Variability in upper layer circulation
The main features of upper layer circulation in the East Sea have investigated through observations and numerical experiments in previous studies (Kawabe 1982; Yoon 1982a,b,c; Hase et al. 1999; Park et al. 2013; Kim et al. 2020). One of the main currents, the TWC inflowing through the KTS, bifurcates into two branches: the EKWC flowing northward along the Korean coast and the Nearshore Branch flowing northeastward along the Japanese coast. In addition, a cyclonic gyre appears in the northern part (Fig. 1). The model results clearly reproduce distinct features of the upper layer circulation. The distributions of the intrinsic variability (Intrinsic-Exp) were large around meandering and eddy activities regions in the southern part of the East Sea (Fig. 3a), similar to previous studies (Trusenkova, 2014; Choi et al. 2018; Kim et al. 2024). Based on the Control Run, the IV_Qnet experiment revealed that the variability in upper layer circulation increases, especially in the Yamato Basin, but does not significantly increase in the Ulleung Basin (Fig. 3b). The surface heat flux plays an important role in the formation of cold water in the northern region. The correlation between the surface heat flux and nonseasonal SST, with a 1-month time lag in the northern region was approximately 0.44. This means that the effects of the surface heat flux on the upper ocean have a time lag of approximately one month. Therefore, to analyze the hydrographic conditions of the upper ocean caused by surface heat flux, winter data such as temperature and density data, were averaged from January to March.
It is necessary to consider how interannual variations in the winter surface heat flux in the northern region could influence the variability in the upper layer circulation in the southern region through which physical processes occur. The surface heat flux not only contributes to cold water formation in the northern region but also enhances ventilation. Consequently, the planetary (topographic) beta effects strengthen (weaken), leading to the formation of the EKWC, one of the TWC branches (Kim et al. 2020). Considering to the role of surface heat flux in the East Sea based on previous researches, two possibilities can be considered for inducing variability in the upper layer circulation due to interannual variation of the surface heat flux: 1) the changes in the meandering path due to the formation and expansion of the cold water in the northern region and 2) how it affects the separation latitude of the EKWC, although the variability in the Ulleung Basin did not increase significantly because of interannual variations in the surface heat flux.
As mentioned earlier, the water temperature and density at a depth of 100 m were composited for periods of strong and weak cooling periods, and the spatial distributions during these periods were analyzed (Fig. 4). During strong cooling in winter, cold water corresponding to temperatures less than 1°C was distributed over the Japan Basin at a depth of 100 m. Furthermore, the cold water extended to the entrance of the Yamato Basin, leading to meandering of the TWC and development of the subpolar front between the southern and northern region. In contrast, during weak cooling, cold water was limited to the eastern Japan Basin. When the 6°C isotherms (white lines in Fig. 4a, b), which represent the northern boundary of the main TWC at a depth of 100 m, with strong cooling and weak cooling conditions were compared, significant differences in the meandering phase of the TWC appeared in the Yamato Basin. In the water density distribution at a depth of 100 m, the 27.13 \(\:{\sigma\:}_{\theta\:}\) isopycnal, corresponding to the center of the ESIW, extended distinctly into the Yamato Basin (white dashed line in Fig. 4c, d). In these cases, the buoyancy frequency (\(\:N=\:\sqrt{-(g/{\rho\:}_{0})(\:\partial\:\rho\:/\partial\:z)}\)) at point A´ peaks at approximately 80 m depth under strong cooling conditions, which is approximately 50 m shallower than that under weak cooling conditions (Fig. 5). This indicates that, under strong cooling conditions, the cold water widely formed in the northern part extends into the southern part and penetrates beneath the warm TWC at relatively shallower depths compared to weak cooling conditions. At shallow depths, cold water subduction can induce baroclinic instability. The intensity of baroclinic instability can be estimated by the eddy growth rate (Stammer 1998; Zhai et al. 2008), which is expressed as follows:
$$\:EGR\:\left(eddy\:growth\:rate\right)=\:\frac{f}{N}\:\frac{dU}{dz}$$
2
where \(\:f\) and \(\:N\) indicate the Coriolis parameter and the buoyancy frequency, respectively. \(\:du/dz\) represents vertical shear. In this study, monthly mean EGR was calculated from the sea surface to a depth of 100 m in the Yamato Basin (Fig. 6), where the large variability in the upper layer circulation appears to be due to the extension of the cold water in the northern region. At this time, the EGR during strong cooling periods was greater than that during weak cooling periods, especially in February. Qiu and Miao (2000) suggested that variations in the Kuroshio path were influenced by baroclinic instability leading to the development of meandering system. In the East Sea, increased baroclinic instability led to the development of the meandering TWC path (Fig. 6). However, over time, the difference in the EGR between the strong and weak cooling periods decreased gradually. By August, the EGR was similar in both cases, and the main path of the TWC was similar.
Another possibility for increasing variability in the southern part of the East Sea is variability in the separation latitude of the EKWC due to interannual variation in surface heat flux. Previous studies have analyzed numerical experiments on wind forcings as the major atmospheric forcings affecting the separation latitude of the EKWC (Kim and Yoon, 1996; Yoon et al. 2005; Kim et al. 2020). However, this study focused on the surface heat flux, highlighted as a significant external forcing influencing the formation of the EKWC, as demonstrated by Kim et al. (2020). We analyzed how interannual variation in surface heat flux affects variability in upper layer circulation, especially the separation latitude of the EKWC. Seung and Nam (1991) and Ito (2014) reported that during strong cooling, cold water in the northern part of the East Sea extended southward along the Korean coast, influencing the separation latitude of the EKWC. However, in the results from IV_Qnet, during strong cooling, the ESIW in the northern region was widely distributed and the cyclonic eddy called the Dok cold eddy is simulated west of the Yamato Rise, but the separation latitude of the EKWC is not significantly different from that under weak cooling conditions (Fig. 4 and Fig. 6b, c). Because Ito (2014) analyzed hydrographic conditions under strong and weak winter cooling conditions using observational data, it was difficult to separate the signals related to wind and surface heat flux independently. Combining these numerical experiments with previous research, it is inferred that only surface heat flux does not influence the separation latitude of the EKWC. However, if a positive wind stress curl acts along with strong winter cooling conditions, cold water in the northern part of the East Sea could extend to the Korean coast, thereby affecting the separation latitude of the EKWC.
Distribution of the temperature from the observation data
The characteristics of the hydrographic conditions in the southern part of the East Sea, depending on the strong cooling and weak cooling periods selected from JRA-55, were analyzed via ship observations from the National Institute of Fisheries Science (NIFS) and Japan Meteorological Agency (JMA) (Fig. 7). Owing to the rarity of both institutions observing data concurrently, we defined winter as the months of January, February, and March. Since the characteristics of the hydrographic conditions derived from observation data reflect the combined influence of both surface heat flux and wind stress, analyzing the effects of only surface heat flux is difficult. Nevertheless, even in observational data, during strong cooling periods, the 2°C isotherm extends as far south as 40°N near the Yamato Basin. In contrast, during weak cooling periods, cold water does not extend southward beyond 40°N.
According to observational data, the cold water extending southward along the Korean coast also expanded much further south during strong cooling periods than during weak cooling periods. These results are consistent with those of previous studies (Seung and Nam 1991; Ito 2014). As mentioned earlier, cold water extending southward along the Korean coast was considered to be influenced not only by surface heat flux but also significantly by wind stress. Therefore, in the next section, we discuss how wind stress affects the separation latitude of the EKWC.
Wind stress response to the latitude of the EKWC
According to the results of IV_Qnet in this study, the variability in the separation latitude of the EKWC seems to be unrelated to the interannual variation in the surface heat flux. Although cold water forms in the northern part of the East Sea during strong winter cooling, if the cyclonic gyre does not develop significantly, cold water cannot expand southward along the Korean coast. Therefore, to analyze the relationship between the interannual variation in wind stress and the separation latitude of the EKWC, an additional experiment, which applied the interannual variation in wind stress (IV_Wind), was conducted. Similar to the IV_Qnet experiment, during the analysis period, winter wind stress curls exceeding 1 standard deviation were classified as strong and weak positive wind stress curl conditions. The time series of the winter wind stress curl in the northern part of the East Sea is similar to that in the entire East Sea (not shown here). From winter to summer, under strong positive wind stress conditions, cold water below 2°C was widely distributed at a depth of 100 m, and the separation latitude of the EKWC was farther south than that under weak positive wind stress conditions (Fig. 8). The difference in the separation latitude of the EKWC between the two conditions distinctly appears in summer. In the vertical structures along the 130°E line in August, during strong positive wind stress conditions, the thickness between 1°C and 2°C isotherms is prominently thickened and the subpolar front is located further south, because cold water in the northern part flows southward in the subsurface (Fig. 9). In other words, when the upper layer circulation in the northern part of the East Sea strengthens due to a positive wind stress curl rather than strong cooling conditions, the 2°C isotherm in the northern part prominently extends further southwestward. At a result, the northward flow of the EKWC was impeded, but the isotherm was not clearly different in the Yamato Basin.