observation of oceanic structure around tosa-bae southeast ... · 544 y. sekine et al. topography...

Journal of OceanographyVol. 50, pp. 543 to 558. 1994

Observation of Oceanic Structure around Tosa-BaeSoutheast of Shikoku

YOSHIHIKO SEKINE, HARUKI OHWAKI and MOTOYA NAKAGAWA

Institute of Oceanography, Faculty of Bioresources, Mie University, 1515 Kamihamachou, Tsu 514, Japan

(Received 15 November 1993; in revised form 23 May 1994; accepted 25 May 1994)

The hydrographic observations in the vicinity of a seamount, the Tosa-Bae,southeast of Shikoku have been carried out two times in summer of 1991 and 1992.The temperature, salinity fields are observed by CTD and velocity fields aremeasured by ADCP. Results of these observation are presented in this paper. It isshown that salinity maximum water at a depth of 100 m is confined to a southeasternarea of the Tosa-Bae, however, salinity minimum water is found in northern sideof the Tosa-Bae. This indicates the westward intrusion of less saline water overnorthern slope. A positive correlation is detected between the estimated Rossbyheight (fL/N) and the observed height of Taylor Column estimated from thevertical change in the isotherms and isohalines. Almost both heights give smallervalue than representative depth of bottom topography of the Tosa-Bae, it isindicated that the topographic effect of the Tosa-Bae is not fully reached to thesurface. From the correlations between the vertical difference of geostrophic flowand that of ADCP velocity, ageostrophic flow component is detected.

1. IntroductionTo estimate the topographic effects of seamounts on temperature, salinity, density and

velocity fields are important and it has been of interest to oceanographers (e.g., Hogg, 1973,1980; Johnson, 1977; Roden, 1987). There have been various observational and theoreticalstudies: if a stationary barotropic geostrophic flow is assumed, a current has no vertical change(Taylor-Proudman theorem) and an isolated eddy with vertically coherent flow (Taylor column)is formed over a seamount. If a density stratification is considered, topographic effect of aseamount is weakened and its topographic effect is confined to a shorter thickness over theseamount, of which the height of Taylor column is estimated from Rossby height defined byfL/N, where f is the Coriolis parameter, L half of representative horizontal scale of the seamountand N is the Brunt-Väisä lä frequency assumed to be constant (e.g., Gill, 1982). In an unstationarystate, topographic waves take important roles on the time evolutions of oceanic structure (e.g.,Huppert and Bryan, 1976).

As oceanic conditions over seamounts may be different in localities, hydrographic obser-vations should be made for each seamount. The present study is directed toward oceanicconditions over Tosa-Bae southeast of Shikoku (Fig. 1). The Tosa-Bae has an elliptic shape witha longer axis in zonal direction. The top of this seamount has a depth of 147 m. It should be noticedthat as the Kuroshio flows over or near Tosa-Bae, the intensity of the topographic effect on theflow may be an important factor for the path dynamics of the Kuroshio. It is well-known that theKuroshio shows bimodal path characteristics between large meander path and non-largemeander path south of Japan (e.g., Nitani, 1972; Taft, 1972). In particular, because the Tosa-Baelocates at the separating area of the large meander path of the Kuroshio from the coastal

544 Y. Sekine et al.

topography of Japan, the separation of the Kuroshio path is enhanced if the topographic effectof Tosa-Bae is relatively large and the main axis is deflected southward over the Tosa-Bae. Fromthese view points, to estimate the topographic effect of the Tosa-Bae on the Kuroshio flow isstrongly needed.

Up to now, some observations have been carried out around the Tosa-Bae. Yoshioka et al.(1986) made STD observations over the Tosa-Bae in April of 1984. They showed that thetemperature field in the layer shallower than 300 m has a frontal structure, whereas a cold dome-like structure is observed below 300 m. Sekine and Matsuda (1987a) carried out observations byCTD and GEK over the Tosa-Bae in November 1985. They showed that a coupled warm and coldwater are observed in upper layer shallower than 500 m and only cold water is observed at greater

Fig. 1. Location of Tosa-Bae southeastern Shikoku (upper). Isoplethes of depth (in meter) are also shown(after Taft, 1972). Isopleth of depth (in meter) in the vicinity of Tosa-Bae (lower).

Oceanic Structure in the Vicinity of Tosa-Bae 545

depth. The coupled warm and cold water are suggested to be almost identical to the results ofnumerical experiment performed by Huppert and Bryan (1976). It is also pointed out that thethickness of a mixed layer over the Tosa-Bae varied by a small horizontal scale of 5–10 km witha depth range of 35–105 m (Sekine and Matsuda, 1987b).

We have made two more detailed observations by CTD and ADCP around Tosa-Bae insummer of 1991 and 1992. Details of the two observations and some noteworthy results areshown in the present paper. The main emphasis is placed on the temperature, salinity, density andvelocity fields that indicates the topographic effect of the Tosa-Bae on the mean flow of theKuroshio.

2. ObservationsTwo observations were carried out on July 21–24 in 1991 and on July 16–19 in 1992 by use

of Training Vessel Seisui-maru of Mie University. The first and second observations arehereafter referred to as KS-91JUL and KS-92JUL, respectively. Temperature and salinity wereobserved by CTD at the observation points shown in Fig. 2. Because of the long time lag of thetotal observation, the observation points were placed to see the vertical field for each observationline rather than over all horizontal field. Unfortunately, because of the rough sea conditions,

Fig. 2. Observational points of CTD for KS-91JUL (upper) and for KS-92JUL (lower). Names of straightobservation lines are also shown by capitals.


observations along the meridional observation line between LINE-B and LINE-C of KS-92JULwere not completed. Density (σt) is calculated from temperature and salinity data. The currentvelocity was measured by acoustic doppler current profiler (ADCP) at the three depths, 50, 100and 150 m in KS-91JUL and 5, 50 and 200 m in KS-92JUL. The mean ADCP velocity over 10minutes are used in the data analysis.

3. Results

3.1 Temperature, salinity and density fieldsThe vertical distributions of temperature, salinity and density (σt) along zonal observation

line A, which is hereafter referred to as LINE-A, of KS-91JUL are shown in Fig. 3. A weakupward shifts of the isotherms of 21°C and 22°C are detected over the top of the Tosa-Bae.However, the upward shifts are unclear for the isotherms warmer than 24°C. In depths below thetop of Tosa-Bae, complicated and various vertical displacement of the isotherms is found overthe side slope of Tosa-Bae, which is common to the results of previous observations (Yoshiokaet al., 1986; Sekine and Matsuda, 1987a). The salinity maximum layer and minimum layer arefound at depths 100 db and 500 db, respectively. Saline water which is more than 34.7 psu is foundat a depth of 100 db in eastern side of the Tosa-Bae, but such a saline water is not found in westernside. In the upper layer, vertical variations in the isotherms are relatively more dominant in downstream (eastern) area rather than those of upstream (western) area. It is suggested that thedominant vertical variations in downstream area are due to the internal lee wave formed by thebottom topography of the Tosa-Bae.

Since the salinity difference is relatively small, density fields have a close resemblance tothose of isotherms (Fig. 3). Because similar density patterns to those of temperature arecommonly observed for all the other observation lines, density distribution are not shown in thefollowing.

The vertical temperature and salinity distributions along meridional observation lines of KS-91JUL are shown in Fig. 4. As for LINE-B, upward shifts of isotherms are detected over the topof Tosa-Bae in depths of 350–500 db. Although the data of LINE-A is used at station 18 with arelatively large observation time lag from other data, this uplift is also suggested from the dataof the stations 23 and 24. However, downward shift of the isotherms over the top of the seamountin LINE-C and LINE-E are due to the one station of LINE-A: 13 and 5, respectively. Thesedownward shifts may be attributed to the time change in the isotherms during the observationperiods.

Upward shifts of the isotherms of 3.8–6°C are depicted over the northern side slope of theTosa-Bae in LINE-D. Because of the upward shift of isotherms of 4.4–12°C with the oppositehorizontal gradient in a southern side, the existence of cold-dome is suggested. As for LINE-E,more dominant uplift of isotherms of 4.6–7°C is found over the ridge of the top of the Tosa-Bae.Since this upward shift shows no evanescent change from the topography, this upward shift ofisotherms are suggested to be due to internal waves.

Maximum and minimum salinity layer are observed at depths 100–200 db and 400–600 db,respectively (Fig. 4b). Saline water more than 34.7 psu is found at the southern side of the ridgeof Tosa-Bae, while such a saline water is not found in the northern side. In contrast to this, lessthan 34.24 psu saline water in the salinity minimum layer is found in both sides. Horizontalsalinity distribution over isopycnal surface of σt = 26.8 corresponding to a depth with salinityminimum layer is shown in Fig. 5. Less than 34.22 psu saline water spreads westward along the


Fig

. 3.

T

empe

ratu

re,

sali

nity

and

den

sity

(σ t

) se

ctio

ns a

long

obs

erva

tion

al l

ine

A o

f K

S-9

1JU

L.

The

loca

tion

s of

the

obse

rvat

iona

l poi

nts

are

show

n on

the

top.


Fig

. 4.

Ver

tica

l dis

trib

utio

n of

(a) t

empe

ratu

re a

nd (b

) sal

init

y al

ong

four

mer

idio

nal o

bser

vati

on li

nes

B–

E o

f K

S-9

1JU

L.


northern slope of Tosa-Bae. There is a possibility that the less saline water intrudes westwardaround the northern topography of Tosa-Bae.

Almost similar characteristics of LINE-A of KS-92JUL to those of KS-91JUL are observed(not shown). Main characteristics along meridional observation lines of KS-92JUL are alsocommon to those of KS-91JUL, however, some new features are observed (Fig. 6). Verticallycoherent uplift of the isotherms is perceived over the ridge in LINE-D. Clear vertical variationsof isotherms in the down stream area (Fig. 4) are not detected in Fig. 6. Northward intrusion ofsaline water with 34.6–34.7 psu is detected in a salinity maximum layer at a depth of 100 db, whilemore saline water than 34.8 psu is confined to the southern side. An isolated saline water is foundat a depth of 500 db on LINES-D and -E.

To examine the intensity of the topographic effect of Tosa-Bae, the estimated Rossby height(fL/N) and the height of observed Taylor column are shown in Fig. 7. Here, the Rossby height(RH) is estimated from observed N and evaluated L, while the observed height of Taylor column(OHT) is evaluated by the coherent or evanescent vertical changes of isotherm and isohaline.Since to estimate constant values of L and N is difficult, their possible ranges are evaluated andthe resulted height range of RH are shown for maximum, intermediate and minimum values ofN and maximum and minimum for L, of which details are tabulated in Table 1. Positive corre-lation which exceeds 95% confidence limit of which correlation coefficient (r) is 0.63 is detectedin Fig. 7. Two large OHT were observed on LINE-D and LINE-E of KS-92JUL, in which depthof the ridge of Tosa-Bae is deeper than 1000 m. Except for these two cases, all OHT and RH aresmaller than 300 m, and this height is smaller than the representative depth of the Tosa-Bae. Thisindicates that the topographic effect of the Tosa-Bae is not fully reached to the surface.

3.2 Velocity fieldsObserved velocity fields by ADCP is shown in Fig. 8. Large eastward velocity is detected

Fig. 5. Salinity distribution (psu) on the depth with σt = 26.8.


Table 1. Estimates of Rossby height (fL/N).

1991observation

line

N(10–3 sec–1 )

L (km) Rossby height (m)

Minimum Maximum Minimum Maximum

A 15.0 4.5 7.5 24 4012.5 29 4810.0 36 60

B 6.0 5.5 10.5 73 1405.0 88 1684.0 110 210

C 15.0 5.0 10.5 27 5612.5 32 6710.0 40 84

D 10.0 4.0 8.8 32 707.5 43 935.0 64 140

E 4.0 5.8 7.0 115 1403.0 153 1872.0 230 280

1992observation

line

N(10–3 sec–1 )

L (km) Rossby height (m)

Minimum Maximum Minimum Maximum

A 15.0 5.1 9.0 58 10312.5 68 12010.0 82 144

B 6.0 6.0 7.9 96 1265.0 120 1574.0 160 209

C 15.0 3.7 5.6 42 6312.5 49 7410.0 59 89

D 10.0 2.8 6.8 56 1357.5 75 1805.0 112 270

E 4.0 7.9 10.2 157 2043.0 209 2722.0 314 408


Fig

. 6.

Sam

e as

in F

ig. 4

but

for

KS

-92J

UL

.


Fig. 7. Correlation between the estimated Rossby height (fL/N) and the observed Taylor column. (a) Caseof possible maximum of Vä isä lä frequency (N), (b) middle N and (c) possible minimum N (for details,see Table 1). Horizontal bars in each data show the possible range of the Rossby height depending onhalf of the horizontal scale of the Tosa-Bae (L), while vertical bars show the possible range of OHTdepending on the estimation of vertical change in isotherms and isohalines.


Fig

. 8.

Obs

erve

d ho

rizo

ntal

vel

ocit

y by

AD

CP

in K

S-9

1JU

L (

left

) an

d in

KS

-92J

UL

(ri

ght)

.


Fig. 9. Correlation between the vertical difference of eastward geostrophic velocity and those of observedvelocity by ADCP, which is estimated as a mean value of all the ADCP velocity data observed betweenthe two neighboring CTD points. Two vertical levels which estimate the vertical velocity differenceare shown on top of each panel and r on the upper right shows the coefficient of correlation.

over the Tosa-Bae. As for KS-91JUL, it is noted that vertically coherent flow is dominant insouthern area of Tosa-Bae, while an eastward velocity at a depth of 150 m is weakened in northernside. As for KS-92JUL, large vertical difference is found between 50 m and 200 m. Cycloniccirculation in an eastern area of the Tosa-Bae is suggested at a depth of 200 m from eastern area.In particular, an eastward flow is observed at the northern area of LINE-E. Although the depthis different between 200 m and 500 m, the cyclonic flow in the eastern side of Tosa-Bae supportsthe westward intrusion of the less saline water along northern slope of Tosa-Bae (Fig. 5).

In order to see the attainment of geostrophic flow balance over the Tosa-Bae, correlationsbetween the vertical difference of geostrophic flow and that of velocity observed by ADCP areshown in Fig. 9. Here, because there are several data of ADCP velocity between two neighboringstations of CTD for the geostrophic flow estimation, three different values of ADCP velocity,which is correlated to the geostrophic velocity, are estimated and their correlation coefficients


Table 2. Coefficient of correlation between the vertical difference of eastward geostrophic velocity andthat of observed ADCP velocity by three different estimation of the ADCP velocity.

Cruise Estiamted depth(m)

One point* Simple mean** Weightedmean***

KS-91JUL 50–100 0.55 0.50 0.52100–150 0.43 0.54 0.5350–150 0.69 0.70 0.70

KS-92JUL 5–50 0.05 0.17 0.1650–200 0.58 0.62 0.635–200 0.56 0.61 0.61

*ADCP velocity estimated midway between the two neighboring CTD stations used to estimate thegeostrophic velocity.

**ADCP velocity is estimated as the simple mean of all ADCP velocity data between the twoneighboring CTD stations.

***ADCP velocity is estimated as a weighted mean value of all the ADCP data observed betweenthe two neighboring CTD stations. The weight is inversely proportional to the distance of the ADCPstation from the mid-point between two CTD stations.

are compared in Table 2. Relatively large difference is found between one point estimation andother two mean estimations, however they show no significant difference in correlationcoefficient. It is shown from Fig. 9 that although positive correlation is totally found, they do notgive common velocity difference and their dispersions are notable. This result agrees with thecase of the seamount, Daini-Kinan Kaizan, located in south of Kii Peninsula (Sekine andHayashi, 1992). It is also recognized from Fig. 9 and Table 2 that significant low coefficient isfound in case with data of 5 m. It is suggested that the velocity at a depth of 5 m is more influencedby the wind-drift current, the Ekman flow and motions of wind waves and/or swells. Ageostrophicflow in deeper layers is suggested to be due to internal tidal flow in ADCP velocity and verticalvariation in density fields by internal waves that influences the estimation of the geostrophicflow.

4. Summary and DiscussionThe hydrographic observations around Tosa-Bae southeast of Shikoku were made by the

Training Vessel Seisui-maru of Mie University two times in summer of 1991 and 1992. Notableresults of the observations are summarized as follows.

(1) A saline water than 34.7 psu is confined to a southeastern area of the Tosa-Bae. Incontrast to this, less saline water than 34.24 psu is found in northern and southern area. Westwardintrusion of less saline water is suggested in the northern slope of the Tosa-Bae.

(2) The estimated Rossby height ( fL/N) and the observed height of Taylor Column werecompared. They show a positive correlation. Almost both heights are smaller than the representativedepth of bottom topography of the Tosa-Bae, indicating that the topographic effect of the Tosa-Bae has not fully reached the surface layer.

(3) Large eastward velocity is detected over the Tosa-Bae. It is noted that large verticaldifference is found between 50 m and 200 m. Cyclonic circulation in an eastern area of the Tosa-Bae is suggested at a depth of 200 m of KS-92JUL.


(4) In order to see the attainment of geostrophic balance, correlations between the verticaldifference of geostrophic flow and that of ADCP velocity were examined. Although positivecorrelation is perceived, they do not give common velocity difference and their dispersion arenotable. In particular, coefficient of the correlation decreases significantly in a case withshallower layer of 5 m, resulting that ageostrophic flow component is not negligible in thisregion.

It should be noted that some different characteristics of temperature and salinity fields areobserved between the two cruise: vertically coherent uplift of the isotherms is perceived over theridge in LINE-D of KS-92JUL. Dominant vertical change of isotherms are observed in KS-91JUL, but enhanced vertical change is not found in KS-92JUL. These observational resultsstrongly suggests that frequent observation is needed to see the detailed oceanic condition overthe Tosa-Bae.

Existence of ageostrophic flow is resulted from (4). Although the topographic effect of theTosa-Bae is not fully reached to the surface layer (2), the topographic effect on the distributionof salinity maximum layer denoted in (1) indicates the existence of the batropopic flow, whichenhances the topographic effect of the Tosa-Bae. Furthermore, the oceanic conditions aroundthis seamount is influenced by the seasonal variation and also by the distance from main axis ofthe Kuroshio. Long term direct current measurements which are able to see ageostrophic andbarotropic flow component are needed to obtain the real velocity fields around Tosa-Bae.

AcknowledgmentsThe authors would like to thank Captain I. Ishikura, officers and crew of the Tranining

Vessel Seisui-maru of Mie University for their excellent help in the observations and Miss A.Miyake now at Okayama City Office for her help in drawing some figures. Thanks are extendedto Dr. K. Taguchi and Mr. T. Hayashi of the Faculty of Bioresources of Mie University for theirhelp during observation.

ReferencesGill, A. E. (1982): Atmosphere-Ocean Dynamics. Academic Press, New York, London, 662 pp.Hogg, N. G. (1973): On the stratified Taylor column. J. Fluid Mech., 58, 517–537.Hogg, N. G. (1980): Effects of bottom topography on ocean currents. Orographic Effects in Planetary Flows, GARP

Publ. Ser., WMO, 23, 167–205.Huppert, H. E. and K. Bryan (1976): Topographically generated eddies. Deep-Sea Res., 23, 655–679.Johnson, E. R. (1977): Stratified Taylor column on a beta-plane. Geophys. Astrophys. Fluid Dyn., 9, 159–177.Nitani, H. (1972): Beginning of the Kuroshio. p. 129–163. In Kuroshio—Its Physical Aspects, ed. by H. Stommel

and K. Yoshida, University of Tokyo Press, Tokyo.Roden, G. I. (1987): Effects of seamounts and seamount chains on oceanic circulation and thermocline structure.

Seamounts, Islands and Atolis. B. Keating eds., Geophys. Monogr., 43, A.G.U., 335–354.Sekine, Y. and Y. Matsuda (1987a): Hydrographic structure around the Tosa-Bae, the bump off Shikoku south of

Japan, in November 1985. La mer, 25, 137–146 (in Japanese with English abstract).Sekine, Y. and Y. Matsuda (1987b): Observation on surface mixed layer around the Tosa-Bae, the bump off Shikoku

south of Japan, in November 1985. Umi to Sora, 63, 1–14 (in Japanese with English abstract).Sekine, Y. and T. Hayashi (1993): Oceanic structure in the vicinity of a seamount, the Daini Kinan Kaizan, south

of Japan. La mer, 30, 17–26.Taft, B. A. (1972): Characteristics of the flow of the Kuroshio south of Japan. p. 165–214. In Kuroshio—Its Physical

Aspects, ed. by H. Stommel and K. Yoshida, University of Tokyo Press, Tokyo.Yoshioka, H., T. Sugimoto, Y. Sekine, S. Serizawa and H. Kunishi (1986): Temperature structures and geostrophic

current around the Bump, Tosa-Bae off Kii Channel south of Japan. Umi to Sora, 61, 101–109 (in Japanese withEnglish abstract).

observation of oceanic structure around tosa-bae southeast ... · 544 y. sekine et al. topography...

Documents