Journal of Oceanography, Vol. 55, pp. 197 to 205. 1999

Long-term Variations of Potential Temperature and Dissolved Oxygen of the Japan Sea Proper Water HIDETO MINAMI1, YUJI KANO2 and KAN OGAWA2 1Maizuru 2Japan

Marine Observatory, 901 Shimofukui, Maizuru 624-0946, Japan Meteorological Agency, 1-3-4 Otemachi, Chiyoda-ku, Tokyo 100-8122, Japan

(Received 23 March 1998; in revised form 9 November 1998; accepted 24 November 1998)

Observed potential temperatures and concentrations of dissolved oxygen are analyzed to elucidate their variations during the period from 1958 to 1996 at Stn. P (37°43′ N, 134°43′ E) and from 1965 to 1996 at Stn. H (40°30′ N, 137°40′ E) in the Japan Sea. At Stn. P, increases of the potential temperature for the period are found below 800 m depth with the largest value of 0.16 ± 0.09°C per century at 800 m depth. At Stn. H, the potential temperature increased below 500 m depth. The increase rate has the largest value of 0.50 ± 0.18°C per century at 500 m depth and it is 0.30 ± 0.09°C per century at 800 m depth. The concentrations of dissolved oxygen increased around 800 m depth at Stn. P. At Stn. H, they increased above 800 m depth. On the other hand, they decreased below 1200 m depth at both stations. The layer of the dissolved oxygen minimum has deepened in these decades. These features appearing in the distributions of temperature and dissolved oxygen are successively simulated by a vertical one-dimensional advection-diffusion model including consumption of dissolved oxygen and termination of the deep water supply. These results suggest that the supply of the Japan Sea Proper Water into the deep layer, which is cold and rich in dissolved oxygen, has been decreasing for the last four decades.

1. Introduction The Japan Sea is enclosed by the Asian continent and the islands of Japan and it is connected with outer seas through four narrow and shallow straits, namely, Tsushima, Tsugaru, Soya and Mamiya Straits. The deep part of the sea floor of the Japan Sea mainly consists of the Japan Basin in the northern part, the Yamato Basin in the southeastern part and Tsushima Basin in the southwestern part around the Yamato Rise in the middle part, as shown in Fig. 1. The deepest sea floor is as deep as 3700 m. The Tsushima Warm Current, whose water flows from the East China Sea through the Tsushima Straits, covers only the upper layer of the Japan Sea, and the deep water, called the Japan Sea Proper Water, is located below the Tsushima Warm Current (Shuto, 1982a). It has been revealed that the Japan Sea Proper Water does not change seasonally by the statistical study of the oceanographic data seasonally observed by the Maizuru Marine Observatory (Minami et al., 1987). The Japan Sea Proper Water exists below about 500 m depth with water temperature lower than 1°C and salinity between 34.0 and 34.1 psu (Shuto, 1982b). Its dissolved oxygen is richer in the northern part of the Japan Sea than in the southern part. The layer of the minimum dissolved oxygen appears between 1000 m and 1500 m depths (Minami et al., 1987). The Japan Sea Proper Water makes up about

80% of the total amount of water in the Japan Sea. Because the deep water in the Japan Sea is separated from other seas by four straits shallower than about 200 m depth, this deep water is not affected directly by the outer seas. Nitani (1972) mentioned that the production of the Japan Sea Proper Water is related to the air temperature at Vladivostok. Martin et al. (1992) suggested that the bottom water in the Japan Sea is formed in proportion to the frequency of severe winter storms by sinking of the surface water which becomes heavier due not only to cooling and evaporation but also to icing. Thus, the formation of the Japan Sea Proper Water is affected by the climate over the Japan Sea. Gamo (1995) and Gamo et al. (1986) showed that the dissolved oxygen in the bottom water has decreased and the oxygen minimum layer has deepened for 15 years from 1969 to 1984 by using data from four hydrographic cruises. Considering these studies, we analyze the vertical distributions of potential temperature and dissolved oxygen, and their long-term tendencies by using the data obtained by seasonal hydrographic observations carried out by the Maizuru Marine Observatory. In addition we propose a conceptual model which can explain the variations of both potential temperature and dissolved oxygen, and discuss their relations to the climate change over the Japan Sea. 197

Copyright  The Oceanographic Society of Japan.

Keywords: ⋅ Japan Sea, ⋅ Japan Sea Proper Water, ⋅ climate change, ⋅ long-term variation, ⋅ one-dimensional model.

Fig. 1. Bottom topography and the station locations of Stns. P and H in the Japan Sea.

Fig. 2. Time series of potential temperature at 100 m and 200 m depths at Stns. P and H with regression lines.

198

H. Minami et al.

2. Observed Long-term Variations In order to analyze the deep water in the Japan Sea, we choose two stations, called Stns. P and H, where the Maizuru Marine Observatory has been carrying out seasonal hydrographic observations down to 2500 m depth. The mean position of Stn. P is 37°43′ N, 134°43′ E and its mean depth is 2960 m. It is located at the southwestern part of the Yamato Basin (see Fig. 1). The position of Stn. H is 40°30′ N, 137°40′ E and its mean depth is 3120 m. It is located near the boundary between the Yamato Basin and the Japan Basin. Figure 2 shows the time series of potential temperature at 100 m and 200 m depths at Stns. P and H. These potential temperatures at fixed depths are calculated by vertical linear interpolation. Because almost all of the water temperatures at 100 m depth at Stn. P in Fig. 2 are warmer than 6°C, Stn. P is located in the area under the influence of the Tsushima

Warm Current (Shuto, 1982a). So the sub-surface above about 500 m depth at this station is considered to be influenced by the Tsushima Warm Current because the current seems to reach about 500 m depth (Shuto, 1982a; Minami et al., 1987). Stn. H is considered to lie within the cold water area north of the sub-polar front, because the water temperature at 200 m depth is lower than about 1°C. The water temperatures and salinities at eight standard depths (i.e. 500, 600, 800, 1000, 1200, 1500, 2000 and 2500 m) are estimated by vertical linear interpolation of the data at observed depths, and potential temperatures are calculated from these interpolated values at the eight standard depths. The analyzed data are from 1958 to 1996 at Stn. P and from 1965 to 1996 at Stn. H. At Stn. P the data are obtained within a circle of 15 miles radius from 1965 to 1996, and within a 45 miles radius from 1958 to 1964. At Stn. H all the data are

Fig. 3. Time series of potential temperature at Stn. P with regression lines, at indicated depths which are shown at lower-right corner in each panel. Long-term Variations of Potential Temperature and Dissolved Oxygen

199

Table 1. The changing rates of potential temperature and dissolve oxygen and their 95% significant ranges. Depth (m)

500 600 800 1000 1200 1500 2000 2500

Potential temperature (°C/century)

Dissolved oxygen (µmol/l/century)

Station P

Station H

Station P

Station H

–0.12 ± 0.18 0.07 ± 0.10 0.16 ± 0.09 0.15 ± 0.07 0.12 ± 0.06 0.09 ± 0.05 0.06 ± 0.04 0.04 ± 0.04

0.50 ± 0.18 0.42 ± 0.10 0.30 ± 0.09 0.27 ± 0.08 0.22 ± 0.06 0.12 ± 0.05 0.04 ± 0.04 0.02 ± 0.05

–4.9 ± 13.9 10.5 ± 13.2 23.2 ± 15.7 2.8 ± 11.9 –13.8 ± 12.5 –33.0 ± 20.0 –56.3 ± 32.2 –54.7 ± 35.2

17.1 ± 14.1 14.5 ± 13.4 27.0 ± 16.0 10.5 ± 12.1 –12.7 ± 12.7 –47.9 ± 20.3 –72.8 ± 32.7 –71.5 ± 35.9

Fig. 4. Same as Fig. 3, but for series of dissolved oxygen at Stn. P.

200

H. Minami et al.

obtained within a 60 miles radius circle. Figure 3 shows time series of deep potential temperature at Stn. P, with the linear regression line. The changing rates of potential temperature per century are shown in Table 1, with 95% confidence limits. Potential temperature increases below 800 m depth, as shown in both the figure and the table. In particular, the warming rate found between 800 m and 1200 m depths exceeds 0.1°C per century, with a maximum value of 0.16 ± 0.09°C per century at 800 m depth. Time series of dissolved oxygen at Stn. P are shown in Fig. 4, with the regression line. The changing rates of dissolved oxygen per century are shown in Table 1, together with 95% confidence limits. Dissolved oxygen increases around 800 m depth, while it decreases below 1200 m depth. The deeper the depth, the clearer the decreasing tendency is. Its largest rate of decrease is –56.3 ± 32.2 µmol/l per century at 2000 m depth.

Time series of potential temperature and dissolved oxygen with the regression line of the long-term variation at Stn. H are shown in Figs. 5 and 6, respectively. Their rates of change, with 95% confidence limits, are also shown in Table 1. Potential temperature increases at every depth deeper than 500 m at a rate of 0.50 ± 0.18°C per century at 500 m depth and 0.30 ± 0.09°C per century even at 800 m depth. Dissolved oxygen decreases below 1200 m depth and appears to increase above 800 m depth. Its largest rate of decrease is –72.8 ± 32.7 µmol/l per century at 2000 m depth. Figure 7 shows the vertical profiles of potential temperature and dissolved oxygen derived from the values of regression lines in Figs. 3–6 in the years 1960 (only for Stn. P), 1970, 1980 and 1990. Figure 7 indicates that potential temperature has increased below 600 m depth at both stations, and the layer of the dissolved oxygen minimum has deepened over the decades; the layer was located at 800 m

Fig. 5. Same as Fig. 3, but for series of potential temperature at Stn. H. Long-term Variations of Potential Temperature and Dissolved Oxygen

201

Fig. 6. Same as Fig. 3, but for series of dissolved oxygen at Stn. H.

in 1960, 1000 m in 1970, 1200 m in 1980, and disappeared in 1990 at Stn. P. The same feature is also found at Stn. H; the minimum layer is located at depths of 1000–1200 m in 1970, 1500 m in 1980 and 1500–2000 m in 1990. Below 1200 m depth, dissolved oxygen has decreased over the decades, while it has increased between 600 m and 800 m depths at both stations. 3. Conceptual Model Here we analyze causes of the long-term variations of potential temperature increase and dissolved oxygen decrease in the deep water at Stns. P and H over the years. Gamo (1995) proposed a hypothesis that the sinking of the sea surface water, which is cold and rich in dissolved oxygen, occurs around the north part of the Japan Sea in winter and the sinking water, called the Japan Sea Proper Water, has spread over a deep layer in the Japan Sea. Thus, 202

H. Minami et al.

the high dissolved oxygen water in the deep layer at Stns. P and H is probably horizontally advected from the source area of the Japan Sea Proper Water. If the supply of this water terminates, dissolved oxygen in the deep water will decrease because its biochemical consumption becomes more dominant. In the same time, potential temperature will increase because diffusion from the upper layer becomes more dominant. 3.1 One-dimensional model Here, in order to explain qualitatively the main features of long-term variations of both potential temperature and dissolved oxygen, we introduce a vertical one-dimensional model with the following equations (Tsunogai, 1973).

∂θ ∂ 2θ ∂θ = D 2 +W ∂t ∂z ∂z

(1)

Fig. 7. The vertical profiles of potential temperature and dissolved oxygen in 1960 (only for Stn. P), 1970, 1980 and 1990 which are estimated from the regression line of each depth at Stns. P and H as shown in Figs. 3–6.

and

∂C ∂ 2C ∂C = D 2 +W − P, ∂t ∂z ∂z

(2)

where t and z are time and depth, and θ and C are potential temperature and dissolved oxygen, respectively. D, W and P denote vertical diffusion coefficient, upward advection velocity and consumption rate of dissolved oxygen, respectively, and they are assumed to be constant in t and z except for W in the deep water. In order to solve the equations, we introduce a ten-box structured model. The top box (No. 1 box) and bottom box (No. 10 box) are positioned at 500–700 m and 2300–2500 m depths, respectively, and the thickness of each box is 200 m, as shown in Fig. 8. Since the bottom water in the Japan Sea has an almost vertically constant potential temperature, as observed from the CTD (conductivity/temperature/depth profiling system) records at Stns. P and H, cold and high dissolved oxygen water is assumed to be injected equally into the deep layers. From the regression lines at Stn. P in 1958 (Figs. 3 and 4), the following conditions are introduced; the specified water

with potential temperature of 0.05°C and dissolved oxygen of 241 µmol/l is constantly injected into the four boxes below No. 7. Above the top box (No. 1 box), another specified water whose potential temperature is 0.45°C and dissolved oxygen is 237 µ mol/l is assumed to exist throughout the calculation. In order to conserve the volume in each box, the water in box i (i = 1 to 10) has to move upward with the following speeds W i (i = 1 to 10); Wi = W0 (i = 1 to 7), W8 = 3/4W0 , W9 = 2/4W0 , and W10 = 1/4W0. In order to decide the values of D, W0 and P, a calculation with a repeat step of one year is carried out by means of tried and error examination, referring to the values given by Tsunogai (1973). The vertical feature becomes a steady state after several hundreds of years. After several examinations with different values of D, W0 and P, the suitable values are chosen to well represent the features of potential temperature curved upward and dissolved oxygen with a minimum layer, as shown in Fig. 7. As a result, the following values are adopted here;

Long-term Variations of Potential Temperature and Dissolved Oxygen

203

D = 1.5 cm2 /s, W0 = 30 m/yr = 9.5 × 10–5 cm/s, and P = 0.17 µmol/l/yr = 3.8 ml/m3/yr. But these values should not be considered as the best fit of the model results to the observation; rather the results of a one-dimensional model should be considered to coincide roughly with the observation through W0 , D and P values in reasonable ranges, as indicated in some references (Tsunogai, 1973; Gamo, 1995).

Fig. 8. The outline of the one-dimensional advection-diffusion model including consumption of dissolved oxygen. The left figure shows the situation of existence of deep inflow, and the right one shows the situation of absence of deep inflow.

3.2 Results of the model calculation To simulate the changes in the distribution of potential temperature and dissolve oxygen described in Section 2, we first set up an equilibrium state with deep water supply into four bottom boxes (No. 7–10) of the model, and then we suddenly terminate the supply of deep water and upward advection (W0 = 0). After this sudden halt of the water supply, the vertical distributions of the potential temperature and dissolved oxygen begin to change, as shown in Fig. 9. The potential temperature rises as a whole, most of all around 1000 m depth where the curvature of the distribution is at a maximum. This feature is also found in the observed distributions at Stn. P (Fig. 7). Dissolved oxygen decreases in the layers deeper than 1000 m, but it changes little or even increases in shallower layers. The minimum layer moves downward with the years. These results of the model, shown in Fig. 9, are consistent

Fig. 9. The results of the advection-diffusion model of potential temperature (left) and dissolved oxygen (right). Thick lines show the equilibrium states under the condition of deep inflow, and thin lines show profiles at every ten years after sudden termination of the deep inflow. 204

H. Minami et al.

with the observational results at Stn. P as shown in Fig. 7. These consistencies between model results and observations strongly suggest that the supply of the Japan Sea Proper Water into the deep layer, which is cold and rich in dissolved oxygen, has been decreasing for the last four decades.

These results show that the Japan Sea is a typical example where climatological change can have clear effects on the formation of the deep water. Thus, the Japan Sea is one of the important seas to monitor climate change, and more comprehensive observations should be carried out in future.

4. Conclusion Long-term variations of potential temperature and dissolved oxygen of the Japan Sea Proper Water are analyzed using hydrographic data for the last four decades. The results obtained have been simulated by a vertical onedimensional diffusion-advection model. The results are as follows: (1) Potential temperature increases below 800 m depth at Stn. P. The warming rate has a maximum value of 0.16 ± 0.09°C per century at 800 m depth. At Stn. H, the potential temperature increases over all the layers below 500 m depth. The increasing rate has the largest value of 0.50 ± 0.18°C per century at 500 m depth and the rate is 0.30 ± 0.09°C per century at 800 m depth. (2) The dissolved oxygen increases around 800 m depth and above at Stns. P and H. On the other hand, it decreases below 1200 m depth at both stations. The layer of minimum dissolved oxygen has deepened during the last four decades. (3) These features are successfully simulated by a vertical one-dimensional advection-diffusion model including consumption of dissolved oxygen and termination of the deep water supply. (4) The results of this conceptual model examination and observations suggest that the supply of the Japan Sea Proper Water into the deep layer, which is cold and rich in dissolved oxygen, has been decreasing for the last four decades.

Acknowledgements The authors would like to thank Drs. T. Uji and I. Kaneko, Meteorological Research Institute of the Japan Meteorological Agency, for their kind and useful comments on the manuscript. The authors also sincerely thank two anonymous reviewers and the editor (Dr. S. Imawaki) for their valuable suggestions concerning the manuscript. References Gamo, T. (1995): Bottom circulation in the Japan Sea. Kagaku, 65, 316–323 (in Japanese). Gamo, T., Y. Nozaki, H. Sakai, T. Nakai and H. Tsubota (1986): Spatial and temporal variations of water characteristics in the Japan Sea bottom layer. J. Mar. Res., 44, 781–793. Martin, S., M. Esther and R. Drucker (1992): The effect of severe storms on the ice cover of the northern Tatarskiy Strait. J. Geophys. Res., 97, 17753–17764. Minami, H., Y. Hashimoto, Y. Konishi and H. Daimon (1987): Sea conditions in the southern part of the Japan Sea. Umi to Sora, 62, 163–175 (in Japanese with English abstract). Nitani, H. (1972): On the deep and the bottom waters in the Japan Sea. Researches in Hydrography and Oceanography, Hydrogr. Dep. of Japan, 151–201. Shuto, K. (1982a): A review of sea conditions in the Japan Sea (I). Umi to Sora, 57, 157–169 (in Japanese with English abstract). Shuto, K. (1982b): A review of sea conditions in the Japan Sea (II). Umi to Sora, 57, 171–186 (in Japanese with English abstract). Tsunogai, S. (1973): Quantification of biological activity by the constituents in sea water. Marine Biochemistry, ed. by A. Hattori, Tokyo University Press, 211 pp. (in Japanese).

Long-term Variations of Potential Temperature and Dissolved Oxygen

205