Journal of Oceanography Vol. 51, pp. 99 to 109. 1995 Thermohaline Staircases in the Warm-Core Ring off Cape Erimo, Hokkaido and Their Fluxes Due to Salt Finger HIDEO MIYAKE, SEIJI SASAKI, HIDEKAZU YAMAGUCHI, KIYOSHI MASUDA, GEN ANMA and YOSHIHIKO KAMEI Faculty of Fisheries, Hokkaido University, 3-1-1 Minato-cho, Hakodate, Hokkaido 041, Japan (Received 1 June 1994; in revised form 25 July 1994; accepted 16 August 1994) Thermohaline staircases off Cape Erimo, Hokkaido are described and their physical properties are compared with those in other seas. The mean fluxes for heat and salt across the interface induced by salt finger were estimated as 105 cal cm 2 d 1 and 0.03 g cm 2 d 1, respectively. These values were in the same order as those in Caribbean Sea. The effective eddy diffusivities were also in the same order as the Caribbean ones. This suggests that the double-diffusive convection plays on important role on the water mass conversions occurring in the interfrontal zone between the Oyashio and the Kuroshio Waters. 1. Introduction Thermohaline staircases have been reported in various oceans; the Northeastern Atlantic (Tait and Howe, 1968), the Caribbean Sea (Lambert and Sturges, 1977), the Mediterranean Sea (Johannessen and Lee, 1974) and the Equatorial Pacific (Miller and Browning, 1974). The Northwestern Pacific is thought to be one of the active regions for double-diffusive mixing, because the warm salty eddies and intrusions develop over the cooler fresher Oyashio Water. Thus thick layers, in which both the water temperature and salinity decrease with increasing depth, spread over in this region. The presence of the layers favors salt finger convection. However, in the Northwestern Pacific, thermohaline staircases have not been reported yet as far as we know. The purpose of this paper is to describe the characteristics of the staircases found off Hokkaido, Japan and to compare them with the results observed in other seas. 2. Oceanographic Background and Data The Tsugaru Warm Water (TW), which is warmer than 6 C and saltier than 33.6 psu (Sugiura, 1958; Ohtani, 1971), flows into the southeastern region off Hokkaido from the Japan Sea after passing through the Tsugaru Strait. Moreover, warm salty eddies which have sometimes detached from the Kuroshio Extension flow into the region from the south of Honshu. These eddies are referred as the Warm-core ring (WCR). Under these warm salty waters, the Oyashio Water (OW), which is colder than 3 C and less salty than 33.3 psu, lies widely in this region throughout the year (Ohtani, 1971). Typical examples of hydrographic condition in this area are shown in Figs. 1 and 2. Since the two source waters, the TW and the water of WCR are inherently warm and salty, both the water temperature and salinity decrease with depth at the lower part of eddies and the TW (see Figs. 3 and 6). The transitional layer which is distributed between the warm salty water and the cool fresh water occupies a significantly large part of the water column, making conditions favorable for salt finger convection. On the other hand, below the dicothermal layer, both the water temperature and salinity increase gradually with increasing depth down to about
100 H. Miyake et al. Fig. 1. Monthly mean horizontal temperature distribution at the 100-m depth in June 1988. TW, KE and WCR indicate the Tsugaru Warm Water, the Kuroshio Extension and the Kuroshio Warm Core Ring, respectively. Solid circles and triangles denote the station positions made in June and September 1988, respectively. The hatched areas show the region of the Oyashio Water where the water temperature was below 5 C. 800 db (Figs. 3 and 6). This thick layer constructs conditions favorable for another doublediffusive convection, diffusive-type convection (Turner, 1973). Thus, staircases caused by the double-diffusive convection are expected to exist in the interfrontal zone of the region. The double-diffusive activity is parameterized by the density ratio which is defined as R ρ = α T Z /β S Z (Turner, 1973). This means the ratio of change in density due to temperature to the change in density due to salinity. Where, α = (1/ρ)( ρ/ T) is the thermal expansion coefficient and β = (1/ρ)( ρ/ S) is the corresponding haline contraction coefficient. T Z and S Z are vertical gradient of temperature and salinity, respectively. The α is calculated by using the state equation and β is put 0.78 10 3 per psu. The salt finger convection occurs when R ρ > 1, and the diffusivetype convection occurs when 0 < R ρ < 1 (Turner, 1973). We analyze the physical properties of the staircases using the density ratio as an index of the double-diffusive convection. Data sets were obtained by using the CTD system of Neil Brown MK III. The CTD was lowered at a rate of about 30 m/min and data were recorded at a 1-db interval in a floppy diskette. The 1-db interval data were obtained by taking the mean of all the data over one decibar except for the maximum and minimum values. The accuracy of the data were 0.1% of full scale for pressure, 0.005 C for temperature and 0.005 mmho for conductivity by precalibration. The
Thermohaline Staircases in the Warm-Core Ring off Cape Erimo 101 Fig. 2. Potential temperature (a), salinity (b) and sigma-θ (c) sections along the latitude of 41 31 N in June 6 to 8, 1988. Thermohaline staircases were observed in the hatched regions.
102 H. Miyake et al. accuracy of salinity was about 0.01 psu when calibrated with the water sampled by Rosette Multi Sampler. 3. Field Observations 3.1 Thermohaline staircases off Cape Erimo in June 1988 We carried out hydrographic casts along 41 30 N latitude in June 6 to 8, 1988 (Fig. 1). Hydrographic conditions in a vertical section off Cape Erimo are shown in Fig. 2. The TW flowed in the layer of 125 300 db at Stn. 26 and intruded horizontally into the layer of 75 125 db at the next location of Stn. 27. Between the TW and the OW, step-like structures were found in the CTD record at Stn. 27 (Fig. 3). From 70 db to 200 db, both the water temperature and salinity decreased with depth, and five steps were found in the layer between 75 db and 140 db. The density ratios in these layers were around 1. Enlarged profiles of the properties at Stn. 27 are shown in Fig. 4. Five homogeneous layers with 10 20 db thickness were separated by thin interfaces. Differences of the physical properties across the interface were from about 0.3 to 0.5 C for temperature, 0.02 to 0.05 psu for salinity Fig. 3. Vertical profiles of potential temperature, salinity, sigma-θ and density ratio at Stn. 27 off Cape Erimo. The density ratio was computed with a vertical interval of 50 db.
Thermohaline Staircases in the Warm-Core Ring off Cape Erimo 103 Fig. 4. Enlarged profiles of potential temperature, salinity and sigma-θ at Stn. 27. Fig. 5. θ-s diagram at Stn. 27. TW and OW denote the Tsugaru Warm Water and the Oyashio Water, respectively. Solid circles indicate the homogeneous layers of the thermohaline staircases. Numerals are the depth in decibar.
104 H. Miyake et al. and 0.003 to 0.005 for sigma-θ, respectively. Salinities at the interface overshot slightly due to the mismatch of the response times between the temperature and conductivity sensors. These biases in salinity caused the corresponding inversions in the density profile. A θ-s diagram at Stn. 27 is shown in Fig. 5. The upper 40-db layer was covered with the modified OW which resulted from the heating of solar radiation. Between 47 db and 70 db, the water property changed rapidly from the OW to near the TW. The water column between the salinity maximum (70 db) and the salinity minimum (185 db) is favorable for the formation of salt finger. Below 185 db the thick layer which was favorable for diffusive-type convection, lay down to about 800 db. The staircases were found on the isopycnal surface of 26.6 sigma-θ from near the TW to the OW. The WCR had a diameter of 150 km and a thickness of 500 m in June 1988 (Yasuda et al., 1992). Stations 42 and 43 were sited on the marginal rims of the ring (Figs. 1 and 2). At these two stations, thermohaline structures similar to those observed at Stn. 27 were distinguished in the layer between 150 db and 400 db. These layers consisted of the transitional water from the warm salty water to the OW, which intruded isopycnally along the bottom of the WCR. Fig. 6. Vertical profiles of potential temperature, salinity, sigma-θ and density ratio at Stn. 173 off Kushiro.
Thermohaline Staircases in the Warm-Core Ring off Cape Erimo 105 Fig. 7. θ-s diagram at Stn. 173. OW denote the Oyashio Water. Solid circles indicate the homogeneous layers of the thermohaline staircases. Numerals are the depth in decibar. Fig. 8. Enlarged vertical profiles of potential temperature, salinity and sigma-θ at Stn. 173.
106 H. Miyake et al. 3.2 Thermohaline staircases off Kushiro in September 1988 Clearest staircases were found at Stn. 173 off Kushiro after three months in the same WCR as that of June. Vertical profiles of the water properties and a θ-s diagram at Stn. 173 are shown in Figs. 6 and 7, respectively. The modified OW which existed in the surface layer altered gradually by the warm salty water of the WCR, and the core water of the WCR occupied the layer between 28 db and 104 db. In the layer between 104 db and 380 db, there was a transitional water produced from the WCR water and the OW, and eight staircases were clearly observed in the upper part of this layer along an isopycnal surface of 26.6. The density ratios in this layer kept nearly 1, and those in the layer between 380 db and 800 db were from 0 to 1 (Fig. 6). Enlarged profiles of the water properties are shown in Fig. 8 and the parameters of the homogeneous layer are listed in Table 1. Mean thickness of the layers is calculated as 19.2 db from the table, and the maximum standard deviations of potential temperature, salinity and sigma-θ in the layers are within the ranges of 0.01 C, 0.002 psu and 0.001, respectively. Thus, these values are well within their accuracy except for temperature. Mean differences in potential temperature, salinity and Table 1. Layer parameters in the thermohaline staircases at Stn. 173 off Kushiro. Layer Mean depth (db) P. Tem. ( C) Sal. (psu) Sig-θ Thickness (db) 1 115 6.591 33.954 26.650 9 2 130 6.404 33.931 26.656 16 3 154 6.241 33.908 26.659 27 4 190 5.979 33.870 26.663 35 5 229 5.776 33.845 26.668 32 6 255 5.536 33.814 26.670 16 7 272 5.104 33.760 26.680 10 8 287 4.739 33.716 26.687 8 Mean 19.2 Table 2. Interfacial parameters across the staircases at Stn. 173. Interface Mean depth (db) θ ( C) S (psu) Sig-θ Thickness (db) 1 121 0.188 0.024 0.006 3 2 140 0.162 0.023 0.003 4 3 170 0.262 0.037 0.004 5 4 210 0.203 0.025 0.005 5 5 246 0.240 0.031 0.002 2 6 265 0.432 0.054 0.010 4 7 280 0.365 0.044 0.007 6 Mean 0.265 0.034 0.0052 4.1
Thermohaline Staircases in the Warm-Core Ring off Cape Erimo 107 sigma-θ across the interface are 0.265 C, 0.034 psu and 0.005, respectively, as shown in Table 2. The layer thicknesses are large at the central part of the step structures in both cases of the Caribbean Sea and off Kushiro. In the case of the Caribbean Sea, the differences in the physical properties across the interface tended to decrease with increasing depth (Boyd and Perkins, 1987), but in the case off Kushiro, they have rather large values at the lower part of the step structures. In Figs. 2(c), 3 and 6, the layers with staircases were very weak in vertical density gradient. This suggests that the homogeneous water was actively produced within these layer. 4. Fluxes Across the Salt Finger Interfaces 4.1 Density ratios across the interface By using the interfacial parameters, we can estimate the density ratios and the vertical fluxes of heat and salt produced by salt finger. First, we calculate the density ratio which is an important indicator of the double-diffusive activity (Turner, 1973). These values are listed in Table 3. The staircases off Kushiro (Stn. 173) show quite small density ratios of 1.16 to 1.29, which correspond to the smallest ones in Boyd and Perkins (1987). 4.2 Fluxes of salt and heat The double-diffusive fluxes were recently formulated by Hebert (1988) and Kelly (1990), but we applied the formula used by Boyd and Perkins (1987) for comparing our results with their ones. The vertical buoyancy flux due to salt across the salt finger interface is expressed as follows: βf S = 0.19 ( R ρ 0.50 gk T ) 1/3 ( β S) 4/3. Where, F S is the salt flux per unit horizontal area (g cm 2 s 1 ), g the acceleration of gravity (cm s 2 ), k T the thermal diffusivity of heat (put 1.5 10 3 cm 2 s 1 ), and S the salinity difference Table 3. Salt (F S ) and heat (F T ) fluxes in the staircases at Stn. 173. R ρ, αf T and βf S are the density ratio, the buoyancy flux for heat and that for salt through the salt finger interface, respectively. Interface α T (10 5 ) β S (10 5 ) R ρ βf S (10 7 cm s 1 ) αf T (10 7 cm s 1 ) F S (g cm 2 d 1 ) F T (cal cm 2 d 1 ) 1 2.41 1.87 1.29 1.38 0.73 0.015 49 2 2.07 1.79 1.16 1.54 0.90 0.017 61 3 3.35 2.89 1.16 2.90 1.70 0.032 115 4 2.50 1.95 1.28 1.45 0.78 0.016 55 5 2.81 2.42 1.16 2.29 1.36 0.025 100 6 5.05 4.21 1.20 4.52 2.57 0.050 190 7 4.02 3.43 1.17 3.60 2.10 0.040 165 Mean 1.20 2.53 1.45 0.028 105
108 H. Miyake et al. across the interface. The buoyancy flux due to heat is also expressed by the following relationship; αf T = βf S R f, R f = 0.44 R ρ 0.60 ( ) 1/2, where F T is the heat flux per unit horizontal area (cal cm 2 s 1 ), and R f the flux ratio of heat to salt. The fluxes calculated across each staircase are listed in Table 3 using the parameters mentioned above. The fluxes per day are ranged between 0.015 to 0.050 g cm 2 d 1 for salt and from 49 to 190 cal cm 2 d 1 for heat. These values are in the same order as those calculated in Caribbean Sea (Boyd and Perkins, 1987). Linden (1974) has shown that a weak uniform shear has little effects on the fluxes, but a non-uniform shear decreases the fluxes. Consequently, the fluxes are overestimated when they are calculated under no shear condition. Salt fluxes at the interfaces of 3, 6 and 7 are larger than 0.03 g cm 2 d 1 in accordance with the large salinity differences. In the lower part of the step structures, background gradients of salinity and density are larger than those in the central part. This may be the reason why the thin but large salinity steps arose there and the salt-finger convection is active. 4.3 Eddy diffusivities Using the same methods as in Lambert and Sturges (1977), we can estimate effective eddy diffusivity in the vertical direction. This is done by assuming βf S = K S βds/dz, αf T = K T αdt/dz. When using the mean salinity and temperature gradients between 104 db and 380 db (Figs. 6 and 7), the effective diffusivities are estimated as 17 cm 2 s 1 for salt and 7 cm 2 s 1 for heat, respectively (Table 4). If we estimate effective eddy diffusivities from Lambert and Sturges data in the Caribbean Sea using the same methods, we get 16 cm 2 s 1 for salt and 5 cm 2 s 1 for heat. In the Caribbean Sea, the staircases arose between the Subtropical Under Water and the Antarctic Intermediate Water (Boyd and Perkins, 1987). On the other hand, those off Kushiro occurred in the WCR. It should be noted that the effective eddy diffusivities are almost the same for both cases although the hydrographic circumstances are quite different. Table 4. Effective eddy diffusivities for salt (K S ) and heat (K T ), and background gradients of temperature (T Z ) and salinity (S Z ). Effective eddy diffusivities Background gradient K S (cm 2 s 1 ) K T (cm 2 s 1 ) S Z ( 10 2 psu m 1 ) T Z ( 10 2 C m 1 ) Boyd and Perkins (1987) 18 6 0.4 2.5 Zhurbas and Ozmidov (1983) 11.5 4.0?? Lambert and Sturges 16 5 0.25 1.85 This study 17 7 0.18 1.62
Thermohaline Staircases in the Warm-Core Ring off Cape Erimo 109 5. Concluding Remarks We described the thermohaline staircases off Cape Erimo and off Kushiro in the Northwestern Pacific. The staircases had the same scales of the physical properties as those of the large scale staircases in other seas. However, the staircases off Kushiro were characterized by their small density ratios. The vertical fluxes of salt and heat are estimated to be in the same order of magnitude as those in the Caribbean Sea. In spite of their small density ratios and background gradients, effective eddy diffusivities in the WCR off Kushiro were estimated to be the same as those in Caribbean Sea. Analysis of the CTD data in this region show that there are thick layers in which the density ratios are nearly 1. Schmitt (1981) suggested that the vertical fluxes of heat and salt should increase exponentially as the density ratio approaches 1.0, and that the existence of staircases become clearer in ocean. Thus, in the Northwestern Pacific where the Oyashio Water is in contact with the Kuroshio Water and mixes with it, the double-diffusive convection seems to play important roles on the mixing processes of the different water masses. One is the process of water mass conversion in the interfrontal zone, and the other is the vertical transport of salt and heat into the intermediate salinity minimum layer. Extensive studies in this field should be carried out in the future. Acknowledgements We wish to thank the crew members, scientists and cadets in cruises 23 and 24 of the T.S. OSHORO MARU for their cooperation in the acquisition of the CTD data. We also wish to thank Dr. J. Yoshida of the Tokyo University of Fisheries and the anonymous reviewers for their helpful comments on the earlier manuscript. This is contribution number 280 from the Research Institute of North Pacific Fisheries of the Faculty of Fisheries, Hokkaido University. References Boyd, J. D. and H. Perkins (1987): Characteristics of the thermohaline steps off the northeast coast of South America, July 1983. Deep-Sea Res., 34, 337 364. Hebert, D. (1988): Estimates of salt-finger fluxes. Deep-Sea Res., 35, 1887 1901. Johannessen, O. M. and O. S. Lee (1974): A deep stepped thermohaline structure in the Mediterranean. Deep-Sea Res., 21, 629 639. Kelly, D. L. (1990): Fluxes through diffusive staircases: A new formation. J. Geophys. Res., 95, 3365 3371. Lambert, R. B. and W. Sturges (1977): A thermohaline staircase and vertical mixing in the thermocline. Deep-Sea Res., 24, 211 222. Linden, P. F. (1974): Salt fingers in a steady shear flow. Geophys. Fluid Dyn., 6, 1 24. Miller, R. R. and D. G. Browning (1974): Thermal layering between Galapagos Islands and south America. Deep- Sea Res., 21, 669 673. Ohtani, K. (1971): Studies on the change of the hydrographic conditions in the Funka Bay. II. Characteristics of the waters occupying the Funka Bay. Bull. Fac. Fish., Hokkaido Univ., 22, 58 66. Schmitt, R. W. (1981): Form of the temperature-salinity relationship in the Central water: Evidence for doublediffusive mixing. J. Phys. Oceanogr., 11, 1015 1026. Sugiura, J. (1958): On the Tsugaru Warm Current. Geophys. Mag., 28, Takematsu Okada s anniversary vol., part 1, 399 409. Tait, R. I. and M. R. Howe (1968): Some observations of thermohaline stratification in the deep ocean. Deep-Sea Res., 15, 275 280. Turner, J. S. (1973): Buoyancy Effects in Fluids. Cambridge University Press, New York, 367 pp. Yasuda, I., K. Okuda and M. Hirai (1992): Evolution of a Kuroshio warm-core ring variability of the hydrographic structure. Deep-Sea Res., 39, suppl., S131 S161. Zhubas, V. M. and R. V. Ozumidov (1983): Formation of stepped fine structure in the ocean by thermocline intrusions. Izuvestiya, Atmospheric and Oceanic Physics, 19, 977 982 (English translation).