• Alford, M. H., and R. Pinkel, 2000: Observations of overturning in the thermocline: The context of ocean mixing. J. Phys. Oceanogr., 30, 805832, doi:10.1175/1520-0485(2000)030<0805:OOOITT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Alford, M. H., and M. Whitmont, 2007: Seasonal and spatial variability of near-inertial kinetic energy from historical moored velocity records. J. Phys. Oceanogr., 37, 20222037, doi:10.1175/JPO3106.1.

    • Search Google Scholar
    • Export Citation
  • Alford, M. H., and Coauthors, 2011: Energy flux and dissipation in Luzon Strait: Two tales of two ridges. J. Phys. Oceanogr., 41, 22112222, doi:10.1175/JPO-D-11-073.1.

    • Search Google Scholar
    • Export Citation
  • Arhan, M., X. Carton, A. Piola, and W. Zenk, 2002: Deep lenses of circumpolar water in the Argentine basin. J. Geophys. Res., 107, doi:10.1029/2001JC000963.

    • Search Google Scholar
    • Export Citation
  • Boyer, T. P., and Coauthors, 2009: World Ocean Database 2009. S. Levitus, Ed., NOAA Atlas NESDIS 66, 216 pp.

  • Broecker, W. S., W. C. Patzert, J. R. Toggweiler, and M. Stuvier, 1986: Hydrography, chemistry, and radioisotopes in the Southeast Asian basin. J. Geophys. Res., 91, 14 34514 354, doi:10.1029/JC091iC12p14345.

    • Search Google Scholar
    • Export Citation
  • Chang, Y.-T., W.-L. Hsu, J.-H. Tai, T.-Y. Tang, M.-H. Chang, and S.-Y. Chao, 2010: Cold deep water in the South China Sea. J. Oceanogr., 66, 183190, doi:10.1007/s10872-010-0016-x.

    • Search Google Scholar
    • Export Citation
  • Dillon, T. M., 1982: Vertical overturns: A comparison of Thorpe and Ozmidov length scales. J. Geophys. Res., 87, 96019613, doi:10.1029/JC087iC12p09601.

    • Search Google Scholar
    • Export Citation
  • Galbraith, P. S., and D. E. Kelley, 1996: Identifying overturns in CTD profiles. J. Atmos. Oceanic Technol., 13, 688702, doi:10.1175/1520-0426(1996)013<0688:IOICP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gargett, A., and T. Garner, 2008: Determining Thorpe scales from ship-lowered CTD density profiles. J. Atmos. Oceanic Technol., 25, 16571670, doi:10.1175/2008JTECHO541.1.

    • Search Google Scholar
    • Export Citation
  • Gordon, A. L., B. A. Huber, E. J. Metzger, R. D. Susanto, H. E. Hurlburt, and T. R. Adi, 2012: South China Sea Throughflow impact on the Indonesian Throughflow. Geophys. Res. Lett., 39, L11602, doi:10.1029/2012GL052021.

    • Search Google Scholar
    • Export Citation
  • Hamilton, P., 2009: Topographic Rossby waves in the Gulf of Mexico. Prog. Oceanogr.,82, 1–31, doi:10.1016/j.pocean.2009.04.019.

  • Hansen, B., and S. Østerhus, 2000: North Atlantic–Nordic Seas exchanges. Prog. Oceanogr., 45, 109208, doi:10.1016/S0079-6611(99)00052-X.

    • Search Google Scholar
    • Export Citation
  • Hogg, N., G. Siedler, and W. Zenk, 1999: Circulation and variability at the southern boundary of the Brazil basin. J. Phys. Oceanogr., 29, 145157, doi:10.1175/1520-0485(1999)029<0145:CAVATS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Høyer, J. L., and D. Quadfasel, 2001: Detection of deep overflows with satellite altimetry. Geophys. Res. Lett., 28, 16111614, doi:10.1029/2000GL012549.

    • Search Google Scholar
    • Export Citation
  • Liu, C.-T., and R.-J. Liu, 1988: The deep current in the Bashi Channel. Acta Oceanogr. Taiwan, 20, 107116.

  • Mercier, H., and K. G. Speer, 1998: Transport of bottom water in the Romanche Fracture Zone and the Chain Fracture Zone. J. Phys. Oceanogr., 28, 779790, doi:10.1175/1520-0485(1998)028<0779:TOBWIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Meredith, M. P., and Coauthors, 2011: Sustained monitoring of the Southern Ocean at Drake Passage: Past achievements and future priorities. Rev. Geophys., 49, RG4005, doi:10.1029/2010RG000348.

    • Search Google Scholar
    • Export Citation
  • Metzger, E. J., and H. E. Hurlburt, 1996: Coupled dynamics of the South China Sea, the Sulu Sea, and the Pacific Ocean. J. Geophys. Res., 101, 12 33112 352, doi:10.1029/95JC03861.

    • Search Google Scholar
    • Export Citation
  • Osborn, T. R., 1980: Estimates of the local rate of vertical diffusion from dissipation measurements. J. Phys. Oceanogr., 10, 8389, doi:10.1175/1520-0485(1980)010<0083:EOTLRO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Qu, T., 2000: Upper-layer circulation in the South China Sea. J. Phys. Oceanogr., 30, 14501460, doi:10.1175/1520-0485(2000)030<1450:ULCITS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Qu, T., Y. Kim, M. Yaremchuk, T. Tozuka, A. Ishida, and T. Yamagata, 2004: Can Luzon Strait transport play a role in conveying the impact of ENSO to the South China Sea? J. Climate, 17, 36443657, doi:10.1175/1520-0442(2004)017<3644:CLSTPA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Qu, T., Y. Du, G. Meyers, A. Ishida, and D. Wang, 2005: Connecting the tropical Pacific with Indian Ocean through South China Sea. Geophys. Res. Lett., 32, L24609, doi:10.1029/2005GL024698.

    • Search Google Scholar
    • Export Citation
  • Qu, T., Y. Du, and H. Sasaki, 2006a: South China Sea Throughflow: A heat and freshwater conveyor. Geophys. Res. Lett., 33, L23617, doi:10.1029/2006GL028350.

    • Search Google Scholar
    • Export Citation
  • Qu, T., J. B. Girton, and J. A. Whitehead, 2006b: Deepwater overflow through Luzon Strait. J. Geophys. Res., 111, C01002, doi:10.1029/2005JC003139.

    • Search Google Scholar
    • Export Citation
  • Rudnick, D., 1997: Direct velocity measurements in the Samoan Passage. J. Geophys. Res., 102, 32933302, doi:10.1029/96JC03286.

  • Smith, W. H. F., and D. T. Sandwell, 1997: Global seafloor topography from satellite altimetry and ship depth soundings. Science, 277, 19561962, doi:10.1126/science.277.5334.1956.

    • Search Google Scholar
    • Export Citation
  • Song, Y. T., 2006: Estimation of interbasin transport using ocean bottom pressure: Theory and model for Asian marginal seas. J. Geophys. Res., 111, C11S19, doi:10.1029/2005JC003189.

    • Search Google Scholar
    • Export Citation
  • Thierry, V., A. M. Treguier, and H. Mercier, 2006: Seasonal fluctuations in the deep central equatorial Atlantic Ocean: A data–model comparison. Ocean Dyn., 56, 581593, doi:10.1007/s10236-005-0045-y.

    • Search Google Scholar
    • Export Citation
  • Thorpe, S., 1977: Turbulence and mixing in a Scottish Loch. Philos. Trans. Roy. Soc. London,A286, 125181, doi:10.1098/rsta.1977.0112.

    • Search Google Scholar
    • Export Citation
  • Tian, J., and T. Qu, 2012: Advances in research on the deep South China Sea circulation. Chin. Sci. Bull., 57, 31153120, doi:10.1007/s11434-012-5269-x.

    • Search Google Scholar
    • Export Citation
  • Tian, J., Q. Yang, X. Liang, L. Xie, D. Hu, F. Wang, and T. Qu, 2006: Observation of Luzon Strait transport. Geophys. Res. Lett., 33, L19607, doi:10.1029/2006GL026272.

    • Search Google Scholar
    • Export Citation
  • Tian, J., Q. Yang, and W. Zhao, 2009: Enhanced diapycnal mixing in the South China Sea. J. Phys. Oceanogr., 39, 31913203, doi:10.1175/2009JPO3899.1.

    • Search Google Scholar
    • Export Citation
  • Torrence, C., and G. P. Compo, 1998: A practical guide to wavelet analysis. Bull. Amer. Meteor. Soc., 79, 6178, doi:10.1175/1520-0477(1998)079<0061:APGTWA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Tozuka, T., T. Qu, and T. Yamagata, 2007: Dramatic impact of the South China Sea on the Indonesian Throughflow. Geophys. Res. Lett., 34, L12612, doi:10.1029/2007GL030420.

    • Search Google Scholar
    • Export Citation
  • Tozuka, T., T. Qu, Y. Masumoto, and T. Yamagata, 2009: Impacts of the South China Sea Throughflow on seasonal and interannual variations the Indonesian Throughflow. Dyn. Atmos. Oceans, 47, 7385, doi:10.1016/j.dynatmoce.2008.09.001.

    • Search Google Scholar
    • Export Citation
  • Wang, J., 1986: Observation of abyssal flows in the northern South China Sea. Acta Oceanogr. Taiwan, 16, 3645.

  • Wyrtki, K., 1961: Physical oceanography of the Southeast Asian waters. Naga Rep. 2, 195 pp.

  • Yang, Q., J. Tian, and W. Zhao, 2010: Observation of Luzon Strait transport in summer 2007. Deep-Sea Res. I, 57, 670676, doi:10.1016/j.dsr.2010.02.004.

    • Search Google Scholar
    • Export Citation
  • Yang, Q., J. Tian, and W. Zhao, 2011: Observation of material fluxes through the Luzon Strait. Chin. J. Oceanol. Limnol., 29, 26–32, doi:10.1007/s00343-011-9952-6.

    • Search Google Scholar
    • Export Citation
  • Yaremchuk, M., and T. Qu, 2004: Seasonal variability of the large-scale currents near the coast of the Philippines. J. Phys. Oceanogr., 34, 844855, doi:10.1175/1520-0485(2004)034<0844:SVOTLC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zenk, W., 2008: Temperature fluctuations and current shear in Antarctic Bottom Water at the Vema Sill. Prog. Oceanogr., 77, 276284, doi:10.1016/j.pocean.2006.05.006.

    • Search Google Scholar
    • Export Citation
  • Zenk, W., K. G. Speer, and N. G. Hogg, 1993: Bathymetry at the Vema Sill. Deep-Sea Res. I, 40, 19251933, doi:10.1016/0967-0637(93)90038-5.

    • Search Google Scholar
    • Export Citation
  • Zenk, W., G. Siedler, B. Lenz, and N. G. Hogg, 1999: Antarctic Bottom Water flow through the Hunter Channel. J. Phys. Oceanogr., 29, 27852801, doi:10.1175/1520-0485(1999)029<2785:ABWFTT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zhao, W., C. Zhou, J. Tian, Q. Yang, B. Wang, L. Xie, and T. Qu, 2014: Deep water circulation in the Luzon Strait. J. Geophys. Res., 119, 790804, doi:10.1002/2013JC009587.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Bottom topography in (a) the SCS and (b) Luzon Strait (Smith and Sandwell 1997). The black stars in Fig. 1b denote the mooring locations.

  • View in gallery

    Bihourly time series of U, V, and temperature observed at 120-m HAB of the (left) BC and the (right) LT. The black lines indicate the daily mean time series.

  • View in gallery

    Histogram of subinertial Va and Vc at 120-m HAB of the (a),(c) BC and (b),(d) LT with the skewness of the subinertial Va or Vc series indicated in each panel.

  • View in gallery

    Vertical profile of Va at the (a) BC and (b) LT. Thick black (red) lines indicate the mean profile from the mooring during the periods October 2010–April 2011 at the BC and April 2011–March 2012 at the LT (April 2011–March 2013 at the BC and March 2012–March 2013 at the LT). Blue bars indicate the standard deviations. Green pentagrams and triangles display the maxima and minima of Va at different depth. Yellow pentagrams indicate the mean Va at 120-m HAB.

  • View in gallery

    15- and 120-day low-passed (a) Va and (b) temperature at 120-m HAB of the BC and LT.

  • View in gallery

    Spectrum analysis of the 20-day low-passed Va at 120-m HAB of the BC and LT.

  • View in gallery

    15–120-day bandpassed Va at 120-m HAB of the BC and LT, with standard deviations of the corresponding series indicated.

  • View in gallery

    Wavelet analysis of Va at the (a) BC and (b) LT. The color-filled contours are the log10-scaled variance (cm2 s−2) of the wavelet transform of normalized Va. The black solid contours are the 95% confidence level, and the black dashed line in each panel indicates that the areas below it are subject to the edge effect.

  • View in gallery

    An example of (top) horizontal velocity distribution and the (bottom) corresponding temperature time series at the BC (blue) and LT (red) when reversals of the overflow take place.

  • View in gallery

    Lag correlation between temperature and Va at the (a) BC and (b) LT and (c) between Va of the two channels during the observation with a time window of 90 days. Positive values indicate the former lags the latter.

  • View in gallery

    15-day low-passed time series of local SLA (gray) compared with Va [(a) BC and (c) LT] and temperature [(b) BC and (d) LT].

  • View in gallery

    Horizontal velocities against depth at the LT [(top) April 2011 ~ March 2012; (bottom) March 2012 ~ April 2013].

  • View in gallery

    Mean seasonal cycle of (a) Va and (b) temperature at the BC and LT based on monthly averaged time series. Standard deviations are indicated by gray bars.

  • View in gallery

    (a) Mean density profiles in summer and winter in the deep Pacific (EA: 21°–23°N, 122°–124°E) and SCS (WA: 18°–20°N, 119°–121°E); (b) Mean diapycnal diffusivity in the deep northeastern SCS (18°–21°N, 117°–120°E).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 289 204 31
PDF Downloads 235 157 24

Variability of the Deep-Water Overflow in the Luzon Strait

View More View Less
  • 1 Physical Oceanography Laboratory/Qingdao Collaborative Innovation Center of Marine Science and Technology, Ocean University of China, Qingdao, China
  • | 2 International Pacific Research Center, School of Ocean and Earth Science and Technology, University of Hawai‘i, Honolulu, Hawai‘i
© Get Permissions
Full access

Abstract

The Luzon Strait, with its deepest sills at the Bashi Channel and Luzon Trough, is the only deep connection between the Pacific Ocean and the South China Sea (SCS). To investigate the deep-water overflow through the Luzon Strait, 3.5 yr of continuous mooring observations have been conducted in the deep Bashi Channel and Luzon Trough. For the first time these observations enable us to assess the detailed variability of the deep-water overflow from the Pacific to the SCS. On average, the along-stream velocity of the overflow is at its maximum at about 120 m above the ocean bottom, reaching 19.9 ± 6.5 and 23.0 ± 11.8 cm s−1 at the central Bashi Channel and Luzon Trough, respectively. The velocity measurements can be translated to a mean volume transport for the deep-water overflow of 0.83 ± 0.46 Sverdrups (Sv; 1 Sv ≡ 106 m3 s−1) at the Bashi Channel and 0.88 ± 0.77 Sv at the Luzon Trough. Significant intraseasonal and seasonal variations are identified, with their dominant time scales ranging between 20 and 60 days and around 100 days. The intraseasonal variation is season dependent, with its maximum strength taking place in March–May. Deep-water eddies are believed to play a role in this intraseasonal variation. On the seasonal time scale, the deep-water overflow intensifies in late fall (October–December) and weakens in spring (March–May), corresponding well with the seasonal variation of the density difference between the Pacific and SCS, for which enhanced mixing in the deep SCS is possibly responsible.

School of Ocean and Earth Science and Technology Publication Number 9196 and International Pacific Research Center Publication Number IPRC-1076.

Corresponding author address: Jiwei Tian, 238 Songling Road, Physical Oceanography Laboratory/Qingdao Collaborative Innovation Center of Marine Science and Technology, Ocean University of China, Qingdao 266100, China. E-mail: tianjw@ouc.edu.cn

Abstract

The Luzon Strait, with its deepest sills at the Bashi Channel and Luzon Trough, is the only deep connection between the Pacific Ocean and the South China Sea (SCS). To investigate the deep-water overflow through the Luzon Strait, 3.5 yr of continuous mooring observations have been conducted in the deep Bashi Channel and Luzon Trough. For the first time these observations enable us to assess the detailed variability of the deep-water overflow from the Pacific to the SCS. On average, the along-stream velocity of the overflow is at its maximum at about 120 m above the ocean bottom, reaching 19.9 ± 6.5 and 23.0 ± 11.8 cm s−1 at the central Bashi Channel and Luzon Trough, respectively. The velocity measurements can be translated to a mean volume transport for the deep-water overflow of 0.83 ± 0.46 Sverdrups (Sv; 1 Sv ≡ 106 m3 s−1) at the Bashi Channel and 0.88 ± 0.77 Sv at the Luzon Trough. Significant intraseasonal and seasonal variations are identified, with their dominant time scales ranging between 20 and 60 days and around 100 days. The intraseasonal variation is season dependent, with its maximum strength taking place in March–May. Deep-water eddies are believed to play a role in this intraseasonal variation. On the seasonal time scale, the deep-water overflow intensifies in late fall (October–December) and weakens in spring (March–May), corresponding well with the seasonal variation of the density difference between the Pacific and SCS, for which enhanced mixing in the deep SCS is possibly responsible.

School of Ocean and Earth Science and Technology Publication Number 9196 and International Pacific Research Center Publication Number IPRC-1076.

Corresponding author address: Jiwei Tian, 238 Songling Road, Physical Oceanography Laboratory/Qingdao Collaborative Innovation Center of Marine Science and Technology, Ocean University of China, Qingdao 266100, China. E-mail: tianjw@ouc.edu.cn

1. Introduction

Deep passages connecting deep basins and marginal seas, including the Vema and Hunter Channels, the Romanche Fracture Zone, the Samoan Passage, the Drake Passage, and the Faroe Bank Channel, have been repeatedly investigated, and the deep-water overflow through these passages is believed to play a significant role in generating the global thermohaline circulation (e.g., Hogg et al. 1999; Mercier and Speer 1998; Rudnick 1997; Meredith et al. 2011; Hansen and Østerhus 2000). Motivated by the need to understand the abyssal dynamics in the northwestern Pacific Ocean and its role in the global thermohaline circulation, this study is focused on the Luzon Strait, a natural geographical constraint of the deep-water overflow from the Pacific to South China Sea (SCS).

The SCS is the largest marginal sea in the northwestern Pacific, with a large, deep (>2000 m) basin of more than 1.0 × 106 km2 and a maximum water depth over 5000 m. Diapycnal mixing in the deep SCS reaches as large as O(10−3) m2 s−1, significantly larger than that in the deep Pacific, because of the energetic internal waves and complicated bathymetry there (e.g., Tian et al. 2009; Alford et al. 2011). The corresponding upwelling in the deep SCS is on the order of 10−6 m s−1, indicative of a resident time of less than 100 yr, making the SCS a potentially important pathway of the global thermohaline circulation (e.g., Broecker et al. 1986; Qu et al. 2006b; Yang et al. 2011). The Luzon Strait, with a sill depth of about 2400 m, is the only deep connection between the SCS and the Pacific. Across the Luzon Strait, there is a persistent pressure gradient that drives a deep-water overflow from the Pacific into the SCS (e.g., Qu et al. 2006b; Tian et al. 2006; Song 2006). After crossing the Luzon Strait, water of Pacific origin sinks to the deep SCS (Wyrtki 1961). It then upwells as a result of enhanced mixing in the deep SCS (Tian et al. 2009) and eventually exits the SCS as part of the SCS Throughflow (e.g., Qu et al. 2005, 2006a), exerting notable impacts on the Indonesia Throughflow and its associated heat and freshwater fluxes from the Pacific to the Indian Ocean (e.g., Tozuka et al. 2007, 2009; Gordon et al. 2012).

Given the key role it plays in the SCS Throughflow, the deep-water overflow through the Luzon Strait has been investigated by several earlier studies. Two kinds of interfaces have usually been used to delimit the deep water from intermediate water, of which one is the bifurcation depth (~1500 m), calculated from mean density profiles on the east and west sides of the Luzon Strait (Qu et al. 2006b). The other is the 36.82 kg m−3 potential density isopycnal referenced to 2000 m (σ2), corresponding to ~2000 m, estimated from repeat-occupation conductivity–temperature–depth (CTD)/lowered acoustic Doppler current profiler (LADCP) profiles (Zhao et al. 2014). These studies have arrived at mean transport estimates ranging from 1.0 to 2.5 Sverdrups (Sv; 1 Sv ≡ 106 m3 s−1), based on diagnostic calculation (e.g., Wang 1986; Qu et al. 2006b; Song 2006), hydrographic data (e.g., Tian et al. 2006; Yang et al. 2010, 2011; Zhao et al. 2014), mooring observations (Liu and Liu 1988; Chang et al. 2010; Tian and Qu 2012), and model simulations (Zhao et al. 2014). By analyzing the ADCP measurements from repeat occupation stations, Zhao et al. (2014) recently provided the first picture of vertical structure of the deep-water overflow in the Luzon Strait, with the along-channel velocity below the 36.82 isopycnal, increasing with depth above the 120-m height above bottom (HAB) and decreasing below the 120-m HAB in the deep Bashi Channel and Luzon Trough (Fig. 1).

Fig. 1.
Fig. 1.

Bottom topography in (a) the SCS and (b) Luzon Strait (Smith and Sandwell 1997). The black stars in Fig. 1b denote the mooring locations.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

However, the temporal variability of the deep-water overflow in the Luzon Strait has been barely examined due to the lack of observations. To our best knowledge, the only continuous observations available so far were from Liu and Liu (1988) and Chang et al. (2010). Liu and Liu (1988) conducted an 82-day mooring observation with one active current meter in the Bashi Channel, which was apparently not long enough to study the subinertial temporal variability of the deep-water overflow. Another mooring observation was reported by Chang et al. (2010), with double current meter moorings lasting for 9 months at the Bashi Channel and Taltung Canyon, as indicated in their work. Energetic variation with a period spanning from 20 to 60 days was revealed, and the visual correlation between the deep-water overflow and sea surface height anomaly was identified in their study, though the processes responsible for this variation remain unknown. Based on a high-resolution regional model, Zhao et al. (2014) suggested that seasonal variation might exist in the deep-water overflow through the Bashi Channel and Luzon Trough, but this has not yet been confirmed by observations.

To better understand the temporal variability of the deep-water overflow and its associated governing processes, we deployed two moorings at two carefully selected sites in the Luzon Strait and acquired 3.5 yr of data. The results from an analysis of these data are reported in this study. The rest of this paper is organized as follows: Section 2 is devoted to description of the experiment configuration. The basic characteristics are presented in section 3, and results on temporal variability of the deep-water overflow are presented in section 4. The results are summarized in section 5.

2. Data

a. Mooring data

As part of the SCS deep circulation experiment, two bottom-anchored moorings were deployed in October 2009 and recovered in April 2013 at two sites of the Luzon Strait (Fig. 1). One (marked as BC in Fig. 1) was positioned in the Bashi Channel at a depth of 2720 m, on approximately the same site as where the mooring measurements were conducted by Liu and Liu (1988) and Chang et al. (2010). The other (marked as LT in Fig. 1) was positioned in the central Luzon Trough, at a depth of 3620 m, where the funnel-shaped topography leads to an intensification of the deep flow. The moorings were recovered and redeployed five times to refurbish and maintain the instruments. During each of the refurbishments, data were interrupted for less than a day. A cubic spline fit was applied to fill the gaps between the refurbishments. Sea-Bird Electronics SBE 37-SM CTDs, Aanderaa Instruments Recording Current Meter (RCM) Seaguard current meters, and Teledyne RD Instruments Workhorse Long Ranger 75-kHz acoustic Doppler current meters (ADCPs) were mounted on the moorings to monitor the salinity, temperature, pressure, and horizontal velocity of the deep flow. The accuracies of the instruments are 0.002°C for temperature, 0.003 mS cm−1 for conductivity, and 0.1% of full-scale range for pressure (which is 7 m for the CTD used in this experiment). Accuracies of velocity measurements were 0.15 cm s−1 for current meters and 0.1% S ± 5 mm s−1 for ADCP (S stands for the water velocity relative to ADCP). The sampling intervals were set to 2 h for all instruments from October 2009 to March 2010 and 1 h thereafter. Current meters were configured with 150 pings per record in burst mode, while the ADCPs sampled 37 ensembles with 24 pings each in burst mode and a bin size of 16 m. Although the designed range of the 75-kHz ADCP amounts to 600 m in long-range mode with a bin size of 16 m, poor scatterer concentration in the deep water weakened the echo intensity and reduced the maximum range to about 300 m from the transducers. Because of resource limitations, each mooring deployed from October 2009 to March 2010 was equipped with only one current meter. As additional resources became available, more instruments were mounted on the moorings during the latter period of the observations. Details pertinent to mooring design and configuration are shown in Table 1. Since the conductivity sensor failed in some segments, the present study only examines the temperature and horizontal velocity data in the Luzon Strait.

Table 1.

Design of moorings shown in Fig. 1.

Table 1.

Strong deep current can cause the moorings to tilt. During the period of observation, over 99.9% of the tilt records of the instruments were below the designed limits, which are 45° for current meters and 15° for ADCPs, suggesting that the velocity measurements were reliable. The tilt of moorings could also cause vertical excursions of the instruments monitored by the pressure sensors of the CTD. Unlike the case in open ocean where moorings tend to exhibit basically circular motion driven by quasi-circular tidal and inertial currents (e.g., Alford and Whitmont 2007), the moorings at BC and LT swing like pendulums because of the strong constraint of bathymetry. The typical pressure fluctuation between sequential records is about 6 dbar, which is the equivalent of a lateral deflection of 37 m at 120-m HAB, suggesting a horizontal velocity error of 0.5 cm s−1, significantly smaller than the velocity of the deep-water overflow in the Luzon Strait.

b. CTD profiles

Historical hydrographic data, including large numbers of CTD profiles from the World Ocean Database 2009 (WOD09; Boyer et al. 2009) and from our own field experiments, are also used in this study. After eliminating those profiles not passing standard deviation checks or flagged as outliers and those extending shallower than 1500-m depth, the hydrographic data used for this study consist of 150 temperature/salinity profiles in the northeastern SCS and northwestern Pacific (see section 4b for further details).

3. General description

Time series of horizontal velocity and temperature at 120-m HAB of the BC and LT are shown in Fig. 2. Notable tidal signals are visible both in horizontal velocity and temperature during the period of observation. A Butterworth bandpass filter is performed to resolve the semidiurnal and diurnal tidal signals, of which the standard deviations in the zonal (meridional) direction are 12.1 and 8.7 (9.4 and 8.7) cm s−1, respectively, at the BC and 4.0 and 6.2 (7.5 and 8.1) cm s−1, respectively, at the LT. This result suggests that both the semidiurnal and diurnal amplitudes at the BC are stronger than those at the LT. The mean magnitudes of the U and V components are −15.2 and −13.0 cm s−1 at the BC and −8.1 and −21.6 cm s−1 at the LT, respectively, representing a strong, deep-water overflow from the Pacific to the SCS through the Luzon Strait. The mean temperature around 120-m HAB is 2.03°C at the BC and 2.22°C at the LT. Once crossing the Bashi Channel, the deep water experiences a substantial descending, with the depth of maximum velocity following the topography and increasing from 2600 to 3500 m (Zhao et al. 2014).

Fig. 2.
Fig. 2.

Bihourly time series of U, V, and temperature observed at 120-m HAB of the (left) BC and the (right) LT. The black lines indicate the daily mean time series.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

Given the strong topography constraint, we remap the horizontal velocity into the along-stream Va and cross-stream Vc components, with the along-stream direction defined as the mean current direction, which is 230° and 201° clockwise from north at the BC and LT, respectively, and is basically parallel to the local isobaths. The cross-stream direction is defined as 90° counterclockwise from the along-stream direction. By this definition we have a mean velocity of 19.9 ± 6.5 cm s−1 at 120-m HAB of the BC and 23.0 ± 11.8 cm s−1 at 120-m HAB of the LT, where the values after the signs represent the standard deviations of the daily mean velocities. Figure 3 shows the histograms of velocities in both directions. While the distributions of Va at the BC and Vc at both the BC and LT basically follow a Gaussian distribution, Va at the LT tends to display obvious negative skewness, indicating that the velocity anomalies are much stronger in the negative direction than those in positive direction. We will return to this point later (section 4a).

Fig. 3.
Fig. 3.

Histogram of subinertial Va and Vc at 120-m HAB of the (a),(c) BC and (b),(d) LT with the skewness of the subinertial Va or Vc series indicated in each panel.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

Further analysis of current measurements at shallower depths allows us to investigate the vertical structure of the deep-water overflow in the Luzon Strait. Similar to what has been revealed by Zhao et al. (2014), results from August 2010 to April 2011, the segment with the largest vertical coverage of current measurements, show that the mean velocity at the BC is about zero at 2000 m and increases gradually with depth to about 22.6 cm s−1 at 120-m HAB (Fig. 4a). Owing to the low-scattering environment in the deep water, the vertical coverage of ADCP measurements at the BC from April 2011 to March 2013 is limited to about 300-m HAB. Despite a relatively minor vertical shear between 2400 and 2600 m, the mean profile during this period shows a similar vertical structure to that from August 2010 to April 2011. Vertical structure of the deep-water overflow at the LT also displays a bottom-intensified structure the same as that at the BC (Fig. 4b), that is, increasing from 1600 to 3500 m or about 120-m HAB. Results from the two segments are basically consistent. In contrast to the vertical structure of the mean current, it is noticeable that the standard deviations of velocity decrease with depth at both the BC and LT, with their maxima found around 2000 m at the BC and 2700 m at the LT, reflecting strong influence of bottom topography. Because of this topographic constraint, the velocity profiles of Vc at both the BC and LT are fairly weak.

Fig. 4.
Fig. 4.

Vertical profile of Va at the (a) BC and (b) LT. Thick black (red) lines indicate the mean profile from the mooring during the periods October 2010–April 2011 at the BC and April 2011–March 2012 at the LT (April 2011–March 2013 at the BC and March 2012–March 2013 at the LT). Blue bars indicate the standard deviations. Green pentagrams and triangles display the maxima and minima of Va at different depth. Yellow pentagrams indicate the mean Va at 120-m HAB.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

According to Zhao et al. (2014), the Rossby radii at the BC and LT, 19 and 20 km, are comparable with the local channel widths, 20 and 22 km, respectively, and so it seems reasonable to discuss the volume transport at the BC and LT with individual profiles. Based on the velocity profiles mentioned above, we employ 2000 and 1600 m to be the upper interface of the deep water at the BC and LT, respectively. To construct the mean profile of deep inflow, we average all the profiles during the period October 2009–March 2013 at the BC and LT and linearly interpolate the velocity vertically between the current meters with an assumption of zero velocity at the sea floor. The cross-stream topography is from Smith and Sandwell (1997), which has been widely used for the region studied (e.g., Qu et al. 2006b). After digitizing the vertical profiles to a vertical grid of 20 m and interpolating the velocity horizontally using a cubic spline with the assumptions of zero velocity at two sidewalls, the volume transport of the deep-water overflow is estimated to be 0.83 ± 0.46 Sv at the BC and 0.88 ± 0.77 Sv at the LT. These estimates are generally comparable with the previous studies (e.g., Liu and Liu 1988; Tian et al. 2006; Qu et al. 2006b; Chang et al. 2010; Yang et al. 2011; Zhao et al. 2014).

4. Variability

To focus on the subinertial signals, a 15-day Butterworth low-pass filter is applied to the time series of remapped velocity and temperature (Fig. 5). Low-passed time series of Va and temperature at the BC (LT) are shown to fluctuate between 1 and 37 cm s−1 (−17 and 36 cm s−1) and between 1.78° and 2.36°C (2.13° and 2.32°C), respectively, visually characterized by energetic variations from intraseasonal to interannual time scales. The result also shows that there is not only a substantial increase in temperature but also a substantial increase in the homogeneousness of the water property from the BC to LT. Because of the strong topography constraint, the low-passed Vc at the BC and LT are weak. Spectrum analysis of the low-passed Va manifests peaks in the frequency field (Fig. 6). Despite some quantitative differences between the BC and LT, both time series are dominated by intraseasonal variations, with their periods ranging between 20 and 60 days. While not obvious at the BC, a peak around 100 days also stands out at the LT. On seasonal time scale, the variation satisfies the 95% confidence level at the LT, but is relatively minor at the BC. In the following, the intraseasonal and seasonal variations are discussed separately.

Fig. 5.
Fig. 5.

15- and 120-day low-passed (a) Va and (b) temperature at 120-m HAB of the BC and LT.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

Fig. 6.
Fig. 6.

Spectrum analysis of the 20-day low-passed Va at 120-m HAB of the BC and LT.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

a. Intraseasonal variations

Intraseasonal variations associated with topographic waves and eddies have been observed in many parts of the deep ocean (e.g., Hamilton 2009; Arhan et al. 2002). High-frequency fluctuations in abyssal passages that connect deep basins have also been reported by previous studies. At the Samoan Passage in the Pacific, for example, Rudnick (1997) claimed that the 30-day fluctuation dominates the variability of transport and temperature beneath 4000 m. Based on a 687-day current meter record in the deep Vema Channel, Zenk (2008) revealed that, although eddy kinetic energy is smaller than the mean kinetic energy, high velocity variance falls in the submesoscale range between 14 and 20 days.

Considering that variations with a period of 20–100 days dominate the fluctuation of the deep-water overflow in the Luzon Strait, a Butterworth bandpass filter with a window of 15–120 days was applied to the Va time series to study the intraseasonal variations (Fig. 7). Because of the constraint of bathymetry, significant correlation is found between the time series at the BC and LT, with a correlation coefficient of 0.49 for the period of observation. The amplitudes of variations are different at the two channels, with a standard deviation of 5.0 cm s−1 at the BC and a standard deviation of 9.1 cm s−1 at the LT, respectively, suggesting that the intraseasonal variations at the LT are more energetic than those at the BC.

Fig. 7.
Fig. 7.

15–120-day bandpassed Va at 120-m HAB of the BC and LT, with standard deviations of the corresponding series indicated.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

It is interesting to note that the intraseasonal variations exhibit obvious nonstationary features (Fig. 7). Wavelet analysis following the method of Torrence and Compo (1998) is conducted to investigate how the deep flow varies in amplitude and frequency during the period of observation. In Fig. 8 it seems obvious that the frequency peaks shown in Fig. 6 are strongly dependent on season. At the LT, the 30-day oscillation is enhanced substantially in boreal spring (March–May), except for 2012, when the most energetic period seemed to shift from 30 to 50 days. A similar oscillation can be seen at the BC, though not as regular as that at the LT.

Fig. 8.
Fig. 8.

Wavelet analysis of Va at the (a) BC and (b) LT. The color-filled contours are the log10-scaled variance (cm2 s−2) of the wavelet transform of normalized Va. The black solid contours are the 95% confidence level, and the black dashed line in each panel indicates that the areas below it are subject to the edge effect.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

The positive skewness shown in the Va distribution at the LT (Fig. 3) could be interpreted by its bandpassed filtered time series (Fig. 7), with stronger anomalies in the negative direction. These negative anomalies are basically found during March–May each year, when short-term reversals of the inflow, with a typical period of 5–10 days, occur at the LT. These reversals are usually accompanied by weakened inflow at the BC. The causes of these reversals are not understood. When a 50-day oscillation was enhanced in March 2012, the flow at the BC was also reversed (Fig. 5). As an example, Fig. 9 shows the details of how this reversal took place. For this period, the temperature fluctuation was anticorrelated with Va both at the BC and LT. The time series of subinertial velocity at the LT indicates that the reversal took place from 5 March to 15 March and was associated with a positive temperature anomaly of 0.04°C. Given the fact that the deep water experiences continuous warming as it penetrates toward the SCS along the deep Luzon Trough, the positive temperature anomaly implies that relatively warm downstream water be pushed back to the northern Luzon Trough and sometimes all the way back to the Bashi Channel, like the case noted above when the mean Va reached −10.4 cm s−1 and the duration of the reversal exceeded 13.5 days. Similar reversals have also been reported in the Vema and Hunter Channels (e.g., Zenk et al. 1993; Hogg et al. 1999; Zenk et al. 1999). The similarities and differences between these reversals in unidirectional deep channel flows need to be investigated further by research.

Fig. 9.
Fig. 9.

An example of (top) horizontal velocity distribution and the (bottom) corresponding temperature time series at the BC (blue) and LT (red) when reversals of the overflow take place.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

Interestingly, a lagged anticorrelation can be found between the temperature and Va at the BC. To further investigate this anticorrelation, we perform a lag correlation analysis within the time window of 90 days. Figure 10a shows that the variation of temperature corresponds well with the variation of velocity during the period of observation at the BC, with the former lagging the latter by 2–6 days. Around March 2012, this time lag increased to about 7 days. No obvious time lag is found at the LT (Fig. 10b). In fact as one can see in Fig. 9, the reversals of velocity at the BC took place at a slightly different time as those at the LT. Following Figs. 10a and 10b, we also calculate the lag correlation of Va between the two channels, as shown in Fig. 10c. Presumably because of the strong constraint of topography, the two time series correspond well most of the time. In boreal spring and sometimes in October, when intraseasonal variations are intensified, the time series at the LT leads that at the BC by about 30 h, implying that the signals propagate against the main stream of the overflow.

Fig. 10.
Fig. 10.

Lag correlation between temperature and Va at the (a) BC and (b) LT and (c) between Va of the two channels during the observation with a time window of 90 days. Positive values indicate the former lags the latter.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

The high correlation of intraseasonal variations between the BC and LT suggests that they are possibly regulated by the same mechanism. Surface detectable eddies have been assumed to have notable influence on fluctuations of deep-water overflows through relatively shallow passages like the Faroe Bank Channel and Denmark Strait (e.g., Høyer and Quadfasel 2001). In the deep Bashi Channel and Taltung Canyon, Chang et al. (2010) indicated that the intraseasonal variations of deep-water overflow could also be related to the mesoscale processes in the upper ocean. Here, based on the merged satellite altimeter data of Jason-2, Jason-1, and EnviSat from Archiving, Validation, and Interpretation of Satellite Oceanographic (AVISO) data, we also compare the local sea level anomaly (SLA) with the along-stream velocity observed at the BC and LT. As one can see in Fig. 11, the local SLA is dominated by seasonal cycle, and during most periods of the observation it does not correspond to the variability of along-stream velocity. So, we suggest that the intraseasonal variations observed in the deep Luzon Strait cannot originate from the surface.

Fig. 11.
Fig. 11.

15-day low-passed time series of local SLA (gray) compared with Va [(a) BC and (c) LT] and temperature [(b) BC and (d) LT].

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

One possible forcing mechanism is related to deep eddies. Based on hydrographic sections, Arhan et al. (2002) have reported the existence of deep eddies centered at the interface between the North Atlantic Deep Water and Antarctic Bottom Water. By examining the current and temperature time series from the mooring observation, they indicated that the occasional reversals spotted at the deep Vema Channel could result from the passage of deep eddies. According to Fig. 4b, eddy kinetic energy of the deep flow reaches its maximum at about 3100 m at the LT, and from there it slightly decreases both upward and downward. Figure 12 shows the vertical structure of bandpassed flow at the LT. The flow beneath 2500 m shows a vertically coherent phase, while the flow above it (1500 m) also shows a good correspondence with that at 2500 m. These are consistent with the characteristics of deep eddies (e.g., Arhan et al. 2002), which, if exist in the Luzon Strait, could lie in the depth range centered at about 3100 m, where the maximum eddy kinetic energy is attained.

Fig. 12.
Fig. 12.

Horizontal velocities against depth at the LT [(top) April 2011 ~ March 2012; (bottom) March 2012 ~ April 2013].

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

b. Seasonal variation

Seasonal variation of the deep-water overflow is also shown in Fig. 5 (the bold lines). By averaging the 3.5 yr of 100-day low-passed time series of Va and temperature for each month, we obtain the mean seasonal cycle of the flow. The flow appears to attain its seasonal maximum in late boreal fall (October–December) and its seasonal minimum in boreal spring (March–May) (Fig. 13a). Since deep water in the northwestern Pacific is significantly cooler than that in the Luzon Strait (e.g., Qu et al. 2006b), the maximum Va is always accompanied by minimum temperature and vice versa (Fig. 13b). Standard deviations of the mean seasonal cycle of Va reach 1.6 cm s−1 at the BC and 3.3 cm s−1 at the LT, indicating that the seasonal variation is more energetic at the LT than at the BC, similar to what has been discussed for the intraseasonal variation. But, the magnitude of the seasonal variation of Va is significantly weaker than that of the intraseasonal variation (Fig. 5a), implying that the velocity variability of deep-water overflow in the Luzon Strait is dominated by fluctuations within the intraseasonal period band, which differs from the case of temperature variability (Fig. 5b). Similar to the intraseasonal variation, the phase of the seasonal variation is also vertically consistent in the depth range from the bottom to about 2000 m at the BC and to about 2500 m at the LT.

Fig. 13.
Fig. 13.

Mean seasonal cycle of (a) Va and (b) temperature at the BC and LT based on monthly averaged time series. Standard deviations are indicated by gray bars.

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

The seasonal cycle is ubiquitous in the deep ocean, especially in the equatorial area. Evidence exists to suggest that downward-propagating Kelvin and Rossby waves can contribute to the variability in the deep equatorial Atlantic and Pacific (e.g., Thierry et al. 2006). In the upper layer of the Luzon Strait, a strong seasonal cycle has been reported relating to variability of the pressure gradient across the Luzon Strait, which is attributed to the pileup of water from the prevailing monsoon (e.g., Metzger and Hurlburt 1996; Qu 2000). The vertically synchronous seasonal cycle, however, does not seem to favor the downward propagation assumption noted above.

Persistent density difference exists between the deep Pacific and SCS, which sustains a baroclinic pressure gradient across the Luzon Strait that in turn drives the deep-water overflow from the Pacific into the SCS (Qu et al. 2006b). Two boxes, one lying east (EA; 21°–23°N, 122°–124°E) of the Luzon Strait and the other lying west (WA; 18°–20°N, 119°–121°E) of it, are chosen to examine this density difference. As shown in Fig. 14a, the deep water is more weakly stratified in the SCS than that in the northwestern Pacific below the bifurcation depth of about 1500 m. By averaging the density profiles in boreal summer and winter separately, it is noticed that the density below the bifurcation depth in boreal winter is notably lower than that in boreal summer (with a σ2 difference ~0.01 in 2400 m) in the WA, while the density in the EA is basically the same for the two seasons. This result suggests that the pressure gradient across the Luzon Strait is larger in boreal winter, and this possibly drives a stronger deep-water overflow than in boreal summer.

Fig. 14.
Fig. 14.

(a) Mean density profiles in summer and winter in the deep Pacific (EA: 21°–23°N, 122°–124°E) and SCS (WA: 18°–20°N, 119°–121°E); (b) Mean diapycnal diffusivity in the deep northeastern SCS (18°–21°N, 117°–120°E).

Citation: Journal of Physical Oceanography 44, 11; 10.1175/JPO-D-14-0113.1

Enhanced mixing of up to 10−2 m2 s−1 is revealed in the deep (>1000 m) SCS (Qu et al. 2006b; Tian et al. 2009), which could explain the weak stratification there. Recently, Zhao et al. (2014) conducted a series of numerical experiments, and their results confirmed the hypothesis that the enhanced mixing in the deep SCS is a major process driving the deep-water overflow in the Luzon Strait (Qu et al. 2006b). The question that may arise immediately is if its seasonal variation (Figs. 13, 14a) is also related to the mixing in the deep SCS.

To address this question, we use one of the most commonly used models for turbulent diapycnal diffusivity κ, suggested by Osborn (1980),
eq1
where Γ is the mixing efficiency and typically taken to be 0.2 (Osborn 1980), N is the buoyancy frequency, and ε is the turbulence dissipation rate that can be estimated from density overturns (Thorpe 1977; Alford and Pinkel 2000):
eq2
where a = 0.8 is a constant of proportionality (Dillon 1982), and LT is the Thorpe scale (Thorpe 1977). It must be noted that both the spatial resolution of the measurements and the noise of the instruments may impose constraints on the overturn detection. In this paper, the overturn size criteria proposed by Galbraith and Kelley (1996) and the modified profile preprocessing method and overturn ratio Ro criterion proposed by Gargett and Garner (2008) are employed to reject those spurious overturns.

Using this Thorpe scale method, diffusivity of the deep SCS is examined based on the available density profiles with fine vertical resolution. Figure 14b shows that diapycnal diffusivity generally increases with depth below 1500 m, from the order of 10−4 m2 s−1 at 1500 m to 10−3 m2 s−1 at 3000 m. From this figure one can see that diffusivity in boreal winter is stronger than in boreal summer. The enhanced mixing in boreal winter may generate a weaker stratification in the deep SCS, resulting in a stronger pressure gradient and consequently a stronger deep-water overflow across the Luzon Strait. The details need to be carefully examined when more observations become available.

5. Summary

Based on 3.5 yr of measurements from two current meter moorings deployed in the Bashi Channel and central Luzon Trough, we have investigated the mean structure and temporal variability of the deep-water overflow in the Luzon Strait. Averaged over the period of observation, the along-stream velocity is at its maximum near 120-m HAB, reaching 19.9 ± 6.5 cm s−1 at the BC and 23.0 ± 11.8 cm s−1 at the LT. As it penetrates from the Bashi Channel to the central Luzon Trough, the deep water experiences a significant water property transformation, with its mean temperature increasing from 2.02° to 2.22°C at 120-m HAB, indicative of strong mixing processes in the deep Luzon Strait. Whereas the vertical stratification is weakened by the enhanced mixing as the deep water penetrates into the Luzon Strait (Zhao et al. 2014), the bottom-intensified vertical structure seems not to be modified substantially, with Va increasing monotonically from around 2000 m (1600 m) to 120-m HAB at the BC (LT), contributing to a transport estimate of 0.83 ± 0.46 Sv at the BC and 0.88 ± 0.77 Sv at the LT.

Significant temporal variability on time scales from intraseasonal to interannual is observed in the along-stream velocity of the deep Luzon Strait. The dominant time scales of this variability are between 20 and 60 days and around 100 days, while the seasonal variation is relatively weaker. The intraseasonal variation is strongly dependent on season, with its maximum amplitude taking place in boreal spring. During this period, the deep-water overflow can sometimes reverse its direction, and associated with these reversals are positive temperature anomalies in both the LT and BC. The causes for these reversals are not clear. They are possibly related to deep eddies, but the details require further investigation.

Seasonal variation is dominant in the upper-layer circulation of the Luzon Strait (e.g., Qu 2000; Yaremchuk and Qu 2004). Though also present in the deep-water overflow, with a seasonal maximum in fall and a seasonal minimum in spring, no downward propagation signals from the surface can be identified. Since the deep-water overflow is primarily forced by a persistent pressure gradient between the Pacific and the SCS (Qu et al. 2006b), it is speculated that the seasonal fluctuation of stratification in the deep SCS is a key process responsible for the seasonal variation of deep-water overflow through the Luzon Strait. With the presence of energetic internal tides in the SCS, it has already been known that mixing in the deep SCS is about two orders larger than that in the deep Pacific (e.g., Tian et al. 2009; Alford et al. 2011). So, the seasonal fluctuation of stratification can be largely attributed to the mixing in the deep SCS.

With a time series of 3.5 yr, we are unable to characterize the interannual variation of the deep-water overflow in the Luzon Strait. Nevertheless, noticeable interannual variation can be identified in the 100-day low-passed time series of along-stream velocity (Fig. 5). During 2012/13, for example, the typical annual cycle was replaced by a double-peak feature: one around late July and the other around January. The amplitudes of intraseasonal variation in spring are noticeably different from year to year. Interannual variability in the Luzon Strait have been examined by several earlier studies (e.g., Qu et al. 2004; Tozuka et al. 2009), indicating that the Luzon Strait transport, especially in the upper layer, have significant correlation with shift in the bifurcation of North Equatorial Current owing to the ENSO events. As for the deep layer, what dominates the interannual variability and if there is any relationship to the ENSO have not been investigated yet. We will leave these for future studies.

Acknowledgments

The authors thank the two anonymous reviewers for their constructive comments and suggestions. The altimeter products used in this work are produced by SSALTO/DUCAS and distributed by AVISO, with support from CNES (available online at www.aviso.oceanobs.com/duacs). Author Jiwei Tian wishes to acknowledge the support from National Natural Science Foundation of China (91028008), author Wei Zhao wishes to acknowledge the support from the National Natural Science Foundation of China (41176010) and the Program for New Century Excellent Talents in University (NCET-10-0764), and author Tangdong Qu wishes to acknowledge the support of the U.S. National Science Foundation (OCE10-29704 and OCE-1130050). The authors are grateful to the crew of the R/V Dongfanghong II for their considerable help during the cruises and also the mooring group at the Ocean University of China.

REFERENCES

  • Alford, M. H., and R. Pinkel, 2000: Observations of overturning in the thermocline: The context of ocean mixing. J. Phys. Oceanogr., 30, 805832, doi:10.1175/1520-0485(2000)030<0805:OOOITT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Alford, M. H., and M. Whitmont, 2007: Seasonal and spatial variability of near-inertial kinetic energy from historical moored velocity records. J. Phys. Oceanogr., 37, 20222037, doi:10.1175/JPO3106.1.

    • Search Google Scholar
    • Export Citation
  • Alford, M. H., and Coauthors, 2011: Energy flux and dissipation in Luzon Strait: Two tales of two ridges. J. Phys. Oceanogr., 41, 22112222, doi:10.1175/JPO-D-11-073.1.

    • Search Google Scholar
    • Export Citation
  • Arhan, M., X. Carton, A. Piola, and W. Zenk, 2002: Deep lenses of circumpolar water in the Argentine basin. J. Geophys. Res., 107, doi:10.1029/2001JC000963.

    • Search Google Scholar
    • Export Citation
  • Boyer, T. P., and Coauthors, 2009: World Ocean Database 2009. S. Levitus, Ed., NOAA Atlas NESDIS 66, 216 pp.

  • Broecker, W. S., W. C. Patzert, J. R. Toggweiler, and M. Stuvier, 1986: Hydrography, chemistry, and radioisotopes in the Southeast Asian basin. J. Geophys. Res., 91, 14 34514 354, doi:10.1029/JC091iC12p14345.

    • Search Google Scholar
    • Export Citation
  • Chang, Y.-T., W.-L. Hsu, J.-H. Tai, T.-Y. Tang, M.-H. Chang, and S.-Y. Chao, 2010: Cold deep water in the South China Sea. J. Oceanogr., 66, 183190, doi:10.1007/s10872-010-0016-x.

    • Search Google Scholar
    • Export Citation
  • Dillon, T. M., 1982: Vertical overturns: A comparison of Thorpe and Ozmidov length scales. J. Geophys. Res., 87, 96019613, doi:10.1029/JC087iC12p09601.

    • Search Google Scholar
    • Export Citation
  • Galbraith, P. S., and D. E. Kelley, 1996: Identifying overturns in CTD profiles. J. Atmos. Oceanic Technol., 13, 688702, doi:10.1175/1520-0426(1996)013<0688:IOICP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gargett, A., and T. Garner, 2008: Determining Thorpe scales from ship-lowered CTD density profiles. J. Atmos. Oceanic Technol., 25, 16571670, doi:10.1175/2008JTECHO541.1.

    • Search Google Scholar
    • Export Citation
  • Gordon, A. L., B. A. Huber, E. J. Metzger, R. D. Susanto, H. E. Hurlburt, and T. R. Adi, 2012: South China Sea Throughflow impact on the Indonesian Throughflow. Geophys. Res. Lett., 39, L11602, doi:10.1029/2012GL052021.

    • Search Google Scholar
    • Export Citation
  • Hamilton, P., 2009: Topographic Rossby waves in the Gulf of Mexico. Prog. Oceanogr.,82, 1–31, doi:10.1016/j.pocean.2009.04.019.

  • Hansen, B., and S. Østerhus, 2000: North Atlantic–Nordic Seas exchanges. Prog. Oceanogr., 45, 109208, doi:10.1016/S0079-6611(99)00052-X.

    • Search Google Scholar
    • Export Citation
  • Hogg, N., G. Siedler, and W. Zenk, 1999: Circulation and variability at the southern boundary of the Brazil basin. J. Phys. Oceanogr., 29, 145157, doi:10.1175/1520-0485(1999)029<0145:CAVATS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Høyer, J. L., and D. Quadfasel, 2001: Detection of deep overflows with satellite altimetry. Geophys. Res. Lett., 28, 16111614, doi:10.1029/2000GL012549.

    • Search Google Scholar
    • Export Citation
  • Liu, C.-T., and R.-J. Liu, 1988: The deep current in the Bashi Channel. Acta Oceanogr. Taiwan, 20, 107116.

  • Mercier, H., and K. G. Speer, 1998: Transport of bottom water in the Romanche Fracture Zone and the Chain Fracture Zone. J. Phys. Oceanogr., 28, 779790, doi:10.1175/1520-0485(1998)028<0779:TOBWIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Meredith, M. P., and Coauthors, 2011: Sustained monitoring of the Southern Ocean at Drake Passage: Past achievements and future priorities. Rev. Geophys., 49, RG4005, doi:10.1029/2010RG000348.

    • Search Google Scholar
    • Export Citation
  • Metzger, E. J., and H. E. Hurlburt, 1996: Coupled dynamics of the South China Sea, the Sulu Sea, and the Pacific Ocean. J. Geophys. Res., 101, 12 33112 352, doi:10.1029/95JC03861.

    • Search Google Scholar
    • Export Citation
  • Osborn, T. R., 1980: Estimates of the local rate of vertical diffusion from dissipation measurements. J. Phys. Oceanogr., 10, 8389, doi:10.1175/1520-0485(1980)010<0083:EOTLRO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Qu, T., 2000: Upper-layer circulation in the South China Sea. J. Phys. Oceanogr., 30, 14501460, doi:10.1175/1520-0485(2000)030<1450:ULCITS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Qu, T., Y. Kim, M. Yaremchuk, T. Tozuka, A. Ishida, and T. Yamagata, 2004: Can Luzon Strait transport play a role in conveying the impact of ENSO to the South China Sea? J. Climate, 17, 36443657, doi:10.1175/1520-0442(2004)017<3644:CLSTPA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Qu, T., Y. Du, G. Meyers, A. Ishida, and D. Wang, 2005: Connecting the tropical Pacific with Indian Ocean through South China Sea. Geophys. Res. Lett., 32, L24609, doi:10.1029/2005GL024698.

    • Search Google Scholar
    • Export Citation
  • Qu, T., Y. Du, and H. Sasaki, 2006a: South China Sea Throughflow: A heat and freshwater conveyor. Geophys. Res. Lett., 33, L23617, doi:10.1029/2006GL028350.

    • Search Google Scholar
    • Export Citation
  • Qu, T., J. B. Girton, and J. A. Whitehead, 2006b: Deepwater overflow through Luzon Strait. J. Geophys. Res., 111, C01002, doi:10.1029/2005JC003139.

    • Search Google Scholar
    • Export Citation
  • Rudnick, D., 1997: Direct velocity measurements in the Samoan Passage. J. Geophys. Res., 102, 32933302, doi:10.1029/96JC03286.

  • Smith, W. H. F., and D. T. Sandwell, 1997: Global seafloor topography from satellite altimetry and ship depth soundings. Science, 277, 19561962, doi:10.1126/science.277.5334.1956.

    • Search Google Scholar
    • Export Citation
  • Song, Y. T., 2006: Estimation of interbasin transport using ocean bottom pressure: Theory and model for Asian marginal seas. J. Geophys. Res., 111, C11S19, doi:10.1029/2005JC003189.

    • Search Google Scholar
    • Export Citation
  • Thierry, V., A. M. Treguier, and H. Mercier, 2006: Seasonal fluctuations in the deep central equatorial Atlantic Ocean: A data–model comparison. Ocean Dyn., 56, 581593, doi:10.1007/s10236-005-0045-y.

    • Search Google Scholar
    • Export Citation
  • Thorpe, S., 1977: Turbulence and mixing in a Scottish Loch. Philos. Trans. Roy. Soc. London,A286, 125181, doi:10.1098/rsta.1977.0112.

    • Search Google Scholar
    • Export Citation
  • Tian, J., and T. Qu, 2012: Advances in research on the deep South China Sea circulation. Chin. Sci. Bull., 57, 31153120, doi:10.1007/s11434-012-5269-x.

    • Search Google Scholar
    • Export Citation
  • Tian, J., Q. Yang, X. Liang, L. Xie, D. Hu, F. Wang, and T. Qu, 2006: Observation of Luzon Strait transport. Geophys. Res. Lett., 33, L19607, doi:10.1029/2006GL026272.

    • Search Google Scholar
    • Export Citation
  • Tian, J., Q. Yang, and W. Zhao, 2009: Enhanced diapycnal mixing in the South China Sea. J. Phys. Oceanogr., 39, 31913203, doi:10.1175/2009JPO3899.1.

    • Search Google Scholar
    • Export Citation
  • Torrence, C., and G. P. Compo, 1998: A practical guide to wavelet analysis. Bull. Amer. Meteor. Soc., 79, 6178, doi:10.1175/1520-0477(1998)079<0061:APGTWA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Tozuka, T., T. Qu, and T. Yamagata, 2007: Dramatic impact of the South China Sea on the Indonesian Throughflow. Geophys. Res. Lett., 34, L12612, doi:10.1029/2007GL030420.

    • Search Google Scholar
    • Export Citation
  • Tozuka, T., T. Qu, Y. Masumoto, and T. Yamagata, 2009: Impacts of the South China Sea Throughflow on seasonal and interannual variations the Indonesian Throughflow. Dyn. Atmos. Oceans, 47, 7385, doi:10.1016/j.dynatmoce.2008.09.001.

    • Search Google Scholar
    • Export Citation
  • Wang, J., 1986: Observation of abyssal flows in the northern South China Sea. Acta Oceanogr. Taiwan, 16, 3645.

  • Wyrtki, K., 1961: Physical oceanography of the Southeast Asian waters. Naga Rep. 2, 195 pp.

  • Yang, Q., J. Tian, and W. Zhao, 2010: Observation of Luzon Strait transport in summer 2007. Deep-Sea Res. I, 57, 670676, doi:10.1016/j.dsr.2010.02.004.

    • Search Google Scholar
    • Export Citation
  • Yang, Q., J. Tian, and W. Zhao, 2011: Observation of material fluxes through the Luzon Strait. Chin. J. Oceanol. Limnol., 29, 26–32, doi:10.1007/s00343-011-9952-6.

    • Search Google Scholar
    • Export Citation
  • Yaremchuk, M., and T. Qu, 2004: Seasonal variability of the large-scale currents near the coast of the Philippines. J. Phys. Oceanogr., 34, 844855, doi:10.1175/1520-0485(2004)034<0844:SVOTLC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zenk, W., 2008: Temperature fluctuations and current shear in Antarctic Bottom Water at the Vema Sill. Prog. Oceanogr., 77, 276284, doi:10.1016/j.pocean.2006.05.006.

    • Search Google Scholar
    • Export Citation
  • Zenk, W., K. G. Speer, and N. G. Hogg, 1993: Bathymetry at the Vema Sill. Deep-Sea Res. I, 40, 19251933, doi:10.1016/0967-0637(93)90038-5.

    • Search Google Scholar
    • Export Citation
  • Zenk, W., G. Siedler, B. Lenz, and N. G. Hogg, 1999: Antarctic Bottom Water flow through the Hunter Channel. J. Phys. Oceanogr., 29, 27852801, doi:10.1175/1520-0485(1999)029<2785:ABWFTT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zhao, W., C. Zhou, J. Tian, Q. Yang, B. Wang, L. Xie, and T. Qu, 2014: Deep water circulation in the Luzon Strait. J. Geophys. Res., 119, 790804, doi:10.1002/2013JC009587.

    • Search Google Scholar
    • Export Citation
Save