- Email: [email protected]
Two processes by which short-period fluctuations in the meander of the Kuroshio affect its countercurrent
Deep-Sea Research I 47 (2000) 745}754
Note
Two processes by which short-period #uctuations in the meander of the Kuroshio a!ect its countercurrent Shingo Kimura*, Takashige Sugimoto Ocean Research Institute, University of Tokyo, 1-15-1, Minamidai, Nakano-ku, Tokyo, 164-8639, Japan Received 9 July 1998; accepted 8 June 1999
Abstract Direct current velocity measurements in the countercurrent of the Kuroshio, south of Japan, were carried out to investigate the in#uence of short-period #uctuations in the small-scale meander of the Kuroshio on its countercurrent. When the Kuroshio took a path having a meander west of the Izu Ridge and approaching the Izu Peninsula, the countercurrent freely intruded into coastal seas with a period of 17 d and a phase velocity almost equal to that of the Kuroshio itself. However, when the Kuroshio did not signi"cantly bend and de#ect o! the Izu Peninsula, even when taking the same path, the velocity of the countercurrent was considerably reduced and the periodic #uctuations propagated into the coastal seas as a continental shelf wave. The results indicate that a small change in the Kuroshio's path can cause a di!erent process of propagation of the small-scale meandering; this di!erence probably explains why there are two kinds of phase velocity in coastal temperature #uctuations. ( 2000 Elsevier Science Ltd. All rights reserved. Keywords: Kuroshio; Small-scale meander; Countercurrent; Continental shelf wave; Gulf Stream
1. Introduction The Kuroshio is a western boundary current in the Paci"c Ocean and is considered to be a counterpart of the Gulf Stream in the Atlantic Ocean. A signi"cant di!erence between the two currents is that the Kuroshio has bimodal stationary #ow patterns
* Corresponding author. Fax: 0081-3-5351-6506. E-mail address: [email protected] (S. Kimura) 0967-0637/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 7 - 0 6 3 7 ( 9 9 ) 0 0 0 6 7 - 9
746
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
(known as the large and small meander paths) o! the southern part of the main island of Japan (Stommel and Yoshida, 1972). The large meander emerges with periods of 20 yr and 7}8.5 yr (Kawabe, 1987), and its large #uctuations are limited to the Enshu-nada Sea, east of the Kii Peninsula (Kawabe, 1985,1986). Small-scale meanders with shorter periods are also dominant in the Kuroshio region, and they are well recognized in satellite images as shown in this paper. In association with the small-scale meander, displacement of the Kuroshio front #uctuates with a period of 17}19 d and 400 km wavelength (Kimura and Sugimoto, 1993). Similar periodic #uctuations of the Kuroshio front are widely recognized south of Japan (Qiu et al., 1990; Maeda et al., 1993; Kasai et al., 1993), and it is well known that cyclonic eddies associated with the frontal meandering induce upwelling (Toda, 1993; Ito et al., 1995). This upwelling supplies nutrients to the euphotic layer in the cyclonic eddy and accelerates primary production in o!shore regions (Kimura et al., 1997), as do similar processes identi"ed in the Gulf Stream system (Lee et al., 1981; Yoder et al., 1981; Atkinson et al., 1996). East of the Kii Peninsula, when a meander of the Kuroshio is located west or over the center of the Izu Ridge, a warm water mass derived from the Kuroshio often intrudes into the coastal areas of the Kumano-nada and Enshu-nada Seas with a dominant period of about 20 d (Kimura and Sugimoto, 1987,1988,1990). The good periodic correspondence between the Kuroshio region and the coastal region indicates that the small-scale meander of the Kuroshio in#uences coastal oceanographic change. However, the observed coastal temperatures along the Kumano-nada Sea suggest that there are two distinct cases where temperature #uctuations propagate, one with a few days time-lag and the other with no time-lag (Kimura and Sugimoto, 1987; Kimura et al., 1994). While variation in the Kuroshio's path may in#uence variation in the time-lag, no physical explanation for the variation has yet been proposed. Accordingly, we carried out observations over an 18 month period, using moored current meters, on the continental shelf of the Kumano-nada Sea to understand the frequency structures of the coastal countercurrent of the Kuroshio and their relationships to the Kuroshio's path type.
2. Observations and data Direct current measurements in the Kumano-nada Sea were made during Period 1 (9 December 1986}22 April 1987) and Period 2 (25 April 1987}28 April 1988). Two mooring stations (CM-4 and CM-7) were placed at a depth of about 800 m on the continental slope east of the Kii Peninsula along the expected direction of the countercurrent (Fig. 1). Each mooring consisted of two current meters, at depths of about 400 m (upper level) and 700 m (lower level), although current meters could not always be situated at predetermined depths. The data were sampled at 30 min intervals during Period 1 and at hourly intervals during Period 2 by Aanderaa RCM-5 current meters. Fig. 2 shows the time series of low-frequency current velocity after "ltering with a 25-h running mean to remove high-frequency #uctuations. Basic information and
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
747
Fig. 1. Observational locations and schematic views of the Kuroshio's paths during Period 1(bold line) and Period 2 (thin line).
statistics of the observations are summarized in Table 1. Time-averaged velocities of the eastward component (E}W velocity) and northward component (N}S velocity) are denoted as ; and <, respectively, and the variances of each velocity are denoted as u@2 and v@2. KE (eddy kinetic energy) and KM (mean kinetic energy) are de"ned as (u@2#v@2)/2 and (;2#<2)/2, respectively. Hereafter, CM4U (CM-4 upper layer), CM7U (CM-7 upper layer), CM4L (CM-4 lower layer) and CM7L (CM-7 lower layer) are used to refer to the current meters.
3. Results 3.1. Time-mean structure of the countercurrent of the Kuroshio Current directions at the upper levels, CM4U and CM7U, were westward and southwestward along the coast, respectively, throughout both observational periods. The tangential components of the velocities were much larger than the normal components. In particular, the ratios of the eddy kinetic energy to the mean kinetic energy of current velocity (KE/KM) in Period 1 were less than 1.0, which indicated
748
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Fig. 2. Time series of current velocity in Period 1 (upper panel) and Period 2 (lower panel). Water depths and current meter depths are shown at right.
that the current at each station #owed in one direction with no large #uctuations. The current #ow pattern at CM4L was similar to those at CM4U and CM7U although the mean velocity was small. However, the current direction at CM7L was strongly a!ected by the local seabed topography and showed a #ow pattern quite di!erent from the other three current meters. During both observational periods, the Kuroshio usually meandered west of the Izu Ridge and approached the Izu Peninsula. Fig. 3b and d show estimated paths of the Kuroshio in each period, de"ned as contours of water temperature of 153C at a depth of 200 m, from biweekly published reports of the Hydrographic Department,
} Apr. 1987) 780 380 860 460 780 680 860 760
} Apr. 1988) 833 464 845 476 833 718 845 730
Period 2 (Apr. 1987 CM-4 (UPPER) CM-7 (UPPER) CM-4 (LOWER) CM-7 (LOWER)
Sensor (m)
Period 1 (Dec. 1986 CM-4 (UPPER) CM-7 (UPPER) CM-4 (LOWER) CM-7 (LOWER)
Depth (m)
9.7 5.1 2.3 0.4
19.9 12.1 3.8 0.3
14.5 8.8 5.6 1.8
21.3 13.5 7.0 1.6
!7.0 !5.1 !1.0 0.3
!16.0 !11.9 !2.9 0.3
; (cm/s)
Vector (cm/s)
Scalar
Velocity
Velocity
!6.7 !0.3 !2.0 0.2
!11.8 !2.1 !2.5 0.2
<
107 84 22 2
79 73 30 1
u@2 (cm/s)2
99 8 19 3
56 15 23 3
v@2
Var. of vel.
Table 1 Basic information and statistics of observations with current meters at CM4 and CM7
226 267 207 56
234 260 229 56
230 290 220 340
230 270 230 0
47.3 12.9 2.6 0.1
198.4 72.7 7.1 0.1
KM (cm/s)2
Mean (deg.)
Dominant
Energy
Direction
103.1 46.1 20.7 2.2
67.5 43.9 26.4 2.0
KE
2.2 3.6 8.1 32.2
0.3 0.6 3.7 37.3
Ratio KE/KM
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754 749
750
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Fig. 3. Satellite images and estimated Kuroshio paths de"ned by contours of water temperature of 153C at a depth of 200 m during Period 1 (a, b) and Period 2 (c, d). The paths in Period 2 are plotted every second contour.
Maritime Safety Agency of Japan; schematic views of their averaged paths are indicated in Fig. 1. When the Kuroshio takes these path types, warm water masses separated from the main stream of the Kuroshio o! the Izu Peninsula intrude into the Enshu-nada Sea and move westward into the Kumano-nada Sea as a countercurrent of the Kuroshio between the Kuroshio and the coast (Kimura and Sugimoto, 1987,1988,1990; Kasai et al., 1993). In#uenced by the circulation, current measurements at CM4U and CM7U showed a westward and southwestward movement, respectively, along the coast. However, vector-averaged velocities during Period 2, of 10 cm s~1 (at CM4U) and 5 cm s~1 (at CM7U), were half those of Period 1 and the eddy kinetic energy was larger than mean kinetic energy in Period 2. According to the estimated paths, the Kuroshio's path bent and de#ected much further to the west o! the Izu Peninsula in Period 1 than in Period 2. A satellite thermal infrared image (Fig. 3a) shows an example of intrusion of the Kuroshio water as a countercurrent during Period 1. While the Kuroshio's path type in Period 2 was similar to that in Period 1, similar intrusions of warm Kuroshio water and de#ection of the path were not recognized in Period 2 and the Kuroshio usually #owed as shown in the satellite image in Fig. 3c. We therefore infer that this small change of the path type probably
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
751
explains the di!erence in strength of the Kuroshio countercurrent between Periods 1 and 2. 3.2. Frequency structure of the current velocity and its phase lag Periodic #uctuations with intervals of a week were readily recognized from the time series of current velocity as shown in Fig. 2, particularly in velocity at CM4L and CM7U in Period 1 with CM-7 leading CM-4 by a time-lag of a few days. Fig. 4 shows power spectra of the tangential component of current velocity along the mean current directions in Periods 1 and 2 and indicates the existence of a dominant 17-d periodicity in CM7U and CM4L throughout both observational periods. At CM4U, the spectral peaks in Period 1 were seen to be similar, and in Period 2 identical. The peaks of the 17-d period were more obvious in Period 1 than in Period 2. However, periods of 7}12 ds, much shorter than 17 days, were signi"cantly more dominant in Period 2. Focusing on the 17-day period observed at CM7U and CM4L, which had also been identi"ed as a period of the small-scale meander of the Kuroshio (Kimura and Sugimoto, 1993), coherence and phase lag between the stations were calculated (Fig. 5). In period 1, signi"cant peaks were recognized at 17 d and 7 d with CM-7 leading CM-4 by 58 and 55 h, respectively. However, the phase lag of the 17-day period in Period 2, 6 h, was much shorter than that in Period 1, and other periods in Period 2, 11 and 9 d, also had considerably shorter phase lags. According to the time-lag between the two stations, the phase velocities of 17-d periods in Periods 1 and 2 were calculated to be 0.3 and 3 m s~1, respectively. Although the phase velocity in Period 1 was of the same order of magnitude as the mean current observed at CM7U and CM4L, that in Period 2 was considerably larger than the current velocity.
Fig. 4. Power spectra of tangential component of current velocity in Period 1 (a) and Period 2 (b).
752
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Fig. 5. Coherence and phase lag between CM7U and CM4L in Period 1 (a) and Period 2 (b).
4. Discussion According to the previous study of Kimura and Sugimoto (1993), periodic #uctuation of 17}19 days is dominant in the main stream region of the Kuroshio with a wavelength of 400 km and a phase velocity of 26 cm s~1. This period and phase velocity correspond well to #uctuations in velocity observed in Period 1, when the Kuroshio took a path type where its countercurrent easily intruded into the Enshunada Sea. This good correspondence between the main stream and countercurrent of the Kuroshio suggests a continuity of physical characteristics from the main stream to the countercurrent. In fact, estimated paths de"ned by water temperature of 153C at a depth of 200 m, satellite images, and the larger averaged current velocity of Period 1 indicate strong intrusion of the countercurrent in Period 1. It could be concluded, therefore, that the short-period meander of the Kuroshio was propagated into the sea directly through the countercurrent of the Kuroshio in Period 1 when the Kuroshio's path signi"cantly bent and de#ected o! the Izu Peninsula. However, the phase velocity observed in Period 2 was ten times larger than that in Period 1 in spite of slower averaged current velocity. This suggests that the high phase velocity cannot be explained in the same way as the results in Period 1. Thus, let us consider the continental shelf wave along a single-step topography de"ned as C " 4 p/kf l(1!h /h ), where p is frequency, k is wave number, f is the Coriolis parameter, 1 2 l is the scale of the shelf in the o!shore direction, h is maximum shelf depth, and h is 1 2 the depth of the deep ocean (Larsen, 1969). With f"8]10~5 s~1, l"30 km, h "200 m and h "2000 m, the corresponding phase velocity is 220 cm s~1 in the 1 2 along stream direction of the countercurrent. This velocity is almost the same as the phase velocity of the 17-day periodic #uctuation in Period 2. This correspondence indicates that the wavelike meandering of the Kuroshio would propagate into the coastal seas as the continental shelf wave when the Kuroshio's path did not signi"cantly bend and de#ect o! the Izu Peninsula.
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
753
Daily coastal temperatures observed at "xed stations and o!shore temperatures observed by volunteer vessels on the Kumano-nada Sea revealed two cases where temperatures #uctuated with time-lag and no time-lag (Kimura and Sugimoto, 1987, Kimara et al., 1994). Since the temperatures were observed only once a day in the survey, no time-lag means that the #uctuation propagated rapidly. Although no concrete physical model was proposed in the previous papers, change in path type of the Kuroshio was suggested as a reason for the di!erence in time-lag. In the present study, two di!erent phase velocities in the o!shore region of the Kumano-nada Sea were recognized quantitatively and the di!erence was found to be dependent on the path type of the Kuroshio o! the Izu Peninsula. The considerably higher phase velocity could be explained by propagation of continental shelf wave. Therefore, the two di!erent mechanisms for propagation of the small-scale meander of the Kuroshio into the coastal seas, depending on the small di!erence of the Kuroshio path type, may explain the two kinds of phase velocities in the coastal temperature #uctuations. Further studies, based on other observations and numerical simulations, should be made to con"rm the mechanism. However, since prediction of the coastal water temperature is important for "sheries in this region (such as yellowtail "shing and marine aqua-culture), a physical explanation contributing to the prediction could be of considerable economic bene"t. In the Gulf Stream region, disturbances with periods of 5}9 days are dominant (Lee and Atkinson, 1983). Since this periodicity has also been recognized in the Kuroshio region (Kimura and Sugimoto, 1993) and some periods observed in the present study were included in the same period band, this periodicity is probably a ubiquitous phenomenon in the western boundary currents. However, a 17-day period is seldom found in the Gulf Stream region. Since the continental shelf is considerably wider in the Gulf Stream region than in the Kuroshio region, this topographic di!erence may be one of the reasons this periodicity is dominant in the Kuroshio region.
References Atkinson, L.P., Miller, J.L., Lee, T.N., Dunstan, W.M., 1996. Nutrients and chlorophyll at the shelf break o! the southeastern United States during the Genesis of Atlantic Lows Experiment: Winter 1986. Journal of Geophysical Research 101, 20565}20578. Ito, T., Kaneko, A., Furukawa, H., Gohda, N., Koterayama, W., 1995. A structure of the Kuroshio and its related upwelling on the East China Sea shelf slope. Journal of Oceanography 51, 267}278. Kasai, A., Kimura, S., Sugimoto, T., 1993. Warm water intrusion from the Kuroshio into the coastal areas south of Japan. Journal of Oceanography 49, 607}624. Kawabe, M., 1985. Sea level variations at the Izu Islands and typical stable paths of the Kuroshio. Journal of Oceanographical Society of Japan 41, 307}326. Kawabe, M., 1986. Transition processes between the three typical paths of the Kuroshio. Journal of Oceanographical Society of Japan 42, 174}191. Kawabe, M., 1987. Spectral properties of sea level and time scales of Kuroshio path variations. Journal of Oceanographical Society of Japan 43, 111}123. Kimura, S., Kasai, A., Sugimoto, T., 1994. Migration of yellowtail in relation to intrusions of warm water from the Kuroshio. Fisheries Science 60, 635}641.
754
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Kimura, S., Kasai, A., Nakata, H., Sugimoto, T., Simpson, J.H., Cheok, J.V.S., 1997. Biological productivity of meso-scale eddies caused by frontal disturbances in the Kuroshio. ICES Journal of Marine Science 54, 179}192. Kimura, S., Sugimoto, T., 1987. Short period #uctuations in oceanographic and "shing conditions in the coastal area of Kumano-nada Sea. Nippon Suisan Gakkaishi 53, 585}593. Kimura, S., Sugimoto, T., 1988. Characteristics of short period #uctuations in oceanographic and "shing conditions in the coastal area of Enshu-nada Sea. Bulletin of the Japanese Society of Fisheries Oceanography 52, 221}228 (in Japanese with English abstract and legends). Kimura, S., Sugimoto, T., 1990. Intrusion processes of warm water mass from the Kuroshio into the coastal area of Kumano-nada and Enshu-nada Seas. Bulletin of the Japanese Society of Fisheries Oceanography 54, 19}31 (in Japanese with English abstract and legends). Kimura, S., Sugimoto, T., 1993. Short-period #uctuations in meander of the Kuroshio path o! Cape Shionomisaki. Journal of Geophysical Research 98, 2407}2418. Larsen, J.C., 1969. Long waves along a single-step topography in a semi-in"nite uniformly rotating ocean. Journal of Marine Research 27, 1}6. Lee, T., Atkinson, L., 1983. Low-frequency current and temperature variability from Gulf Stream frontal eddies and atmospheric forcing along the southeast U.S. outer continental shelf. Journal of Geophysical Research 88, 4541}4567. Lee, T., Atkinson, L., Legeckis, R., 1981. Observations of a Gulf Stream frontal eddy on the Georgia continental shelf, April 1977. Deep-Sea Research 28, 347}378. Maeda, A., Yamashiro, T., Sakurai, M., 1993. Fluctuation in volume transport distribution accompanied by the Kuroshio frontal migration in the Tokara Strait. Journal of Oceanography 49, 231}245. Qiu, B., Toda, T., Imasato, N., 1990. On Kuroshio front #uctuations in the East China Sea using satellite and in situ observational data. Journal of Geophysical Research 95, 18191}18204. Stommel, H., Yoshida, K., 1972. Kuroshio } Its Physical Aspects, University of Tokyo Press, Tokyo, 517pp. Toda, T., 1993. Movement of the surface front induced by Kuroshio frontal eddy. Journal of Geophysical Research 98, 16331}16339. Yoder, J., Atkinson, L., Lee, T., Kim, H., McLain, C., 1981. Role of Gulf Stream frontal eddies informing phytoplankton patches on the outer southeast shelf. Limnology and Oceanography 26, 1103}1110.
Note
Two processes by which short-period #uctuations in the meander of the Kuroshio a!ect its countercurrent Shingo Kimura*, Takashige Sugimoto Ocean Research Institute, University of Tokyo, 1-15-1, Minamidai, Nakano-ku, Tokyo, 164-8639, Japan Received 9 July 1998; accepted 8 June 1999
Abstract Direct current velocity measurements in the countercurrent of the Kuroshio, south of Japan, were carried out to investigate the in#uence of short-period #uctuations in the small-scale meander of the Kuroshio on its countercurrent. When the Kuroshio took a path having a meander west of the Izu Ridge and approaching the Izu Peninsula, the countercurrent freely intruded into coastal seas with a period of 17 d and a phase velocity almost equal to that of the Kuroshio itself. However, when the Kuroshio did not signi"cantly bend and de#ect o! the Izu Peninsula, even when taking the same path, the velocity of the countercurrent was considerably reduced and the periodic #uctuations propagated into the coastal seas as a continental shelf wave. The results indicate that a small change in the Kuroshio's path can cause a di!erent process of propagation of the small-scale meandering; this di!erence probably explains why there are two kinds of phase velocity in coastal temperature #uctuations. ( 2000 Elsevier Science Ltd. All rights reserved. Keywords: Kuroshio; Small-scale meander; Countercurrent; Continental shelf wave; Gulf Stream
1. Introduction The Kuroshio is a western boundary current in the Paci"c Ocean and is considered to be a counterpart of the Gulf Stream in the Atlantic Ocean. A signi"cant di!erence between the two currents is that the Kuroshio has bimodal stationary #ow patterns
* Corresponding author. Fax: 0081-3-5351-6506. E-mail address: [email protected] (S. Kimura) 0967-0637/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 7 - 0 6 3 7 ( 9 9 ) 0 0 0 6 7 - 9
746
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
(known as the large and small meander paths) o! the southern part of the main island of Japan (Stommel and Yoshida, 1972). The large meander emerges with periods of 20 yr and 7}8.5 yr (Kawabe, 1987), and its large #uctuations are limited to the Enshu-nada Sea, east of the Kii Peninsula (Kawabe, 1985,1986). Small-scale meanders with shorter periods are also dominant in the Kuroshio region, and they are well recognized in satellite images as shown in this paper. In association with the small-scale meander, displacement of the Kuroshio front #uctuates with a period of 17}19 d and 400 km wavelength (Kimura and Sugimoto, 1993). Similar periodic #uctuations of the Kuroshio front are widely recognized south of Japan (Qiu et al., 1990; Maeda et al., 1993; Kasai et al., 1993), and it is well known that cyclonic eddies associated with the frontal meandering induce upwelling (Toda, 1993; Ito et al., 1995). This upwelling supplies nutrients to the euphotic layer in the cyclonic eddy and accelerates primary production in o!shore regions (Kimura et al., 1997), as do similar processes identi"ed in the Gulf Stream system (Lee et al., 1981; Yoder et al., 1981; Atkinson et al., 1996). East of the Kii Peninsula, when a meander of the Kuroshio is located west or over the center of the Izu Ridge, a warm water mass derived from the Kuroshio often intrudes into the coastal areas of the Kumano-nada and Enshu-nada Seas with a dominant period of about 20 d (Kimura and Sugimoto, 1987,1988,1990). The good periodic correspondence between the Kuroshio region and the coastal region indicates that the small-scale meander of the Kuroshio in#uences coastal oceanographic change. However, the observed coastal temperatures along the Kumano-nada Sea suggest that there are two distinct cases where temperature #uctuations propagate, one with a few days time-lag and the other with no time-lag (Kimura and Sugimoto, 1987; Kimura et al., 1994). While variation in the Kuroshio's path may in#uence variation in the time-lag, no physical explanation for the variation has yet been proposed. Accordingly, we carried out observations over an 18 month period, using moored current meters, on the continental shelf of the Kumano-nada Sea to understand the frequency structures of the coastal countercurrent of the Kuroshio and their relationships to the Kuroshio's path type.
2. Observations and data Direct current measurements in the Kumano-nada Sea were made during Period 1 (9 December 1986}22 April 1987) and Period 2 (25 April 1987}28 April 1988). Two mooring stations (CM-4 and CM-7) were placed at a depth of about 800 m on the continental slope east of the Kii Peninsula along the expected direction of the countercurrent (Fig. 1). Each mooring consisted of two current meters, at depths of about 400 m (upper level) and 700 m (lower level), although current meters could not always be situated at predetermined depths. The data were sampled at 30 min intervals during Period 1 and at hourly intervals during Period 2 by Aanderaa RCM-5 current meters. Fig. 2 shows the time series of low-frequency current velocity after "ltering with a 25-h running mean to remove high-frequency #uctuations. Basic information and
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
747
Fig. 1. Observational locations and schematic views of the Kuroshio's paths during Period 1(bold line) and Period 2 (thin line).
statistics of the observations are summarized in Table 1. Time-averaged velocities of the eastward component (E}W velocity) and northward component (N}S velocity) are denoted as ; and <, respectively, and the variances of each velocity are denoted as u@2 and v@2. KE (eddy kinetic energy) and KM (mean kinetic energy) are de"ned as (u@2#v@2)/2 and (;2#<2)/2, respectively. Hereafter, CM4U (CM-4 upper layer), CM7U (CM-7 upper layer), CM4L (CM-4 lower layer) and CM7L (CM-7 lower layer) are used to refer to the current meters.
3. Results 3.1. Time-mean structure of the countercurrent of the Kuroshio Current directions at the upper levels, CM4U and CM7U, were westward and southwestward along the coast, respectively, throughout both observational periods. The tangential components of the velocities were much larger than the normal components. In particular, the ratios of the eddy kinetic energy to the mean kinetic energy of current velocity (KE/KM) in Period 1 were less than 1.0, which indicated
748
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Fig. 2. Time series of current velocity in Period 1 (upper panel) and Period 2 (lower panel). Water depths and current meter depths are shown at right.
that the current at each station #owed in one direction with no large #uctuations. The current #ow pattern at CM4L was similar to those at CM4U and CM7U although the mean velocity was small. However, the current direction at CM7L was strongly a!ected by the local seabed topography and showed a #ow pattern quite di!erent from the other three current meters. During both observational periods, the Kuroshio usually meandered west of the Izu Ridge and approached the Izu Peninsula. Fig. 3b and d show estimated paths of the Kuroshio in each period, de"ned as contours of water temperature of 153C at a depth of 200 m, from biweekly published reports of the Hydrographic Department,
} Apr. 1987) 780 380 860 460 780 680 860 760
} Apr. 1988) 833 464 845 476 833 718 845 730
Period 2 (Apr. 1987 CM-4 (UPPER) CM-7 (UPPER) CM-4 (LOWER) CM-7 (LOWER)
Sensor (m)
Period 1 (Dec. 1986 CM-4 (UPPER) CM-7 (UPPER) CM-4 (LOWER) CM-7 (LOWER)
Depth (m)
9.7 5.1 2.3 0.4
19.9 12.1 3.8 0.3
14.5 8.8 5.6 1.8
21.3 13.5 7.0 1.6
!7.0 !5.1 !1.0 0.3
!16.0 !11.9 !2.9 0.3
; (cm/s)
Vector (cm/s)
Scalar
Velocity
Velocity
!6.7 !0.3 !2.0 0.2
!11.8 !2.1 !2.5 0.2
<
107 84 22 2
79 73 30 1
u@2 (cm/s)2
99 8 19 3
56 15 23 3
v@2
Var. of vel.
Table 1 Basic information and statistics of observations with current meters at CM4 and CM7
226 267 207 56
234 260 229 56
230 290 220 340
230 270 230 0
47.3 12.9 2.6 0.1
198.4 72.7 7.1 0.1
KM (cm/s)2
Mean (deg.)
Dominant
Energy
Direction
103.1 46.1 20.7 2.2
67.5 43.9 26.4 2.0
KE
2.2 3.6 8.1 32.2
0.3 0.6 3.7 37.3
Ratio KE/KM
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754 749
750
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Fig. 3. Satellite images and estimated Kuroshio paths de"ned by contours of water temperature of 153C at a depth of 200 m during Period 1 (a, b) and Period 2 (c, d). The paths in Period 2 are plotted every second contour.
Maritime Safety Agency of Japan; schematic views of their averaged paths are indicated in Fig. 1. When the Kuroshio takes these path types, warm water masses separated from the main stream of the Kuroshio o! the Izu Peninsula intrude into the Enshu-nada Sea and move westward into the Kumano-nada Sea as a countercurrent of the Kuroshio between the Kuroshio and the coast (Kimura and Sugimoto, 1987,1988,1990; Kasai et al., 1993). In#uenced by the circulation, current measurements at CM4U and CM7U showed a westward and southwestward movement, respectively, along the coast. However, vector-averaged velocities during Period 2, of 10 cm s~1 (at CM4U) and 5 cm s~1 (at CM7U), were half those of Period 1 and the eddy kinetic energy was larger than mean kinetic energy in Period 2. According to the estimated paths, the Kuroshio's path bent and de#ected much further to the west o! the Izu Peninsula in Period 1 than in Period 2. A satellite thermal infrared image (Fig. 3a) shows an example of intrusion of the Kuroshio water as a countercurrent during Period 1. While the Kuroshio's path type in Period 2 was similar to that in Period 1, similar intrusions of warm Kuroshio water and de#ection of the path were not recognized in Period 2 and the Kuroshio usually #owed as shown in the satellite image in Fig. 3c. We therefore infer that this small change of the path type probably
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
751
explains the di!erence in strength of the Kuroshio countercurrent between Periods 1 and 2. 3.2. Frequency structure of the current velocity and its phase lag Periodic #uctuations with intervals of a week were readily recognized from the time series of current velocity as shown in Fig. 2, particularly in velocity at CM4L and CM7U in Period 1 with CM-7 leading CM-4 by a time-lag of a few days. Fig. 4 shows power spectra of the tangential component of current velocity along the mean current directions in Periods 1 and 2 and indicates the existence of a dominant 17-d periodicity in CM7U and CM4L throughout both observational periods. At CM4U, the spectral peaks in Period 1 were seen to be similar, and in Period 2 identical. The peaks of the 17-d period were more obvious in Period 1 than in Period 2. However, periods of 7}12 ds, much shorter than 17 days, were signi"cantly more dominant in Period 2. Focusing on the 17-day period observed at CM7U and CM4L, which had also been identi"ed as a period of the small-scale meander of the Kuroshio (Kimura and Sugimoto, 1993), coherence and phase lag between the stations were calculated (Fig. 5). In period 1, signi"cant peaks were recognized at 17 d and 7 d with CM-7 leading CM-4 by 58 and 55 h, respectively. However, the phase lag of the 17-day period in Period 2, 6 h, was much shorter than that in Period 1, and other periods in Period 2, 11 and 9 d, also had considerably shorter phase lags. According to the time-lag between the two stations, the phase velocities of 17-d periods in Periods 1 and 2 were calculated to be 0.3 and 3 m s~1, respectively. Although the phase velocity in Period 1 was of the same order of magnitude as the mean current observed at CM7U and CM4L, that in Period 2 was considerably larger than the current velocity.
Fig. 4. Power spectra of tangential component of current velocity in Period 1 (a) and Period 2 (b).
752
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Fig. 5. Coherence and phase lag between CM7U and CM4L in Period 1 (a) and Period 2 (b).
4. Discussion According to the previous study of Kimura and Sugimoto (1993), periodic #uctuation of 17}19 days is dominant in the main stream region of the Kuroshio with a wavelength of 400 km and a phase velocity of 26 cm s~1. This period and phase velocity correspond well to #uctuations in velocity observed in Period 1, when the Kuroshio took a path type where its countercurrent easily intruded into the Enshunada Sea. This good correspondence between the main stream and countercurrent of the Kuroshio suggests a continuity of physical characteristics from the main stream to the countercurrent. In fact, estimated paths de"ned by water temperature of 153C at a depth of 200 m, satellite images, and the larger averaged current velocity of Period 1 indicate strong intrusion of the countercurrent in Period 1. It could be concluded, therefore, that the short-period meander of the Kuroshio was propagated into the sea directly through the countercurrent of the Kuroshio in Period 1 when the Kuroshio's path signi"cantly bent and de#ected o! the Izu Peninsula. However, the phase velocity observed in Period 2 was ten times larger than that in Period 1 in spite of slower averaged current velocity. This suggests that the high phase velocity cannot be explained in the same way as the results in Period 1. Thus, let us consider the continental shelf wave along a single-step topography de"ned as C " 4 p/kf l(1!h /h ), where p is frequency, k is wave number, f is the Coriolis parameter, 1 2 l is the scale of the shelf in the o!shore direction, h is maximum shelf depth, and h is 1 2 the depth of the deep ocean (Larsen, 1969). With f"8]10~5 s~1, l"30 km, h "200 m and h "2000 m, the corresponding phase velocity is 220 cm s~1 in the 1 2 along stream direction of the countercurrent. This velocity is almost the same as the phase velocity of the 17-day periodic #uctuation in Period 2. This correspondence indicates that the wavelike meandering of the Kuroshio would propagate into the coastal seas as the continental shelf wave when the Kuroshio's path did not signi"cantly bend and de#ect o! the Izu Peninsula.
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
753
Daily coastal temperatures observed at "xed stations and o!shore temperatures observed by volunteer vessels on the Kumano-nada Sea revealed two cases where temperatures #uctuated with time-lag and no time-lag (Kimura and Sugimoto, 1987, Kimara et al., 1994). Since the temperatures were observed only once a day in the survey, no time-lag means that the #uctuation propagated rapidly. Although no concrete physical model was proposed in the previous papers, change in path type of the Kuroshio was suggested as a reason for the di!erence in time-lag. In the present study, two di!erent phase velocities in the o!shore region of the Kumano-nada Sea were recognized quantitatively and the di!erence was found to be dependent on the path type of the Kuroshio o! the Izu Peninsula. The considerably higher phase velocity could be explained by propagation of continental shelf wave. Therefore, the two di!erent mechanisms for propagation of the small-scale meander of the Kuroshio into the coastal seas, depending on the small di!erence of the Kuroshio path type, may explain the two kinds of phase velocities in the coastal temperature #uctuations. Further studies, based on other observations and numerical simulations, should be made to con"rm the mechanism. However, since prediction of the coastal water temperature is important for "sheries in this region (such as yellowtail "shing and marine aqua-culture), a physical explanation contributing to the prediction could be of considerable economic bene"t. In the Gulf Stream region, disturbances with periods of 5}9 days are dominant (Lee and Atkinson, 1983). Since this periodicity has also been recognized in the Kuroshio region (Kimura and Sugimoto, 1993) and some periods observed in the present study were included in the same period band, this periodicity is probably a ubiquitous phenomenon in the western boundary currents. However, a 17-day period is seldom found in the Gulf Stream region. Since the continental shelf is considerably wider in the Gulf Stream region than in the Kuroshio region, this topographic di!erence may be one of the reasons this periodicity is dominant in the Kuroshio region.
References Atkinson, L.P., Miller, J.L., Lee, T.N., Dunstan, W.M., 1996. Nutrients and chlorophyll at the shelf break o! the southeastern United States during the Genesis of Atlantic Lows Experiment: Winter 1986. Journal of Geophysical Research 101, 20565}20578. Ito, T., Kaneko, A., Furukawa, H., Gohda, N., Koterayama, W., 1995. A structure of the Kuroshio and its related upwelling on the East China Sea shelf slope. Journal of Oceanography 51, 267}278. Kasai, A., Kimura, S., Sugimoto, T., 1993. Warm water intrusion from the Kuroshio into the coastal areas south of Japan. Journal of Oceanography 49, 607}624. Kawabe, M., 1985. Sea level variations at the Izu Islands and typical stable paths of the Kuroshio. Journal of Oceanographical Society of Japan 41, 307}326. Kawabe, M., 1986. Transition processes between the three typical paths of the Kuroshio. Journal of Oceanographical Society of Japan 42, 174}191. Kawabe, M., 1987. Spectral properties of sea level and time scales of Kuroshio path variations. Journal of Oceanographical Society of Japan 43, 111}123. Kimura, S., Kasai, A., Sugimoto, T., 1994. Migration of yellowtail in relation to intrusions of warm water from the Kuroshio. Fisheries Science 60, 635}641.
754
S. Kimura, T. Sugimoto / Deep-Sea Research I 47 (2000) 745}754
Kimura, S., Kasai, A., Nakata, H., Sugimoto, T., Simpson, J.H., Cheok, J.V.S., 1997. Biological productivity of meso-scale eddies caused by frontal disturbances in the Kuroshio. ICES Journal of Marine Science 54, 179}192. Kimura, S., Sugimoto, T., 1987. Short period #uctuations in oceanographic and "shing conditions in the coastal area of Kumano-nada Sea. Nippon Suisan Gakkaishi 53, 585}593. Kimura, S., Sugimoto, T., 1988. Characteristics of short period #uctuations in oceanographic and "shing conditions in the coastal area of Enshu-nada Sea. Bulletin of the Japanese Society of Fisheries Oceanography 52, 221}228 (in Japanese with English abstract and legends). Kimura, S., Sugimoto, T., 1990. Intrusion processes of warm water mass from the Kuroshio into the coastal area of Kumano-nada and Enshu-nada Seas. Bulletin of the Japanese Society of Fisheries Oceanography 54, 19}31 (in Japanese with English abstract and legends). Kimura, S., Sugimoto, T., 1993. Short-period #uctuations in meander of the Kuroshio path o! Cape Shionomisaki. Journal of Geophysical Research 98, 2407}2418. Larsen, J.C., 1969. Long waves along a single-step topography in a semi-in"nite uniformly rotating ocean. Journal of Marine Research 27, 1}6. Lee, T., Atkinson, L., 1983. Low-frequency current and temperature variability from Gulf Stream frontal eddies and atmospheric forcing along the southeast U.S. outer continental shelf. Journal of Geophysical Research 88, 4541}4567. Lee, T., Atkinson, L., Legeckis, R., 1981. Observations of a Gulf Stream frontal eddy on the Georgia continental shelf, April 1977. Deep-Sea Research 28, 347}378. Maeda, A., Yamashiro, T., Sakurai, M., 1993. Fluctuation in volume transport distribution accompanied by the Kuroshio frontal migration in the Tokara Strait. Journal of Oceanography 49, 231}245. Qiu, B., Toda, T., Imasato, N., 1990. On Kuroshio front #uctuations in the East China Sea using satellite and in situ observational data. Journal of Geophysical Research 95, 18191}18204. Stommel, H., Yoshida, K., 1972. Kuroshio } Its Physical Aspects, University of Tokyo Press, Tokyo, 517pp. Toda, T., 1993. Movement of the surface front induced by Kuroshio frontal eddy. Journal of Geophysical Research 98, 16331}16339. Yoder, J., Atkinson, L., Lee, T., Kim, H., McLain, C., 1981. Role of Gulf Stream frontal eddies informing phytoplankton patches on the outer southeast shelf. Limnology and Oceanography 26, 1103}1110.