• For Contributors +
• Journal Search +
Journal Search Engine
ISSN : 1225-6692(Print)
ISSN : 2287-4518(Online)
Journal of the Korean earth science society Vol.41 No.4 pp.323-343
DOI : https://doi.org/10.5467/JKESS.2020.41.4.323

# Status of Observation Data at Ieodo Ocean Research Station for Sea Level Study

MyeongHee Han*
Research Institute of Oceanography, Seoul National University, Seoul 08826, Korea
*Corresponding author: skiing1@snu.ac.kr Tel: +82-2-880-4139
August 16, 2020 August 22, 2020 August 24, 2020

## Abstract

Observation data measured at Ieodo Ocean Research Station (IORS) have been utilized in oceanographic and atmospheric studies since 2003. Sea level data observed at the IORS have not been paid attention as compared with many other variables such as aerosol, radiation, turbulent flux, wind, wave, fog, temperature, and salinity. Total sea level rises at the IORS (5.6 mm yr−1 ) from both satellite and tide-gauge observations were higher than those in the northeast Asian marginal seas (5.4 mm yr−1 ) and the world (4.6 mm yr−1 ) from satellite observation from 2009 to 2018. The rates of thermosteric, halosteric, and steric sea level rises were 2.7-4.8, −0.7-2.6, 2.3-7.4 mm yr−1 from four different calculating methods using observations. The rising rate of the steric sea level was higher than that of the total sea level in the case with additional data quality control. Calculating the non-steric sea level was not found to yield meaningful results, despite the ability to calculate non-steric sea level by simply subtracting the steric sea level from total sea level. This uncertainty did not arise from the data analysis but from a lack of good data, even though tide, temperature, and salinity data were quality controlled two times by Korea Hydrographic and Oceanography Agency. The status of the IORS data suggests that the maintenance management of observation systems, equipment, and data quality control should be improved to facilitate data use from the IORS.

## Introduction

Ieodo Ocean Research Station (IORS) is located in the southwest of Jeju Island at (125.182 °E, 32.123 °N) and 149 km southwest of Mara Island, the southernmost island of Korea (Shim et al., 2004;KHOA, 2020a). Oceanic and atmospheric parameters have been observed there since 2003 because it is a suitable location for observing various phenomena such as freshwater discharge from Changjiang River, Tsushima warm current path, and air-sea interaction away from lands (e.g., Ha et al., 2019;Yeo and Nam, 2020). It is also a good location to observe the precursor of Changma (rainy season in Korea) and typhoons before they arrive in the Korean Peninsula (e.g., Moon et al., 2010;Oh et al., 2014). Also, aerosol, solar radiation, turbulent flux, wind, wave, temperature, salinity were used for various purposes of atmospheric and oceanic researches (e.g., Hwang et al., 2008;Kang et al., 2017).

There have been researches using the data observed at the IORS, as above but sea level (the level of the sea’s surface) change has not been highlighted there. The sea level, especially in a marginal sea where the IORS was built, heavily influences on the coastal areas where there are many human and living beings (Schenatto et al., 2017). Sea level is composed of steric and non-steric sea levels largely and they can be calculated using tide, water temperature, and salinity (Gill and Niller, 1973;Calafat et al., 2010). The main goal of this study is to show the status of observation and its application at the IORS since 2003. Additionally, the rising rates of sea levels are calculated to examine the observed data from the IORS as one example of applications. Especially, the present state and trends of water temperature, salinity, and density at the IORS are tested and those of total and steric sea levels from 2009 to 2018 are estimated.

## Data and Methods

Hourly tide, wave, current, water temperature, salinity, wind, air temperature and pressure, relative humidity, solar radiation, precipitation, etc. observed at the IORS were analyzed. These data were downloaded from Korean Hydrographic and Oceanographic Agency (KHOA) website (KHOA, 2020b) from 2005 to 2018 and separately obtained by email from a person in charge of the IORS at KHOA from 2009 to 2018. The former was plotted to check usable data periods, and the latter was used to calculate total and steric sea levels.

Monthly mean sea levels from satellite observation at a grid nearest to the IORS at (125.125 °E, 32.125 °N) were analyzed from 2009 to 2018 using the daily absolute dynamic topography (ADT) at Copernicus Marine Environment Monitoring Service (CMEMS). The ADT data provided by the CMEMS were gridded with a horizontal resolution of 0.25° from all satellite altimeter missions (CMEMS, 2020).

Serial oceanographic observation data collected four times a year off the south coast of Korea in National Institute of Fisheries Sciences (NIFS) were downloaded from the NIFS website (NIFS, 2020) from 2009 to 2018. The station is nearest to the IORS and its station number is 316-19 at (125.285 °E, 32.000 °N).

The Hybrid Coordinate Ocean Model (HYCOM) Global Ocean Forecasting System 3.1 (GOFS3.1) monthly mean reanalysis and analysis product (GLBb0.08) were used after being downloaded from their website (HYCOM_FTP, 2020) from 2009 to 2018.

Fourteen years (2005-2018) worth of oceanic and atmospheric observation data at the IORS were plotted to check the periods and quality of the data with a normalization method such as standard score (Schenatto et al., 2017) using Eq. (1):

$Z = X − X ¯ σ$
(1)

Here, Z, X, X, and σ are standard score, original data value, sample average, and standard deviation.

Thermosteric, halosteric, and steric sea levels (TSL, HSL, and SSL) were calculated (Gill and Niller, 1973;Levitus et al., 2005;Pinardi et al., 2014) using Eq. (2):

$SSL = − ∫ − H 0 ρ − ρ 0 ρ 0 d z = TSK + HSL = ∫ − H 0 α ( T − T 0 ) d z − ∫ − H 0 β ( S − S 0 ) d z$
(2)

Here, T, S, ρ, z, H, T0, S0, ρ0 are potential temperature, salinity, potential density, depth, bottom depth, reference potential temperature, reference salinity, reference potential density, and $α = 1 ρ 0 ∂ ρ ∂ T$ and $β = 1 ρ 0 ∂ ρ ∂ s$ are the thermal expansion and the haline contraction coefficients, respectively.

In-situ temperature and salinity were converted to potential temperature and absolute salinity, and potential density was calculated from in-situ temperature and absolute salinity using Thermodynamic Equation Of Seawater-2010 (TEOS-10, 2020).

## Results

### Status of observation and its applications for 14 years (2005-2018)

More than ten parameters have been observed at the IORS for 14 years (2005-2018) including tide, water temperature, and salinity. The hourly data were plotted using Eq. (1) with a ratio of observation period to a year (Figs. 2a-n). Although it is hard to distinguish bad data from the graph, it is easy to find some blanks such as water temperature, conductivity, and salinity data in 2005, and some peaks such as wind speed, air temperature and pressure in July 2005 in Fig. 2a.

Tide and wave have been observed for 14 years, which could be used to study for change of sea level and wave over a decade, because the percentages of observation periods per year were high (percentages at y-axis in Figs. 2a-n). Note that the percentages of yaxis in Figs. 2a-n are ratios of observation periods to a year and they do not mean the percentages of good data, but the percentages of all observed data. There was no water speed and direction, temperature, conductivity, and salinity data for the first 3 years (2005-2007, Figs. 2a-c). Water speed and direction data were collected for less than 4 years (2008-2010, and 2012, Figs. 2d-f, 2h) and have not been observed since 2013. Water temperature, conductivity, and salinity have been observed for 11 years since 2008, thus they could be used to study changes of temperature, salinity, density, and steric sea level. Wind speed and direction, air temperature and pressure, relative humidity have been observed for almost 14 years (2005-2018), thus they could be used to study atmospheric phenomena such as typhoons and fine dust pollution.

There have been atmospheric and oceanographic studies using observation data at the IORS. In atmospheric studies, the concentration and compositions of fine aerosol (Hwang et al., 2008;Park et al., 2012;Han et al., 2013;Lee et al., 2017c), and diurnal and seasonal patterns of ozone (Shin et al., 2007;Han et al., 2015) were observed and analyzed at the IORS. The onset date of Changma in Korea was studied using wind observed at the IORS (Oh et al., 2014) and seasonal characteristics and drag coefficient of surface turbulent fluxes (Oh et al., 2007;Oh et al., 2010;Yun et al., 2015;Yeo and Nam, 2020) were observed and calculated at the IORS. Also, surface energy and CO2 fluxes (Lee et al., 2004), solar radiation (Lee et al., 2017b), and photosynthetically active radiation (Byun and Cho, 2006) were measured and calculated at the IORS. The wind speed observed at the IORS was converted to that at 10 m (Byun et al., 2018) and compared with sea fog (Heo et al., 2010). The sea surface height and wind speed at the IORS were compared with satellite altimeters, scatterometers, and passive microwave radiometers (Kim et al., 2005;Jeong et al., 2008;Choi et al., 2018;Lee and Kim, 2019).

In oceanographic studies, temperature and salinity observed at the IORS were analyzed to find the characteristics of the water in the East China Sea, the Changjiang diluted water, and the influence of typhoons around the IORS (Oh et al., 2006;Moon and Kwon, 2012;Jeong et al., 2013;Lee et al., 2017a). Significant wave height observed at the IORS was compared with those from satellite altimeter and radar, and numerical simulation (Kim, 2012;Moon et al., 2016;Yang et al., 2016;Woo et al., 2018). The sea surface temperature (SST) observed at the IORS were compared with the four types of SSTs from satellite observations (Woo et al., 2019), and real-time thermal infrared SST observation system on the IORS have been established to compared with SSTs from other methods (Kang et al., 2017). Underwater ambient noise observed using a hydrophone at the IORS was analyzed for detecting environmental variation (Kim and Choi, 2006).

### Sea level rise using quality-controlled data for 10 years (2009-2018)

Even though there have been plenty of data and studies since 2003, the data quality was not good (Figs. 2a-n). Thus, I used another quality controlled data for 10 years (2009-2018, Figs. 3-6) which I got by email to investigate water temperature, salinity, and sea level, and to calculate potential density and steric sea level.

First, the sea levels from satellite observation and tide at the IORS were compared in Fig. 3 (black circles and dots). Although their data intervals were different such as one month (satellite) and one hour (tide-gauge) and there were blank data in case of tidegauge, sea level trends of satellite observation (5.555 mm yr−1 , black solid line, insignificant at 95% confidence level) and tide-gauge data (5.629 mm yr−1 , red dashed line, insignificant at 95% confidence level) were similar. The sea level trend at the IORS (about 5.6 mm yr−1 from above) was larger than those in the northeast Asian marginal seas (5.4 mm yr−1 , CMEMS (2020)) and the world (4.6 mm yr−1 , Beckley et al. (2017)) from 2009 to 2018. Moreover, it was larger than those in the northeast Asian marginal seas (3.6 mm yr−1 ) and the world (3.4 mm yr−1 ) from 1993 when satellite observation started to 2018 (Han et al., 2020). The differences between them might come from the different time periods and averaged regions of the sea level. Also, there was no meta data about tide observation at the IORS, although the sea level from satellite observation was provided after removing the inverted barometer effect.

Second, the sea level at the IORS (Fig. 3) could be divided into steric and non-steric sea levels. Thus, I calculated the steric part such as TSL, HSL, and SSL using Eq. (2), and they were plotted with temperature, salinity, and density in Fig. 4. The trends of in-situ temperature, salinity, and potential density at the surface, middle, and bottom were 0.140, 0.310, and 0.167 °C yr−1 (Fig. 4a), and −0.007, −0.043, and −0.027 psu yr−1 (Fig. 4b), and −0.0499, −0.130, and −0.101 kg m−3 yr−1 (Fig. 4c). The trends of TSL, HSL, and SSL were 3.525, 1.836, and 5.361 mm yr−1 . The trend of SSL (5.361 mm yr−1 in Fig. 4e) was similar to those of sea levels from satellite observation and tide data (5.6 mm yr−1 in Fig. 3). Thus, sea level at the IORS could be controlled almost by steric sea level based on Fig. 4e even though there was vertical potential density inversion in Fig. 4d, and TSL, HSL, and SSL could be calculated only when there were data at every three depths such as the surface, middle and bottom.

Usually the drift of the conductivity sensor was larger than that of temperature (Ando et al., 2005) and there was big drift of salinity (conductivity) at the middle depth (blue in Fig. 4b). To reduce the error of salinity (conductivity) drift, I removed the data at the middle depth and calculated TSL, HSL, and SSL only with data at the surface and bottom in Fig. 5. The trends of in-situ temperature, salinity, and potential density at the surface and bottom were 0.082 and 0.301 °C yr−1 (Fig. 5a), and −0.076 and −0.084 psu yr−1 (Fig. 5b), and −0.078 and −0.130 kg m−3 yr−1 (Fig. 5c). The trends of TSL, HSL, and SSL were 2.676, 2.332, and 5.008 mm yr−1 (Fig. 5e). The trend of SSL (5.008 mm yr−1 in Fig. 5e) from two depths was similar to that of SSL from 3 depths (5.361 mm yr−1 in Fig. 4e). Thus, sea level at the IORS could be controlled almost by steric sea level based on Figs. 4e and 5e, even though vertical potential density inversion data were removed in Figs. 5d-e.

These 10 year data were quality controlled two times by the KHOA, but there were still some bad data especially in Figs. 4b-c. Thus, additional data quality control procedures were adopted such as removing temperature data which were lower than 0 °C and upper than 34.9 °C, and salinity data which were lower than 20 psu and upper than 35 psu. After the additional data quality control, I calculated TSL, HSL, and SSL in Fig. 6. The trends of in-situ temperature, salinity, and potential density at the surface, middle, and bottom were 0.038, 0.376, and 0.310 °C yr−1 (Fig. 6a), and −0.066, −0.086, and −0.082 psu yr− 1 (Fig. 6b), and −0.061, −0.149, and −0.130 kg m−3 yr− 1 (Fig. 6c). The trends of TSL, HSL, and SSL were 4.801, 2.552, and 7.353 mm yr−1 (Fig. 6e). The trend of SSL (7.353 mm yr−1 in Fig. 6e) was larger than those of sea level from satellite and tide-gauge observations (5.6mmyr−1 ), and those of SSL from three depths (5.361 mm yr−1 in Fig. 4e) and two depths (5.008 mm yr−1 in Fig. 5e). It was probably because there were not enough data after removing the bad data from the additional data quality control in Fig. 6a-b and the vertical density inversion data in Fig. 6d, and because steric sea level could not be calculated if there was no data at any of the three depths.

The TSL, HSL, and SSL were calculated using the NIFS serial oceanographic observation (NSOO). The nearest station of the NSOO to the IORS is 316-19 and the TSL, HSL, and SSL had been calculated using data at the four depths (0, 10, 20 and 30-m depths) in Fig. 7. The trends of in-situ temperature, salinity, and potential density at 0, 10, 20, and 30-m depths were 0.139. 0.155, 0.283, and 0.157 °C yr−1 (Fig. 7a), and −0.007, −0.009, 0.011, and −0.006 psu yr−1 (Fig. 7b), and −0.038, −0.044, −0.063, and −0.043 kg m−3 yr−1 (Fig. 7c). The trends of TSL, HSL, and SSL with (Fig. 7d) and without (Fig. 7f) vertical density inversion data were 2.087, 0.143, and 2.231 mm yr−1 (Fig. 7e), and 3.012, −0.686, and 2.326 mm yr−1 (Fig. 7g),. The trends of SSL (2.231 mm yr−1 in Fig. 7e and 2.326 mm yr−1 in Fig. 7g) from the NSOO were much smaller than those at the IORS (5.361 mm yr−1 in Fig. 4e, 5.008 mm yr−1 in Fig. 5e, and 7.353 mm yr−1 in Fig. 6e). Thus, steric sea level calculated using data at the IORS and from the NSOO were different and it might be because of observation time intervals such as one hour at the IORS and three months from the NSOO.

The TSL, HSL, and SSL were calculated using global reanalysis data (HYCOM GOFS 3.1), high temporal and spatial resolution numerical model with data assimilation in Fig. 8. The trends of potential temperature, salinity, and potential density at the surface (2-m depth), middle (20-m depth), and bottom (40-m depth) were 0.106, −0.026, and −0.052 °C yr−1 (Fig. 8a), and −0.034, −0.024, and −0.004 psu yr−1 (Fig. 8b), and −0.054, −0.011, and 0.009 kg m−3 yr−1 (Fig. 8c). The trends of TSL, HSL, and SSL were 0.009, 0.662, and 0.671 mm yr−1 (Fig. 8e). The trend of SSL (0.671 mm yr−1 in Fig. 8e) was smaller than those at the IORS (5.361 mm yr−1 in Fig. 4e, 5.008 mm yr−1 in Fig. 5e, and 7.353 mm yr−1 in Fig. 6e) and from the NSOO (2.231 mm yr−1 in Fig. 7e and 2.326 mm yr−1 in Fig. 7g). It might be because of observation time intervals such as hourly snapshot data at the IORS, three monthly snapshot data from the NSOO, and monthly mean data of HYCOM GOFS 3.1. Also, HYCOM GOFS 3.1 used data assimilation once a day and its data assimilative sea level was constrained to be annually averaged climatology approximately.

## Summary and Discussion

The freely downloadable data from the website of the KHOA were 14-year data (2005-2018), which were raw data without enough metadata and flags of data quality control. The observation periods of tide, wave, wind speed and direction, air temperature and pressure, relative humidity were long enough such as 14 years, therefore, they might be useful for researchers. There were water speed and direction data only less than four years, and it was hard to use it for longterm trend and variation such as 10 years. There were enough water temperature and salinity data for 11 years, and the change of temperature, salinity, density, and steric sea level could be studied with them.

The IORS is an important atmospheric and oceanic research station in the East China Sea (Shim et al., 2004). There have been variety of observations and researches including aerosol, ozone, CO2, Changma, solar radiation, turbulent flux, wind, wave, fog, sea surface height, temperature, salinity, SST, underwater ambient noise, and typhoons since 2003. The oceanic and atmospheric studies using data observed at the IORS have been previously conducted (e.g., Moon et al., 2010;Oh et al., 2014;Ha et al., 2019;Yeo and Nam, 2020), but sea level has not been widely calculated and investigated. The data downloaded from the website of KHOA has been plotted (Figs. 2a-n) and it was found that there were not enough data to study sea level rise. Thus, the new quality controlled data were obtained by email to investigate the sea level, TSL, HSL, and SSL changes (Figs. 3-6).

The rising rate of sea level at the IORS (5.6mmyr−1 , 2009-2018) has higher value than those in the northeast Asian marginal seas and the world (5.4 mm yr−1 and 4.6 mm yr−1 , respectively, 2009-2018). The different rising rates might come from the differences of observation methods and averaged spatial area. The TSL, HSL, and SSL were calculated with quality controlled data. Their trends were 3.525, 1.836, and 5.361 mm yr−1 using quality controlled data by the KHOA at the three depths (surface, middle, and bottom), 2.676, 2.332, and 5.008 mm yr−1 using quality controlled data by the KHOA at the two depths (surface and bottom) without vertical density inversion, 4.801, 2.552, and 7.353mmyr−1 using quality controlled data by the KHOA and the additional quality controlled data at the three depths (surface, middle, and bottom) without vertical density inversion. The trends of these three results were different and their differences might not depend on cutting-edge data quality control but depend on the number of data and maintenance of system and equipment at the IORS.

The trends of TSL, HSL, and SSL with and without vertical density inversion data from the NSOO were 2.087, 0.143, and 2.231 mm yr−1 and 3.012, −0.686, and 2.326 mm yr−1 . Those of HYCOM reanalysis were 0.009, 0.662, and 0.671 mm yr−1 . The data from the NSOO had been obtained once per three months and its observation time interval was very larger than that of the IORS. HYCOM GOFS 3.1 data were monthly mean data and it was data assimilated once a day. Its global data assimilative sea level was constrained to be the climatology of annual mean, thus the users should be careful to use it especially used in sea level.

## Conclusions and Suggestions

The sea level rising rate at the IORS was 5.6 mm yr−1 from 2009 to 2018 and it was larger than those in the marginal seas around the IORS and the world. This sea level rising rate should be considered when the IORS would be rebuilt in multiple decades. The rising rates of TSL, HSL, and SSL were 2.7-4.8, −0.7-2.6, 2.3-7.4 mm yr−1 from the observations, and that of SSL was higher than that of sea level from satellite and tide observations in the case with the additional data quality control. The range of rising rates in SSL was too wide, thus, it was hard to distinguish the non-steric sea level from the sea level.

The IORS is located far southwest of Korea, although it is closer to Korea than China and Japan. Thus, it was hard to maintain the observation system and equipment frequently and adequately. Most researchers have used the observed data for a short period because there were not good and long enough data at the IORS. They would like to have more good quality data when they tried to use the data at the IORS. Even though they have an up-to-date data processing methods such as artificial intelligence technique or computer programming, they could not get reliable results from short and bad data taken from inadequately maintained sensors and equipment. Public servants in charge of the IORS data and system at the KHOA should know everything about both data and system at the IORS. Also, they should be able to calculate and plot the oceanic and atmospheric data at the IORS to produce practical and useful data for students and researchers. The data would be useless if they did not know how to manage and control those properly. I hope this study would be a good starting point to use the data observed at the IORS widely and appropriately, and would spotlight the significant effect of maintenance of data and system on the researches which use its observation data. This study could be used to raise the responsibility and obligation of public officials at ocean agencies.

## Acknowledgments

Ocean and atmospheric data at the IORS (http:// www.khoa.go.kr/oceangrid/khoa/koofs.do), NIFS serial oceanographic observation (http://www.nifs.go.kr/kodc/ soo_list.kodc), Ssalto/Duacs altimeter products which were produced and distributed by the Copernicus Marine and Environment Monitoring Service (CMEMS) (http://www.marine.copernicus.eu), HYCOM GOFS 3.1 (https://www.hycom.org) were used in this study. I thank Y-. K. Cho and S. J. Lyu, who provided me advice and suggestion of data analysis.

## Figure

Domain of the northeast Asian marginal seas and location of the Ieodo Ocean Research Station (IORS, magenta pentagram). Schematics of the regional circulation patterns, with major surfaces (red or blue) and intermediate (gray) currents transporting warm (red) and cold (blue) waters in the upper left box (Park et al., 2013). These current data are also available from Korea Hydrographic and Oceanography Agency (KHOA, http://www.khoa.go.kr). KOR, CHN, JPN, and RUS are Korea, China, Japan, and Russia, respectively. KS, TAS, TSS, and SS are Korea, Taiwan, Tsugaru, and Soya Straits, respectively.

Timeseries of observed parameters such as tide, significant wave height and frequency, maximum wave height, wave direction, water speed and direction at several depths, water temperature, conductivity, and salinity at surface, middle, and bottom depths, wind speed and direction, wind instantaneous maximum speed, air temperature, relative humidity, air pressure, and visibility from (a) 2005 to (n) 2018 at the Ieodo Ocean Research Station (IORS). Number in parenthesis after parameter was a percentage of observation period to a year, which means that it was a percentage of observed data, not a percentage of good data

Monthly mean sea level (m, black circles, upper panel) from all satellite altimeter mission observation at the nearest grid to the IORS and hourly tide-gauge observation (m, black dots, lower panel) at the IORS are plotted. The trends of sea level from monthly mean satellite observation (black solid line, upper panel) and hourly tide-gauge observation (red dashed line) are drawn, respectively.

Three layer (surface, middle and bottom) data with twice data quality controls. Hourly (a) in-situ water temperature (°C), (b) salinity (psu), (c) potential density (kg m−3 ) at the surface (3-m depth, red), middle (21-m depth, blue), and bottom (38-m depth, black) after data quality control by the KHOA are plotted for 10 years (2009-2018) at the IORS. (d) potential density differences (surface minus middle (red), surface minus bottom (blue), and middle minus bottom (black)). (e) Thermosteric (m, red), halosteric (m, blue), and steric (m, black) sea levels calculated using temperature and salinity are plotted. Steric sea level can be calculated only when there are temperature and salinity data at every three depths. Dashed lines (a, b, c, and e) are their trends.

Two layer (surface and bottom) data with twice data quality controls. Hourly (a) in-situ water temperature (°C), (b) salinity (psu), (c) potential density (kg m−3) at the surface (3-m depth, red) and bottom (38-m depth, black) after data quality control by the KHOA are plotted for 10 years (2009-2018) at the IORS. (d) Potential density difference (surface minus bottom (blue)) is removed if it is minus, which is vertical density inversion. (e) Thermosteric (m, red), halosteric (m, blue), and steric (m, black) sea levels calculated using temperature and salinity at 2 depths (surface and bottom). Dashed lines (a, b, c, and e) are their trends.

Three layer (surface, middle and bottom) data with twice and additional data quality controls. Hourly (a) in-situ water temperature (°C), (b) salinity (psu), (c) potential density (kg m−3) at the surface (3-m depth, red), middle (21-m depth, blue), and bottom (38-m depth, black) after data quality control by the KHOA and additional data quality control are plotted for 10 years (2009-2018) at the IORS. (d) Potential density differences (surface minus middle (red), surface minus bottom (blue), and middle minus bottom (black)). The potential density difference is removed if it is minus, which is vertical density inversion. (e) Thermosteric (m, red), halosteric (m, blue), and steric (m, black) sea levels calculated using temperature and salinity without vertical density inversion data are plotted. Dashed lines (a, b, c, and e) are their trends.

Three monthly (a) in-situ water temperature (°C), (b) salinity (psu), and (c) potential density (kg m−3) at 0-m (red), 10-m (green), 20-m (blue), and 30-m (black) depths after data quality control by the NIFS are plotted for 10 years (2009-2018) at the station of 316-19 nearest to the IORS. Potential density differences (0 m minus 10 m (red), 10 m minus 20 m (green), 20 m minus 30 m (blue), and 0 m minus 30 m (black)) (d) with and (f) without vertical density inversion data. Thermosteric (m, red), halosteric (m, blue), and steric (m, black) sea levels calculated using temperature and salinity (e) with and (g) without vertical density inversion data are plotted. Steric sea levels can be calculated only when there are temperature and salinity data at every four depths. Dashed lines are their trends.

Monthly mean (a) potential water temperature (°C), (b) salinity (psu), (c) potential density (kg m−3) at the surface (2-m depth, red), middle (20-m depth, blue), and bottom (40-m depth, black) from HYCOM GOFS3.1 are plotted for 10 years (2009-2018) at the nearest grid to the IORS. (d) potential density differences (surface minus middle (red), surface minus bottom (blue), and middle minus bottom (black)). (e) Thermosteric (m, red), halosteric (m, blue), and steric (m, black) sea levels calculated using temperature and salinity are plotted. Dashed lines (a, b, c, and e) are their trends.

## Reference

1. Ando, K. , Matsumoto, T. , Nagahama, T. , Ueki, I. , Takatsuki, Y. , and Kuroda, Y. ,2005, Drift characteristics of a moored conductivity-temperature-depth sensor and correction of salinity data. Journal of Atmospheric and Oceanic Technology, 22, 282-291.
2. Beckley, B. , Zelensky, N. , Holmes, S. , Lemoine, F. , Ray, R. , Mitchum, G. , Desai, S. , and Brown, S. ,2017, Global mean sea level trend from integrated multimission ocean altimeters TOPEX/Poseidon Jason-1 and OSTM/Jason-2 Version 4. Ver. 4. PO. DAAC, CA, USA.
3. Byun, D.-S. and Cho, Y.-K. ,2006, Estimation of the PAR irradiance ratio and its variability under clear-sky conditions at Ieodo in the East China Sea. Ocean Science Journal, 41, 235-244.
4. Byun, D.-S. , Kim, H. , Lee, J. , Lee, E. , Park, K.-A. , and Woo, H.-J. ,2018, Converting Ieodo Ocean Research Station Wind Speed Observations to Reference Height Data for Real-Time Operational Use. The Sea, 23, 153-178.
5. Calafat, F.M. , Marcos, M. , and Gomis, D. ,2010, Mass contribution to Mediterranean Sea level variability for the period 1948-2000. Global and Planetary Change, 73, 193-201.
6. Choi, D.-Y. , Woo, H.-J. , Park, K. , Byun, D.-S. , and Lee, E. ,2018, Validation of Sea Surface Wind Speeds from Satellite Altimeters and Relation to Sea State Bias- Focus on Wind Measurements at Ieodo, Marado, Oeyeondo Stations. Journal of the Korean earth science society, 39, 139-153.
7. CMEMS,2020, GLOBAL OCEAN GRIDDED L4 SEA SURFACE HEIGHTS AND DERIVED VARIABLES REPROCESSED (1993-ONGOING). http://marine.copernicus.eu/services-portfolio/access-to-products/?option=com_csw&view=details&product_id=SEALEVEL_GLO_PHY_L4_REP_OBSERVATIONS_008_047 (March 4th 2020).
8. Gill, A. and Niller, P. ,1973, The theory of the seasonal variability in the ocean. In Deep Sea Research and Oceanographic Abstracts 20, 141-177.
9. Ha, K.-J. , Nam, S. , Jeong, J.-Y. , Moon, I.-J. , Lee, M. , Yun, J. , Jang, C.J. , Kim, Y.S. , Byun, D.-S. , and Heo, K.-Y. ,2019, Observations utilizing Korea Ocean Research Stations and their applications for process studies. Bulletin of the American Meteorological Society, 100, 2061-2075.
10. Han, J. , Kim, J. , Kang, E. , Lee, M. , and Shim, J.-S. ,2013, Ionic Compositions and Carbonaceous Matter of PM 2.5 at Ieodo Ocean Research Station. Journal of Korean Society for Atmospheric Environment, 29, 701-712.
11. Han, J. , Shin, B. , Lee, M. , Hwang, G. , Kim, J. , Shim, J. , Lee, G. , and Shim, C. ,2015, Variations of surface ozone at Ieodo Ocean Research Station in the East China Sea and the influence of Asian outflows. Atmospheric Chemistry & Physics, 15.
12. Han, M. , Cho, Y.-K. , Kang, H.-W. , Nam, S. , Byun, D.-S. , Jeong, K.-Y. , and Lee, E. ,2020, Impacts of Atmospheric Pressure on the Annual Maximum of Monthly Sea-Levels in the Northeast Asian Marginal Seas. Journal of Marine Science and Engineering, 8, 425.
13. Heo, K.-Y. , Ha, K.-J. , Mahrt, L. , and Shim, J.-S. ,2010, Comparison of advection and steam fogs: From direct observation over the sea. Atmospheric research, 98, 426-437.
14. Hwang, G. , Lee, M. , Shin, B. , Lee, G. , Lee, J. , and Shim, J. ,2008, Mass concentration and ionic composition of PM 2.5 observed at Ieodo Ocean Research Station. Journal of Korean Society for Atmospheric Environment, 24, 501-511.
15. HYCOM_FTP,2020, HYCOM monthly mean data. ftp://ftp.hycom.org/datasets/GLBb0.08/ (August 1st 2020).
16. Jeong, J.-Y. , Shim, J.-S. , Lee, D.-K. , Min, I.-K. , and Kwon, J.-I. ,2008, Validation of QuikSCAT wind with resolution of 12.5 km in the vicinity of Korean Peninsula. Ocean and Polar Research, 30, 47-58.
17. Jeong, Y.Y. , Moon, I.-J. , and Kim, S.-H. ,2013, A study on upper ocean response to typhoon Ewiniar (0603) and its impact. Atmosphere, 23, 205-220.
18. Kang, K.-M. , Kim, D.-J. , Hwang, J.-H. , Choi, C. , Nam, S. , Kim, S. , Cho, Y.-K. , Byun, D.-S. , and Lee, J. ,2017, Establishment of Thermal Infrared Observation System on Ieodo Ocean Research Station for Time-series Sea Surface Temperature Extraction. The Sea, 22, 57-68.
19. KHOA,2020a, About Ieodo Ocean Research Station. http:// www.khoa.go.kr/eng/kcom/cnt/selectContentsPage.do?cntId=31080200 (August 1st 2020).
20. KHOA,2020b, Yearly data at Ieodo Ocean Research Station. http://www.khoa.go.kr/oceangrid/khoa/koofs.do (August 1st 2020).
21. Kim, B.-C. and Choi, B.K. ,2006, Variation of underwater ambient noise observed at IORS station as a pilot study. Ocean Science Journal, 41, 175-179.
22. Kim, C.-O. , Shim, J.-S. , Hwang, J.-S. , Lee, J.-H. , Kim, S. , and Kim, J.W. ,2005, Mutual Adjustment of Oceanographic Measurements from leodo Station and Satellite Data. Korean Journal of Remote Sensing, 21, 113-120.
23. Kim, T.-R. ,2012, Examination of altimeter wave data in the sea around Ieodo ocean research station. Korean Journal of Remote Sensing, 28, 95-100.
24. Lee, H.-C. , Lee, B.-Y. , Kim, J. , and Shim, J.-S. ,2004, Tower-based Flux Measurement Using the Eddy Covariance Method at Ieodo Ocean Research Station. Ocean and Polar Research, 26, 145-154.
25. Lee, J. and Kim, T. ,2019, Comparison Study of HY-2A, WindSat, ASCAT Satellite Wind Data with In-situ Measurement Data. Journal of Coastal Research, 91, 341-345.
26. Lee, J.H. , Moon, I.J. , Moon, J.H. , Kim, S.H. , Jeong, Y.Y. , and Koo, J.H. ,2017a, Impact of typhoons on the Changjiang plume extension in the Yellow and East China Seas. Journal of Geophysical Research: Oceans, 122, 4962-4973.
27. Lee, S.-H. , Zo, I.-S. , Lee, K.-T. , Kim, B.-Y. , Jung, H.-S. , Rim, S.-H. , Byun, D.-S. , and Lee, J.-Y. ,2017b, A strategy for environmental improvement and internationalization of the IEODO ocean research station’s radiation observatory. The Sea, 22, 118-134.
28. Lee, S. , Han, C. , Shin, D. , Do Hur, S. , Jun, S.J. , Kim, Y.- T. , Byun, D.-S. , and Hong, S. ,2017c, Characteristics of elemental and Pb isotopic compositions in aerosols (PM10-2.5) at the Ieodo Ocean Research Station in the East China Sea. Environmental Pollution, 231, 154-164.
29. Levitus, S. , Antonov, J. , Boyer, T. , Garcia, H. , and Locarnini, R. ,2005, Linear trends of zonally averaged thermosteric, halosteric, and total steric sea level for individual ocean basins and the world ocean, (1955-1959)-(1994-1998). Geophysical Research Letters, 32.
30. Moon, I.-J. and Kwon, S.J. ,2012, Impact of upper-ocean thermal structure on the intensity of Korean peninsular landfall typhoons. Progress in oceanography, 105, 61-66.
31. Moon, I.-J. , Shim, J.-S. , Lee, D.Y. , Lee, J.H. , Min, I.-K. , and Lim, K.C. ,2010, Typhoon researches using the Ieodo Ocean Research Station: Part I. Importance and present status of typhoon observation. Atmosphere, 20, 247-260.
32. Moon, I.-J. , Kim, M. , Joh, M. , Ahn, J. , Shim, J.-S. , and Jung, J. ,2016, Recent record-breaking high ocean waves induced by typhoons in the seas adjacent to Korea. Journal of Coastal Research, 1397-1401.
33. NIFS,2020, NIFS Serial Oceanographic observation. http://www.nifs.go.kr/kodc/eng/eng_soo_list.kodc (August 1st 2020).
34. Oh, H.-M. , Ha, K.-J. , Shim, J.S. , Hyun, Y.-K. , and Yun, K.-S. ,2007, Seasonal characteristics of turbulent fluxes observed at leodo Ocean Research Station. Atmosphere, 17, 421-433.
35. Oh, H.-M. , Ha, K.-J. , Heo, K.-Y. , Kim, K.-E. , Park, S.-J. , Shim, J.-S. , and Mahrt, L. ,2010, On drag coefficient parameterization with post processed direct fluxes measurements over the ocean. Asia-Pacific Journal of Atmospheric Sciences, 46, 513-523.
36. Oh, H. , Ha, K.-J. , and Shim, J.-S. ,2014, Analysis for onset of Changma using Ieodo ocean research station data. Atmosphere, 24, 189-196.
37. Oh, K.-H. , Park, Y.-G. , Lim, D.-I. , Jung, H.-S. , and Shim, J.-S. ,2006, Characteristics of temperature and salinity observed at the Ieodo ocean research station. Journal of the Korean Society for Marine Environment & Energy, 9, 225-234.
38. Park, K. , Park, J.-E. , Choi, B.-J. , Byun, D.-S. , and Lee, E.-I. ,2013, An oceanic current map of the East Sea for science textbooks based on scientific knowledge acquired from oceanic measurements. The Sea, 18, 234-265.
39. Park, S.-H. , Jang, S.-M. , Lee, D.-I. , Jung, W.-S. , Jeong, J.-H. , Jung, S.-A. , Jung, C.H. , Kim, K. , and Kim, K.-E. ,2012, The variation of aerosol number concentrations in relation with 3D wind components in the Ieodo Ocean Research Station. Atmosphere, 22, 97-107.
40. Pinardi, N. , Bonaduce, A. , Navarra, A. , Dobricic, S. , and Oddo, P. ,2014, The mean sea level equation and its application to the Mediterranean Sea. Journal of climate, 27, 442-447.
41. Schenatto, K. , de Souza, E.G. , Bazzi, C.L. , Gavioli, A. , Betzek, N.M. , and Beneduzzi, H.M. ,2017, Normalization of data for delineating management zones. Computers and Electronics in Agriculture, 143, 238-248.
42. Shim, J.-S. , Chun, I.-S. , and Min, I.-K. ,2004: Construction of Ieodo Ocean Research Station and its operation. The Fourteenth International Offshore and Polar Engineering Conference, International Society of Offshore and Polar Engineers.
43. Shin, B.-C. , Lee, M.-H. , Lee, J.-H. , and Shim, J.-S. ,2007, Seasonal and Diurnal Variations of Surface Ozone at Ieodo in the East China Sea. Journal of Korean Society for Atmospheric Environment, 23, 631-639.
44. TEOS-10,2020, The Gibbs SeaWater (GSW) Oceanographic Toolbox of TEOS-10. http://www.teos-10.org/pubs/gsw/html/gsw_contents.html (August 1st 2020).
45. Woo, H.-J. , Park, K.-A. , Byun, D.-S. , Lee, J. , and Lee, E. ,2018, Characteristics of the Differences between Significant Wave Height at Ieodo Ocean Research Station and Satellite Altimeter-measured Data over a Decade (2004~2016). The Sea, 23, 1-19.
46. Woo, H.-J. , Park, K. , Choi, D.-Y. , Byun, D.-S. , Jeong, K.-Y. , and Lee, E.-I. ,2019, Comparison of Multi-Satellite Sea Surface Temperatures and In-situ Temperatures from Ieodo Ocean Research Station. Journal of the Korean earth science society, 40, 613-623.
47. Yang, Y. , Park, S. , Kwon, S. , and Park, J. ,2016: A Development of Ocean Surface Wave and Wind Measurement From X-Band Marine Radar. Offshore Technology Conference Asia, Offshore Technology Conference.
48. Yeo, D.-E. and Nam, S. ,2020, Seasonal and spatial variations of air-sea heat exchange in the seas around the Korean Peninsula: Based on the observations and reanalysis products from 2011 to 2016. Progress in Oceanography, 181, 102239.
49. Yun, J. , Oh, H. , and Ha, K.-J. ,2015, Observation and Analysis of Turbulent Fluxes Observed at Ieodo Ocean Research Station in Autumn 2014. Atmosphere, 25, 707-718.