Sea Surface Current Circulation Pattern in the South China Sea Derived from Satellite
Altimetry
Nurul Hazrina Idris and Mohd Ibrahim Seeni Mohd
Department of Remote Sensing
Faculty of Geoinformation Science and Engineering
Universiti Teknologi Malaysia
81310 Skudai, Johor Darul Takzim
Malaysia.
Email: [email protected]
[email protected]
Contact no: +6075530745
KEY WORDS: Altimetry, Geostrophic current, South China Sea, Tidal current, Wind-driven surface current.
ABSTRACT: Sea surface currents and circulation patterns can be estimated globally and in a continuous manner using
satellite radar altimeters. This study is concerned with the surface current circulation pattern in the South China Sea.
The main data utilized are the tidal amplitudes from co-tidal charts, the sea level anomaly and the wind speed from the
Jason-1 satellite altimetry data. Algorithms used to estimate the surface current are the geostrophic current algorithms,
the wind-driven current algorithms and the tidal current algorithms. The derived surface currents from these models
were combined to produce the total surface current. Maps of surface current circulation pattern are produced for four
periods, which are the northeast monsoon, the southwest monsoon and the inter monsoon periods of April and October
in 2004 and in 2005. In order to validate the derived surface current, the correlation coefficient with field measurement
data was determined. The field data are measured using Lagrangian buoys under the Data Buoy Cooperation Panel
programme (DBCP) and also from the Japanese Oceanography Data Center (JODC). In this study, it was found that the
correlation of total surface current speed and direction is reasonable with coefficient of 0.9 and 0.8, respectively. As for
the total surface current circulation pattern, the currents in the South China Sea move southward during the northeast
monsoon and reversed during the southwest monsoon period. In general, during the northeast and southwest monsoon,
the water moved in anticlockwise and clockwise directions, respectively.
1.0
INTRODUCTION
In the work described in this paper, Jason-1 satellite altimetry data have been used together with field data to
study the surface currents in the South China Sea. The South China Sea is a large marginal sea situated at the western
side of the tropical Pacific Ocean. It is a semi-closed ocean basin surrounded by South China, Philippines, Borneo
Island, Indo-China Peninsula and Peninsular Malaysia. This water body connects with the East China Sea, the Pacific
and the Indian Ocean through the Taiwan Straits, the Luzon Straits and the Straits of Malacca, respectively (Chung et
al., 2000). The study area of this research is from 20 to 150 north and from 1000 to 1200 east. The area of the South
China Sea is shown in Figure 1.
Since the capability of radar altimetry to provide the surface current has been established (Challenor et al.,
1996), this study is concerned with the surface current in the southern part of the South China Sea during four monsoon
periods, which are the northeast monsoon (NE), the southwest monsoon (SW) and the inter monsoon periods of April
and October in 2004 and in 2005. The main data utilized in this study includes the tidal amplitudes from co-tidal chart,
the sea level anomaly and the wind speed data from satellite altimetry. Besides field measurement, data obtained from
the Coriolis ftp site and from the Japanese Organization Data Center (JODC) were also used for validation. The data are
available via the Coriolis ftp site; ftp://ftp.ifremer.fr/ifremer/coriolis/lagrangian_buoy and JODC site;
http://jdoss1.jodc.go.jp/cgi-bin/1997/ocs. The algorithms used to estimate the surface current are the geostrophic current
algorithms, the wind-driven current algorithms and the tidal current algorithms.
2.0
SEA SURFACE CURRENT ALGORITHMS
Algorithms used to estimate the surface currents include the geostrophic current algorithms (Chung et al.,
2000), the wind-driven current algorithms (Chung et al., 2000; Yanagi, 1999) and the tidal current algorithms (AlRabeh et al., 1990).
2.1
Geostrophic Current Algorithms
Algorithms of geostrophic current are expressed in Equation (1) to Equation (6).
ug = -g / f (δζ / δy)
(1)
vg = g / f (δζ / δx)
(2)
θg = tan-1 (vg / ug)
(3)
f = 2Ω cos φ
(4)
x = R (λ – λ0) cos φ
(5)
y = Rφ
(6)
where ug and vg are the geostrophic current in zonal and meridional direction, respectively, θg is the geostrophic current
direction, g is the gravitational acceleration, f is the Coriolis force, ζ is the sea surface height relative to the geoid, x and
y is the local east coordinate and the local north coordinate, respectively, Ω is the Earth rotation rate = 7.27 x 10-5 s-1, φ
is the latitude in radian, λ is the longitude in radian and R is the Earth radius = 6371 kilometers.
In order to implement the equations, firstly, the Coriolis force, the local east coordinate and the local north
coordinates are determined using Equation (4), Equation (5) and Equation (6). Then, derivatives of zonal and
meridional direction of sea surface height relative to the geoid are calculated. Then, geostrophic currents in zonal and
meridional directions are calculated using Equation (1) and Equation (2). In order to estimate the direction of the
geostrophic current, Equation (3) is used.
2.2
Wind-Driven Current Algorithms
Algorithms of wind-driven current are expressed in Equation (7) to Equation (10).
ue = V0 exp(az) cos(π/4 + az)
(7)
ve = V0 exp(az) sin(π/4 + az)
(8)
V0 = τ / (√ ρ2 f Az)
(9)
a = √ (f / 2 Az)
(10)
where ue and ve are the wind-driven current in zonal and meridional direction, respectively, Az is the vertical eddy
viscosity, f is the Coriolis force, ρ is the water density, τ is the wind stress = ρaCdW2 where ρa is the air density, W is the
wind speed 10 m above the sea surface and Cd is the drag coefficient = 2.6 x 10-3.
In order to implement the equations, firstly, the wind-driven current components in zonal and meridional
directions for water level, z = 1 m and z = 15 m are determined. Then, the mean of these two velocities is computed and
the vector sums of the components are calculated to obtain the current speeds. This speed is called the wind-driven
surface current. In this study, wind-driven current velocities at the water level of 15 m from the sea surface are
important because the current movement in this level varies significantly. For example, if wind speed blowing in the
water surface is 400 cm s-1, the wind-driven current speed at the water level of 1 m, 15 m and 30 m are 8.5 cm s-1, 2.2
cm s-1 and 1.5 cm s-1, respectively. This shows that the difference of wind-driven velocities in 15 m from surface water
is 6.3 cm s-1 which is significant. However, below the 15 m of water level from the sea surface, the current velocity
does not change significantly whereby the difference is less than 1 cm s-1.
The equations above indicate that the speed of the wind-driven surface current is proportional to wind stress.
According to Ekman spiral theory, the surface water movement is deflected by π/4 to the wind direction. In the northern
hemisphere, the wind-driven surface current is deflected in a clockwise direction. This is because of the force balance
between the wind drag and the water drag (Yanagi, 1999). Therefore, in this study, the direction of wind-driven current
(θe) is estimated by adding 450 clockwise to the direction of the surface wind.
2.3
Tidal Current Algorithms
Algorithms of tidal current are expressed in Equation (11) to Equation (13).
δ ut/ δt –(δ/ δz(Av δu/ δz)) – fv = -g δη/ δx – (1/ρ(δP/ δx))
(11)
δ vt /δt –(δ/ δz(Av δv/ δz)) + fu = -g δη/ δy – (1/ρ(δP/ δy))
(12)
θt = tan-1 (vt / ut)
(13)
where ut and vt are the zonal and meridional tidal currents, respectively, f is the Coriolis force, g is the gravitational
acceleration, η is the tidal amplitudes, Av is the vertical eddy viscosity, x and y is the local east coordinate and the local
north coordinate, respectively, t is the time, P is the atmospheric pressure, ρ is the water density, z is the water level and
θt is the tidal current direction.
In order to implement the algorithms of tidal current, firstly, tidal current in zonal and meridional direction is
determined using Equation (11) and Equation (12). Then, the direction of tidal current is determined using Equation
(13).
2.4
Total Sea Surface Current Algorithms
In the work described in this paper, total surface current is used to present the combination of geostrophic
current, wind-driven current and tidal current. In order to get the total surface current velocities, the vector sum of
geostrophic velocities, wind-driven velocities and tidal velocities are computed. As far as the total surface current
directions are concerned, means of the geostrophic current, the wind-driven current and the tidal current are estimated.
Equation (14) to Equation (16) is the formulae to compute the total surface current velocity and Equation (17) is the
formula to compute the total surface current direction.
utotal = ug + ue + ut
(14)
vtotal = vg + ve + vt
(15)
Vtotal = √ (utotal2 + vtotal2)
(16)
θtotal =( θg + θe + θt) / 3
(17)
where utotal and vtotal are the total surface current in zonal and meridional direction, respectively, Vtotal is the vector sum
of total surface current and θtotal is the total surface current direction.
3.0
RESULTS
The sea surface circulation patterns over the four monsoon periods in 2004 and in 2005 are shown in Figure 2.
These are presented in vector plots where the arrows represent the directions of the total surface current and the length
of arrows represent the magnitudes of the total surface current in cm s-1. The big arrows present the general circulation
pattern of the sea surface current.
Generally, during the inter monsoon periods of April in 2004 (Figure 2 (a)), water moves in a clockwise
circulation over the sea. On the coast of Peninsular Malaysia, water moves towards the north and then deflects to the
right on the northern part of the sea. On the coast of Borneo Island, water moves along the coast towards the south. In
the Gulf of Thailand, the surface current moves in an anticlockwise direction. In addition, there are two clockwise
eddies near the Natuna Islands. As for the surface current speed, the mean surface current is 52.9 cm s-1. The minimum
and the maximum speed over the period are 2.6 cm s-1 and 286.9 cm s-1, respectively. Strong current movements are
found on the Gulf of Thailand, on the south coast of Indochina and on the east coast of Peninsular Malaysia. As for the
surface circulation pattern during the inter monsoon periods of April in 2005 (Figure 2 (b)), similar circulation pattern is
found with minor changes over the period. There are two eddies near the Natuna Islands that move in anticlockwise
directions. As for the surface current speed, the mean speed over the period is 41.7 cm s-1. In addition, the minimum
and the maximum speed are 1.8 cm s-1 and 220.4 cm s-1, respectively. Strong current movements are found on the Gulf
of Thailand and on the southern part of the sea. Means during the two-year period indicates that surface current during
the inter monsoon periods of April in 2004 is stronger than in 2005.
As for the surface circulation pattern during inter monsoon period of October, there was no significant
difference between inter monsoon period of October in 2004 and in 2005 (Figure (c) and Figure (d)). Generally,
surface water moved in anti-clockwise direction towards the south over the period. Water in the coast of Indochina
moved southwards along the coast and then entered the Gulf of Thailand. In the Gulf of Thailand, the water was
deflected and moved in an anti-clockwise direction. Along the coast of Peninsular Malaysia and the coast of Borneo
Island, the surface current moved parallel to the land towards the south. In 2004 and 2005, an anti-clockwise eddy was
found in the middle of the sea. As for the current speed, surface current speed in 2004 is greater than in 2005 whereby
the mean speed was 52.8 cm s-1 and 40.1 cm s-1, respectively. In 2004, the minimum and maximum current speed was
1.5 cm s-1 and 264 cm s-1, respectively. Strong current movements were found in the south coast of Indochina and the
west coast of Borneo Island. In 2005, the minimum and maximum speed was 2.1 cm s-1 and 238.9 cm s-1, respectively.
However, strong current movements were found in the east coast of Peninsular Malaysia, the Natuna Islands, the south
coast of Indochina and the east coast of Indochina.
As for the surface current circulation pattern during the SW monsoon in 2004 (Figure 2 (e)), the surface
circulation pattern shows not much difference from the inter monsoon periods of April in 2004 whereby surface water
moves in a clockwise circulation over the periods. Similarly, water on the east coast of Peninsular Malaysia and on the
east coast of Indochina moves along the coast towards the north. However, the direction deflects to the right on the
northern part of the sea over the period. Besides, surface water on the coast of Borneo Island moves along the coast
towards the south. Over the period, strong surface current movements are found on the Gulf of Thailand, near the
Natuna Islands and on the south coast of Indochina. Mean, minimum and maximum surface current speeds are 55.2 cm
s-1, 3.1 cm s-1 and 280.4 cm s-1, respectively. In 2005 (Figure (f)), a clockwise circulation was found in the South China
Sea over the period. In the coast of Peninsular Malaysia, Indochina and Borneo Island, the water moved northwards
and in the northern part of the sea, water deflected to the right. Water circulation pattern in the Gulf of Thailand in 2005
is similar to 2004 where the surface current moved in anti-clockwise direction over the period. As for the surface
current speed, the current speed in 2005 is weaker than in 2004 where the mean, minimum and maximum speed was
43.4 cm s-1, 3 cm s-1 and 233.7 cm s-1, respectively. Strong surface current was found in the east coast of Indochina and
the southern part of the sea.
Generally, surface circulation pattern during the NE monsoon is opposite to the SW monsoon. In 2004 (Figure
2 (g)), it is found that surface water moves in anticlockwise direction during the NE monsoon. In the coast of Borneo
Island, the water moved along the coast towards the south. In addition, surface water in the middle of the sea deflected
to the left and then moved southwards along the coast of Indochina. As for the surface circulation pattern in the Gulf of
Thailand, surface water moved in a clockwise direction and then moved to the south along the coast of Peninsular
Malaysia. In the middle of the sea, a gyre move in anticlockwise direction is found. In 2005 (Figure 2 (h)), there is a
minor change on the surface circulation pattern whereby a clockwise eddy appears in the middle of the sea. As for the
surface current speed, it is found that the current movement in 2004 is stronger than in 2005 whereby the means of
current speed are 53.2 cm s-1 and 49.2 cm s-1, respectively. As for the speed in 2004, the minimum and the maximum
speeds are 1 cm s-1 and 299.7 cm s-1, respectively. Strong current movements are found in the Gulf of Thailand and on
the east coast of Peninsular Malaysia. In 2005, the minimum and the maximum speeds are 1.9 cm s-1 and 265.9 cm s-1,
respectively. Strong current movements are found in the Gulf of Thailand, near the Natuna Islands, along the east coast
of Peninsular Malaysia and along the south coast of Indochina.
4.0
ANALYSIS OF RESULTS
In order to assess the accuracy of the total surface current derived from the equations, the correlation
coefficient between these satellite derived values and field measurement values was computed. It was found that the
correlation of total surface current speed and direction is reasonable with correlation coefficients of 0.9 and 0.8,
respectively. Figure 3 and Figure 4 represent the graph of relationship between derived value and field measurement for
speed and direction, respectively. In addition, the mean of the derived total surface current is not significantly different
from the mean of the field measurement values. The standard deviations of derived current speed and direction are
improved significantly to 5 cm s-1 and 760 compared to the standard deviation of the field measurement which are 10
cm s-1 and 880. Table 1 illustrates the mean and the standard deviation of derived wind-driven current and field
measurement current values.
5.0
CONCLUSIONS
Total surface current can be estimated using the sea surface height data and wind data derived from satellite
altimeter in conjunction with field data. It was found that the derived total surface current is well correlated with the
field data whereby the correlation coefficient for the speed and the direction are 0.9 and 0.8, respectively. The means of
derived total surface current are not significantly different from the field measurement values and the standard
deviations of derived total surface current are improved significantly compared to the field measurement values. This
leads to the conclusion that the surface current in the South China Sea can be estimated using the combination of
geostrophic current algorithms, wind-driven current algorithms and tidal current algorithms.
ACKNOWLEDGEMENTS
The Jason-1 satellite data were produced by SSALTO/DUACS and distributed by Archiving, Validation, and
Interpretation of Satellite Data in Oceanography (AVISO), with support from National d’Etudes Spatiales (CNES),
France. The field measurement data were provided by the Malaysian Meteorological Services, Japanese Organization
Data Centre and Coriolis Data Centre. We are also grateful to Professor T. Yanagi and Japan Society for Promotion of
Science (JSPS) for their contributions.
REFERENCES
Al-Rabeh, A. H., Gunay, N., and Cekirge, H. M., 1990, A hydrodynamic model for wind-driven and tidal circulation in
the Arabian Gulf. Applied Mathematical Modelling, 14, pp 410 – 419.
Challenor, P. G., Read, J. F., and Tokmakin, R. T., 1996, Measuring surface currents in Drake passage from satellite
altimetry and hydrography. Journal of Physical Oceanography, 26. pp 2748-2758.
Chung, R. H., Zheng, Q., Soong, Y. S., Nan, J. K., and Jian, H. H., 2000, Seasonal variability of sea surface height in
the South China Sea observed with TOPEX/Poseidon altimeter data. Journal of Geophysical Research, 105, pp
13,981-13,990.
Yanagi, T., 1999, Coastal oceanography. Terra Scientific Publishing Company, Tokyo and Kluwer Academic Publisher,
Dordrecht, London, Boston.
14
INDOCHINA
12
GULF OF
THAILAND
PALAWAN
ISLAND
10
SOUTH CHINA SEA
8
BALABAC STRAIT
AR
UL
NS IA
NI Y S
P E A LA
M
6
4
2
100
102
104
NATUNA ISLANDS
BORNEO
ISLAND
106
108
110
112
114
116
118
Figure 1. The South China Sea region.
(a)
(b)
14
14
INDOCHINA
INDOCHINA
12
12
PALAWAN
ISLAND
8
2
100
102
BORNEO
ISLAND
104
106
108
110
112
114
116
4
2
100
118
AR
UL
NS IA
NI YS
PE ALA
M
4
8
6
AR
UL
NS IA
NI YS
PE ALA
M
6
PALAWAN
ISLAND
10
Latitude
Latitude
10
102
BORNEO
ISLAND
104
106
(c)
PALAWAN
ISLAND
BORNEO
ISLAND
104
106
108
110
112
114
116
4
2
100
118
AR
UL
NS IA
NI YS
PE ALA
M
6
AR
UL
NS IA
NI YS
PE ALA
M
102
8
102
BORNEO
ISLAND
104
106
(e)
110
112
114
116
118
(f)
14
14
INDOCHINA
INDOCHINA
12
12
PALAWAN
ISLAND
8
104
106
108
110
Longitude
112
114
116
4
118
2
100
AR
UL
NS I A
NI YS
PE ALA
M
AR
UL
NS IA
NI YS
PE ALA
M
102
8
6
BORNEO
ISLAND
PALAWAN
ISLAND
10
Latitude
10
Latitude
108
Longitude
Longitude
2
100
118
PALAWAN
ISLAND
10
Latitude
Latitude
8
4
116
12
10
6
114
INDOCHINA
INDOCHINA
2
100
112
(d)
12
4
110
14
14
6
108
Longitude
Longitude
102
104
BORNEO
ISLAND
106
108
110
Longitude
112
114
116
118
(h)
(g)
14
14
INDOCHINA
INDOCHINA
12
12
PALAWAN
ISLAND
8
2
100
102
BORNEO
ISLAND
104
106
108
110
112
114
116
AR
UL
NS IA
NI YS
PE ALA
M
4
8
6
AR
UL
NS IA
NI YS
PE ALA
M
6
PALAWAN
ISLAND
10
Latitude
Latitude
10
4
2
100
118
102
BORNEO
ISLAND
104
106
108
110
112
114
116
118
Longitude
Longitude
Figure 2. Total sea surface current circulation pattern on (a) inter monsoon periods of April 2004, (b) inter monsoon
periods of April 2005, (c) inter monsoon periods of October 2004, (d) inter monsoon periods of October 2005,
(e) SW 2004, (f) SW 2005, (g) NE 2004 and (h) NE 2005.
Satellite
derived
total
surface
current
speed
-1
(cm s )
40.0
350
35.0
300
30.0
250
25.0
10.0
Satellite 200
derived total
surface 150
current
direction 100
5.0
50
20.0
15.0
(0)
0.0
0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
-1
Surface current speed, field data (cm s )
Figure 3. Relationship between field data
and derived total surface current speed.
0
50
100
150
200
250
300
350
400
0
Surface current direction, field data ( )
Figure 4. Relationship between field data
and derived total surface current direction.
Table 1. Means and standard deviations of sea surface current.
Satellite derived total surface
Field measurement values
current (cm s-1)
(cm s-1)
-1
0
Speed (cm s )
Direction ( )
Speed (cm s-1)
Direction (0)
Mean
15
194
10
215
Standard deviation
10
88
5
76