• Bower, A. S., and et al. , 2002: Directly measured mid-depth circulation in the northeastern North Atlantic Ocean. Nature, 419, 603607, doi:10.1038/nature01078.

    • Search Google Scholar
    • Export Citation
  • Bower, A. S., , M. S. Lozier, , S. F. Gary, , and C. W. Boning, 2009: Interior pathways of the North Atlantic meridional overturning circulation. Nature, 459, 243248, doi:10.1038/nature07979.

    • Search Google Scholar
    • Export Citation
  • Chang, K. I., and et al. , 2004: Circulation and currents in the southwestern East/Japan Sea: Overview and review. Prog. Oceanogr., 61, 105156, doi:10.1016/j.pocean.2004.06.005.

    • Search Google Scholar
    • Export Citation
  • Choi, Y. J., 2015: Comparative simulation study of effects of eddy-topography interaction in the East/Japan Sea deep circulation. Acta Oceanol. Sin., 34, 118, doi:10.1007/s13131-015-0693-1.

    • Search Google Scholar
    • Export Citation
  • Choi, Y. J., , and J. H. Yoon, 2010: Structure and seasonal variability of the deep mean circulation of the East Sea (Sea of Japan). J. Oceanogr., 66, 349361, doi:10.1007/s10872-010-0031-y.

    • Search Google Scholar
    • Export Citation
  • Danchenkov, M. A., , S. C. Riser, , and J. H. Yoon, 2003: Deep currents of the central Sea of Japan. Pacific Oceanogr. J., 1, 611.

  • Danchenkov, M. A., , V. B. Lobanov, , S. C. Riser, , K. Kim, , M. Takematsu, , and J.-H. Yoon, 2006: A history of physical oceanographic research in the Japan/East Sea. Oceanography, 19, 1831, doi:10.5670/oceanog.2006.41.

    • Search Google Scholar
    • Export Citation
  • Freeland, H. D., and et al. , 2010: Argo—A decade of progress. Proc. OceanObs'09: Sustained Ocean Observations and Information for Society, Venice, Italy, IOC/UNESCO/ESA, 21–25, doi:10.5270/OceanObs09.cwp.32.

  • Gamo, T., 1999: Global warming may have slowed down the deep conveyor belt of a marginal sea of the northwestern Pacific: Japan Sea. Geophys. Res. Lett., 26, 31373140, doi:10.1029/1999GL002341.

    • Search Google Scholar
    • Export Citation
  • Gamo, T., , Y. Nozaki, , H. Saki, , T. Nakai, , and H. Tsubota, 1986: Spatial and temporal variations of water characteristics in the Japan Sea bottom layer. J. Mar. Res., 44, 781793, doi:10.1357/002224086788401620.

    • Search Google Scholar
    • Export Citation
  • Hogan, P. J., , and H. E. Hurlburt, 2000: Impact of upper ocean–topographical coupling and isopycnal outcropping in Japan/East Sea models with 1/8° to 1/64° resolution. J. Phys. Oceanogr., 30, 25352561, doi:10.1175/1520-0485(2000)030<2535:IOUOTC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kim, K., , R.-R. Kim, , D.-H. Min, , Y. Volkov, , J.-W. Yoon, , and M. Takematsu, 2001: Warming and structural changes in the East (Japan) Sea: A clue to future changes in global oceans. Geophys. Res. Lett., 28, 32933296, doi:10.1029/2001GL013078.

    • Search Google Scholar
    • Export Citation
  • Kim, K., , K.-R. Kim, , Y.-G. Kim, , Y.-K. Cho, , D.-J. Kang, , M. Takematsu, , and Y. Volkov, 2004: Water masses and decadal variability in the East Sea (Sea of Japan). Prog. Oceanogr., 61, 157174, doi:10.1016/j.pocean.2004.06.003.

    • Search Google Scholar
    • Export Citation
  • Kim, K.-R., , G. Kim, , K. Kim, , V. Lobanov, , V. Ponomareyv, , and A. Salyuk, 2002: A sudden bottom-water formation during the severe winter 2000–2001: The case of the East/Japan Sea. Geophys. Res. Lett., 29, doi:10.1029/2001GL014498.

    • Search Google Scholar
    • Export Citation
  • Lavender, K. L., , R. E. Davis, , and W. B. Owens, 2000: Mid-depth recirculation observed in the interior Labrador and Irminger Seas by direct velocity measurements. Nature, 407, 6669, doi:10.1038/35024048.

    • Search Google Scholar
    • Export Citation
  • Mitchell, D. A., and et al. , 2005: Upper circulation patterns in the Ulleung basin. Deep-Sea Res. II, 52, 16171638, doi:10.1016/j.dsr2.2003.09.005.

    • Search Google Scholar
    • Export Citation
  • Park, J. J., , and K. Kim, 2013: Deep currents obtained from Argo float trajectories in the Japan/East Sea. Deep-Sea Res. II, 85, 169181, doi:10.1016/j.dsr2.2012.07.032.

    • Search Google Scholar
    • Export Citation
  • Park, J. J., , K. Kim, , B. A. King, , and S. C. Riser, 2005: An advanced method to estimate deep currents from profiling floats. J. Atmos. Oceanic Technol., 22, 12941304, doi:10.1175/JTECH1748.1.

    • Search Google Scholar
    • Export Citation
  • Park, Y., , H.-R. Shin, , J.-H. Yoon, , C.-H. Kim, , and Y. Yoshikawa, 2015: Simulation of eddy-driven deep circulation in the East/Japan Sea by using a three-layer model with wind, throughflow and deep water formation forcings. J. Mar. Syst., 150, 4155, doi:10.1016/j.jmarsys.2015.05.006.

    • Search Google Scholar
    • Export Citation
  • Roemmich, D., and et al. , 2009: The Argo Program: Observing the global ocean with profiling floats. Oceanography, 22, 3443, doi:10.5670/oceanog.2009.36.

    • Search Google Scholar
    • Export Citation
  • Takematsu, M., , Z. Nagano, , A. G. Ostrovskii, , K. Kim, , and Y. Volkov, 1999: Direct measurements of deep current in the northern Japan Sea. J. Oceanogr., 55, 207216, doi:10.1023/A:1007842013257.

    • Search Google Scholar
    • Export Citation
  • Talley, L. D., , V. Lobanov, , V. Ponomarev, , A. Salyuk, , P. Tishchenko, , and I. Zhabin, 2003: Deep convection and brine rejection in the Japan Sea. Geophys. Res. Lett., 30, 11591162, doi:10.1029/2002GL016451.

    • Search Google Scholar
    • Export Citation
  • Yanagimoto, D., , and K. Taira, 2003: Current measurements of the Japan Sea Proper Water and the Intermediate Water by ALACE floats. J. Oceanogr., 59, 359368, doi:10.1023/A:1025572112019.

    • Search Google Scholar
    • Export Citation
  • Yoon, J. H., , K. Abe, , T. Ogata, , and Y. Wakamatsu, 2005: The effects of wind-stress curl on the Japan/East Sea circulation. Deep-Sea Res. II, 52, 18271844, doi:10.1016/j.dsr2.2004.03.004.

    • Search Google Scholar
    • Export Citation
  • Yoshikawa, Y., 2012: An eddy-driven abyssal circulation in a bowl-shaped basin due to deep water formation. J. Oceanogr., 86, 971983, doi:10.1007/s10872-012-0148-2.

    • Search Google Scholar
    • Export Citation
  • Yoshikawa, Y., , T. Awaji, , and K. Akitomo, 1999: Formation and circulation processes of intermediate water in the Japan Sea. J. Phys. Oceanogr., 29, 17011722, doi:10.1175/1520-0485(1999)029<1701:FACPOI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Trajectories of the selected nine floats presented in different colors. Eight floats are deployed in the UB, with one in the eastern location of Primorye, and the deployment locations are marked with stars. See Table 1 for more details on the nine floats.

  • View in gallery

    (a) Trajectories of Argo floats associated with the BW-gyre. The starting point of each trajectory is marked with a large, solid circle and the end point is marked with a triangle. Floats are the circled numbers 1 through 6, and numerals corresponding to year and month are marked along the trajectory of each float. The same notations are used in following figures. (b) Trajectories of Argo floats associated with the first E-gyre group. They turn toward deeper regions off the isobaths on the northern edge of the EJB north of 43°N. Floats are numbered from 1 to 6. (c) As in Fig. 2b, but for the second E-gyre group. They (floats 7–10) turn toward deeper regions off the isobaths on the eastern side of the EJB roughly between 42.5° and 43°N. Note that the region of negative wind stress curl reaches far offshore (138°–139.5°E) in November (Fig. A1).

  • View in gallery

    Quasi-Eulerian vectors obtained from Argo data for (a) summer (June to September) and (b) winter (December to March). Strong along-isobathic currents are present in region PS in summer, whereas strong southward currents are present in the region PW in winter. Red (yellow) vectors indicate the exceptional trajectory associated with float 6 of E-gyre (float 6 of BW-gyre), as explained in the text.

  • View in gallery

    Wind stress vectors (arrows) and wind stress curl (colored shading) in winter (November–February) overlaid with the trajectories associated with the E-gyre (Figs. 2b and 2c). In region WF (the rectangular area off Primorye), negative wind stress curl is presumed to drive water columns toward deeper regions, hence enhancing the E-gyre. Note that the distribution of wind stress curl in November is somewhat different from the winter average, especially on the eastern side of the EJB (138°–139.5°E), as seen in Fig. A1.

  • View in gallery

    Time of occurrence of each float found in region WF (dots in red and blue) marked along the time series plot of wind stress curl averaged over the region WF. They subsequently trace out a trajectory of type BW-gyre (blue) or E-gyre (red).

  • View in gallery

    Climatological monthly mean wind stress curl from SCOW in the East/Japan Sea. During winter, the wind stress curl patterns are very similar to each other, with a dipole pattern in the northern EJS. The patterns are altogether different and weaker in other seasons.

  • View in gallery

    The 2-month mean wind stress curl during (a) March and April and (b) November and December in 2013 in the East/Japan Sea. Fig. A2a shows the presence of obvious negative wind stress curl region (42°N, 139.5°E; yellow circled region) in early spring. During November and December, the negative wind stress curl is weak, south of Primorye (Fig. 1; yellow circled region), compared with that of Fig. 4.

  • View in gallery

    The track of float 2900440 showing all locations between 28 Nov 2004 and 26 May 2013, with yellow points representing at surface locations during its communication with satellites. The surface drifting effect on middepth circulation is seen to be minor, even though this is the most extreme surface drifting case.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 133 133 19
PDF Downloads 99 99 9

Seasonal Variability in Middepth Gyral Circulation Patterns in the Central East/Japan Sea as Revealed by Long-Term Argo Data

View More View Less
  • 1 Korea Institute of Ocean Science and Technology, Ansan, South Korea
  • | 2 Geo System Research Corporation, Gunpo, South Korea
  • | 3 Kyungpook National University/Kyungpook Institute of Oceanography, Daegu, South Korea
  • | 4 Korea Institute of Ocean Science and Technology, Ansan, South Korea
© Get Permissions
Full access

Abstract

Trajectories of Argo floats deployed in the East/Japan Sea from 2001 to 2014 reveal that the middepth gyral circulation pattern of the Japan basin, the central part of the East/Japan Sea, undergoes a seasonal variation. The middepth circulation of the Japan basin is found to be characterized usually by the gyres trapped to the east of the Bogorov Rise (E-gyres) and those extending farther westward into the whole basin (BW-gyres). The E-gyre trajectories are generally associated with the turning of the floats toward deeper regions off the isobaths. This occurs in winter either on the northern or eastern side of the eastern Japan basin. It seems that the upstream part of the otherwise BW-gyre is subject to a strong negative wind stress curl in winter, and there the circulating water columns are driven toward the deeper region, thus triggering the formation of the E-gyre. The topographic effect associated with the Bogorov Rise seems to interfere thereafter in the process of determining the passage of the E-gyre. Otherwise, the water columns continue to flow along the isobaths, hence maintaining the BW-gyre. To the knowledge of the authors, this is the first observational evidence of seasonal variability in the middepth gyral circulation pattern in the East/Japan Sea. It suggests that oceanic middepth circulation, usually known to be quasi steady or slowly varying on climatological time scales, might also undergo a significant seasonal variation as it does in the East/Japan Sea.

Denotes Open Access content.

Current affiliation: Inha University, Incheon, South Korea.

Corresponding author address: Sok Kuh Kang, Korea Institute of Ocean Science and Technology, 787 Haeanro, Ansan 426-744, South Korea. E-mail: skkang@kiost.ac.kr

Abstract

Trajectories of Argo floats deployed in the East/Japan Sea from 2001 to 2014 reveal that the middepth gyral circulation pattern of the Japan basin, the central part of the East/Japan Sea, undergoes a seasonal variation. The middepth circulation of the Japan basin is found to be characterized usually by the gyres trapped to the east of the Bogorov Rise (E-gyres) and those extending farther westward into the whole basin (BW-gyres). The E-gyre trajectories are generally associated with the turning of the floats toward deeper regions off the isobaths. This occurs in winter either on the northern or eastern side of the eastern Japan basin. It seems that the upstream part of the otherwise BW-gyre is subject to a strong negative wind stress curl in winter, and there the circulating water columns are driven toward the deeper region, thus triggering the formation of the E-gyre. The topographic effect associated with the Bogorov Rise seems to interfere thereafter in the process of determining the passage of the E-gyre. Otherwise, the water columns continue to flow along the isobaths, hence maintaining the BW-gyre. To the knowledge of the authors, this is the first observational evidence of seasonal variability in the middepth gyral circulation pattern in the East/Japan Sea. It suggests that oceanic middepth circulation, usually known to be quasi steady or slowly varying on climatological time scales, might also undergo a significant seasonal variation as it does in the East/Japan Sea.

Denotes Open Access content.

Current affiliation: Inha University, Incheon, South Korea.

Corresponding author address: Sok Kuh Kang, Korea Institute of Ocean Science and Technology, 787 Haeanro, Ansan 426-744, South Korea. E-mail: skkang@kiost.ac.kr

1. Introduction

The East/Japan Sea (EJS) is an almost landlocked sea, without the exchange of intermediate and deep waters with the surrounding open oceans (Gamo et al. 1986). However, the EJS is often compared to a miniature ocean (Gamo 1999): it has both a subtropical-like and a subpolar-like region separated by a subpolar front. In the former, the surface current is anticyclonic, driven by an inflow and outflow, and is western intensified, whereas in the latter, the surface current is cyclonic with wintertime deep convection that is known to have been affected by recent global warming (Kim et al. 2001). Bottom water formation (Kim et al. 2002; Talley et al. 2003) in 2000/01 was reported to have taken place, but the corresponding thermohaline “conveyor belt system” has slowed down since the mid-1980s (Gamo 1999). On the other hand, the intermediate and central water masses have been more actively formed in wintertime, increasing their volumes in the EJS (Kim et al. 2004).

The upper part (generally shallower than 200 m) of the East/Japan Sea can be divided into a warm region to the south and a cold region to the north. The warm region is occupied by the warm Tsushima Current Water originating from the Kuroshio and entering the basin through the inlet located on the southern part of the basin and flowing out through the outlets located on the northeastern part of the basin. The cold region is not covered by the Tsushima Current Water and is separated from the warm region by a subpolar front. In the warm region, there is a strong stratification between the upper layer occupied by the Tsushima Current Water and the lower layer underlying it. The lower layer is generally divided into the intermediate (or middepth) and deep layers. However, the water density and current are known to vary very little throughout the whole lower layer (Kim et al. 2004; Chang et al. 2004). There is an indication that the currents in the lower layer of the warm region do not undergo a significant seasonal variation (Mitchell et al. 2005). In the cold region, where stratification is much weaker than in the warm region, the current is expected to be largely barotropic in the middepth and deep layers, especially in winter.

Many attempts have been made to understand the middepth circulation of the EJS (Danchenkov et al. 2003; Takematsu et al. 1999; Yoshikawa et al. 1999; Hogan and Hurlburt 2000; Yanagimoto and Taira 2003; Danchenkov et al. 2006; Choi and Yoon 2010; Choi 2015; Park et al. 2015). Generally, they show that it is cyclonic and flows along the isobaths. However, many questions still remain unanswered. Among others, it is quite interesting to know whether the middepth or deep circulation pattern undergoes a seasonal variation. In the warm region, the middepth or deep circulation does not seem to undergo a seasonal variation as described above. However, this may not be the case for the cold region because it has weaker stratification and stronger wind forcing. Analyzing the seasonal behavior of the middepth or deep circulation has now become possible because of 15 yr of Argo float data.

In this study, we use autonomous quasi-Lagrangian float trajectories to show that the middepth gyral circulation of the EJS can be classified into two patterns: one occupying the whole central basin and the other occupying only the eastern Japan basin (EJB). There is an indication that the latter is probably driven by negative wind stress curl prevailing in the northern/eastern part of the eastern Japan basin. The results of this study suggest that oceanic middepth gyral circulation, usually known to be quasi-steady or varying slowly on climatological time scales (Bower et al. 2002, 2009; Lavender et al. 2000), might also change its pattern seasonally.

2. Data and methods

The detailed descriptions of Argo and wind data used in this study are given as follows.

a. Argo data

The compiled Argo float data were obtained from the Global Data Assembly Center (GDAC; ftp://ftp.ifremer.fr/ifremer/argo/) (Roemmich et al. 2009; Freeland et al. 2010). From 2001 to 2014, nearly 150 Argo floats have been deployed at depths of 700–800 m in the EJS, mostly around the Ulleung basin (UB), and they have subsequently moved around the whole EJS basin (Fig. 1). Most of the deployed Argo floats survived for 36–60 months, with some of them still active, and moved northward and penetrated into the Japan basin. Argo floats occupy the depths 700–800 m. Hence, the corresponding trajectories may represent the middepth or deep circulation because, as noted earlier, the whole water column below the upper layer (~200 m) is nearly barotropic in the cold region (Kim et al. 2001; Chang et al. 2004). Since we are interested in the circulation within the JB, only those seven floats that have traveled sufficient distances within the JB are selected (Table 1). Each of these nine floats shows very complicated trajectories of various length scales (Fig. 1), indicating that the middepth circulation of the JB can be characterized by cyclonic gyres of various sizes. A very interesting point is that the gyres are quite often trapped to the east of the Bogorov Rise off Primorye (see Fig. 2 for location), which otherwise extend westward on various scales. Additionally, the Autonomous Profiling Explorer (APEX) float data obtained from 2000 to 2004 (Danchenkov et al. 2003) were also used in constructing the time series that measures the relative strength of the trapping gyres, as will be presented shortly.

Fig. 1.
Fig. 1.

Trajectories of the selected nine floats presented in different colors. Eight floats are deployed in the UB, with one in the eastern location of Primorye, and the deployment locations are marked with stars. See Table 1 for more details on the nine floats.

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

Table 1.

Argo data details used for analysis.

Table 1.
Fig. 2.
Fig. 2.

(a) Trajectories of Argo floats associated with the BW-gyre. The starting point of each trajectory is marked with a large, solid circle and the end point is marked with a triangle. Floats are the circled numbers 1 through 6, and numerals corresponding to year and month are marked along the trajectory of each float. The same notations are used in following figures. (b) Trajectories of Argo floats associated with the first E-gyre group. They turn toward deeper regions off the isobaths on the northern edge of the EJB north of 43°N. Floats are numbered from 1 to 6. (c) As in Fig. 2b, but for the second E-gyre group. They (floats 7–10) turn toward deeper regions off the isobaths on the eastern side of the EJB roughly between 42.5° and 43°N. Note that the region of negative wind stress curl reaches far offshore (138°–139.5°E) in November (Fig. A1).

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

b. Wind data

Wind data are also needed because the observed circulations may be related to the wind stress field. For this purpose, climatological wind stress curl (Fig. A1) is obtained using the Scatterometer Climatology of Ocean Winds (SCOW; ftp://numbat.coas.oregonstate.edu/pub/scow/) data based on 10 yr (September 1999–October 2009) of QuikSCAT scatterometer measurements. On the other hand, the time series of wind stress curl is obtained using QuikSCAT (ftp://ftp.ifremer.fr/ifremer/cersat/products/gridded/MWF/L3/QuikSCAT/Monthly/) and Advanced Scatterometer (ASCAT; ftp://ftp.ifremer.fr/ifremer/cersat/products/gridded/MWF/L3/ASCAT/Monthly/) data. QuikSCAT data are for the period from 2000 to 2007, and ASCAT data are for the period from 2008 to 2014. ASCAT data are also used to check the magnitude of the wind stress curl during March and April (Fig. A2a) and during November and December 2013 (Fig. A2b). More details on the wind data are described in the appendix.

Finally, a special comment may be needed about possible errors in estimating the middepth circulation using the Argo float trajectories. The estimated circulations are inevitably affected by surface currents, while the floats periodically move up and down between the middepth and the surface, thereby contaminating the quality of estimation. This problem, however, is not considered to be serious for the following reasons: First, the average time period of staying above the middepth for each float is about 6 h (Park et al. 2005), which is only 10% of the total floating period (Fig. B1). Second, the currents in this region are known to be nearly barotropic.

3. Results

As noted earlier, the trajectories of the nine floats indicate that the middepth circulation in the JB can generally be characterized by the gyres trapped to the east of the Bogorov Rise, referred to as “E-gyres” (Fig. 2b), and those extending farther westward into the whole basin, referred to as “BW-gyres” (Fig. 2a). The BW-gyres can generally be considered to have length scales on the order of the basin scale and complete their circulations in about 430–970 days. The E-gyres have length scales of 100–400 km and time scales of 180–420 days. It is interesting to note that the smaller-scale E-gyres tend to be centered to the south compared with the larger-scale E-gyres (Fig. 1).

a. BW-gyre in central basin of the East/Japan Sea

The trajectories of six BW-gyres (Fig. 2a) show the circulation over the whole basin. Most of the BW-gyres (floats 1–5) show that they pass through the northern EJB during times other than winter. Float 1, traveling northward, arrives at its northernmost position of 43.8°N in August 2008, moves westward/southwestward during autumn months in the northern EJB, and continues to move southwestward during November to January 2009. Finally, in January and February 2009, it moves southward off Vladivostok. Float 2 arrives at its northernmost position of 43.7°N in March 2010 and then moves southwestward along isobaths in the northern EJB from March to May 2010. Finally, it arrives at a location off Vladivostok and turns southward there. Float 3 arrives at its northernmost position of 43.7°N in March 2010 and then moves southwestward along isobaths in the northern EJB from March to June 2010, arrives at a location off Vladivostok during autumn, and turns southward there in January 2011. Float 4 arrives at its northernmost position of 43.5°N in late June 2011, moves westward roughly along the 3600-m isobath in the northern EJB during July and August, and then continues to move southwestward during August and September 2011 before it arrives between Vladivostok and Primorye. Float 5 arrives at its northernmost position of 43.6°N in early May 2010, moves southwestward during May and June along the 3600-m isobath in northern EJB, and then keeps moving southwestward during August 2010 before it reaches 134°E between Vladivostok and Primorye. All these floats are seen to move along isobaths, especially when they are in the northern part of the EJB.

Float 6 is an exception in that it passes through the northern EJB in winter. A close examination indicates that it slightly crosses isobaths in the northern EJB, while floats 1–5 follow the isobaths there. For example, float 6 arrives at its northernmost position of 44.0°N in September 2013, moves southwestward across the 3000-m isobath in October, crosses the 3500-m isobath in November and finally moves across the 3600-m isobath in December. Afterward, it meets the northwestern slope of the Bogorov Rise, follows the western side of the Bogorov Rise, and finally joins to the BW-gyre. The cross-isobathic movement toward deeper regions is likely due to negative wind stress curl, as will be explained later. Also, it will become clear shortly that the float 6 trajectory would form an E-gyre if the negative wind stress curl were sufficiently strong.

The BW-gyre trajectories shown above look very similar to various numerical model results obtained with annual mean or seasonal wind (Hogan and Hurlburt 2000; Yoshikawa 2012; Park et al. 2015; Choi 2015).

b. E-gyre

Trajectories in the JB also show shorter-scale gyres mostly centered in the EJB, as shown in Fig. 1. The presence of the E-gyre has already been remarked by Choi and Yoon (2010) but has not been rigorously discussed. To examine the behavior of E-gyres more closely, the corresponding trajectories made by 10 floats are shown. For visual convenience, they are separated into two groups (Figs. 2b,c). The first group (floats 1–6) consists of midsize gyres generally extending to the north of 43°N and nearly occupying the whole EJB. The second group (floats 7–10) consists of small gyres occupying the southern EJB with a limited northward extension.

In general, the E-gyre floats in the first group are around their northernmost position in winter. Float 1 traveling southward arrives at the 3000-m isobath in early December 2009 and keeps moving southward during December 2009, traveling east of Bogorov Rise during January 2010. Float 2 arrives at its northernmost position of 43.7°N in mid-November 2011 and then crosses the isobaths southward during December 2011 and January 2012. Float 3 arrives at its northernmost position of 43.4°N in late-February 2013, moving slightly southward during March and April and then moving rapidly southward during May. Float 4 arrives at its northernmost position of 43.3°N in mid-November 2012, after crossing isobaths, and then moves southwestward during December 2012 along the eastern side of Bogorov Rise.

Float 5, traveling northwestward, arrives at its northernmost position of 42.9°N in late December 2010, moves southward during January and February 2011, and then begins moving eastward during mid-February, forming the smallest E-gyre. After completing this gyral circulation, this float continues to move, forming a different type of trajectory that, for convenience, is named float 6. Unlike the other E-gyres in the first group, however, float 6 happens to arrive at its northernmost position of 43°N in early August 2011, not in winter. Afterward, it travels southward following isobaths during October 2011 to February 2012 and forming a larger E-gyre. It seems that float 6 has been already trapped deep inside the basin before it begins its gyral motion and hence does not need to follow the isobaths.

The floats (7, 9, and 10) in the second group generally tend to turn toward deeper regions off the isobaths in early winter around November. Float 7 moves along isobaths in the eastern EJB during September and October 2013, turns westward into deeper regions off the 3600-m isobath at 42.8°N in early November 2013, travels farther westward in December, and turns southward in January 2014 before completing its E-gyre. Float 8 is an exception in that it moves along the isobaths in the eastern EJB, turns westward into deeper regions during late March and April 2013, reaches its northernmost position, travels westward, and then moves southward after showing a small cyclonic loop during June and July. In general, the turning of floats into deeper regions off the isobaths is associated with negative stress curl. In fact, the exceptional behavior of float 8 seems to be due to an unusual development of negative wind stress curl in the early spring of 2013 (see Fig. A2a in comparison with the climatological wind stress curl in Fig. A1). Float 9, after crossing the 3600-m isobaths in the eastern EJB in early November 2012, arrives at its northernmost position of 42.8°N in mid-December and then flows southwestward during December and January 2013, forming an E-gyre. Float 10, after crossing the 3600-m isobath at 42.3°N in the eastern EJB in early November 2004, arrives at its northernmost position of 42.9°N in mid-December 2004, moves southwestward during January 2005, and goes southward during February 2005. The climatological wind stress curl in November (Fig. A1) shows that the region of negative wind stress curl reaches far offshore, extending to 138°–139.5°E, probably triggering the formation of the second group of E-gyres in early winter.

Overall, the formation of the E-gyres seems to be correlated with negative wind stress curl in winter. In fact, as described in the next section, negative wind stress curl may effectively push the water column into deeper regions and across isobaths, triggering the formation of an E-gyre. In the absence of the wind stress curl, the middepth circulation would extend westward, following the isobaths, into the whole basin, hence maintaining the BW-gyre rather than the E-gyre. Note that the middepth circulation of the JB has been traditionally known to take place over the whole basin, similar to the BW-gyre defined here, with currents generally following the isobaths (Hogan and Hurlburt 2000; Park et al. 2005; Choi and Yoon 2010; Park and Kim 2013).

Figures 3a and 3b show quasi-Eulerian current vectors obtained from all the floats present during summer and winter, respectively. It is seen that in summer, there is a strong along-isobathic flow through the region 41.5°–42.5°N, 134°–136°E, referred to as region PS (Fig. 3a), whereas in winter, there is a strong southward flow through the region 42°–43°N, 136°–137°E, referred to as region PW (Fig. 3b). An exception is that westward currents are present in region PS in winter (yellow vectors in Fig. 3b). This seems to be because the wind stress curl in this region was weaker in November and December 2013 (Fig. A2b) than the climatological average (Fig. 4). An exception present in region PW in summer (red vectors in Fig. 3a) is related to float 6 of the BW-gyre due to the trapped motion inside the basin in summer, as discussed in the BW-gyre.

Fig. 3.
Fig. 3.

Quasi-Eulerian vectors obtained from Argo data for (a) summer (June to September) and (b) winter (December to March). Strong along-isobathic currents are present in region PS in summer, whereas strong southward currents are present in the region PW in winter. Red (yellow) vectors indicate the exceptional trajectory associated with float 6 of E-gyre (float 6 of BW-gyre), as explained in the text.

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

Fig. 4.
Fig. 4.

Wind stress vectors (arrows) and wind stress curl (colored shading) in winter (November–February) overlaid with the trajectories associated with the E-gyre (Figs. 2b and 2c). In region WF (the rectangular area off Primorye), negative wind stress curl is presumed to drive water columns toward deeper regions, hence enhancing the E-gyre. Note that the distribution of wind stress curl in November is somewhat different from the winter average, especially on the eastern side of the EJB (138°–139.5°E), as seen in Fig. A1.

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

4. Discussion

An interesting question that arises is why the flow changes its direction, from along isobathic to cross isobathic, in winter. It is likely that wind forcing plays an important role in this process, either wind stress itself or its curl. To answer this question, trajectories of the E-gyre (from Figs. 2b,c) are overlaid with the horizontal fields of climatological winter wind stress and its curl (Fig. 4). The dipole pattern of wind stress curl in the northern EJS, which has already been known to be due to orographic effects (Yoon et al. 2005), is evident. In the region of the E-gyres, no apparent relationship can be found between wind drag and trajectories. Instead, the turning of flow direction inherent to the E-gyre of the first group seems to take place in the region of negative stress curl, referred to as region WF (rectangular area off Primorye in Fig. 4). For the E-gyres in the second group, the turning occurs earlier around November on the eastern side of the EJB. The presence and location of negative wind stress curl at this time of year is not clear in the seasonal-mean wind field (Fig. 4) but can be seen in the monthly mean November wind field (Fig. A1). It is not quite clear whether similar processes occur in the region farther west. However, in the region around 41°–42°N, 132°–134°E, cross-isobath movements appear to be stronger in winter than in summer (cf. Figs. 3a and 3b). The fact that stress curl is more related to circulation than stress itself may be further supported by the following considerations. First, the Ekman depth does not usually extend downward beyond a few hundred meters, and second, the water column is strongly subjected to the Coriolis’ effect. Assume that (i) the ocean is nearly barotropic, (ii) wind stress dominates over the bottom frictional stress, and (iii) the topographic beta effect is locally much larger than the planetary beta effect. Then, negative wind stress curl drives water columns toward regions of smaller background potential vorticity and vice versa. In the study area, the geostrophic contours generally run in the same direction as the isobaths, that is, roughly in the east–west direction on the northern side and north–south direction on the eastern side. Hence, one can expect that water columns flowing westward (northward) north of the region PW (on the eastern side of the EJB) in winter are driven southward (westward) toward deeper regions by negative wind stress curl. It seems that a topographic effect might interfere thereafter in the process of determining the passage of the E-gyre because, for the E-gyres in the first group, floats usually go along the eastern side of the Bogorov Rise, that is, through the region PW. The circulation patterns in spring and fall are similar to those in summer (Fig. 3a), although they are not presented here.

Consider now the floats passing through the region WF. Many of them, including the nine floats considered earlier (Table 1), make trajectories that can be classified into either BW- or E-gyres or others that do not make any gyral shape trajectories. On the other hand, recalling the important role played by wind stress curl in region WF, it is expected that the floats found in region WF in winter will later trace out E-gyre trajectories, while those found in the same region in other seasons will later trace out BW-gyre trajectories. In other words, the time when floats are found in region WF, referred to as “time of occurrence,” may subsequently determine the type of trajectories they make. This presumption can be checked by investigating the time of occurrence of each float found in region WF and the type of trajectory it subsequently makes. The time of occurrence and the type of trajectory thus obtained are marked along a time series plot of wind stress curl averaged over the region WF (Fig. 5). In obtaining these values, the data from 2000 to 2004 are adopted from a previous work (Danchenkov et al. 2003), and those floats whose trajectories do not show any clear gyral shape are excluded. Indeed, there is a tendency that the floats found in the region WF in winter are subject to a strong negative wind stress curl and subsequently trace out E-gyre trajectories. Those found in WF in other seasons, trace out BW-gyres. The above explanation applies more to E-gyres in the first group of floats. For the second group, the same explanation can be applied. In the eastern part of the EJB, the water column is driven westward toward deeper regions by a negative wind stress curl, which otherwise continues to move northward along isobaths to form BW-gyres.

Fig. 5.
Fig. 5.

Time of occurrence of each float found in region WF (dots in red and blue) marked along the time series plot of wind stress curl averaged over the region WF. They subsequently trace out a trajectory of type BW-gyre (blue) or E-gyre (red).

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

5. Concluding remarks

Trajectories of Argo floats deployed at depths of 700–800 m in the EJS from 2001 to 2014 reveal that the middepth gyral circulation pattern of the JB undergoes a seasonal variation. The middepth circulation can usually be characterized by the gyres trapped to the east of the Bogorov Rise, referred to as E-gyres, and those extending farther westward into the whole basin, referred to as BW-gyres. The E-gyre trajectories are generally associated with the turning of the floats toward deeper regions off the isobaths. This occurs in winter either on the northern or eastern side of the eastern Japan basin. It seems that in the upstream part of the otherwise BW-gyre, a strong negative wind stress curl drives the circulating water columns toward deeper regions, triggering the formation of the E-gyre. The topographic effect caused by the Bogorov Rise seems to interfere thereafter in the process of determining the passage of an E-gyre. Otherwise, the water columns continue to flow along the isobaths, hence maintaining the BW-gyre. To our knowledge, this may be the first report of seasonal variability of middepth gyral circulation pattern. This result suggests that oceanic middepth circulations, usually known to be quasi steady or varying slowly on climatological time scales, might also undergo significant seasonal variations as in the EJB.

Although it is evident that the middepth circulation undergoes a seasonal variation in the JB, the dynamics of how wind enhances the E-gyre and how the topographic effect associated with the Bogorov Rise interferes thereafter in the process of determining the passage of the E-gyre are not clearly understood. Direct field measurements, including the mooring of current meters at the locations of interest, together with further float tracking may help to resolve this problem. It is quite unfortunate that there has not been any numerical model result supporting the seasonal variation of the middepth gyral circulation pattern found in this study. It is quite probable that there are still some processes, like the formation of the E-gyres considered here, that are too complicated to be reproduced accurately by numerical models. Perhaps, the dynamics may involve various interactions among the eddy, topography, mean current, and even wind forcing.

Acknowledgments

We express our thanks to Dr. M. G. G. Foreman of the Institute of Ocean Sciences (IOS), Canada, for reading the manuscript and providing English corrections. We thank two anonymous reviewers for many constructive comments and suggestions. The Argo data were made available in the public domain. This work was partially supported by grants from KIOST Research programs (PE99396 and PE99392) and was performed as a collaborative research project of project No (Development of HPC-based management system against national-scale disaster) and supported by the Korea Institute of Science and Technology Information (KISTI). J. H. L. and Y. H. K. were supported by the National Research Foundation of Korea grant funded by the South Korean government (NRF-2009-C1AAA001-0093065).

APPENDIX A

Climatological and Time Series Wind Data

a. Climatology data of ocean winds (SCOW)

The SCOW data are based on 10 yr (September 1999–October 2009) of QuikSCAT scatterometer data. Monthly wind stress curl data from these data are presented in Fig. A1 from January to December. Winter wind stress curl (Fig. 4) is averaged for the three winter months (December to February).

Fig. A1.
Fig. A1.

Climatological monthly mean wind stress curl from SCOW in the East/Japan Sea. During winter, the wind stress curl patterns are very similar to each other, with a dipole pattern in the northern EJS. The patterns are altogether different and weaker in other seasons.

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

b. Quick Scatterometer

The Quick Scatterometer (a SeaWinds instrument placed in orbit quickly) was launched in June 1999 and operated until November 2009. QuikSCAT provided measurements of wind speed and direction referenced to 10 m above the sea surface at a spatial resolution of 25 km (operator: NASA/JPL; FTP: ftp://ftp.ifremer.fr/ifremer/cersat/products/gridded/MWF/L3/QuikSCAT/Monthly/).

c. Advanced Scatterometer

Gridded monthly wind fields have been estimated over the global ocean for the period of April 2007 to the present. ASCAT wind fields are estimated as equivalent to neutral stability 10-m daily winds and have spatial resolutions of 0.25° in longitude and latitude. The 2-month (March and April 2013) means of the wind stress curl in EJS are also presented in Fig. A2a, along with 2-month mean wind stress curl data (Fig. A2b) during November and December 2013 (FTP: ftp://ftp.ifremer.fr/ifremer/cersat/products/gridded/MWF/L3/ASCAT/Monthly/.)

Fig. A2.
Fig. A2.

The 2-month mean wind stress curl during (a) March and April and (b) November and December in 2013 in the East/Japan Sea. Fig. A2a shows the presence of obvious negative wind stress curl region (42°N, 139.5°E; yellow circled region) in early spring. During November and December, the negative wind stress curl is weak, south of Primorye (Fig. 1; yellow circled region), compared with that of Fig. 4.

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

QuikSCAT data were used for 2000 to 2007, and ASCAT data were used for 2008 to 2014.

APPENDIX B

Surface Drifting Distance

The mean duration of the Argo floats at the sea surface is about 4.5–6.5 h (4–6 h for the communication with satellites and 0.5 h of unknown surface drift). All the drifting locations in the surface are plotted in Fig. B1, which demonstrates their negligible effect on the pattern of each gyre track. Thus, such an effect is too small to change the middepth circulation patterns.

Fig. B1.
Fig. B1.

The track of float 2900440 showing all locations between 28 Nov 2004 and 26 May 2013, with yellow points representing at surface locations during its communication with satellites. The surface drifting effect on middepth circulation is seen to be minor, even though this is the most extreme surface drifting case.

Citation: Journal of Physical Oceanography 46, 3; 10.1175/JPO-D-15-0157.1

REFERENCES

  • Bower, A. S., and et al. , 2002: Directly measured mid-depth circulation in the northeastern North Atlantic Ocean. Nature, 419, 603607, doi:10.1038/nature01078.

    • Search Google Scholar
    • Export Citation
  • Bower, A. S., , M. S. Lozier, , S. F. Gary, , and C. W. Boning, 2009: Interior pathways of the North Atlantic meridional overturning circulation. Nature, 459, 243248, doi:10.1038/nature07979.

    • Search Google Scholar
    • Export Citation
  • Chang, K. I., and et al. , 2004: Circulation and currents in the southwestern East/Japan Sea: Overview and review. Prog. Oceanogr., 61, 105156, doi:10.1016/j.pocean.2004.06.005.

    • Search Google Scholar
    • Export Citation
  • Choi, Y. J., 2015: Comparative simulation study of effects of eddy-topography interaction in the East/Japan Sea deep circulation. Acta Oceanol. Sin., 34, 118, doi:10.1007/s13131-015-0693-1.

    • Search Google Scholar
    • Export Citation
  • Choi, Y. J., , and J. H. Yoon, 2010: Structure and seasonal variability of the deep mean circulation of the East Sea (Sea of Japan). J. Oceanogr., 66, 349361, doi:10.1007/s10872-010-0031-y.

    • Search Google Scholar
    • Export Citation
  • Danchenkov, M. A., , S. C. Riser, , and J. H. Yoon, 2003: Deep currents of the central Sea of Japan. Pacific Oceanogr. J., 1, 611.

  • Danchenkov, M. A., , V. B. Lobanov, , S. C. Riser, , K. Kim, , M. Takematsu, , and J.-H. Yoon, 2006: A history of physical oceanographic research in the Japan/East Sea. Oceanography, 19, 1831, doi:10.5670/oceanog.2006.41.

    • Search Google Scholar
    • Export Citation
  • Freeland, H. D., and et al. , 2010: Argo—A decade of progress. Proc. OceanObs'09: Sustained Ocean Observations and Information for Society, Venice, Italy, IOC/UNESCO/ESA, 21–25, doi:10.5270/OceanObs09.cwp.32.

  • Gamo, T., 1999: Global warming may have slowed down the deep conveyor belt of a marginal sea of the northwestern Pacific: Japan Sea. Geophys. Res. Lett., 26, 31373140, doi:10.1029/1999GL002341.

    • Search Google Scholar
    • Export Citation
  • Gamo, T., , Y. Nozaki, , H. Saki, , T. Nakai, , and H. Tsubota, 1986: Spatial and temporal variations of water characteristics in the Japan Sea bottom layer. J. Mar. Res., 44, 781793, doi:10.1357/002224086788401620.

    • Search Google Scholar
    • Export Citation
  • Hogan, P. J., , and H. E. Hurlburt, 2000: Impact of upper ocean–topographical coupling and isopycnal outcropping in Japan/East Sea models with 1/8° to 1/64° resolution. J. Phys. Oceanogr., 30, 25352561, doi:10.1175/1520-0485(2000)030<2535:IOUOTC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kim, K., , R.-R. Kim, , D.-H. Min, , Y. Volkov, , J.-W. Yoon, , and M. Takematsu, 2001: Warming and structural changes in the East (Japan) Sea: A clue to future changes in global oceans. Geophys. Res. Lett., 28, 32933296, doi:10.1029/2001GL013078.

    • Search Google Scholar
    • Export Citation
  • Kim, K., , K.-R. Kim, , Y.-G. Kim, , Y.-K. Cho, , D.-J. Kang, , M. Takematsu, , and Y. Volkov, 2004: Water masses and decadal variability in the East Sea (Sea of Japan). Prog. Oceanogr., 61, 157174, doi:10.1016/j.pocean.2004.06.003.

    • Search Google Scholar
    • Export Citation
  • Kim, K.-R., , G. Kim, , K. Kim, , V. Lobanov, , V. Ponomareyv, , and A. Salyuk, 2002: A sudden bottom-water formation during the severe winter 2000–2001: The case of the East/Japan Sea. Geophys. Res. Lett., 29, doi:10.1029/2001GL014498.

    • Search Google Scholar
    • Export Citation
  • Lavender, K. L., , R. E. Davis, , and W. B. Owens, 2000: Mid-depth recirculation observed in the interior Labrador and Irminger Seas by direct velocity measurements. Nature, 407, 6669, doi:10.1038/35024048.

    • Search Google Scholar
    • Export Citation
  • Mitchell, D. A., and et al. , 2005: Upper circulation patterns in the Ulleung basin. Deep-Sea Res. II, 52, 16171638, doi:10.1016/j.dsr2.2003.09.005.

    • Search Google Scholar
    • Export Citation
  • Park, J. J., , and K. Kim, 2013: Deep currents obtained from Argo float trajectories in the Japan/East Sea. Deep-Sea Res. II, 85, 169181, doi:10.1016/j.dsr2.2012.07.032.

    • Search Google Scholar
    • Export Citation
  • Park, J. J., , K. Kim, , B. A. King, , and S. C. Riser, 2005: An advanced method to estimate deep currents from profiling floats. J. Atmos. Oceanic Technol., 22, 12941304, doi:10.1175/JTECH1748.1.

    • Search Google Scholar
    • Export Citation
  • Park, Y., , H.-R. Shin, , J.-H. Yoon, , C.-H. Kim, , and Y. Yoshikawa, 2015: Simulation of eddy-driven deep circulation in the East/Japan Sea by using a three-layer model with wind, throughflow and deep water formation forcings. J. Mar. Syst., 150, 4155, doi:10.1016/j.jmarsys.2015.05.006.

    • Search Google Scholar
    • Export Citation
  • Roemmich, D., and et al. , 2009: The Argo Program: Observing the global ocean with profiling floats. Oceanography, 22, 3443, doi:10.5670/oceanog.2009.36.

    • Search Google Scholar
    • Export Citation
  • Takematsu, M., , Z. Nagano, , A. G. Ostrovskii, , K. Kim, , and Y. Volkov, 1999: Direct measurements of deep current in the northern Japan Sea. J. Oceanogr., 55, 207216, doi:10.1023/A:1007842013257.

    • Search Google Scholar
    • Export Citation
  • Talley, L. D., , V. Lobanov, , V. Ponomarev, , A. Salyuk, , P. Tishchenko, , and I. Zhabin, 2003: Deep convection and brine rejection in the Japan Sea. Geophys. Res. Lett., 30, 11591162, doi:10.1029/2002GL016451.

    • Search Google Scholar
    • Export Citation
  • Yanagimoto, D., , and K. Taira, 2003: Current measurements of the Japan Sea Proper Water and the Intermediate Water by ALACE floats. J. Oceanogr., 59, 359368, doi:10.1023/A:1025572112019.

    • Search Google Scholar
    • Export Citation
  • Yoon, J. H., , K. Abe, , T. Ogata, , and Y. Wakamatsu, 2005: The effects of wind-stress curl on the Japan/East Sea circulation. Deep-Sea Res. II, 52, 18271844, doi:10.1016/j.dsr2.2004.03.004.

    • Search Google Scholar
    • Export Citation
  • Yoshikawa, Y., 2012: An eddy-driven abyssal circulation in a bowl-shaped basin due to deep water formation. J. Oceanogr., 86, 971983, doi:10.1007/s10872-012-0148-2.

    • Search Google Scholar
    • Export Citation
  • Yoshikawa, Y., , T. Awaji, , and K. Akitomo, 1999: Formation and circulation processes of intermediate water in the Japan Sea. J. Phys. Oceanogr., 29, 17011722, doi:10.1175/1520-0485(1999)029<1701:FACPOI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
Save