Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

YECSS4680_proof ■ 12 February 2015 ■ 1/14

Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Contents lists available at ScienceDirect

Estuarine, Coastal and Shelf Science journal homepage: www.elsevier.com/locate/ecss

Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea Q4

Wendong Fang a, Pu Guo a, *, Changjian Liu b, Guohong Fang c, Shujiang Li c a

Q1

State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 West Xingang Road, Haizhu District, Guangzhou 510301, China b South China Sea Marine Engineering Surveying Center, South China Sea Branch of State Oceanic Administration, 155 West Xingang Road, Haizhu District, Guangzhou 510300, China c Laboratory of Marine Science and Numerical Modeling, First Institute of Oceanography, State Oceanic Administration, 6 Xian-Xia-Ling Road, Hi-Tech Industry Park, Qingdao 266061, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 September 2014 Accepted 1 February 2015 Available online xxx

Sub-inertial currents (SICs) over the continental shelf of the northern South China Sea (NSCS) are investigated using the in-situ current observations of acoustic Doppler current profiler (ADCP) mooring arrays off the Pearl River Estuary in 2006 and 2007. The survey was carried out in four separate time periods: summer, winter, spring (before the onset of the southwesterly monsoon), and fall (after the establishment of the northeasterly monsoon). The observations showed that the current directions were generally along the shelf, consistent with the directions of monsoonal winds. The currents were also affected by a few of mesoscale eddy events. In summer 2006, the volume transport was northeastward with a mean magnitude of 1.4 Sv through a cross-shelf section from the site of the depth of 135 m to the coast; in winter 2006/2007, spring 2007 and fall 2007, the volume transports were all southwestward with magnitudes of 2.0, 2.1, and 0.9 Sv, respectively, through a cross-shelf section from the site of the depth of 290 m to the coast. The standard deviations of the SICs were generally smaller than the velocities of the mean currents, and the variability of SICs showed significant correlation with the local sea surface winds. No persistent counter-wind currents were observed in the study area during the fall and winter observational periods. © 2015 Published by Elsevier Ltd.

Keywords: current observations variability wind fields volume transport continental shelves

1. Introduction

Q2

The northern South China Sea (NSCS) (Fig. 1) has a broad continental shelf (shallower than 200 m) and a large deep basin (with the maximum depth greater than 4000 m). It is connected to the western Pacific Ocean to the east through the Luzon Strait where the Kuroshio intrudes. The monsoonal winds dominate over the NSCS, with the northeasterly in winter and the southwesterly in summer. The oceanic circulation over the continental shelf in the NSCS is under the influence of wind forcing, the Kuroshio intrusion, and the mesoscale processes (e.g., Guo et al., 1985; Qu, 2000; Yang et al., 2002; Metzger 2003; Centurioni et al., 2004; Chapman et al., 2004; Qu et al., 2004; Caruso et al., 2006). The influence of monsoonal winds on the ocean circulation over the continental

* Corresponding author. E-mail address: [email protected] (P. Guo).

shelf in the NSCS is evident: wind-forced upper layer currents primarily flow in the along-shelf direction, that is, mainly northeastward in summer and southwestward in winter. Guan (1978) proposed, however, that a counter-wind current flowing to the northeast existed in winter in the NSCS, and called it the South China Sea Warm Current (SCSWC) (see also Hu et al., 2000; Guan and Fang, 2006). Qiu et al. (1984) and Guo et al. (1985) proposed that the effect of the Kuroshio intrusion on the NSCS circulation is important and there is a southwestward current known as the SCS Branch of the Kuroshio (SCSBK) existing near the shelf break. It has been suggested in the following studies that the Kuroshio intrusion into the NSCS varies on seasonal timescales. In particular, the greatest strength of the intrusion occurs in winter, and can split into several components after entering the South China Sea. Some studies (e.g. Shaw, 1991; Xue et al., 2004; Wang et al., 2010) suggested that the SCSBK provides a source for the SCSWC. Chao et al. (1995) proposed that the SCSWC could appear only when the northeasterly winds are relaxed. Fang et al. (1998) believed that, in

http://dx.doi.org/10.1016/j.ecss.2015.02.001 0272-7714/© 2015 Published by Elsevier Ltd.

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 2/14

2

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Fig. 1. Map of the northern South China Sea. Locations of ADCP moorings are denoted by solid dots. Measurement was carried out in summer at stations SA, SB, and SC, and in winter, spring and fall at stations SD, SE, and SF. Two parallel lines indicate the cross-shelf sections for estimating volume transports; W1 and W2 indicate the crosspoints of these sections with the islands near the mainland coast (see text of Section 4.4 for details).

term of mean state, the SCSWC could not exist in the western part of the NSCS but might exist in the eastern part of the NSCS shelf. Zheng et al. (2006) and Fang et al. (2009) further speculated that the SCSWC is normally only present in the north-easternmost area of the NSCS. Therefore, the existence of the SCSWC in wintertime on the NSCS shelf is still an unsolved issue. There have been many studies on current variability based on mooring observations over the continental shelf/slope in the NSCS (e.g. Fang et al., 2000, 2005; Liang et al., 2005; Guo et al., 2012; Lee et al., 2012), which contribute to a better understanding of the physical processes on the NSCS continental shelf/slope regime. Observed near-inertial oscillations and their relation to passing typhoons or tropical storms were also reported in a number of studies (e.g. Chu et al., 2000; Sun et al., 2011; Liu et al., 2011). However, these investigations were focused on the near-inertial, internal tidal processes, and were mostly based on individual mooring observations. Our knowledge of the NSCS circulation is mainly based on numerical modeling (e.g. Shaw and Chao, 1994; Metzger and Hurlburt, 1996; Fang et al., 1996; Hsueh and Zhong, 2004; Xue et al., 2004; Wang et al., 2010), calculations of geostrophic current from water density, and individual (or shortterm) direct current observations (e.g. Guan, 1978: Qui et al. 1984; Guo et al., 1985). As compared to the other shelf areas such as the Taiwan Strait and the Korean Strait (Teague et al., 2002, 2003), reports on the low-frequency currents in the NSCS are rare, mainly due to the lack of long-term mooring observations. In particular, no observation-based estimates of volume transport along the shelf have been reported so far. To estimate the magnitudes of volume transport in various seasons, and to understand longer-term current structure and temporal variability over the continental shelf in the NSCS, in-situ observations were conducted in four seasons of 2006 and 2007 by an array of bottom mounted acoustic Doppler current profiler (ADCP) moorings. Based on these observations we are able to provide estimates of the along-shelf volume transport, and to reveal the current variability, as well as to examine the existence of the SCSWC in the area. Our study may contribute to the community for better understanding the current variability in the NSCS and particularly for model validation and development. This paper is organized as follows. Section 2 describes the data and analysis methods. Section 3 presents sea surface winds during

the measurement periods. Section 4 shows the SIC variability and estimated volume transport. Section 5 discusses the effect of mesoscale processes on the current variability. Section 6 provides a discussion and a summary. 2. Data and methods The current moorings were deployed southeast of the Pearl River Estuary, roughly in the middle of the NSCS shelf (see Fig. 1). Observations were conducted by the South China Sea Branch of the State Oceanic Administration of China in four periods: summer 2006, winter 2006/2007, spring 2007, and fall 2007. During each time period, an array consisting of three bottom-mounted RDI ADCPs was deployed across the shelf on the sea floor from southeast to northwest. In summer 2006, the depths of mooring sites (SA, SB, and SC in Fig. 1) were from 55 to 135 m. In the other three seasons the depths of mooring sites (SD, SE, and SF in Fig. 1) varied from 95 to 290 m. The horizontal distances of SA-SB and SB-SC were about 80 and 85 km, respectively, and those of SD
SE and SE-SF were about 37 and 64 km, respectively. Current profiles were recorded by ADCPs at 10-min time interval over nearly the full water column. The shortest continuous observational period, during which three ADCPs worked simultaneously, was 33 days (in summer), and the longest was 59 days (in winter). The ADCPs worked well and provided high quality current data. The mooring site depths, observational periods, and instrument configurations of the ADCPs are given in Table 1. The obtained current records were nearly complete and very little corrections were required. Few unrealistic extreme values in the raw data were first rejected and replaced by linearly interpolated values. The tidal analyses had been performed and the results have been reported in another paper by Guo et al. (2012). Since this work is aimed at studying the current variability with time scales longer than a few days, the high frequency currents, such as tidal and inertial currents, were removed from the current records by applying a fourth-order Butterworth low-pass filter. With the range of inertial frequencies at mooring sites (from 0.70 to 0.74 cpd) considered, the cutoff frequency of the filter was selected as 0.60 cpd, corresponding to the period of 40 h. The filtered currents are called sub-inertial currents (SICs).

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 3/14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

3

Table 1 Information about the ADCP moorings. ADCP Mooring sites

Water depth (m)

Start time (m-d-y)

End time (m-d-y)

Top bin depth (m)

Bottom bin depth (m)

Acoustic frequency (kHz)

SA SB SC SD

55 105 135 95

105

SF

290

09/09/2006 09/08/2006 09/08/2006 02/05/2007 05/07/2007 11/13/2007 02/05/2007 05/07/2007 11/17/2007 02/05/2007 04/27/2007 11/17/2007

4 11 15 8 11 11 9 10 11 53 26 25

48 91 111 84 87 87 93 94 95 269 266 265

300

SE

08/07/2006 08/06/2006 08/06/2006 12/08/2006 03/16/2007 10/09/2007 12/08/2006 03/16/2007 09/28/2007 12/08/2006 03/16/2007 09/28/2007

Bin size (m) 2 4 8 4

Sampling interval (min) 10

150

8 12

To reveal the scatter characteristics of observed current vectors, a principal axes of variance (PAV) analysis can be used, which produces a mean velocity and a standard deviation ellipse from a set of current observations as described by Emery and Thomson (2001). This method is often used to identify a new coordinate system in which most of the velocity fluctuations are along one axis (called major principal axis) and the remaining fluctuations are along the other one (called minor principal axis). In this process, both the maximum standard deviation in the major principal axis direction and the minimum standard deviation in the minor principal axis direction are obtained. The principal angles denoting the orientations of the two axes are also determined. Furthermore, the fluctuations can be presented as a standard deviation ellipse of which the major and minor axes correspond to the square root of the variances along the major and minor axes, respectively. Empirical orthogonal function (EOF) analysis is applied to the ADCP current decomposition. The EOF results will show the characteristics of vertical modes and the corresponding temporal fluctuations of SICs. Finally, the volume transports through the two sections shown in Fig. 1 are calculated for each measurement period.

shown in Fig. 2. It can be seen that the time-averaged sea surface winds over the study area were northward in the summer measurement period governed by southwesterly monsoon, and southwestward in all other three periods governed by northeasterly monsoon, with relatively weaker power in spring. Furthermore, the wind was weaker in summer period and stronger in winter and fall. In the study area the offshore winds were stronger than the near-shore winds, thus, the wind stress curls in winter, spring and fall were all anti-cyclonic, and those in summer were cyclonic. The magnitudes in winter and fall were about seven times larger than those in summer. The winds were strong and steady in winter and were weak and variable in summer (Fig. 3). The magnitudes of the mean wind vectors for summer, winter, spring, and fall observational periods were 1.4, 10.1, 4.4, and 9.5 m/s, respectively. The maximum standard deviations along the major principal axes of variance for the corresponding periods were 3.9, 1.9, 4.8, and 4.3 m/s, respectively. The corresponding directions of major principal axes of variance were 0.9,
79.2, 42.7, and 20.5 measured anticlockwise from east.

3. Sea surface winds during the measurement periods

4.1. Time series of sub-inertial currents

The primary forcing of the SCS circulation is the SCS monsoon, which is the central portion of the Asian-Australian monsoon. In winter, northeasterly winds prevail over the SCS, and the onset of the southwesterly monsoon is abrupt generally occurring in midMay (e.g. Wang et al., 2009). In contrast, the monsoon transition from southwesterly to northeasterly is generally gradually completed: the northeasterly winds prevails over the SCS roughly north of 19 N in late September, and roughly north of 10 N in late October. From November, the northeasterly monsoon begins to control the entire SCS. 2006e2007 was a weak El Nino year (the maximum NINO3.4 temperature anomaly was 1.3 , appearing in December 2006), so the climatological conditions during the observational period should be close to the normal conditions. From Table 1 we see that the summer observation was carried out before the establishment of northeasterly monsoon; the winter observation was conducted in the mature stage of northeasterly monsoon; the spring observation was performed before the onset of southwesterly monsoon, that is, in the declining stage of northeasterly monsoon; the fall observation was carried out in the beginning stage of northeasterly monsoon. The CCMP (Cross-calibrated, multiplatform ocean surface wind velocity product, see Atlas et al., 2011) winds downloaded at http:// apdrc.soest.hawaii.edu/datadoc/ccmp_6hourly.php are used in this study. The CCMP winds and deduced wind stress curls over the study area are averaged in time for each observational period, as

Daily SIC vectors of five layers during the measurement periods are plotted in Fig. 3. In summer, the currents at SA, which is located nearest to the coast, displayed the typical features of coastal currents in the NSCS under the influence of the summer monsoon. The currents went northeastward and fluctuated. The currents at midshelf station SB were mainly toward the northeast in the upper and middle layers, but tended to flow offshore in the deepest layer. The currents in the upper four layers of SC, which is located near shelf edge, were mainly northward during the first two-thirds of the measurement period, and then they turned southeastward during the last third of the measurement period. The currents at the deepest layer of this station were mostly southwestward, possibly indicating the existence of a subsurface current, originating from the Luzon Strait as proposed by Fang et al. (2009). The currents were stronger at SA than those at SB and SC, possibly as a result of the westward intensification of the summer western boundary current. In winter, due to the northeasterly winds, the SICs at SD and SE were mainly southwestward in the most of the observational period, and so did the upper layer currents at SF (see Fig. 3). The SICs in the deeper layers at SF changed direction frequently, indicating less influence of the wind forcing on the deeper layer currents. Northwestward or northward currents might occasionally occur at SF. In general, the southwestward currents at SD were weaker than those at SE and those in the upper layers (roughly

4. Sub-inertial currents and volume transports

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 4/14

4

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Fig. 2. Mean sea surface wind fields calculated from CCMP data during four observational periods. The contours indicate deduced wind stress curls (in 10
7 Pa m
1). The rectangular boxes denote the areas for calculating the area-averaged time series of wind velocity over the mooring sites.

upper 100 m) of SF. Some weak northward currents occasionally appeared at SD (mainly from 30 Dec 2006 to 7 Jan 2007), but no persistent counter-wind (e.g. northeastward) currents resembling the SCSWC were shown to exist. The synoptic event of weak northward currents was associated with the appearance of a mesoscale cyclonic eddy northwest of site SD (see Section 5 for further discussion). The tendency of stronger northward currents and weaker southwestward currents at SD relative to SE and SF may be partially attributed to the anti-cyclonic wind stress curl, as seen in Fig. 2. During the spring measurement period the northeasterly winds still prevailed, but they were weaker and more variable than in winter, and southward winds sometimes occurred (Fig. 3). The currents were basically southwestward at SF, more variable at SE, and much more variable at SD. In addition, the north to northeastward currents at SD were stronger and occurred more frequently, especially from 25 March to 5 April and from 15 April to the last day of the spring observational period. The maximum velocity of the northward current reached a maximum value of 19.5 cm/s. In fall, the southwestward currents prevailed in the area. Currents at SD and SE were quite uniform in the vertical, and they demonstrated significant fluctuations in time. The strong southwestward currents were measured around November 6, which was associated with a mesoscale eddy appearing in the southwestern Taiwan Strait, and extending the positive sea level anomaly southwestward to the measurement sites (see Section 5 and Fig. 9d

below). The maximum velocities of the SICs at the inner stations SD and SE were 38 and 56 cm/s, respectively, both flowing southwestward. The SICs at the shelf edge station SF were mainly westward in upper layers and weaker in deeper layers. Time series of depth-averaged currents (Fig. 4) show that the temporal variations of depth-averaged SICs at SD and SE were similar during the winter, spring, and fall periods. The depthaveraged SICs at SF were much smaller than those at SD and SE, resulted from weaker currents in deep layer. In fall period, the depth-averaged SICs at SE showed a sudden increase in southwestward transport on November 4 and 6. 4.2. Statistical features of the SIC Simple statistics is of fundamental importance in revealing the basic features of the current variability (Poulian and Niiler 1989; Teague et al., 2002). In the present study, we apply the PAV analysis described in Section 2 to the SIC observations. The results, including mean velocity, maximum and minimum standard deviation, and associated major principle direction for five equally spaced layers (Table 2a,2b,2c,2d), are shown as vertical profiles (Fig. 5) and vertical means (Fig. 6) respectively. The angle of the major principle axis has two values, with a difference of 180 from each other. Here we only give the one within ±90 . In summer, the temporal mean currents were strongest at the near-shore station SA, with velocities exceeding 20 cm/s in the upper half water column, and exceeding 10 cm/s in the lower layers

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 5/14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

5

Fig. 3. Daily sea surface winds and daily mean Sub-inertial currents at mooring sites during the four observational periods. Note that water depth H changes with a different mooring site (55, 105, and 135 m at SA, SB, and SC, respectively; 95, 105, and 290 m at SD, SE, and SF, respectively).

(Table 2a, Fig. 5). The mean currents at SB and SC were greater than 10 cm/s only in the surface or near-surface layers, and smaller than 10 cm/s in the other layers. The mean current directions were basically northeastward along the isobaths. It is worth noting, however, that the mean current at the lowest layer of the outer shelf site SC turned to southward, indicating the influence of the southwestward subsurface current (Fang et al., 2009). This influence could also be observed at the lowest layer of SB, though to a lesser extent. The maximum standard deviations were comparable in magnitude with the mean current velocities, and the associated major principle direction was basically along the shelf. The minimum standard deviations were smaller, but still had a definite

magnitude, indicating significant cross-shelf fluxes could be impacted by current fluctuations. The temporal mean currents observed in the fall, winter and spring periods were mainly toward the southwest along the shelf isobaths (Tables 2b, c, d, Figs. 5 and 6). In fall, when the northeasterly monsoon was well established, the mean currents were almost uniformly southwestward along the shelf. The mean current velocities were also quite uniform, centering around 15 cm/s at the layers shallower than about 100 m. In winter, the mean currents were also southwestward along the shelf. The current velocities became greater than in the fall at the two outer stations SE and SF, but they became smaller than in fall at the mid-shelf station SD,

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 6/14

6

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Fig. 3. (continued).

possibly indicating the effect of the pressure pattern that tends to drive a northeastward current (Hsueh and Zhong, 2004). In spring before the onset of the summer monsoon, the currents generally flowed southwestward with lower speeds than those observed in fall and winter, except that currents flowed basically northward with quite low speeds at SD (Table 2c). The standard deviations along the major principal axes in summer were generally smaller than mean current velocities at SA and SB, and they were greater than the mean current velocities at SC. The standard deviations along the major principal axes in winter were around 10 cm/s in upper layers and slightly less than 10 cm/s in lower layers. They were greater, up to near 15 cm/s at SE, in fall and spring. The directions of major principal axes were basically along the shelf. The standard deviations along the minor principal axes mostly had magnitudes of 40%e50% of those in the major principal axes directions at SD and SE, and up to about 80% at SF.

4.3. EOF analysis For each of the four measurement periods, EOF analysis is applied to all through-section (along-shelf) SICs at the three moorings (SA, SB and SC; or SD, SE and SF). The first EOF mode accounts for over 36.5%, 49.3%, 64.0%, and 67.0% of the total variance in the summer, winter, spring, and fall periods, respectively. Vertical patterns and temporal amplitudes of the EOF mode-1 are shown in Fig. 7. The along-shelf (normal to the sections) wind components are also plotted on the EOF amplitudes panels for comparison. It can be seen that the vertical structures of this mode at SD and SE in winter, spring and fall are quasi-barotropic; while those at SF and in summer show significant vertical variations. The temporal variations in EOF mode-1 have significant correlation with the along-shelf wind component: the correlation coefficients are 0.57,

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 7/14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

7

Fig. 4. Depth-averaged time series of zonal component (u) and meridional component (v) of sub-inertial currents. Blue, red, and green colors correspond to the currents at mooring site SA, SB, and SC in summer, respectively; or those at mooring site SD, SE, and SF in the other three seasons, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

0.52, 0.72, and 0.80 for summer, winter, spring and fall respectively. The analysis shows that the maximum correlation appears when current values lag behind winds by about 1 day for summer, winter and fall, and 3 days for spring. 4.4. Volume transport Volume transport values are of fundamental importance for understanding general ocean circulation. However, no observationbased volume transport estimates for the NSCS shelf area are available so far. In this study we set two parallel cross-shelf sections: Section 1 along the ADCP sites of SA, SB and SC; Section 2 along sites SD, SE, and SF, as shown in Fig. 1. The solid lines in Fig. 1 indicate the segments within the observational sites, while the dashed lines are extended from the solid lines to the points W1 and W2 (for Sections 1 and 2, respectively) at southeast coasts of two small islands near the mainland coast. The currents at the projection of SA and SB on Section 1 (denoted by SA0 and SB0 respectively) and the projection of SE on Section 2 (denoted by SE0 ) are assumed to be equal to those at SA, SB, and SE, respectively. The volume transport F can be estimated using the following formula:

Z F¼

A

un dA;

(1)

where A denotes the area of the vertical section, dA is the area element of the section, and un is the current component normal to the section (positive for northeastward). We divide F into two parts:

F ¼ Fo þ Fx ;

(2)

where Fo represents the volume transport through the solid segments in Fig. 1 (that is, from SC to SA0 for Section 1; from SF to SD for Section 2); Fx represents the volume transport through the dashed segments (that is, from SA0 to W1 for Section 1; from SD to W2 for Section 2). The values of un in solid segments are calculated through linear interpolation along terrain-following (i.e., sigma coordinate) levels, and those outside the ADCP sites to the coast (from SA0 to W1 for Section 1; from SD to W2 for Section 2) are calculated through linear interpolation by assuming that the currents at the coast (W1 and W2) are equal to zero. The obtained values of Fo can be regarded as the volume transport estimates on the basis of observations, and they are thus more reliable. The values of Fx are extrapolated from the observations at the innermost mooring station, and they are thus less reliable. Since F represents the volume

transport through the entire section from the shelf edge to the coast, it is thus more desired for understanding the transport on the NSCS shelf. In the calculation of Fx, the influence of the Pearl River discharge is neglected. The discharge rate is the largest in summer. The 12-year average rate from June to August is equal to 0.0165 Sv (Gan et al., 2009, Section 3.1). The rates for the rest seasons are much smaller. The time-mean current components normal to the sections and daily mean values of volume transport are shown in Fig. 8, with the wind velocity components normal to the sections also plotted for comparison. From 8 August to 6 September 2006 (representing summer), the southwesterly winds prevailed, and the transport was mostly to the northeast. The southwestward transport only occurred once around 11 August, when a relatively stronger southwestward wind appeared. Moreover, the northeastward currents in the near-shore area were usually stronger than those in the off-shore area. It is worth noting that a southwestward current was observed in the column below 100 m at SC (Fig. 8b), indicating possible existence of the subsurface current in summer suggested by Fang et al. (2009). During both observational periods from 9 December 2006 to 4 February 2007 (representing winter) and from 10 October to 12 November 2007 (representing fall) the northeasterly winds were quite strong, between 7 and 14 m/s, and the transports were always toward the southwest (Fig. 8c, g). It can also be seen that the strongest southwest currents tended to prevail near the shelf break area, with time-mean velocity exceeding 18 cm/s (Fig. 8d, h). In these two measurement periods, the counter-wind transport did not appear on the shelf southeast of the Pearl River Estuary. During the time period from 16 March to 26 April 2007 (representing spring) before the onset of the southwesterly monsoon, the northeasterly winds still prevailed, and the transport was mostly toward the southwest accordingly. The northeastward transport could appear when the winds reversed (Fig. 8e), and the currents in the upper layer on inner shelf tended to flow northeastward (Fig. 8f). From the calculated daily volume transport, we obtain the mean transport and standard deviation for each measurement period (Table 3). It can be seen that the transport was northeastward with a mean of 1.1 Sv in the segment from 135 to 55 m, and a mean of 1.4 Sv from the depth of 135 m to the coast in summer 2006. In winter 2006/2007 and fall 2007, the southwestward volume transport reached 1.6 Sv within the segment with depths from 290 to 95 m, and it reached 2.0e2.1 Sv on the shelf from the depth of 290 m to the coast. A mean rate of about 0.9 Sv was transported toward the southwest along the northern SCS shelf from a depth of 290 m to a depth of 95 m or to the coast in spring 2007.

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 8/14

8

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Fig. 5. Vertical profiles of mean velocity (top), standard deviation (middle), and major principal angle (bottom) of sub-inertial currents. In the top panel thick (thin) lines denote eastward (northward) components; in the middle panel thick (thin) lines denote maximum (minimum) standard deviation.

Regression analysis indicates that the deviations of the volume transport are closely correlated with the wind component normal to the section. The correlation coefficients between F and the normal wind component are 0.73, 0.55, 0.76, and 0.88 for summer, winter, spring and fall, respectively. 5. The effect of mesoscale processes Although winds were dominant in driving the along-shelf transports, the current records showed a few episodic events of mesoscale processes that cannot be explained by wind variability alone. These events were possibly related with water density variation. Since no simultaneous three-dimensional hydrographic

survey was carried out in the study area, here we employ a sea level anomaly (SLA) product Ssalto/Duacs, which were produced by Aviso, France based on satellite altimeter measurements, to examine the mesoscale processes. The data used here is the Duacs 2014 version which has one-day temporal resolution and 1/4 spatial resolution. The sea surface geostrophic current anomalies (GCA) are also available in the product. By visual inspection we find that the most distinguished mesoscale processes are eddy activities. In the following we discuss the influence of the most important eddy in each observational period on the flow field. At the beginning of the summer observational period (6 Aug 2006) a warm eddy appeared south by southeast of the measurement site SC, and moved westward slowly, causing northwestward

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 9/14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Fig. 6. Mean velocities and standard deviation ellipses of depth-averaged sub-inertial current at mooring sites. The mean velocities and deviation ellipses corresponding to upper 100 m averages at SF are also shown in a box on the right. The arrows pointing to the centers of ellipses indicate the time-and-depth-mean velocities. Violet, blue, red, and green colors correspond to the results in summer, winter, spring, and fall periods, separately. Bottom depths are in meters. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

9

On 27 October 2007, an anti-cyclonic eddy with a radius of about 90 km began to appear in the southwestern Taiwan Strait (not shown). The frontal edge of this eddy extended southwestward and gradually became a coastal band of high SLA with strong southwestward currents occurring at mid-shelf during 3e9 November 2007 (Fig. 9d). The southwestward currents began to weaken at SD and SE from 10 November 2007 due to the withdrawal of the coastal band of high SLA. The strengthened southwestward currents can also be seen in the current vector plots in Fig. 3. The relation between currents from ADCP observation and GCA from altimeter observation indicates that the mesoscale process is one of important factors that influence the observed flow variability. Thus we make multivariable (two-variable in the present case) regression analysis by taking into account of GCA time series. In this analysis the wind component normal to the observational sections (Fig. 1) is taken as the first independent variable as done above; the averaged value of GCA component normal to the observational sections at three observational sites is taken as the second independent variable. The multiple correlation coefficients obtained from the analysis are 0.70, 0.79, 0.91 and 0.94 for summer, winter, spring and fall, respectively, for EOF mode-1; and 0.77, 0.79, 0.90 and 0.96 respectively for volume transport. 6. Summary and discussion

GCA first and then turning to the northeast and further to the east subsequently. A typical SLA/GCA field on 17 August 2006 is shown in Fig. 9a. During 1e8 January 2007, a cyclonic eddy existed northwest of SD (Fig. 9b). The vertical profile of this eddy was captured well in current records at the mooring site SD. The current at SD turned northeastward against the northeasterly winds on 1 January and remained in this direction until 8 January (Fig. 3), during which the southeastern portion of the cyclonic eddy passed. This eddyassociated northeastward velocity dominated throughout the upper and middle water layers. In spring of 2007, a weak clockwise eddy started to form northeast of the site SD on 24 March, then slightly moved southward, and disappeared on 7 April. The joint effect of the weakened winds and eddy induced flow resulted in the northward currents at SD as seen in Fig. 3. Fig. 9c displays the SLA/GCA field on 2 April 2007.

Current observations with ADCP mooring arrays were carried out in four seasons separately during 2006e2007 on the NSCS shelf. The summer measurement period corresponded with the southwesterly monsoon phase; while the winter, spring, and fall observational periods corresponded with the mature, terminal, and early phases of the northeasterly monsoon, respectively. Overall, the current directions were consistent with the monsoon wind directions. The currents were also affected by a few of mesoscale eddies. In summer 2006, the currents were generally northeastward, with a mean volume transport of 1.1 Sv in the portion of depths from 135 to 55 m, and 1.4 Sv from a depth of 135 m to the coast. Based on this, we anticipate that the summer volume transport will eventually join the Taiwan Strait through flow, which has a maximum volume transport of 2.5e3.1 Sv in this season (Fang et al., 1991; Wang et al., 2003; Hu et al., 2010).

Fig. 7. The EOF mode-1 of the through-section currents and the along-shelf wind speed (denoted with red curves) in four measurement periods. Blue lines denote mode structures in the left-hand panels and amplitudes in the right-hand panels. The percentages of variance contained in this mode are also labeled. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 10/14

10

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Fig. 8. Volume transport time series and mean through-section current velocities in summer (a and b), winter (c and d), spring (e and f), and fall (g and h). The summer measurement was conducted along Section 1, and the winter, spring, and fall measurements were done along Section 2 (see Fig. 1). In panels a, c, e, and g, the thick solid curves denote F0: volume transports through the section segments within observational stations; the thick dashed curves denote F: those through the entire sections; the thin solid curves denote Wu: wind component normal to the sections.

From fall to spring (before onset of the summer monsoon), the currents were normally southwestward. In fall 2007 and winter 2006/2007, the southwestward along-shelf volume transport reached about 1.6 Sv through the section with depths ranging from 290 to 95 m, and it reached 2.0e2.1 Sv from 290 m to the

coast. The corresponding volume transports in spring 2007 were also southwestward with relatively smaller magnitude (about 0.9 Sv) due to the weakening of the monsoon winds. The direct current measurement in September 28 e December 14, 1999 in the Taiwan Strait by Teague et al. (2003) showed that the mean

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 11/14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

11

Fig. 9. Sea level anomalies (m) and geostrophic velocity anomalies (cm/s) showing the episodic appearances of mesoscale processes. (a) 17 August 2006, (b) 6 January 2007, (c) 2 April 2007, and (d) 6 November 2007 (Data are available from http://www.aviso.altimetry.fr/duacs/).

northward volume transport was 0.62 Sv through the eastern strait, and the southward mean volume transport was 0.49 Sv along the western strait. We can thus speculate that the southwestward transport of about 2 Sv on the NSCS shelf during northeasterly monsoon should originate partially (roughly one quarter) from the Taiwan Strait, and mostly (roughly three quarters) from the Luzon Strait (e.g., Xue et al., 2004; Fang et al., 2009).

In the near-shore region, freshwater discharge from the Pearl River can influence the coastal current, especially in summer. Since the mean discharge rate in summer is only 0.0165 Sv, its direct influence on volume transport estimates is small. However, the indirect influence (i. e., the buoyancy effect) is an issue subject to further investigation. In addition to the relatively steady mean currents, the observed currents displayed significant variability. The standard deviations

Table 2a Current statistics: summer. Station

Depth

SA

0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical

SB

SC

H H H H H mean H H H H H mean H H H H H mean

u

su

v

sv

V

w

smax

j

smin

Vmax

Dmax

20.1 19.8 17.9 14.8 9.3 16.4 13.9 10.7 4.3 2.8 4.9 7.3 11.1 8.5 4.5 0.6
0.2 4.9

19.3 6.9 5.9 6.4 6.3 7.7 9.3 8.6 7.5 6.6 4.4 5.7 13.1 12.3 9.6 8.5 6.7 9.3

11.2 10.3 10.4 9.5 8.8 10.0 1.3 2.3 5.0 3.0
1.1 2.1
0.2 1.6 2.5 0.2
2.5 0.3

8.9 5.9 5.5 5.1 6.8 4.3 4.8 3.5 4.5 5.2 4.7 2.6 11.6 11.3 8.9 7.0 6.3 7.6

23.0 22.3 20.7 17.6 12.8 19.2 14.0 10.9 6.6 4.1 5.1 7.6 11.1 8.6 5.1 0.6 2.5 4.9

29 27 30 33 43 31 5 12 49 47
13 16
1 11 29 23
94 4

19.5 7.4 6.8 7.2 8.6 8.1 9.4 8.6 7.5 6.7 4.8 5.7 15.1 14.5 10.3 8.6 8.1 9.6


9 32 41 33 48 19
11
4
5 14 66 1
38
40
36 18 42
24

8.5 5.2 4.4 3.9 3.4 3.5 4.6 3.5 4.5 5.0 4.3 2.6 8.8 8.3 8.1 6.8 4.4 7.1

58.3 35.9 37.0 34.5 30.3 33.1 39.0 29.9 21.8 30.5 25.6 19.9 32.4 31.0 21.6 24.8 23.7 19.9

20 40 31 14 41 24 3
3 9 22
36 4
29
35
162
144
152
52

u e mean zonal component of currents (cm/s, the positive is eastward), su e standard deviation of zonal component of current, v e mean meridional component of current (cm/ s, the positive is northward), sv e standard deviation of meridional component of current, V e mean current speed (cm/s), w e mean current direction (degrees, measured anticlockwise from east), smax e standard deviation in major principal direction (cm/s), j e major principal direction (degrees, measured anticlockwise from east), smin e standard deviation in direction perpendicular to major principal direction (cm/s). Vmax e maximum current speed, Dmax e directions of maximum current, H e water depth (see Table 1).

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 12/14

12

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Table 2b Current statistics: winter. Station

Depth

SD

0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical

SE

SF

H H H H H mean H H H H H mean H H H H H mean

u

su

v

sv

V

w

smax

j

smin

Vmax

Dmax


13.9
7.7
6.6
6.7
7.2
8.4
18.7
16.5
15.3
13.8
9.8
14.8
16.9
17.6
9.6
4.7
0.6
9.9

9.5 10.5 9.6 8.2 6.4 8.4 11.4 10.4 9.5 7.9 6.5 8.6 8.0 7.4 8.3 8.3 5.7 5.3


6.0
3.1
4.0
6.2
8.8
5.6
7.2
7.1
8.1
10.0
13.3
9.1
8.8
9.8
6.3
1.3
2.2
5.7

5.7 5.7 4.9 4.8 6.1 4.6 7.1 6.7 7.0 7.1 6.3 6.1 9.8 9.6 7.9 5.2 4.5 5.2

15.2 8.3 7.7 9.1 11.4 10.1 20.0 17.9 17.3 17.0 16.5 17.4 19.1 20.1 11.5 4.9 2.3 11.4


157
158
149
137
129
146
159
157
152
144
127
148
153
151
147
165
104
150

9.6 10.5 9.6 8.5 7.8 8.5 11.7 10.6 9.6 8.3 7.2 8.9 9.8 10.2 9.8 8.4 6.1 6.3


10 2 5 17 43 13 16 16 11 30 42 20 88 65 42 8 27 44

5.6 5.7 4.9 4.3 4.2 4.3 6.5 6.3 6.9 6.5 5.5 5.5 8.0 6.6 6.0 5.1 4.0 4.0

35.4 31.1 29.4 29.8 30.7 28.9 53.1 45.6 38.5 38.3 35.8 39.6 40.5 45.3 48.8 27.6 16.2 31.1

148
160
160
145
130
154
165
161
151
151
143
155
141
138
142
149
117
140

along the major principal axes were generally smaller than the corresponding mean current speeds, showing the steadiness of the currents. However, in a few cases (such as at SC in summer and at SD in spring) when mean currents were relatively weak, the

standard deviations exceeded the mean current speeds. The EOF and volume transport results indicate that the wind forcing is a dominant dynamic factor inducing the SIC variability. The temporal variations in EOF mode-1 and volume transports are strongly

Table 2c Current statistics: spring. Station

Depth

SD

0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical

SE

SF

u

su

v

sv

V

w

smax

j

smin

Vmax

Dmax

H H H H H mean H H H H H mean H H H H H mean


0.3 0.9 0.6
0.5
0.3 0.1
9.4
7.8
8.6
6.8 0.8
6.4
16.3
14.9
10.5
8.4
4.5
10.9

11.4 11.3 8.9 7.4 5.7 8.3 14.4 12.4 10.5 8.3 7.5 10.0 7.5 7.3 7.5 6.2 5.3 4.4

5.9 5.4 3.3 1.7
0.6 3.1
4.4
3.4
3.1
1.7
2.6
3.0
7.0
3.7
3.1
4.3
3.0
4.2

6.8 7.2 6.3 7.3 6.9 6.1 8.4 7.2 8.2 10.0 9.7 7.0 5.6 6.1 7.3 5.7 4.1 3.2

5.9 5.5 3.3 1.8 0.6 3.1 10.4 8.5 9.2 7.0 2.7 7.1 17.7 15.3 11.0 9.4 5.4 11.7

93 80 80 107
113 88
155
157
160
166
72
154
157
166
164
153
147
159

11.7 11.9 9.8 9.7 8.3 9.5 14.7 12.8 11.8 12.2 11.7 11.6 7.7 7.6 8.1 6.3 5.5 4.4

16 21 29 44 53 32 13 17 33 52 54 32 16
24
39
20 22
7.7

6.2 6.3 4.8 4.0 3.4 3.9 7.9 6.5 6.1 4.5 3.7 3.9 5.4 5.7 6.7 5.6 3.8 3.2

24.7 38.6 28.9 28.6 31.5 27.8 44.6 45.4 39.3 37.4 40.0 36.8 39.3 35.3 32.0 32.0 17.5 25.2


173
168
157
127
116
145 176
173
122
132
126
158
161
133
132
128
139
140

u

su

v

sv

V

w

smax

j

smin

Vmax

Dmax

H H H H H mean H H H H H mean H H H H H mean


12.6
12.1
11.3
10.1
6.4
10.5
14.2
12.9
14.4
11.4
4.7
11.5
16.2
14.2
5.3
5.0
3.2
8.8

5.3 8.6 7.9 7.0 5.5 6.4 11.6 11.2 10.0 9.2 8.0 9.3 7.5 5.2 4.0 4.3 4.2 3.4


7.3
8.2
8.4
9.4
11.8
9.0
6.0
5.8
6.5
10.1
12.6
8.2 1.8
3.0
1.3
3.3
2.7
1.7

7.8 8.2 7.0 7.5 8.1 7.1 11.6 11.1 12.2 12.6 10.6 10.5 6.6 6.3 4.4 4.0 3.0 2.9

14.5 14.6 14.0 13.8 13.4 13.8 15.5 14.1 15.8 15.2 13.5 14.1 16.3 14.5 5.4 6.0 4.2 8.9


150
146
143
137
119
139
157
156
156
138
110
145 174
168
166
146
140
169

8.3 10.8 9.5 9.3 9.3 8.9 14.5 14.4 14.3 14.5 12.7 13.3 7.5 6.8 4.9 4.8 4.4 3.8

64 43 39 48 59 49 45 45 54 57 55 49 13 59 55 40 19 35

4.4 4.9 4.5 4.4 3.2 3.4 7.7 6.4 6.6 5.7 3.9 4.3 6.6 4.5 3.4 3.4 2.8 2.4

24.4 36.7 34.3 32.5 32.0 32.0 53.4 52.0 48.9 48.7 46.7 47.9 33.0 36.9 26.8 16.7 14.5 20.3


146
134
134
132
126
133
160
158
116
121
124
118 150
129
128
144
153
140

Table 2d Current statistics: fall. Station

Depth

SD

0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical 0.0e0.2 0.2e0.4 0.4e0.6 0.6e0.8 0.8e1.0 Vertical

SE

SF

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

YECSS4680_proof ■ 12 February 2015 ■ 13/14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14 Table 3 Estimated volume transport through Section 1 in summer, and Section 2 in winter, spring, and fall. Season

Mean volume transport ¶Fo with standard deviation, Sv

Summer* Wintery Springz Fallx

1.05
1.57
0.92
1.55

± ± ± ±

0.55 0.65 0.79 0.85

Mean volume transport jjFx with standard deviation, Sv 0.34
0.38 0.06
0.54

± ± ± ±

0.12 0.27 0.37 0.35

Mean volume transport **F with standard deviation, Sv 1.39
1.95
0.86
2.09

± ± ± ±

0.63 0.86 1.14 1.15

*

Observational period: Aug 8 e Sep 6, 2006. Dec 9, 2006eFeb 4, 2007. Mar 16 e Apr 26, 2007. x Oct 10 e Nov 12, 2007. ¶ Fo represents the volume transport through the section segment within ADCP stations (solid line in Fig. 1). jj Fxthe volume transport through the section segment outside ADCP stations (dashed line in Fig. 1). ** F the volume transport through the entire section (solid plus dashed line in Fig. 1). y z

correlated with the along-shelf wind component, and the correlation coefficients can be further increased when the effect of mesoscale processes, especially eddies, is taken into consideration. The time series show that no persistent counter-wind currents were observed in either fall or winter. This indicates that there was no SCSWC passing through the study area during these northeasterly monsoon periods. The observational results presented here can be used for model verification in the future numerical investigations of the NSCS circulation, including the SCSWC. Q3

Uncited references Chiang et al., 2008; Fang et al., 2010; Gawarkiewicz et al., 2004. Acknowledgments This work was supported by the National Science Foundation of China (No.41176025, 40876008) and Project 908 (No. 908-01-ST07). G. Fang and S. Li are supported by the National Basic Research Program grant 2011CB403502, and the International Cooperation Program of China grant 2010DFB23580. The authors wish to thank the investigators, Captains and crew members for the productive cruises of mooring deployments and recoveries. The altimeter product used in this study was produced by Ssalto/Duacs and distributed by Aviso, with support from Cnes (http://www.aviso. altimetry.fr/duacs/). We wish to thank Dr. Courtney Harris and two anonymous reviewers for their useful comments in improving our manuscript. References Atlas, R., Hoffman, R.N., Ardizzone, J., Leidner, S.M., Jusem, J.C., Smith, D.K., Gombos, D., 2011. A cross-calibrated, multiplatform ocean surface wind velocity product for meteorological and oceanographic applications. Bull. Am. Meteorol. Soc. 92, 157e174. http://dx.doi.org/10.1175/2010BAMS2946.1. Caruso, M.J., Gawarkiewicz, G.G., Beardsley, R.C., 2006. Interannual variability of the Kuroshio intrusion in the South China Sea. J. Oceanogr. 62, 559e575. Chao, S.Y., Shaw, P.T., Wang, J., 1995. Wind relaxation as a possible cause of the South China Sea Warm Current. J. Oceanogr. 51, 111e132. Chapman, D.C., Ko, D.S., Preller, R.H., 2004. High-resolution numerical modeling study of subtidal circulation in the Northern South China Sea. IEEE J. Ocean. Eng. 29, 1087e1104. Centurioni, L.R., Niiler, P.P., Lee, D.-K., 2004. Observations of inflow of Philippine sea water into the South China sea through the Luzon Strait. J. Phys. Oceanogr. 34, 113e121. Chiang, T.L., Wu, C.R., Chao, S.R., 2008. Physical and geographical origins of the South China Sea warm current. J. Geophys. Res. 113, C08028. http://dx.doi.org/ 10.1029/2008JC004794.

13

Chu, P.C., Veneziano, J.M., Fan, C., Carron, M.J., Liu, W.T., 2000. Response of the South China sea to tropical Cyclone Ernie 1996. J. Geophys. Res. 105 (C6), 13991e14009. Emery, W.J., Thomson, R.E., 2001. Data Analysis Methods in Physical Oceanography. Elsevier. Fang, G.H., Zhao, B.R., Zhu, Y.H., 1991. Water volume transport through the Taiwan Strait and the continental shelf of the East China Sea measured with current meters. In: Oceanography of Asian Marginal Seas. Elsevier Oceanography Series, vol. 54, pp. 345e358. Fang, G., Fang, W., Fang, Y., Wang, K., 1998. A survey of studies on the South China Sea upper ocean circulation. Acta Oceanogr. Taiwanica 37, 1e16. Fang, G., Wang, Y., Wei, Z., Fang, Y., Qiao, F., Hu, X., 2009. Interocean circulation and heat and freshwater budgets of the South China Sea based on a numerical model. Dyn. Atmos. Oceans 47, 55e72. http://dx.doi.org/10.1016/ j.dynatmoce.2008.09.003. Fang, G., Susanto, R.D., Wirasantosa, S., Qiao, F., Supangat, A., Fan, B., Wei, Z., Sulistiyo, B., Li, S., 2010. Volume, heat and freshwater transports form the South China Sea to Indonesian seas in the boreal winter of 2007e2008. J. Geophys. Res. 115, C12020. http://dx.doi.org/10.1029/2010JC006225. Fang, W., Shi, P., Mao, Q., Gan, Z., 2000. Upper-ocean variation in the northern South China Sea from mooring station observations. Acta Oceanologica Sinica 19 (4), 22e35. Fang, W., Shi, P., Long, X., Mao, Q., 2005. Internal solitons in the northern South China Sea from insitu observations. Chin. Sci. Bull. 50 (15), 1627e1631. Fang, Y., Fang, G., Yu, K., 1996. ADI barotropic ocean model for simulation of Kuroshio intrusion into China southeastern waters. Chin. J. Oceanol. Limnol. 14, 357e366. Gan, J., Li, L., Wang, D., Guo, X., 2009. Interaction of a river plume with coastal upwelling in the northeastern South China Sea. Cont. Shelf Res. 29, 728e740. Gawarkiewicz, G., Wang, J., Caruso, M., Ramp, S.R., Brink, K.H., Bahr, F., 2004. Shelf break circulation and thermohaline structure in the Northern South China Seadcontrasting spring conditions in 2000 and 2001. IEEE J. Ocean. Eng. 29, 1131e1143. Guan, B.X., 1978. The warm current in the South China Seada current flowing against the wind in winter in the open sea off Guangdong Province. Oceanol. Limnol. Sinica 9, 117e127 (in Chinese with English abstract). Guan, B., Fang, G., 2006. Winter counter-wind currents off the Southeastern China coast: a review. J. Oceanogr. 62, 1e24. Guo, P., Fang, W., Liu, C., Qiu, F., 2012. Seasonal characteristics of internal tides on the continental shelf in the northern South China Sea. J. Geophys. Res. 117, C04023. Guo, Z., Yang, T., Qiu, D., 1985. The South China sea warm current and the SW-ward current on its right side in winter. Trop. Oceanol. 4, 1e9 (in Chinese with English abstract). Hsueh, Y., Zhong, L., 2004. A pressure-driven South China Sea warm current. J. Geophys. Res. 109, C09014. http://dx.doi.org/10.1029/2004JC002374. Hu, J., Kawamura, H., Hong, H., Qi, Y., 2000. A review on the currents in the South China Sea: seasonal circulation, South China Sea warm current, and Kuroshio intrusion. J. Oceanogr. 56, 607e624. Hu, J.Y., Kawamura, H., Li, C.Y., Hong, H.S., Jiang, Y.W., 2010. Review on current and seawater volume transport through the Taiwan Strait. J. Oceanogr. 66, 591e610. Lee, I.H., Wang, Y.H., Yang, Y., Wang, D.P., 2012. Temporal variability of internal tides in the northeast South China Sea. J. Geophys. Res. 117, C02013. Liang, X.F., Zhang, X.Q., Tian, J.W., 2005. Observation of internal tides and nearinertial motions in the upper 450 m layer of the northern South China Sea. Chin. Sci. Bull. 50, 2890e2895. Liu, J.L., Cai, S.Q., Wang, S.A., 2011. Observations of strong near-bottom current after the passage of Typhoon Pabuk in the South China Sea. J. Mar. Syst. 87, 102e108. Metzger, E.J., Hurlburt, H., 1996. Coupled dynamics of the South China sea, Sulu sea, and the Pacific ocean. J. Geophys. Res. 101, 12331e12352. Poulain, P.M., Niiler, P.P., 1989. Statistical analysis of the surface circulation in the California current system using satellite-tracked drifters. J. Geophys. Res. 19, 1588e1603. Qiu, D.Z., Yang, T.H., Guo, Z.X., 1984. A west-flowing current in the northern part of the South China Sea in summer. Trop. Oceanol. 3, 65e73 (in Chinese with English abstract). Qu, T., 2000. Upper layer circulation in the South China Sea. J. Phys. Oceanogr. 30, 1450e1460. Qu, T., Kim, Y.Y., Yaremchuk, M., Tozuka, T., Ishida, A., Yamagata, T., 2004. Can Luzon Strait transport play a role in conveying the impact of ENSO to the South China Sea? J. Clim. 17 (18), 3643e3656. Shaw, P.T., 1991. The seasonal variation of the intrusion of the Philippine Sea Water into the South China Sea. J. Geophys. Res. 96, 821e827. Shaw, P.T., Chao, S.Y., 1994. Surface circulation in the South China sea. Deep-Sea Res. 41, 1663e1683. Sun, Z.Y., Hu, J.Y., Zheng, Q.A., Li, C.Y., 2011. Strong near-inertial oscillations in geostrophic shear in the northern South China Sea. J. Oceanogr. 67, 377e384. Teague, W.J., Jacobs, G.A., Perkins, H.T., Book, J.W., Chang, K.I., Suk, M.S., 2002. Low frequency current observations in the Korea Strait. J. Phys. Oceanogr. 32, 1621e1641. Teague, W.J., Jacobs, G.A., Ko, T.Y., Chang, K.I., Suk, M.S., 2003. Connectivity of the Taiwan, Cheju, and Korea Straits. Cont. Shelf Res. 23, 63e77. Wang, B., Huang, F., Wu, Z., Yang, J., Fu, X., Kikuchi, K., 2009. Multi-scale climate variability of the South China Sea monsoon: a review. Dyn. Atmos. Oceans 47, 15e37. http://dx.doi.org/10.1016/j.dynatmoce.2008.09.004.

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6

YECSS4680_proof ■ 12 February 2015 ■ 14/14

14

W. Fang et al. / Estuarine, Coastal and Shelf Science xxx (2015) 1e14

Wang, D., Hong, B., Gan, J., Xu, H., 2010. Numerical investigation on propulsion of the counter-wind current in the northern South China Sea in winter. Deep-Sea Res. I. http://dx.doi.org/10.1016/j.dsr.2102.06.07. Wang, Y.H., Jan, S., Wang, D.P., 2003. Transports and tidal current estimates in the Taiwan Strait from shipboard ADCP observations (1999e2001). Estuar. Coast. Shelf Sci. 57, 193e199. Xue, H., Chai, F., Pettigrew, N., Xu, D., Shi, M., Xu, J., 2004. Kuroshio intrusion and the circulation in the South China Sea. J. Geophys. Res. 109, C02017. http:// dx.doi.org/10.1029/2002JC001724.

Yang, H., Liu, Q., Liu, Z., Wang, D., Liu, X., 2002. A general circulation model study of the dynamics of the upper ocean circulation of the South China Sea. J. Geophys. Res. 107 (C7), 3085. http://dx.doi.org/10.1029/2001JC001084. Zheng, Q., Fang, G., Song, Y.T., 2006. Introduction to special section: dynamic processes and circulation in Yellow sea, East China sea, and South China sea. J. Geophys. Res. 111, C11S01. http://dx.doi.org/10.1029/2005JC003261.

Please cite this article in press as: Fang, W., et al., Observed sub-inertial current variability and volume transport over the continental shelf in the northern South China Sea, Estuarine, Coastal and Shelf Science (2015), http://dx.doi.org/10.1016/j.ecss.2015.02.001

7 8 9 10 11 12