1. Introduction
The East Sea Current Measurement 1 (EC1) station (blue diamond in Fig. 1), located between Ulleung Island (UlleungDo) and Dok Island (DokDo), has played a crucial role in longterm monitoring of deep ocean currents in the East Sea (ES) since its establishment in 1996 (Chang et al., 2002; Chang et al., 2004; Lee and Kim, 2019; Lee and Nam, 2021; Kim et al., 2024). The station is strategically situated at coordinates 131.383°E, 37.347°N in the Ulleung Interplain Gap (UIG), connecting the Japan Basin with the Ulleung Basin. This region, where EC1 is situated, is significant due to the formation of bottom water in the Japan Basin, which flows into the Ulleung Basin (Han et al., 2020; Kim et al., 2024). The deep current at the UIG, located between Ulleung Island and Dok Island, originates from the Japan Basin, where deep water forms in the East Sea. The southward movement and velocity of this current are influenced by both regional and global climate patterns. Understanding these dynamics is essential for assessing environmental impact locally in Korea and on a global scale (Yoon et al., 2018; Han et al., 2020).
Data collected from EC1 have provided critical insights into mean flow and variability (Chang et al., 2002), intermediatelevel circulation (Park et al., 2004), observed deep circulation (Teague et al., 2005), and upper circulation patterns (Mitchell et al., 2005). These data have also been used to study the monitoring system (Kim et al., 2005), the vertical structure of low-frequency currents (Kim et al., 2009), deep flow and transport through the UIG (Chang et al., 2009), variability of the Dokdo abyssal current (Kim et al., 2013), decadal changes in meridional overturning circulation in the ES (Han et al., 2020), changes in the circulation of the deep ES (Lee and Nam, 2023).
Additionally, EC1 has proven valuable for observing other oceanic phenomena, such as observing sediment transport, dissolved materials, and internal waves. For example, studies have observed dissolved oxygen at the bottom boundary layer of the Ulleung Basin (Kang et al., 2010), N2 production through denitrification and anammox (Na et al., 2018), intraseasonal abyssal current variability of bottom-trapped topographic Rossby waves (Shin et al., 2020), observations of enhanced internal waves (Noh and Nam, 2020), non-seasonal variations in near-inertial kinetic energy (Noh and Nam, 2021), sediment resuspension (Kim et al., 2020), and oceanic carbon cycling with resuspended sediments (Kim, 2021).
The primary objective of this paper is to provide an overview of the long-term observational data collected at the EC1 mooring station, with a focus on deep ocean velocities. The analysis examines the implications of these observations in relation to global climate change, particularly in the context of deep water formation in the Japan Basin within the East Sea, which has broader significance for global climate patterns (Yoon et al., 2018; Han et al., 2020). Special attention is given to addressing data inconsistencies and highlighting the importance of improved data management practices to ensure the future usability of these valuable datasets.
2. Data and Methods
The data used in this study were obtained from the EC1 station, which has been collecting minute-level observational data on zonal and meridional velocities since 1996. The data set, spanning 28 years (1996-2023), includes various parameters such as velocity, temperature, salinity, and dissolved oxygen. This extensive dataset was downloaded from the SEA scieNtific Open data Edition (SEANOE) platform (Lee and Nam, 2021; Kim et al., 2024).
To analyze long-term trends in deep ocean currents, the data were filtered and plotted to identify periods of reliable observations. Previously quality-controlled data were retrieved, and only data with quality flags indicating ‘good data’ and ‘probably good data’ were selected for use in this study. Several instruments, including rotor current meters (RCMs), aquadopp current meters, and acoustic Doppler current profilers (ADCPs), were used for velocity measurements. Velocity at a single depth can be measured using an RCM, while velocities at multiple depths can be measured using an ADCP. Although the former is more cost-effective and easier to deploy, the latter allows for the collection of more comprehensive data. The velocities measured between 1,800 m and 2,500 m depths were treated as bottom velocities. Monthly mean velocities for the zonal and meridional components were subsequently calculated, following the methodology of the previous study (Fig. 4c from Han et al., 2020). If there was any velocity data within a month, those data were calculated as a monthly mean. After getting monthly mean data, the trend, coefficient of determination (R2) and root mean square error (RMSE) were calculated without the gaps between the data. The bottom velocities in November 1996 (the starting month of EC1 observation), November 2020 (the end of observation period from Lee and Nam, 2021), and April 2023 (the end of observation period from Kim et al., 2024) were calculated.
During data processing, an anomaly was detected in the velocity data from 10:30 on July 28, 2012 to 16:30 August 28, 2014. These values were significantly larger than expected due to an inconsistency in units, which recorded as cm s-1 instead of m s-1. To correct this, the data were divided by 100, yielding corrected monthly mean velocities.
3. Results
Over the 28-year period from 1996 to 2023, EC1 has recorded a wealth of data on deep ocean currents, providing insights into the behavior of zonal and meridional velocities at various depths. The data were recorded at intervals ranging from 15 minutes to hourly (Fig. 2). However, data gaps were identified during certain periods, particularly in the surface layer (1999- 2002, 2005-2006, 2012-2014, and 2017), in the middle layer (1997-1998, 2002-2004, 2012-2015, and 2020-2023), and in the bottom layer (1997-1998, 2002-2005, 2017) (Fig. 2).
The observed data were categorized into three layers to assess data availability and enhance analysis of the bottom ocean current (Fig. 3) These layers were defined as the surface (0 m to 900 m), middle (900 m to 1,800 m), and bottom (1,800 m to 2,500 m) layers. Variations in depth were noted, corresponding to the placement of the observation equipment. The surface layer (0 to 900 m) shows sporadic data availability between 1996 and 2023, with several gaps in the observational record. There is an evident variation in the depth at which the velocity measurements were taken, particularly between 0 and 400 m, with dense measurements noted around 100 m and 400 m during certain periods with ADCP. The raw data are plotted alongside monthly averages, showing consistency in the longterm trend, although certain periods (e.g., 2012-2014) exhibit higher variability. The variations in depth and measurement gaps suggest that equipment availability and placement affected data continuity. In the middle layer (900 to 1,800 m), velocity measurements show more consistent depth coverage between 1,300 m and 1,500 m. There is less variability in the depth of observations compared to the surface layer, though some data gaps still occur, particularly before 2000 and after 2020. The longer periods of consistent depth and data collection in this layer provide more stable data than in the surface layer. The bottom layer (1,800 to 2,500 m) demonstrates the most stable measurement depth. Data collection here is more consistent, with less variability in depth compared to the other layers. The velocity measurements in the bottom layer exhibit relatively stable, with fewer gaps in the record. The comparison of raw and monthly-averaged data highlights a smooth long-term velocity observation, with minimal fluctuations, indicating stable bottom current observations over the observed period.
During the analysis, it was discovered that velocity data collected between 2012 and 2014 required corrections (Fig. 4). Specifically, the velocity values during this period were recorded using incorrect units, with values reported in cm s-1 rather the standard m s-1. This discrepancy resulted in significant anomalies in the dataset, particularly in file such as ‘OS_EC1_201207_D_RCM11-2200m.nc’ from SEANOE platform (Lee and Nam, 2021). To resolve this issue, the values were converted from cm s-1 to m s-1, ensuring consistency with the rest of the record. Fig. 4 illustrates the corrections made to the zonal and meridional velocities at a depth of 2,200 m. The zonal velocity (upper part of Fig. 4) shows raw and monthly-averaged data. Prior to correction, the data from 2012 to 2014 displayed highly erratic behavior with unusually high peaks, which were consistent with the unit conversion error. After correction (upper part of Fig. 5), the data show trends more in line with the longterm velocity patterns observed in other periods. Similarly, the meridional velocity (lower part of Fig. 4) displayed substantial variability during the same timeframe, which was also corrected by adjusting the unit conversion. Post-correction (lower part of Fig. 5), the meridional velocity trends returned to more reasonable levels, aligning with the overall dataset. The corrections ensured the integrity of the long-term trends, particularly in the bottom layer, where consistent measurements at 2,200 m provided valuable insights into deep ocean currents.
Once corrected, the monthly mean zonal and meridional velocities were recalculated (Fig. 5). Fig. 5 presents the corrected zonal and meridional velocities at 2,200 m from 1996 to 2023, illustrating both the raw and monthly-averaged data. In the case of the zonal velocity (upper part of Fig. 5), the corrected data reveal a clear periodic trend, with several notable peaks occurring between 1999 and 2022. Before corrections, the data from 2012 and 2014 exhibited large spikes due to the incorrect unit conversion (upper part of Fig. 4). Post-correction (upper part of Fig. 5), the dataset shows more consistent variations, particularly around 0 m s-1, with occasional deviations that align with expected long-term deep current variability. Similarly, in the meridional velocity (lower part of Fig. 5), the corrected data exhibit more stable trends, with a significant reduction in the previously observed spikes (lower part of Fig. 4). The long-term trend in the meridional velocity is characterized by moderate fluctuations, with values hovering around zero, interspersed with both negative and positive deviations that indicate southward and northward flows, respectively (lower part of Fig. 5).
Long-term trends in the zonal and meridional velocities were calculated using the corrected data. For zonal velocity (upper part of Fig. 6), the trend shows a slight positive increase over time, with an average trend of 0.30 mm s-1 yr-1. The data exhibit moderate variability around zero, with some significant positive and negative peaks, notably in 1999, 2010, and 2019, followed by a more consistent pattern of minor oscillations around the zero between 2004 and 2007. The upward trend suggests a gradual decrease in westward zonal flow throughout the study period, potentially indicating long-term shifts in the deep ocean circulation at this depth. In contrast, the meridional velocity (lower part of Fig. 6) shows a slightly negative trend of -0.02 mm s-1 yr-1, pointing to a very minor increase in the southward flow over the same period. Despite this negative trend, the data remains clustered near zero, with moderate variability over the years. Notably, larger deviations are observed around the early 2000s and between 2010 and 2020, though the overall meridional velocity remains relatively stable with minimal long-term change. The overall trends indicate a slight increase in zonal velocity (0.30 mm s-1 yr-1) and a near-zero trend for the meridional velocity (-0.02 mm s-1 yr-1) (Fig. 6).
The zonal velocity decreased from -1.06 cm s-1 in November 1996 to -0.33 cm s-1 in November 2020, and further to -0.25 cm s-1 in April 2023. Meanwhile, the meridional velocity remained stable, with only minor fluctuations over the same period, changing from -1.05 cm s-1 in November 1996 to -1.11 cm s-1 in both November 2020 and April 2023. Overall velocities shifted from 1.49 cm s-1 in November 1996 to 1.16 cm s-1 in November 2020 and 1.14 cm s-1 in April 2023, with the current direction changing from southwestward in November 1996 to predominantly southward in April 2023 (Fig. 7). These trends indicate a gradual weakening of the southward bottom current in the region, which may be linked to broader shifts in oceanic circulation patterns in the ES and the effects of global warming (Chang et al., 2009; Yoon et al., 2018; Han et al., 2020). This suggests that the deep water formation in the Japan Basin, along with its southward flow through UIG and the trough between Ulleung Island and Dok Island, has been decreasing, consistent with findings from previous studies.
4. Summary and Discussion
This paper presents a detailed analysis of long-term deep ocean currents between Ulleung Island and Dok Island in the ES from 1996 to 2023, using data collected from the EC1 mooring station. The study examines both zonal and meridional velocity trends at depths up to 2,500 m, particularly the bottom current at around 2,200 m. The long-term data collected at EC1 over the 28-year period (1996-2023) have offered invaluable insights into deep ocean current dynamics in the ES. Although there is an abundance of observation data, including those from the SEANOE website (Lee and Nam, 2021; Kim et al., 2024), inconsistencies in unit measurement-particularly between 2012 and 2014-highlight the need for improved data management practices.
Over the 28-year period, the zonal velocity exhibited a slight positive trend (+0.30 mm s-1 yr-1), indicating a reduction in westward flow, while the meridional velocity showed a nearzero trend (-0.02 mm s-1 yr-1), pointing to a stable but slowly increasing southward flow. The overall velocity decreased from 1.49 cm s-1 in 1996 to 1.14 cm s-1 in 2023, with the current direction shifting from southwestward to predominantly southward. The gradual weakening of the southward bottom current near 2,200 m consistent with prior studies, suggesting reduced deep water formation in the Japan Basin and a decline in southward flow through the UIG. These trends indicate a progressive weakening of the southward bottom current in the region, potentially related to broader shifts in oceanic circulation patterns in the ES and the impacts of global warming. Changes in bottom water formation in the Japan Basin, likely driven by climate change warming, may be contributing to these shifts.
Given the observed weakening of the southward bottom current, future research should prioritize quantifying the effect of global warming on deep water formation in the Japan Basin. This will clarify how variations in temperature, salinity, and circulation patterns in the ES influence long-term shifts in oceanic currents. The weakening of the bottom current could have profound implications for regional climate and marine ecosystems. Future studies should explore how changes in circulation affect nutrient transport, biological productivity, and the overall health of the marine ecosystem in the ES. The study emphasizes the importance of reliable, accurate, and consistent long-term data management. Future efforts should focus on creating robust data management systems and maintaining continuous monitoring to track changes in ocean circulation and their influence on regional and global climate systems.
The detected weakening of the southward bottom current suggests that oceanic circulation in the ES is undergoing significant changes, likely due to global climate change. This shift has direct consequences for the regional climate, marine ecosystems, and future deep water formation processes in the Japan Basin. Ongoing long-term monitoring and improved data management will enable future research to build on these findings, enhancing our understanding of the broader implications of these oceanographic changes and their potential impact on climate systems and marine life in the region.