycled paper 11- 7

The Twelfth OMISAR Workshop on Ocean Models

Commerce, National Oceanographic and Atmospheric Administration, Tech. Rep.

13: 173 pp.
Luyten, P.J., Jones, J.E., Proctor, R., Tabor, A., Tett, P., and Wild–Allend, K. (1999)
COHERENS ⎯ A Coupled Hydrodynamical–Ecological Model for Regional and
Shelf Seas: User Documentation. MUMM Report, Management Unit of the
Mathematical Models of the North Sea, 914 pp.
Philander, S. G. (1990) El Niño, La Niña, Southern Oscillation. Academic Press, 293 pp.
Stewart, R. H. (2002) Introduction to Physical Oceanography. Spring 2002 Edition, 341
pp.
Tomczak, M. and Godfrey, J. S. (1994) Regional Oceanography: An Introduction, Butler
& Tanner Ltd, Frome and London, 422 pp.
Yan, X., He, Y., Liu, W.T, Zheng, Q., Ho, C. (1997) Centroid Motion of the Western
Pacific Warm Pool during Three Recent El Niño Southern Oscillation Events. J.
Phys. Oceanogr. 27: 837-845.
Wyrtki, K. (1989) Some Thoughts about the west Pacific Warm Pool. Proc. Western
Pacific Int. Meeting and Workshop on TOGA COARE, New Caledonia, ORSTUM,
Centre de Nouma, pp. 99-109.

January February

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

March April

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

May June

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W
Longitude (deg) Longitude (deg)

11 – 8 Printed on recycled paper

 The Twelfth OMISAR Workshop on Ocean Models

July August

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

September October

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

November December

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W
120E 130E 140E 150E 160E 170E 180E 170W 160W 150W
0.0 – 0.5 m/det
0.5 – 2.0 m/det

Figure 1. Monthly averaged surface distributions of currents (m s-1) and temperature

(oC) during 1993. The contours of the 26, 27, 28, and 28.5 oC isotherms are
superimposed.

January 0.35
July 0.45
-100 -100 0.40
0.30
-300 0.25 -300 0.35
0.30
-500 0.20 -500
0.25
0.15
-700 -700 0.20
AN (M)

0.10
(M)

0.15
(meters)

(meters)

-900 0.05
-900
0.10
KEDALAMAN

-1100 0.01 -1100 0.05

KEDALAM

-1300 0.00 -1300 0.00

Depth

Depth

-0.05 -0.05
-1500 -1500 -0.10
-0.10
-1700 -1700 -0.15
-0.15
-0.20
-1900 -0.20 -1900
-0.25
-0.25
-2100 -2100 -0.30
-0.30 -0.35
-2300 -2300
-0.35 -0.40
20N 15N 10N 5N 0 5S 10S 15S 20S -0.40 20N 15N 10N 5N 0 5S 10S 15S 20S -0.45

April 0.40 October 0.65

-100 0.35 -100 0.60
0.30 0.55
-300 -300
0.50
0.25
-500 -500 0.45
0.20
0.40
-700 0.15 -700 0.35
(M)
(M)

0.10
(meters)

(meters)

-900 -900 0.30

KEDALAMAN
KEDALAMAN

0.05 0.25
-1100 0.00 -1100 0.20
-0.05 -1300 0.15
-1300
0.10
Depth

Depth

-0.10
-1500 -1500 0.05
-0.15
0.00
-1700 -0.20 -1700
-0.05
-0.25 -0.10
-1900 -1900
-0.30 -0.15
-2100 -0.35 -2100 -0.20
-0.40 -2300 -0.25
-2300
-0.45 -0.30
20N 15N 10N 5N 0 5S 10S 15S 20S -0.50 20N 15N 10N 5N 0 5S 10S 15S 20S -0.35

Printed on recycled paper 11- 9

The Twelfth OMISAR Workshop on Ocean Models

Figure 2. Monthly averaged zonal velocity (m s-1) at 170 oW for January, April, July, and
October 1993. Positive values are eastward currents, whereas negative
values are westward currents.

January 29 July 29
28 28
27 27
26 26
25
25
-100 -100 24
24
23
23

KEDA LAM A N (M )
Depth (meters)

Depth (meters)
KEDA LAM AN (M )

22
22 21
21 20
-200 -200
20 19
19 18
18 17
17 16
-300 -300
16 15
15 14
14 13
13 12
-400 -400
12 11
11 10
20N 15N 10N 5N 0 5S 10S 15S 20S 10 20N 15N 10N 5N 0 5S 10S 15S 20S 9
9 7

April 29 October 28
28 27
27 26
26 25
25 24
-100 24 -100 23
23 22
Depth (meters)
K EDA LA MAN (M )

K EDALA M AN (M )
Depth (meters)
22 21
21 20
-200 20 -200 19
19 18
18 17
17 16
-300 16 -300 15
15 14
14 13
13 12
-400 12 -400 11
11 10
20N 15N 10N 5N 0 5S 10S 15S 20S 10 20N 15N 10N 5N 0 5S 10S 15S 20S 9
8 5

Figure 3. Monthly averaged temperature (oC) for January, April, July, and October 1993
along a vertical transect at 170 oW.

January February

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

March April

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

May June

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

PSU

Figure 4. Monthly averaged distributions of surface salinity during the months of

January – June 1993. The thin solid lines indicate the 26, 27, 28, and 28.5 oC

11 – 10 Printed on recycled paper

 The Twelfth OMISAR Workshop on Ocean Models

isotherms, whereas the thin dash line is the 34.6 isohaline. The delineation of
the convergence zone is approximated by the thick solid line with arrows on
the both corners (close to the 28.5 oC isotherm), the dash solid line with
arrows on the both corners (close to the 28 oC isotherm), the square dot line
with arrows on the both corners (close to the 27 oC isotherm), and the long
dash line (close to the 26 oC isotherm). For the month of June we imposed the
location of US purse-seiners catch per unit effort (CPUE) data >10 mt d-1 of
skipjack tuna, which is published by Lehodey et al. (1998) and approximated
by the ellipse shapes.

July August

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

September October

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

November December

120E 130E 140E 150E 160E 170E 180E 170W 160W 150W 120E 130E 140E 150E 160E 170E 180E 170W 160W 150W

PSU

Figure 5. As for Fig. 4, except during the months of July – December 1993.

Printed on recycled paper 11- 11