Earth and Planetary Science Letters 430 (2015) 477–485

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

www.elsevier.com/locate/epsl

In situ observation of contour currents in the northern South China

Sea: Applications for deepwater sediment transport
Yulong Zhao a,∗ , Zhifei Liu a , Yanwei Zhang a , Jianru Li a , Meng Wang a , Wenguang Wang a ,
Jingping Xu b
a
State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, China
b
College of Marine Geosciences, Ocean University of China, Qingdao 266100, China

a r t i c l e i n f o a b s t r a c t

Article history: Deepwater currents and related suspended sediment concentration were obtained by an Acoustic Doppler
Received 29 March 2015 Current Profiler (ADCP) mooring system in the northern South China Sea from September 2011 to
Received in revised form 30 August 2015 May 2013 to characterize the occurrence of contour currents and to evaluate their sediment transport
Accepted 6 September 2015
capacity. Magnitude of the current velocity generally varied in the range of 0–2 cm/s, with a dominant
Available online 19 September 2015
flow direction of ∼250◦ (southwestward). The observed contour current, defined as the along-slope
Editor: M. Frank
component of the deepwater currents, has a tunnel-like vertical structure with the largest velocity
Keywords: occurring in the middle of the “tunnel” and decreasing outwards. Both the magnitude and the depth
sub-inertial current range of the maximum velocity display evident inter-seasonal variations, with the strongest velocity in
contour current summer and the weakest in spring, while the thickness of the contour currents was the highest in winter
mesoscale eddies and the lowest in spring. We also found that passing-through of the deep-reaching mesoscale eddies
suspended sediment concentration significantly affected the magnitude and direction of the contour currents. The suspended sediment
sediment transport
concentration (SSC) estimated from echo intensities of the ADCP is the highest at the near-bottom
South China Sea
(>400 μg/L) and decreases upwards to <10 μg/L at water depth shallower than 1750 m. High SSC is
mostly observed during periods of low magnitude of the contour currents, suggesting resuspension of
sediment from the seafloor is not the major controlling factor of these high-SSC events. Our observation
also suggests that the major role that contour currents play is to transport sediment from the sources
through keeping sediment suspended above the lower continental slope of the South China Sea.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction sition of sediments (Rebesco et al., 2008). Contourite, the marine

A contour current is a persistent along-slope flow of water mass
in the deep sea that is controlled mostly by topography (Rebesco
et al., 2008; Zenk, 2008). Deflection against a side wall of deep-
sea basin is an elementary property of the contour currents (Zenk,
2008). At the continental margin, the contour current usually runs
parallel to the topographic isobaths, but it is not necessary that
a contour current has to be in contact with the side wall or the
seafloor (Zenk, 2008). In physical oceanography, the contour cur-
rent is frequently considered as a particular type of bottom cur-
rent, although most contour currents investigated so far occur at
intermediate levels (Zenk, 2008). Contour current has drawn atten-
tions of the marine geologists since the pioneering work of Heezen
and Hollister (1964), because it has the capability to affect the sea
floor by re-suspending, transporting, and/or controlling the depo-
* Corresponding author.
E-mail address: yeoloon@tongji.edu.cn (Y. Zhao).
sediment that has been deposited from or significantly affected by
the contour currents, is known to construct large accumulation of
sediments in the deep-sea environment and to cover large areas
of the present-day seafloor beneath the modern bottom-current
systems (Stow et al., 2008). In situ oceanographic observation of
modern contour currents is the most fundamental method for in-
vestigating contourite depositions in the deep sea (Howe, 2008).
However, such measurements are seldom undertaken in most cases
of contourite researches for reasons of cost. The deficiency of ob-
servational datasets makes the concept of “contour current” not
well accepted in the community of the physical oceanographers,
largely jeopardizing the further development of the research of the
contourites.
The South China Sea (SCS) is the largest marginal sea in the
western Pacific, with a maximum water depth of >5000 m. The
only significant deep connection between the SCS and the open
ocean is the Luzon Strait located between Taiwan and Luzon, with
the deepest sill at around 2400 m (Qu et al., 2006). Recent field ob-

http://dx.doi.org/10.1016/j.epsl.2015.09.008
0012-821X/© 2015 Elsevier B.V. All rights reserved.
478 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485

Fig. 1. Up: Bathymetric map of the northern SCS showing location of the Station TJ-A-1 and vertical structure of the mooring (left side). The contour line at 2000 m is shown
in yellow line. Down: Reflection seismic profile showing structure of the sediment drift, with locations of ODP Site 1144 and Core 17940 (Shao et al., 2007).

servations (Tian et al., 2006) as well as numerical modeling (Lan et
al., 2013) have confirmed that water exchange in the Luzon Strait
is featured by a sandwich-like vertical structure: westward into the
SCS in the upper layer (<500 m), eastward into the Pacific Ocean
in the intermediate layer (500–1500 m), and westward flow again
in the deep layer (>1500 m). The westward deepwater current
is closely related with intrusion of the North Pacific Deep Water
(NPDW) into the SCS (Qu et al., 2006). Mooring observation in the
Luzon Strait reveals that deep water flows into the Strait through
the Bashi Channel and the Taltung Canyon, then turns southward
along the Luzon Trough, and finally enters the SCS through two
gaps on the southern Hengchun Ridge (Zhao et al., 2014). Upon en-
tering the SCS, the deepwater current (2000–2500 m) is enhanced
by the SCS deep western boundary currents, and it flows firstly
northwestward and then turns southwestward along the continen-
tal margins (Qu et al., 2006). In the deeper part of the SCS beneath
the Luzon sill at 2400 m, the water circulation is quasi-barotropic,
with average horizontal current velocity reaching maximum speed
of 1.09 and 1.35 cm/s along the northern and western boundaries,
respectively (G. Wang et al., 2011). In the northern SCS, the deep-
water circulation can occasionally be affected by the activities of
deep-penetrating mesoscale eddies (Z. Zhang et al., 2013; Y. Zhang
et al., 2014). Passing-through of a pair of eddies in early 2012 led
to an increase of the mean magnitude of velocity from 1.7 cm/s
to 3.0 cm/s at >3500 m water depth (Z. Zhang et al., 2013). The
same pair of eddies brought ∼0.52 Mt of suspended sediment
from southeast of Taiwan to the northern SCS basin (Y. Zhang et
al., 2014).
Reflection seismic profiles in the northern SCS show that intru-
sion of the NPDW into the SCS has led to the formation of a series
of discontinuous sediment drifts along the trajectory of the bot-
tom currents (Shao et al., 2007). Among all the contourite drifts
in the northern SCS, the high-accumulation-rate sediment drift at
southeast of the Dongsha Islands, on which several deep-sea sedi-
ment cores are located, including ODP Site 1144, Core MD05-2905
and Core 17940, has drawn attentions of many researchers (Fig. 1;
Sarnthein et al., 1994; Wong et al., 1994; Shao et al., 2001, 2007;
Laj et al., 2005; Lüdmann et al., 2005). The sediment drift pos-
sesses the highest ever-known linear sediment rate (LSR) in the
SCS deep basin, with the average LSR for the last 12 kyr reach-
ing 56–58 cm/kyr on the upper and middle part of drift (L. Wang
et al., 1999; Bühring et al., 2004). Based on results of the reflec-
tion seismic profiles, two hypotheses were proposed to interpret
the formation of the sediment drift at southeast of the Dongsha Is-
lands. Lüdmann et al. (2005) proposed that sediment re-suspended
off Southeast Taiwan was brought into the interior of the SCS by
the bottom currents, when encountering the continental slope off
the Dongsha Islands, the currents started flowing upward the slope
and slowing down, leading to the formation of this sediment drift
(Lüdmann et al., 2005). It was suggested by Shao et al. (2007),
 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485
however, that formation of the sediment drifts in the northern SCS
was mainly due to leakage of sediments from the northern SCS
bottom currents on the levees along the trajectory channels, with
propagation of the drifts following the main direction of the bot-
tom currents. Nevertheless, owing to poor understanding of the
characteristics of deepwater circulation in the northern SCS, it is
difficult to evaluate these hypotheses involving the formation of
the sediment drift.
In order to investigate the patterns and features of the contour
currents in the northern SCS and their potential links with for-
mation of the sediment drift at southeast of the Dongsha Islands,
one bottom-anchored sub-surface mooring system was deployed
on the sediment drift for a period of 20 months. Using an Acoustic
Doppler Current Profiler (ADCP) equipped on the mooring sys-
tem, for the first time in the northern SCS, we have obtained the
long-term variability of the contour current velocity. The echo in-
tensities of the ADCP permit to assess the relative concentration of
suspended sediment in the water column (Deines, 1999; Gostiaux
and van Haren, 2010). Combined with the velocity dataset, we
have evaluated the capacity of sediment transport by the con-
tour currents, and further, the potential role of contour currents
in interpreting the formation of the sediment drift at southeast
of the Dongsha Islands. Through analyzing changes of the con-
tour currents during passing-through of a pair of well-investigated
deep-reaching mesoscale eddies, we have assessed the influence of
eddies on the contour currents.
2. Data and methods
A downward-looking Teledyne RDI’s 75-kHz Long Ranger ADCP
was mounted on a sub-surface mooring (TJ-A-1) in the northern
SCS (117.42◦ E, 20.05◦ N, water depth 2100 m) at southeast of the
Dongsha Islands (Fig. 1). Current velocity (zonal velocity u 0 and
meridional velocity v 0 ) in the depth range of 1685–2015 m was
obtained at a sampling interval of 1 h for a total calendar du-
ration of 635 days (27 August 2011 to 22 May 2013), excluding
a short pause of ∼11 h on 17 October 2012 when the instru-
ments were maintained for redeployment. The raw velocity data
were assessed and screened according to the thresholds created in
a QA/QC model built by the Teledyne RDI’s engineers (Symonds,
2006). All those data marked “bad” were removed to avoid possi-
ble masking of small fluctuations in the deep-sea currents.
Sub-inertial current velocities (u s and v s ) were obtained by
low-pass filtering the ADCP-derived velocity via a fourth-order But-
terworth filter with a cut-off frequency of 0.125 cycles per day. By
doing this, the influence of those processes with periods less than
8 calendar days, such as the tides, the internal waves as well as
the inertial flows caused by the Coriolis force, were eliminated. The
sub-inertial along-slope and cross-slope velocities (u and v) were
obtained by rotating the ENU (East–North–Up) coordinate system
adopted by the ADCP counterclockwise to parallel the northern SCS
shelf break, using the following geometrical relationships,
u = u s cos θ + v s sin θ
v = −u s sin θ + v s cos θ
where θ ≈ 28◦ is the acute angle between the shelf break and the
ENU coordinates (Y. Zhang et al., 2014). Positive/negative values of
the along-slope currents stand for northeastward/southwestward
flows, while positive and negative values of the cross-slope cur-
rents stand for onshore and offshore currents, respectively.
To investigate the potential impact of mesoscale eddies on the
bottom currents, a gridded satellite product of mean sea level
anomaly (MSLA) produced and distributed by AVISO (Archiving,
Validation and Interpretation of Satellite Oceanographic data, http:
//www.aviso.altimetry.com/) based on TOPEX/Poseidon, Jason-1,
479
ERS-1, and ERS-2 data (Ducet et al., 2000) was used to identify sur-
face mesoscale eddies in the study area. The merged MSLA prod-
ucts adopted herein are the global quarter-degree near-real-time
data with a temporal resolution of 1 day and a spatial resolution
of 0.25◦ at both latitude and longitude dimensions.
Acoustic echo intensity of the ADCP, a by-product of the ve-
locity measurements, reflects mostly the backscattering strength of
water, which depends highly on concentration of particles in wa-
ter and acoustic frequency of the ADCP (Deines, 1999). The volume
backscatter strength calculated from the ADCP echo intensities can
provide useful information on sediment delivery and planktonic
activities in the ocean (Gartner, 2004; Z. Wang et al., 2014). Herein,
the volume backscatter strength was calculated from the echo in-
tensities using the equation developed by Gostiaux and van Haren
(2010):
 
Sv = 10kc E /10 –10kc E r /10 × 102 log10 R +0.2α R − A /10
where Sv is the volume backscatter strength, E is the time- and
depth-dependent matrices of echo intensities and E r is the noise
level of echo intensities, R is the slant range of depth bins of the
ADCP, α is the attenuation coefficient of sound in water, kc is the
scale factor to convert echo intensity counts to decibels, and A is
a constant which is independent of time and space. The values
of α = 0.0245 dB/m, kc = 0.5 dB/counts and A = 105 dB, which
were suggested by Gostiaux and van Haren (2010) for the Long
Ranger ADCP, were adopted for calculation. Noticed that altering
of these values changes only the absolute values of Sv but not its
trend. In order to calibrate the values of Sv to be actual SSC (in
μg/L), selective suspended sediments were collected at the moor-
ing site at 1800 m and 2000 m water depth in April 2012 by the
Large Volume Water Transfer System (WTS-LV04) produced by the
McLane Research Laboratories. The linear relationship was attained
using the least square fitting method.
3. Results and discussions
3.1. Variations of the sub-inertial bottom currents
Variations of the sub-inertial along-slope current u and the
cross-slope current v at Station TJ-A-1 are shown respectively in
Fig. 2a and Fig. 2b. The velocity of the along-slope current u gen-
erally varies in the range of −5 to 3 cm/s, with the main direction
dominated by southwestward flows (inward the SCS). Northeast-
ward flows occurred in late February to mid-March, early Septem-
ber and late October 2012, and late February to mid-May 2013,
respectively. High-speed current events, both northeastward and
southwestward, were observed. Strong southwestward along-slope
currents up to −8 to −11 cm/s took place in February and August
2012, while strong northeastward currents at 5–6 cm/s were found
in early to mid-May 2013. The sub-inertial cross-slope currents
were almost perennial onshore-ward during the observed period
(Fig. 2b), with exceptional full-water column offshore-ward flow
occurred only in late November 2011, early May 2012 and early
September 2012. The magnitude of cross-slope current v is rela-
tively low, generally in the range −1 to 2.5 cm/s.
Correspondingly, magnitude and direction of the sub-inertial
currents are shown in Figs. 2c and 2d, where direction of the cur-
rents is defined as the clockwise angle from the true north. The
magnitude of the bottom current velocity generally varies in the
range of 0–2 cm/s. The period between late April and mid-August
2012 possessed relatively persistent high absolute current speed at
around 4–6 cm/s. Elevated velocity up to 8–11 cm/s occurred sev-
eral times, respectively during early November 2011, February and
August 2012. The dominant direction of the deep sub-inertial cur-
rent is ∼250◦ to the true north, i.e. southwestward. Reversals of
480 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485
Fig. 2. Variations of the sub-inertial (a) along-slope and (b) cross-slope currents, and the (3) magnitude and (4) direction of the sub-inertial currents at Station TJ-A-1. The
units of current direction are the clockwise degree from true north.
current direction to northeastward (∼70–80◦ ) were observed sev-
eral times, lasting from less than 10 days to 2–3 months.
3.2. Spatial and temporal variations of the contour currents
The sub-inertial along-slope current observed at Station TJ-A-1
is a quasi-persistent deepwater current that is parallel to the topo-
graphic isobaths. Such a flow is conceptually analogue to the con-
tour current in the sedimentological discipline, and thus is adopted
hereafter as an indicator of contour current in the northeastern
SCS. In order to have a detailed envisage of the vertical structure
of the contour current, monthly averages of the contour currents
from September 2011 to April 2013 were calculated (Fig. 3). In
general, we found that the spatial structure of the monthly av-
eraged contour current can be classified into three patterns. Pat-
tern I: southwestward currents are observed all through the wa-
ter column covered by the ADCP, with the maximum magnitude
of current velocity occurring at the topmost layer and decreasing
gradually downward (Fig. 4 left). The periods that can be assigned
to Pattern I include September 2011 to February 2012, and Novem-
ber 2012 to January 2013, with an exception in December 2011,
when the currents were the strongest at 1950 m water depth (ca.
−2 cm/s) and decreased slightly upward and downward (Fig. 3).
Pattern II: southwestward currents are also dominated all through
the observed water column, but the depth level with the maxi-
mum magnitude of current velocity occurs at the middle of the
observed water column and decreasing both upward and down-
ward (Fig. 4 middle). The depth level with the maximum current
velocity magnitude varies in the range of 1850–1950 m water
depth. The period from April to August 2012 can be classified into
Pattern II. In general, the maximum current velocity magnitude of
Pattern II is greater than that of Pattern I. For example, the maxi-
mum magnitude of the contour currents in May to July 2012 was
∼3.5–4.5 cm/s, and reached 6.2 cm/s in August 2012; while the
maximum magnitude of current velocity in Pattern I never ex-
ceeded 3 cm/s. Pattern III: northeastward currents are observed in
the upper part, while southwestward currents are observed in the
lower part of the water column, with reversal of the current direc-
tion occurring in the middle of the observed water column (Fig. 4
right). The magnitude of current velocity is usually the smallest
at the depth of direction reversal and increasing both upward and
downward. The depth at which the reversal of current direction oc-
curred varied greatly with time: it was at around 1900 m in March
and September 2012, March and April 2013, but at 1720 m in Oc-
tober 2012 and at 1820 m in February 2013 (Fig. 3). The months
that belong to this pattern include March, September and October
2012, and February to April 2013 (Fig. 3). The maximum magni-
tude of current velocity of Pattern III seldom reaches 1.5 cm/s, and
thus is the weakest in the three patterns.
In situ observation of the deep western boundary currents along
the North American margin shows that the deepwater currents
in the western North Atlantic have a tunnel-like structure, usu-
ally several tens of kilometers in width and hundreds of me-
ters in height, with the maximum velocity occurring in the cen-
ter of the “tunnel” and decreasing outwards (Toole et al., 2011;
Meinen et al., 2013). The deepwater current in the North Atlantic
is a steady or quasi-steady flow of water that runs roughly parallel
to the continental margins. Such a current resembles the concept
 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485 481

Fig. 3. Monthly average of the contour currents (in cm/s) at Station TJ-A-1 showing vertical structure of the contour currents. The blue dashed lines show the positions of
zero velocity. Positive values indicate northeastward currents while negative values indicate southwestward currents. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)

Fig. 4. A schematic map showing the spatial patterns of the contour currents. Rightward arrows show northeastward currents, while leftward arrows show southwestward
currents. The yellow arrows show patterns of the flows in the water column covered by the ADCP. The blue, red and green shaded areas show respectively the ranges of the
intermediate water (IW), the contour currents (CC), and the benthic boundary layer (BBL). (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)

of contour current (Rebesco et al., 2008; Zenk, 2008), and can be ilar spatial structure as those in the Atlantic, it well explains the
envisaged as an example of the contour currents. Despite defi- results observed in this study. The currents observed in Pattern II
ciency of the in situ observation of the bottom current in the SCS, have an ideal tunnel-like vertical structure, with the center of the
if the contour currents in the northeastern SCS possess the sim- “tunnel” corresponding to the depth level with the maximum mag-
482 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485
Fig. 5. Power spectral density of the contour current u. The power spectral densities were estimated using the Welch method and a Hamming window (length = 1280,
non-overlap = 640; NFFT = 5000) (Welch, 1967).
nitude of velocity, while the downward and upward decreasing of
velocity magnitude corresponding to the outward decreasing of ve-
locity from the center of the “tunnel”. As thickness and depth of
the center of the contour currents vary drastically with time, only
the lower part of the “tunnel” was observed in Pattern I. The cur-
rents observed in Pattern III are a combination of two processes.
The lower part of the pattern (the southwestward currents) corre-
sponds to the contour current, but with diminished thickness and
strength (Fig. 4). The upper part of Pattern III (the northeastward
currents) can be explained by a downward migration of the inter-
mediate water in the sandwich structure, the lower limit of which
corresponding to the upper limit of the contour currents (Fig. 4).
Combined together the results of the three patterns (Fig. 4), we
get two conclusions: (1) the contour currents observed at Station
TJ-A-1 is dominated by the southwestward, i.e. inward-SCS, flows;
and (2) the spatial distribution of the current velocity possesses
a tunnel-like structure, with both the thickness and the strength
of the current varying greatly with time. In Pattern I, the thick-
ness of the contour currents is very large, much larger than the
thickness of the observed water column. Relative to Pattern I, the
thickness of the contour currents is decreased in Pattern II, but
the magnitude of the current is largely increased. The magnitude
of the current can reach 4–6 cm/s in the center of the “tunnel”
(Fig. 3). In Pattern III, the thickness of the contour currents is
greatly decreased to less than 350 m, and the strength of the cur-
rent is the weakest in the three patterns (Fig. 4). Due to variation
of the thickness of the contour currents, the center of the “tunnel”,
i.e. the depth level with the maximum magnitude of velocity, also
varies greatly with time. In Pattern I, the center of the “tunnel” is
shallower than ∼1650 m, the upper limit of our observation. In
Pattern II, the center of the “tunnel” usually varies in the range
of 1850–1950 m water depth. The center of the “tunnel” for Pat-
tern III is the deepest in the three patterns, and is usually deeper
than 1950 m water depth. Overlaid on the contour currents is the
intermediate water of the northern SCS, which flows northeast-
ward, or outward the SCS (Fig. 4). Nonetheless, this water mass
was observed only in Pattern III, when the thickness of the con-
tour currents is the smallest.
In addition to their variations in spatial structure, temporal
variations of the contour currents is also remarkable. According to
the power spectral analysis, temporal variations of the contour cur-
rent velocity is dominated by periods of ∼3 months (85–105 days),
indicating significant inter-seasonal variability (Fig. 5). We thus ob-
tain the spring pattern, the summer pattern and the winter pattern
of the contour currents according to the monthly average of the
contour currents. The autumn pattern is difficult to illuminate us-
ing the present dataset because the autumns of 2012 and 2013
observed herein possess distinctive features from each other. Varia-
tion of the contour currents in autumn 2011 resembled the winter
pattern, while its variation in autumn 2012 was similar to the
spring pattern. The spring pattern (March 2012 and February to
April 2013) is generally corresponding to Pattern III of the contour
currents described above, with relatively thin vertical thickness
and low magnitude of velocity. In 2012, the spring pattern was
only observed in March, with April 2012 analogue to the summer
pattern and February 2012 analogue to the winter pattern. The
summer pattern is featured by a strong phase (max. 4.5–6.5 cm/s)
with a moderate thickness of ∼400–500 m. It can generally be rep-
resented by Pattern II of the contour currents described above. In
winter (November 2011 to February 2012 and November 2012 to
January 2013), the contour currents in the northeastern SCS are
characterized by very large thickness and moderate magnitude of
velocity. The winter pattern can be represented by Pattern I of the
contour currents.
Intra-seasonal variability of the contour currents at the peri-
ods of ∼45 days and ∼30–35 days were also observed (Fig. 5).
Distribution of the spectral density at these two period bands
displays similar vertical pattern. The most important ocean dy-
namic process that has a period of 30–45 days is the oceanic
mesoscale eddies, which are of great importance to the northern
SCS (G. Wang et al., 2003; Xiu et al., 2010). One pair of deep-
penetrating mesoscale eddies generated at southeast of Taiwan in
November 2011 led to an increase of mean (maximum) magni-
tude of velocity from 1.7 cm/s (4.5 cm/s) to 3.0 cm/s (6.5 cm/s) at
3525 m water depth in the northeastern SCS (Z. Zhang et al., 2013).
Passing-through of the anticyclonic and cyclonic eddies also caused
a temperature anomaly of 0.012 ◦ C and −0.017 ◦ C, respectively, at
2530 m water depth (Z. Zhang et al., 2013). This pair of eddies also
resulted in an increase of deepwater suspended sediment concen-
tration when they passed through Station TJ-A-1 (Y. Zhang et al.,
2014).
We find that this pair of well-investigated mesoscale eddies
have also great impact on the contour currents observed at Sta-
tion TJ-A-1 (Fig. 6). Based on the satellite-derived MSLA dataset,
we have determined, using the method developed by Chelton et
al. (2011), that sea surface of the studied site is affected by this
pair of mesoscale eddies between 1 February and 2 March, 2012
(Fig. 6b). As the anticyclonic eddy approached TJ-A-1 at the end of
January 2012, the southwestward currents at the full observed wa-
ter column increased rapidly to a magnitude of ∼10 cm/s (Fig. 6a).
After the core of the anticyclonic eddy passed the station, a re-
versal of current direction to northeastward occurred on around
21 February 2012. Noticeably, the current reversal in the observed
water column started almost simultaneously, but the process was
more rapid at deeper water levels: at 1700 m water depth, the
current direction gradually shifted clockwise from southwestward
to northeastward and the whole direction reversal spent ∼5 days;
at 1800 m water depth, the reversal process ended ∼1 day ear-
lier; while at 1900 m water depth, the current reversal was the
most straightforward and ended in less than 1 day (Fig. 6a). In
general, the current reversal observed at TJ-A-1 started when the
core of the anticyclonic eddy had passed the station at the sea
surface, and ended prior to the approaching of the core of the cy-
clonic eddy (Fig. 6). It was reported that this pair of eddies were
dominated by the first baroclinic mode, implying that the deep
and the surface currents in the eddies were in opposite direction
 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485
Fig. 6. Correlation between passage of the anticyclonic eddy (grey bar) and velocity of the bottom currents in the period from January to March 2012. (a) Vector plots of
sub-inertial currents at 1700, 1800, and 1900 m water depth, respectively. Modulus and direction of the vectors indicate the magnitude of the velocity and the direction of
the currents. (b) The satellite-derived MSLA during January to March 2012 showing passage of the anticyclonic eddy (grey bar); (c) Hovmöller diagram at 117.4◦ E showing
the latitudinal variation of MSLA (in centimeter) from January to March 2012. The dotted curve indicate the 0-cm contour of MSLA. The dashed line shows the latitude of
Station TJ-A-1.
(Z. Zhang et al., 2013). Thus the reversal of current direction from
southwestward to northeastward occurred on around 21 February
2012 can be explained by passing of the tail of the anticyclonic
eddy and the sequent passing of the front of the cyclonic eddy.
The causes for the temporal inconsistence of current reversal at
different water depth remains unclear and need further investiga-
tion, especially numerical modeling.
3.3. SSC in the water column and its role in the forming of the sediment
drift at southeast of the Dongsha Islands
The acoustic echo intensities measured by the ADCP may pro-
vide continuous and nonintrusive estimate of SSC profiles in vari-
ous environments (Gartner, 2004; Hoitink, 2004; Quaresma et al.,
2007; Traykovski et al., 2007). Using the equation developed by
Gostiaux and van Haren (2010), we obtained the volume backscat-
ter strength Sv to reflect the relative variation of the SSC at Station
TJ-A-1 during the observed interval. The values of Sv were then
calibrated to be actual SSC (in μg/L) using in situ filtered suspended
sediments at 1800 and 2000 m water depth at the mooring site
in April 2012. In general, the SSC at Station TJ-A-1 presents in the
forms of pulse-like high-SSC events (Fig. 7b). The SSC is usually the
highest at the bottom (>400 μg/L) and decreases upward along the
water column. One high-SSC event lasts several days to around one
month. Onsets of the high-SSC events usually occur at the bottom,
extending gradually to a certain depth range covering the low-
ermost 50–150 m of the water column. The depth range that is
affected by the high-SSC events starts to wane after reaching its
maximum, declining gradually to a range covering only the bottom
again (Fig. 7b). Two high-SSC events, one in mid-February to mid-
March 2012 and the other in mid-January to early February 2013,
cover the whole observed water column, both events correspond-
ing to passing-through of the mesoscale eddies (Y. Zhang et al.,
2014). We find that variations of the SSC are generally controlled
by the magnitude of the contour currents. Statistically, the high-
SSC events occurred at the intervals when the magnitude of the
483
contour currents was low (Fig. 7d, e). The SSC in the water profile
is determined by (1) the volume of sediment delivered from the
sources, (2) sinking of sediment particles from the overlaying wa-
ter column and/or (3) resuspension of sediment from the seafloor,
all three processes linking to variations of the contour currents.
However, the relationship between the magnitude of the contour
currents and the SSC in water column is complex. The magnitude
of the contour currents has to be high enough to transport sed-
iment from the source regions to the deep-sea basin, but strong
currents will prevent the sinking of sediment particles from the
water column. High magnitude of the contour currents also causes
resuspension of sediment from the seafloor, leading to increase of
SSC at the bottom of the water column. The finding that the high-
SSC events generally occurred during intervals of low magnitude of
the contour currents indicates that resuspension of particles from
the seafloor is not the main cause of those high-SSC events. Rather,
the high-SSC events are probably caused by the intensified sinking
of particles from the water column during intervals of relatively
weak contour currents. Direction of the contour currents can also
affect the SSC in the water column to some extent. We find that
the SSC is generally low during intervals of northeastward flows
(Fig. 7c, e). Because the Taiwanese rivers are the major sources of
fine sediment to the region (Liu et al., 2010), the low SSC during
northeastward flows can thus be explained by the decreased con-
tribution from the sources.
As the mooring is located exactly on the high-accumulation-
rate sediment drift, the SSC at Station TJ-A-1 can thus be employed
to evaluate the roles of contour currents in forming the drift.
The average LSR at the top and the middle of the drift reaches
56–58 cm/kyr in the Holocene (L. Wang et al., 1999; Bühring et al.,
2004). Given an average LSR of 57 cm/kyr and a dry bulk den-
sity of 0.6 g/cm3 (average value for the topmost 10 m of ODP
Site 1144, Shipboard Scientific Party, 2000), we calculate the sed-
iment accumulation on the sediment drift to ∼340 g/m2 /yr or
∼0.94 g/m2 /day. On the assumption that the overwhelming ma-
484 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485

Fig. 7. Correlation between the contour currents and the SSC in the water column. (a) Contour current velocity u; (b) the SSC (in μg/L) calculated from echo intensities of the
ADCP; (c) direction of the contour currents at 1900 m water depth; (d) magnitude of the contour currents velocity at 1900 m; (e) the vertical averaged SSC.

jority of sediment particles are accumulated in the observed water
column, we then obtain the minimum SSC requested for forming
the drift is ∼2.3 μg/L. This value is only one thirty-fourth of the
average SSC in the observed water column during the whole ob-
servational period (∼77.8 μg/L), implying that the SSC in the water
profile is much higher than requested for forming the sediment
drift. Considering that seawater is an open system, the high SSC
observed at Station TJ-A-1 is unlikely confined only to the ob-
servational site. However, the sediment drifts are formed only at
specific locations, e.g. southeast of the Dongsha Islands. Therefore,
we suggest that the SSC in water column is not the major factor
controlling the formation of the drifts. Rather, the major role that
the contour currents play is to transport sediment from its sources
to the deep-sea basin and in maintaining the SSC in the water
column. Thus, the suspended sediments can continuously deposits
from the water column to form the drift mostly by local morpho-
logical and hydrodynamic factors.
4. Conclusions
Based on the in situ observation undertaken by an ADCP
mounted on a sub-surface mooring in the northern SCS, we have
obtained records of the contour current velocity and the SSC in
water column (in μg/L) at a depth range of 1650 m to 2000 m.
This unprecedented integrated sub-surface mooring observation
provides us an invaluable chance to investigate the relationship
between distributions of the deep contour currents and sediment
delivery in the SCS, which is crucial to understand the formation
of the contourite drifts. Owing to the dataset, we find that the
magnitude of the bottom currents generally varies in the range
of 0–2 cm/s. The dominant direction of the currents is ∼250◦
to the true north, i.e. southwestward. The contour currents in
the northeastern SCS possess a tunnel-like vertical structure, with
the maximum magnitude of velocity corresponding to the cen-
ter of the “tunnel” and decreasing both downward and upward.
Both the strength and the thickness of the contour currents varied
greatly with time. The upper limit of the contour currents contacts
the lower limit of the intermediate water of the northern SCS.
Besides the spatial variations, the contour currents also display
great temporal variations. The most significant temporal variation
is the inter-seasonal changes. In spring, the contour currents are
characterized by very deep residence depth, very thin thickness
and relative low magnitude. Summer pattern of the contour cur-
rents is featured by extreme high southwestward velocity (max.
4.5–6.5 cm/s) with moderate thickness. In winter, the contour
currents are marked by very large thickness and moderate mag-
nitude. Besides the inter-seasonal variability, the contour currents
also show high power spectral density at the bands of ∼45 days
and ∼30–35 days. These periods are probably related to activi-
ties of the deep-reaching mesoscale eddies. Passing-through of the
 Y. Zhao et al. / Earth and Planetary Science Letters 430 (2015) 477–485
mesoscale eddies in early 2012 led to the occurrence of a high-
speed event and a sequent current reversal.
The SSC at Station TJ-A-1 is in general the highest at the bot-
tom (>400 μg/L) and decreases upward along the water profile.
Pulse-like high-SSC events are observed frequently during the ob-
servational interval. These high-SSC events are usually correlated
to periods of weak contour currents. The SSC in the water column
is much higher than requested for forming the sediment drift, in-
dicating that the major role that the contour currents play is to
transport sediment from sources to deep basin.
Acknowledgements
This work is supported by the National Natural Science Foun-
dation of China (91128206 and 41306041). The altimeter products
used in this study are produced by Ssalto/Duacs and distributed
by AVISO, with support from Cnes (http://www.aviso.altimetry.fr/
duacs/). Shaohua Zhao, Xiajing Li, Quan Chen, Qin Zhang, Ke Wen
and Yanli Li are appreciated for their assistance during mooring
deployment and recovery. Two anonymous reviewers are thanked
for their constructive comments.
References
Bühring, C., Sarnthein, M., Erlenkeuser, H., 2004. Toward a high-resolution stable
isotope stratigraphy of the last 1.1 m.y.: Site 1144, South China Sea. In: Prell,
W.L., Wang, P., Blum, P., Rea, D.K., Clemens, S.C. (Eds.), Proceedings of the Ocean
Drilling Program, Scientific Results, vol. 184. Ocean Drilling Program, College
Station, TX, pp. 1–29.
Chelton, D.B., Schlax, M.G., Samelson, R.M., 2011. Global observations of nonlinear
mesoscale eddies. Prog. Oceanogr. 91, 167–216.
Deines, K.L., 1999. Backscatter estimation using broadband acoustic Doppler current
profilers. In: Anderson, S.P., Terray, E.A., Rizoli White, J.A., Williams 3rd, J. (Eds.),
Proceedings of the IEEE Sixth Working Conference on Current Measurement.
IEEE, San Diego, pp. 249–253.
Ducet, N., Traon, P.Y.L., Reverdin, G., 2000. Global high resolution mapping of
ocean circulation from TOPEX/Poseidon and ERS-1 and -2. J. Geophys. Res. 105,
477–478.
Gartner, J.W., 2004. Estimating suspended solids concentrations from backscatter
intensity measured by acoustic Doppler current profiler in San Francisco Bay,
California. Mar. Geol. 211, 169–187.
Gostiaux, L., van Haren, H., 2010. Extracting meaningful information from uncali-
brated backscattered echo intensity data. J. Atmos. Ocean. Technol. 27, 943–949.
Heezen, B.C., Hollister, C.D., 1964. Deep sea current evidence from abyssal sedi-
ments. Mar. Geol. 1, 141–174.
Hoitink, A.J.F., 2004. Tidally-induced clouds of suspended sediment connected to
shallow-water coral reefs. Mar. Geol. 208, 13–31.
Howe, J.A., 2008. Methods for contourite research. In: Rebesco, M., Camerlenghi, A.
(Eds.), Contourites. Elsevier, Amsterdam, pp. 19–33.
Laj, C., Wang, P., Balut, Y., 2005. Les rapports de campagnes à la mer –
MD147/MARCO POLO-IMAGES XII. Institut Polaire Français. 43 pp.
Lan, J., Zhang, N., Wang, Y., 2013. On the dynamics of the South China Sea deep cir-
culation. J. Geophys. Res. 118, 1206–1210. http://dx.doi.org/10.1002/jgrc.20104.
Liu, Z., Colin, C., Li, X., Zhao, Y., Tuo, S., Chen, Z., Siringan, F.P., Liu, J.T., Huang, C.-Y.,
You, C.-F., Huang, K.-F., 2010. Clay mineral distribution in surface sediments
of the northeastern South China Sea and surrounding fluvial drainage basins:
source and transport. Mar. Geol. 277, 48–60.
Lüdmann, T., Wong, H.K., Berglar, K., 2005. Upward flow of North Pacific Deep Water
in the northern South China Sea as deduced from occurrence of drift sediments.
Geophys. Res. Lett. 32, L05614. http://dx.doi.org/10.1029/2004GL021967.
Meinen, C.S., Johns, W.E., Garzoli, S.L., van Sebille, E., Rayner, D., Kanzow, T., Baringer,
M.O., 2013. Variability of the Deep Western Boundary Current at 26.5◦ N during
2004–2009. Deep-Sea Res., Part 2, Top. Stud. Oceanogr. 85, 154–168.
485
Qu, T., Girton, J.B., Whitehead, J.A., 2006. Deepwater overflow through Luzon Strait.
J. Geophys. Res. 111, C01002. http://dx.doi.org/10.1029/2005jc003139.
Quaresma, L.S., Vitorino, J., Oliveira, A., da Silva, J., 2007. Evidence of sediment re-
suspension by nonlinear internal waves on the western Portuguese mid-shelf.
Mar. Geol. 246, 123–143.
Rebesco, M., Camerlenghi, A., Van Loon, A.J., 2008. Contourite research: a field in
full development. In: Rebesco, M., Camerlenghi, A. (Eds.), Contourites. Elsevier,
Amsterdam, pp. 3–10.
Sarnthein, M., Pflaumann, U., Wang, P.X., Wong, H.K. (Eds.), 1994. Preliminary Re-
port on SONNE-95 Cruise ‘Monitor Monsoon’ to the South China Sea. Berichte-
Reports, Geologisch-Paläeontologisches Institut der Universität Kiel, vol. 48.
225 pp.
Shao, L., Li, X., Wei, G., Liu, Y., Fang, D., 2001. Provenance of a prominent sediment
drift on the northern slope of the South China Sea. Sci. China, Ser. D, Earth
Sci. 44, 919–925.
Shao, L., Li, X., Geng, J., Pang, X., Lei, Y., Qiao, P., Wang, L., Wang, H., 2007. Deep
water bottom current deposition in the northern South China Sea. Sci. China,
Ser. D, Earth Sci. 50, 1060–1066.
Shipboard Scientific Party, 2000. Site 1144. In: Wang, P., Prell, W.L., Blum, P., et al.
(Eds.), Proceedings of the Ocean Drilling Program, Initial Reports, vol. 184. Ocean
Drilling Program, College Station, TX, pp. 1–97.
Stow, D.A.V., Hunter, S., Wilkinson, D., Hernández-Molina, F.J., 2008. The nature of
contourite deposition. In: Rebesco, M., Camerlenghi, A. (Eds.), Contourites. Else-
vier, Amsterdam, pp. 143–156.
Symonds, D.R., 2006. QA/QC Parameters for Acoustic Doppler Current Profilers. Tele-
dyne RD Instruments, San Diego, CA. 17 pp.
Tian, J., Yang, Q., Liang, X., Xie, L., Hu, D., Wang, F., Qu, T., 2006. Observation of
Luzon Strait transport. Geophys. Res. Lett. 33, L19607. http://dx.doi.org/10.1029/
2006GL026272.
Toole, J.M., Curry, R.G., Joyce, T.M., McCartney, M., Peña-Molino, B., 2011. Trans-
port of the North Atlantic Deep Western Boundary Current about 39◦ N, 70◦ W:
2004–2008. Deep-Sea Res., Part 2, Top. Stud. Oceanogr. 58, 1768–1780.
Traykovski, P., Wiberg, P.L., Geyer, W.R., 2007. Observations and modeling of wave-
supported sediment gravity flows on the Po prodelta and comparison to prior
observations from the Eel shelf. Cont. Shelf Res. 27, 375–399.
Wang, G., Su, J., Chu, P.C., 2003. Mesoscale eddies in the South China Sea ob-
served with altimeter data. Geophys. Res. Lett. 30. http://dx.doi.org/10.1029/
2003GL018532.
Wang, G., Xie, S.-P., Qu, T., Huang, R., 2011. Deep South China Sea circulation. Geo-
phys. Res. Lett. 38, L05601. http://dx.doi.org/10.1029/2010GL046626.
Wang, L., Sarnthein, M., Erlenkeuser, H., Grimalt, J., Grootes, P., Heilig, S., Ivanova,
E., Kienast, M., Pelejero, C., Pflaumann, U., 1999. East Asian monsoon climate
during the Late Pleistocene: high-resolution sediment records from the South
China Sea. Mar. Geol. 156, 245–284.
Wang, Z., DiMarco, S.F., Ingle, S., Belabbassi, L., Al-Kharusi, L.H., 2014. Seasonal and
annual variability of vertically migrating scattering layers in the northern Ara-
bian Sea. Deep-Sea Res., Part 1, Oceanogr. Res. Pap. 90, 152–165.
Welch, P.D., 1967. The use of Fast Fourier Transform for the estimation of power
spectra: a method based on time averaging over short, modified periodograms.
IEEE Trans. Audio Electroacoust. AU-15, 70–73.
Wong, H.K., Lüdmann, T., Wollschlager, M., 1994. Seismic reflection profiling at the
northern continental margin of the South China Sea (SONNE-95 Cruise). In:
Sarnthein, M., Pflaumann, U., Wang, P., Wong, H.K. (Eds.), Preliminary Report on
Sonne-95 Cruise “Monitor Monsoon” to the South China Sea. In: Geol.-Palaont.
Inst. Univ. Kiel Reports, vol. 68, pp. 46–53.
Xiu, P., Chai, F., Shi, L., Xue, H., Chao, Y., 2010. A census of eddy activities in
the South China Sea during 1993–2007. J. Geophys. Res. 115, C0312. http://
dx.doi.org/10.1029/2009JC005657.
Zenk, M., 2008. Abyssal and contour currents. In: Rebesco, M., Camerlenghi, A.
(Eds.), Contourites. Elsevier, Amsterdam, pp. 37–57.
Zhang, Y., Liu, Z., Zhao, Y., Wang, W., Li, J., Xu, J., 2014. Mesoscale eddies transport
deep-sea sediments. Sci. Rep. 4. http://dx.doi.org/10.1038/srep05937.
Zhang, Z., Zhao, W., Tian, J., Liang, X., 2013. A mesoscale eddy pair southwest
of Taiwan and its influence on deep circulation. J. Geophys. Res. 118, 1–16.
http://dx.doi.org/10.1002/2013JC008994.
Zhao, W., Zhou, C., Tian, J., Yang, Q., Wang, B., Xie, L., Qu, T., 2014. Deep water
circulation in the Luzon Strait. J. Geophys. Res. 119, 790–804. http://dx.doi.org/
10.1002/2013JC009587.