Near-inertial oscillations in the northern South China Sea, close to the Xisha Islands, during the passage of typhoon Conson, were investigated using 52-day observational data and a damped slab model. Using spectral analyses, we found that these near-inertial oscillations were dominated by clockwise components. The observed, as well as simulated, inertial currents matched well before and after the passage of the typhoon, which revealed that significantly enhanced oscillations were induced by the typhoon. They have a peak frequency of 0.0237 cycles per hour, which is 2% lower than the local inertial frequency, meaning a red-shift phenomenon exists, even in shallow water. The red-shift might be attributed to the background negative vorticity. Additionally, we found that the barotropic near-inertial currents were more significant after the passage of the typhoon. The position of the maximum intensity of near-inertial current shear with a period of 20.8 h was found to propagate upward slowly with time before the passage of the typhoon.

Introduction

Near-inertial oscillations (NIOs) in the ocean comprise nearly half of the surface kinetic energy (Pollard and Millard, 1970) and play a crucial role in mixing at the mixed layer base (D'Asaro, 1985). Furthermore, NIOs are considered a primary energy source for internal waves in the abyssal ocean, which enhance mixing (Gill, 1984; Garrett, 2001). The typical manifestation of NIOs is in an intensified oscillating rotary current with a peak frequency near, but not exactly the same as, the local inertial frequency, f. In most cases, the NIOs exhibit peak frequencies slightly above f, meaning they are super-inertial or blue-shift (Perkins, 1972), while frequencies slightly below f, meaning sub-inertial or red-shift, do exist (Perkins, 1976). The NIO frequency shift might be induced by mesoscale processes, such as eddies and loop currents, which could modify the propagation of NIOs (Kunze, 1985; Zhai et al., 2007). The cyclonic (anticyclonic) eddy with the positive (negative) vorticity could lead to the lower bound of the frequency band of NIOs above (below) f, i.e. to induce a blue-shift (red-shift) in the frequency. The wind on the ocean surface, which changes dramatically in both space and time, is the main mechanism of the generation of NIOs. Many analytical and numerical treatments of NIOs generated by typhoons show that the energy initially concentrates in the mixed layer, where most of it dissipates directly, while the remaining part propagates into the interior of the ocean or far away from the source (Watanabe and Hibiya, 2002; Alford, 2003). However, few studies are based on observations of typhoons because it is difficult to acquire the in situ data from anchored buoys which are located in the route of typhoons.

Typhoons are very active in the South China Sea (SCS), but observational data on responses of currents to typhoons are very rare. Ke et al. (1987) studied the typhoon Georgia (1983) in the northern SCS, but the observational data were just the current data at four different depths and the study was limited to the variations of the current velocity before and after the typhoon. Zhang (2009) found that the oceanic turbulent mixing didn't increase significantly after the passage of the storm. Moreover, they found that the sub-inertial current was the main factor in the oceanic mixing, which was more significant than the diurnal and the semi-diurnal tidal currents. Liu et al. (2010, 2011) used current data collected over a month to investigate the oceanic turbulent mixing and the near-bottom strong currents induced by typhoon Pabuk (2007) in the SCS. They found that, after the passage of typhoon, oceanic turbulent mixing increased by about an order of magnitude, which differs from the Zhang's conclusion in 2009. Sun et al. (2011a,b) studied the interaction between the NIOs induced by typhoons and background currents. They found that red-shifts of NIOs were induced by the geotrophic shear.

NIOs are highly horizontal variables and exhibit different characteristics in deep versus shallow water (Shay et al., 1998; Chant, 2001; Peters et al., 2002; Hisaki and Naruke, 2003). Past studies on the influence of background flows on NIOs (Zhang et al., 2005; Sun et al., 2011a,b) based on observations in the SCS are still relatively rare, and most of them are focused on influences in deep water. Therefore, there remains an interesting question to ask: In shallow water, does the NIO frequency shift still appear? Fortunately, typhoon Conson passed our observational site near the Xisha Islands in the continental shelf of the SCS, and the corresponding observational current and wind data were used to study the question mentioned above. In the following, the data and computational methods are presented in the data and methods section. The middle section presents the results, and the final section contains the conclusions and discussion.

Data and methods

Data

A 614.4-kHz Acoustic Doppler Current Profiler (ADCP) was attached at a depth of about 1 m under the water to a buoy (16°51.374′N, 112°19.118′E) on the continental shelf of the northern SCS (Figure 1). The depth of the water was about 50 m. The ADCP measures the current velocity from 2 to 43 m with a depth interval of 1 m, an accuracy of 10−3 m s−1, and a ping rate of 2 s. The effective data range is from 5 to 37 m because the data outside this range may be contaminated by the surface or bottom reflection of echoes from the side slopes (Liu and Weisberg, 2005). The duration of the observation was about 52 days, from 1 June to 21 July 2010. However, we only focus on the NIOs induced by a typhoon in July. Therefore, in the following, except in the calculation of the red-shift phenomenon of NIOs based on the 52-day observational data, we focus on the results based on the observational data from 1 to 21 July. The current data were averaged every 2 min for further analysis. Local wind data were observed with a sensor installed on a nearby island, which is about 3 km away from the buoy. During the observational period, the second typhoon of 2010, named Conson, passed over the observational site at about 23:00 on 15 July. At that time, the central pressure was 975 hPa, the maximum wind speed was 33 m s−1, and the moving speed was 20 km h1. The seven-day averaged geostrophic current data were obtained from the Archiving, Validatino and Interpretation of Satellite Oceanographic data (AVSIO2). The bottom topography near the observational site and the propagation route of Conson are also shown in (Figure 1). Since there is no synchronous in situ observational temperature and salinity data, the 1°×1° resolution climatological temperature and salinity data at a standard depth by the National Oceanographic Data Center (NODC, Corkright et al., 2002) were used to compute the upper mixed layer depth and sea water density in July at (17.5°N, 112.5°E) near the observational site. It was found that the upper mixed layer depth was about 25 m.

Figure 1.

Map of the observational site near the Xisha Islands in the SCS. The star denotes the observational site, the thick solid lines denote the propagation route of typhoon Conson, while the time of each annotated date is at 23:00 local Beijing time, the unit for the contours of the bottom topography denoted by thin lines is m.

Figure 1.

Map of the observational site near the Xisha Islands in the SCS. The star denotes the observational site, the thick solid lines denote the propagation route of typhoon Conson, while the time of each annotated date is at 23:00 local Beijing time, the unit for the contours of the bottom topography denoted by thin lines is m.

Methods

Data processing

In order to reduce the influence of the mooring buoy's motion, the raw current data obtained by the ADCP were processed by a low-pass-filter first. The pass band frequency of the low-pass filter was 1.2 times that of the surface gravity wave frequency, which had a value of 0.1 Hz. Subsequently, the data for local near-inertial currents were obtained by a band-pass-filter from the processed observed current data. The upper and lower cut-off frequencies of the band-pass-filter for the near-inertial currents were 0.8 and 1.2 times that of the local inertial frequency f, respectively. The rotary spectral method was also applied to the calculations of clockwise and anticlockwise components of the currents at inertial frequencies (Leaman and Sanford, 1975).

The current, U(z, t), at depth z and time t can be decomposed into a barotropic current, Ubt(t), and a baroclinic current, Ubc(z, t), as shown here:
formula
(1)
with the barotropic current defined as the depth-averaged current of
formula
(2)
where H is the water depth. In the following, the observed barotropic and baroclinic currents, and near-inertial currents are calculated, respectively.
The current velocity shear used in this study could be calculated by,
formula
(3)
where, u and v are the eastward and the northward components of the current velocity. To obtain the accurate peak frequency of the NIOs, we calculated the power spectral density of the eastward and northward current components based on the 52-day observational data, with a spectral resolution of about 0.0008 cph. If the resultant peak frequency, fp, is near the local inertial frequency, f, which has a value of 0.0242 cph, we define the oscillations as NIOs.

Damped slab model

For inertial frequency, f, mixed layer depth, Hm, density, ρ, wind stress components, τx and τy, mixed layer velocity components, u and v, the following damped slab model (Pollard and Millard 1970) is used to simulate the wind-induced mixed layer inertial currents,
formula
(4)
where R is an artificial damped constant that parameterizes the transfer of energy from the mixed layer to the deeper ocean. Realistic simulations clearly require f2 to be much greater than R2. The model is not stratified, so that the inertial frequency, f, is the only natural frequency. The eastward and northward components of the wind stress, τx and τy, could be calculated by Equation (5),
formula
(5)
where ρo is the density of the air, dc is the drag coefficient, Uw is the 10-m-height wind speed, and uw and vw are the eastward and northward components of the wind velocity, respectively. The drag coefficient can be calculated by the following (Gill, 1982), when Uw < 6 m s−1
formula
(6)
when Uw ≥ 6 m s−1
formula
(7)

Some sensitive experiments were carried out. Since the resulting upper mixed layer depth was 25 m, as stated above, we ran the model with a range of choices for R, with the value of 1/R ranging from 2–10 days. We found that when 1/R was equal to 10 days, the simulated results best matched the observational data.

Results and discussion

NIOs

In order to study the response of the NIOs during the whole observation period, we selected three different stages with the same duration of 7 days with stage 1 spanning from 1 to 7 July, stage 2 from 8 to 14 July and stage 3 from 15 to 21 July. Stages 1 and 2 (stage 3) are chosen as the durations before (after) the passage of typhoon Conson. For convenience, depths of 15 and 35 m are chosen to represent the upper and lower layers of the observed ocean, respectively.

(Figure 2) shows the time series of u and v components of the local near-inertial currents and the observed currents versus depth (hereafter u and v represent the eastward and northward components of the current speed, respectively). We found that at stages 1 and 2, before the passage of the typhoon, the near-inertial currents with a speed of no more than 0.1 m s−1 were relatively weak (Figures 2a and b). However, they started to increase in strength when the eye of typhoon arrived at the observational site on 15 July. The largest near-inertial current with a speed of about 0.5 m s−1 was observed on 19 July, which suggests the typhoon strengthened the kinetic energy of the local NIOs. Meanwhile, the total observed currents (Figures 2c and d) also increased sharply during the passage of the typhoon. The rotary spectra of the total observed currents at a whole water depth were calculated. It was found that the energy densities of the currents all had a maximum of 1 cpd (Figures 3a and b), which revealed that the diurnal currents are dominant in the SCS, as suggested by previous studies. Meanwhile, the observed currents in the inertial frequency band were mainly dominated by clockwise components.

Figure 2.

(a) and (b) are the time series of u and v components of the local near-inertial currents versus depth, respectively; (c) and (d) are the time series of u and v components of the observed currents versus depth, respectively.

Figure 2.

(a) and (b) are the time series of u and v components of the local near-inertial currents versus depth, respectively; (c) and (d) are the time series of u and v components of the observed currents versus depth, respectively.

Figure 3.

Rotary spectra of the currents at a depth of (a) 15 m and (b) 35 m. Here and subsequently, inertial frequency (f), diurnal frequency (D1), and semi-diurnal frequency (D2) are denoted by bold black lines in the figures. Comparisons of the (c) u and (d) v components of the observed and simulated inertial currents, respectively.

Figure 3.

Rotary spectra of the currents at a depth of (a) 15 m and (b) 35 m. Here and subsequently, inertial frequency (f), diurnal frequency (D1), and semi-diurnal frequency (D2) are denoted by bold black lines in the figures. Comparisons of the (c) u and (d) v components of the observed and simulated inertial currents, respectively.

In order to prove whether the enhanced local inertial currents were induced by the typhoon or not, the damped slab model set up by Pollard and Millard (1970) were employed. The 21-day in situ wind data were used to force the damped slab model to simulate the averaged inertial currents, and the simulated inertial currents were compared with the observed ones at the observation site (Figures 3c and d). It is clear that the damped slab model is capable of reproducing the evolutions of the inertial currents in the mixed layer, although the forecasted inertial currents were much stronger than the observed ones when the eye of typhoon arrived. This difference might be because this model can't resolve the sudden reversion of the current direction (about a 180° phase shift) at that time (figure omitted). The time-averaged inertial currents found using the total observed winds before and after the passage of typhoon (from 10 to 15 July and from 16 to 21 July, respectively) were 0.08 m s−1 and 0.28 m s−1, respectively. The significantly enhanced inertial currents after the passage of typhoon are considered to be the response to typhoon Conson.

In order to examine whether a red-shift in NIOs exists or not in this shallow water, the power spectra density (PSD) of the observed u and v components during the observational period were calculated. (Figure 4) shows the PSD at depths from 15 m and 35 m. There were three main frequencies at which the PSD was concentrated, representing NIOs, diurnal, and semi-diurnal tidal currents. The peak frequency of the NIOs had no significant variation vertically and an average value of 0.0237 cph for both the u and v components of the observed currents (note that the nearby resolved PSD has a frequency of 0.0245 cph). It was slightly smaller than the local inertial frequency f (0.0242 cph), which suggests that there is a red-shift in NIOs.

Figure 4.

Power spectral density of the observed current at the depth of (a) 15 m and (b) 35 m during the whole observational period. The gray (black) line denotes the u (v) component.

Figure 4.

Power spectral density of the observed current at the depth of (a) 15 m and (b) 35 m during the whole observational period. The gray (black) line denotes the u (v) component.

In fact, from June to July the observation site was mainly dominated by anticyclonic eddies. The 7-day averaged background geostrophic current data at stage 2, before the passage of typhoon Conson, is shown in (Figure 5). One can see that the observation site is located between two adjacent anticyclonic eddies. According to the shear flow mechanism, the background negative vorticity could reduce the lower limit of admissible frequency band for NIOs, causing the red-shift (Kunze, 1985). Thus, it is possible that the red-shift of the observed NIOs might be attributed to the background negative vorticity.

Figure 5.

Seven-day averaged background geostrophic flows at stage 2 before the passage of the typhoon Conson (the star denotes the observational site).

Figure 5.

Seven-day averaged background geostrophic flows at stage 2 before the passage of the typhoon Conson (the star denotes the observational site).

In order to study the NIOs further, we calculated the barotropic and baroclinic near-inertial currents and observed currents from 1–21 July before and after the passage of the typhoon at our observational site. One can see that the barotropic near-inertial currents (Figures 6a and b) have the same variations as the near-inertial currents (Figures 2a and b). At stages 1 and 2, before the passage of typhoon, the u and v components of the barotropic near-inertial currents, with a speed of no more than 0.1 m s−1, were relatively small. However, after the passage of typhoon Conson, the barotropic near-inertial currents started to increase, and the u (v) component reached a peak value of nearly 0.4 (0.5) m s−1 on 19 July. However, the baroclinic near-inertial currents changed little, even after the passage of the typhoon at stage 3. Meanwhile, the variations of the barotropic and baroclinic observed currents (Figures 6e–h) were similar to those of the near-inertial currents (Figures 6a–d).

Figure 6.

(a) Eastward and (b) northward components of the barotropic near-inertial currents; (c) eastward and (d) northward components of the baroclinic near-inertial currents versus depth; (e) eastward and (f) northward components of the barotropic currents; (g) eastward and (h) northward components of the baroclinic currents versus depth.

Figure 6.

(a) Eastward and (b) northward components of the barotropic near-inertial currents; (c) eastward and (d) northward components of the baroclinic near-inertial currents versus depth; (e) eastward and (f) northward components of the barotropic currents; (g) eastward and (h) northward components of the baroclinic currents versus depth.

The PSD of the barotropic currents at stages 1–3 are shown in (Figures 7a–c). One can see that at stages 1 and 2, before the passage of typhoon Conson, the barotropic diurnal tidal currents are dominant, followed by the barotropic semi-diurnal tidal currents, while the barotropic near-inertial currents are quite weak. Using the u component as an example, the PSD at the NIO, diurnal, and the semi-diurnal frequencies at stage 1 (stage 2) are 0.96 (2.50), 200.64 (81.20), and 4.25 (6.79) (m s−1)2 cph−1, respectively. The PSD ratio at the three frequencies is about 0.23:47.21:1 (0.37:11.96:1) at stage 1 (stage 2), implying that the diurnal tidal currents are dominant. However, at stage 3, after the passage of typhoon Conson, the peak PSD value at the NIO frequency reached 88.36 (m s−1)2 cph−1, i.e. about 92.04 (35.34) times higher than at stage 1 (stage 2). Meanwhile, the peak PSD value at the diurnal and semi-diurnal frequencies was 87.48 and 3.16 (m s−1)2 cph−1, respectively. The PSD ratio at the three frequencies is 27.96:27.68:1, implying that the typhoon increases the kinetic energy of the barotropic near-inertial currents significantly.

Figure 7.

Power spectral density of the barotropic observed currents at stages 1–3 (a–c). Power spectral density of the baroclinic observed currents at depths of 15 m and 35 m at stages 1–3. (d), (f) and (h) represent the power spectral density at a depth of 15 m at stages 1—3. (e), (g) and (i) represent the power spectral density at depth of 35 m at stages 1—3. The gray (black) line denotes the u (v) component.

Figure 7.

Power spectral density of the barotropic observed currents at stages 1–3 (a–c). Power spectral density of the baroclinic observed currents at depths of 15 m and 35 m at stages 1–3. (d), (f) and (h) represent the power spectral density at a depth of 15 m at stages 1—3. (e), (g) and (i) represent the power spectral density at depth of 35 m at stages 1—3. The gray (black) line denotes the u (v) component.

The PSD of the baroclinic currents at a depth of 15 m and 35 m during stages 1–3 are shown in (Figures 7d–i). Taking the u components as an example, it was found that, at 15 m (35 m) depth, the peak PSD ratios at the NIO frequency during stages 1–3 were 8.92:1.42:1 (7.2:0.4:1), implying the passage of typhoon Conson didn't enhance the kinetic energy of the baroclinic near-inertial currents significantly at our observational site. Thus it seems to suggest that, the response of the barotropic near-inertial currents to the passage of typhoon is more significant than that of the baroclinic near-inertial currents in this shallow water.

Near-inertial current shear

The time series of the near-inertial current shear versus depth during the observation were calculated and are shown in (Figure 8). It was found that before 6 July, the local near-inertial current shear with a value of no more than 0.002 s−1 of the whole water depth was very weak. However, the local near-inertial current shear in the lower layer became stronger below the depth of 20 m on 6 July, and it is interesting that the local near-inertial current shear strength in the upper layer changed very little. The shear became stronger over time and reached a peak value of about 0.012 s−1 on 7 July at 20–30 m. It became weaker in the lower layer, while it became stronger in the upper layer over time, and the position of the maximum intensity of near-inertial current shear seemed to propagate upward slowly. After 12 July, the strong signal reached a depth of about 10 m, with a peak value of about 0.009 s−1. The near-inertial current shear before the passage of typhoon Conson was found to have a period of about 20.8 h, nearly half of the local inertial period (about 41.3 h) in our study. However, the same phenomenon doesn't exist in the observed current shear field (figure omitted). After the passage of the typhoon, the upward propagation signal disappeared. This seems to suggest that it might be due to the vertical motion of the pycnocline. Since there were no in situ salinity and temperature data on hand, we could not investigate this problem further.

Figure 8.

Time series of the near-inertial current shear versus depth from 1 to 21 July.

Figure 8.

Time series of the near-inertial current shear versus depth from 1 to 21 July.

Conclusions

In this article, based on the spectral analysis methods, the digital filters, and the damped slab model, the NIOs near the Xisha Islands of the SCS during the passage of typhoon Conson were studied with 52-day in situ observational current and wind data. The following conclusions can be drawn:

First, before the passage of the typhoon the diurnal tidal current was dominant, followed by the semi-diurnal tidal current, while the near-inertial current was quite weak. The speed of the near-inertial current was only about 0.1 m s−1. However, after the passage of typhoon, the near-inertial current was significantly enhanced. The speed of the near-inertial current reached 0.4–0.5 m s−1. Meanwhile, it was found that the barotropic near-inertial current was more significant than the baroclinic near-inertial current after the passage of the typhoon. The barotropic near-inertial current increased significantly after the passage of typhoon, while the baroclinic near-inertial current changed little during the whole observation. The inertial currents were simulated by a damped slab model driven by the in situ wind data. It was found that the observed and the simulated inertial currents matched well before and after the passage of typhoon at our observation site. The NIOs had a peak frequency of 0.0237 cph during the whole observational period, which suggests that there is a red-shift phenomenon of the NIOs. The red-shifts might be caused by the background negative vorticity.

Second, the position of the maximum intensity of near-inertial current shear with a period of 20.8 h, was found to propagate from the lower to the upper layer slowly before the passage of typhoon Conson, which might be due to the vertical motion of the pycnocline.

Funding

This work was jointly supported by the Innovation Group Program of State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences (No. LTOZZ1201), the Knowledge Innovation Program of the Chinese Academy of Sciences (No. SQ201206), the National Basic Research Program (Nos. 2013CB956101 and 2011CB013701), NSFC Grant Nos. 41025019, 41406023, 41430964 and KZCX2-EW-Y040.

Notes

References

Alford, M. H.,
2003
.
Fine-structure contamination: Observations and a model of a simple two-wave case
.
J. Phys. Oceanogr.
31
,
2645
2649
.
Chant, R. J.,
2001
.
Evolution of near-inertial waves during an upwelling event on the New Jersey inner shelf
.
J. Phys. Oceanogr.
31
,
746
764
.
Corkright, M. E., Lorcanini, R. A., Garcia, H. E., O'Brien, T. D., Boyer, T. P., Stephens, C., Antonov, J. I.,
2002
. World Ocean Atlas 2001: Objective Analyses, Data Statistics, and Figures, CD-ROM Documentation.
National Oceanographic Data Center
,
Silver Spring
,
MD
,
17
.
D'Asaro, E. A.,
1985
.
The energy flux from the wind to near-inertial motions in the mixed layer
.
J. Phys. Oceanogr.
15
,
943
959
.
Garrett, C.,
2001
.
What is the “near-inertial” band and why is it different from the rest of the internal wave spectrum?
J. Phys. Oceanogr.
31
,
962
971
.
Gill, A. E.,
1982
.
Atmosphere-Ocean Dynamics
.
Academic
,
San Diego, CA.
Gill, A. E.,
1984
.
On the behavior of inertial waves in the wakes of storms
.
J. Phys. Oceanogr.
14
,
1129
1151
.
Hisaki, Y., Naruke, T.,
2003
.
Horizontal variability of near-inertial oscillations associated with the passage of a typhoon
.
J. Geophys. Res.
108
(
C12
),
3382
.
Ke, P. H., Huang, Q. Z., Chen, F. P., Huang, H. W., Feng, S. Y.,
1987
.
Preliminary analysis of the continental shelf current in northern South China Sea driven typhoon
.
Sci. Atmos. Sinica
11
,
441
445
, (In Chinese, English abstract).
Kunze, E.,
1985
.
Near-inertial wave propagation in geostrophic shear
.
J. Phys. Oceanogr.
15
,
544
565
.
Leaman, K. D., Sanford, T. B.,
1975
.
Vertical energy propagation of internal waves: a vector spectral analysis of velocity profiles
.
J. Geophys. Res.
80
,
1975
1978
.
Liu, J. L., Cai, S. Q., Wang, S. A.,
2010
.
Current and mixing on the Northern South China Sea
.
Chin. J. Oceanol. Limnol.
28
,
974
980
.
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
,
102
108
.
Liu, Y., Weisberg, R. H.,
2005
.
Momentum balance diagnoses for the West Florida Shelf
.
Cont. Shelf Res.
25
,
2054
2074
.
Perkins, H.,
1972
.
Inertial oscillations in the Mediterrranean
.
Deep Sea Res.
19
,
289
296
.
Perkins, H.,
1976
.
Observed effect of an eddy on inertial oscillations
.
Deep Sea Res.
23
,
1037
1042
.
Peters, H., Shay, L. K., Mariano, A. J., Cook, T. M.,
2002
.
Current variability on a narrow shelf with large ambient vorticity
,
J. Geophys. Res.
107
(
C8
),
3087
, doi:.
Pollard, R. T., Millard, R. C.,
1970
.
Comparison between observed and simulated wind-generated inertial oscillations
.
Deep-Sea Res.
17
,
813
821
.
Shay, L. K., Lee, T. N., Williams, E. J., Graber, H. G., Rooth, C. G.,
1998
.
Effects of low-frequency current variability on near-inertial submesoscale vortices
.
J. Geophys. Res.
103
,
18691
18714
.
Sun, L., Zheng, Q. A., Wang, D. X., Hu, J. Y., Tai, C. K., Sun, Z. Y.,
2011a
.
A case study of near-inertial oscillation in the South China Sea using mooring observations and satellite altimeter data
.
J. Oceanogr.
56
(
6
),
677
687
.
Sun, Z. Y., Hu, J. Y., Zheng, Q. A., Li, C. Y.,
2011b
.
Strong near-inertial oscillations in geostrophic shear in the northern South China Sea
.
J. Oceanogr.
67
,
377
384
.
Watanabe, M., Hibiya, T.,
2002
.
Global estimates of the wind-induced energy flux to inertial motions in the surface mixed layer
.
Geophys. Res. Lett.
29
, doi:.
Zhai, X., Greatbatch, R. J., Eden, C.,
2007
.
Spreading of near-inertial energy in a 1/12 model of the North Atlantic Ocean
.
Geophys. Res. Lett.
34
:
L10609
.
Zhang, X., Liang, X., Tian, J.,
2005
.
Study on internal tides and near-inertial motions in the upper 450 m ocean in the northern South China Sea
.
Chin. Sci. Bull.
50
,
2027
2031
.
Zhang, Y. W.,
2009
.
Study on wind-induced turbulent mixing on the South China Sea shelf and long-term mixing observation
. Ph. D. dissertation,
Ocean University of China
.