Eddy-and wind-sustained moderate primary productivity in the temperate East Sea ( Sea of Japan )

Introduction Conclusions References


Introduction
Oceanic circulation is characterized not only by large-scale currents, but also by mesoscale eddies.Now, with the advent of ocean remote sensing, we can observe the mesocale processes that dominate ocean energetics throughout most of the ocean (reviewed by Yoder et al., 2010).The infrared, microwave and visible imagery from various satellites have been used to estimate sea surface temperature (SST), chlorophyll a concentration, and sea surface height (SSH).Moored buoys or towers equipped sea can measure detailed temporal variations at fixed locations for longer time periods, which would be difficult to observe by ship or satellite (Dickey, 1991;Dickey et al., 1991Dickey et al., , 1993)).These time-series oceanographic stations offer insight into short-and long-term variations in meteorological, hydrographic, and biogeochemical variables at selected sites in the ocean.In recent years, there has also been increased use of in vivo fluorescence measurements to study phytoplankton.The East Sea/Japan Sea is a marginal sea located in a temperate zone of the Northwest Pacific Ocean.Warm, saline water enters from the Kuroshio and flows through the southern part of the sea.From the high temporal (8 times a day) and spatial resolution (500 m×500 m) Geostationary Ocean Color Imager (GOCI) launched by the Korea Institute of Ocean Science and Technology in 2010, numerous (over 60) eddies were identified from the chlorophyll a distribution in the period of 10 March to 17 May 2011 in the East/Japan Sea, with larger eddies occurring in the south and smaller ones in the north.The mean radii of the eddies varied from 10 km to 75 km, and the chlorophyll a concentration was higher by as much as three times at the periphery or outside than inside the eddies (Park et al., 2012).
One of the mechanisms responsible for nutrient input into the euphotic zone is uplifting of the nutricline by cyclonic eddies (termed as eddy pumping).Falkowski et al. (1991) reported that eddy pumping injects nutrients into the photic zone and results in enhanced primary production inside a cyclonic eddy.McGillicuddy et al. (1999) also found that a number of physical processes caused by eddies could lead to a region having a significantly higher input of nutrients to its euphotic zone than in surrounding waters.Mesoscale eddies related to primary production have been intensively investigated in the Sargasso Sea (McGillicuddy, 2007;McNeil et al., 1999).Siegel et al. (1999) estimated that eddy-induced transport accounts for nearly 50 % of the new nutrients reaching the euphotic zone at the Bermuda Atlantic Time Series Study site.McNeil et al. (1999) reported changes in chemical, bio-optical, and physical characteristics of the upper ocean due to the mesoscale eddy passing across the Bermuda Test Bed Mooring site.They used temperature and current measurements along with biogeo-Introduction

Conclusions References
Tables Figures

Back Close
Full chemical measurements at the moored buoy to show the biological response of the upper ocean associated with the introduction of nitrate into the euphotic layer by the uplifting of the isotherm.However, the mean distribution of chlorophyll a associated with eddies estimated from thousands of cyclonic and anticyclonic eddies in the Sargasso Sea showed that higher chlorophyll a values were detected not in the center of cyclonic eddies but at the west side of cyclonic eddies and to the east side of anticyclonic eddies (Yoder et al., 2010).Yoder et al. also showed that eddies can strongly upwell nutrient-rich subsurface waters, thus stimulating phytoplankton productivity.Using a three-dimensional primitive-equation numerical model, Lee and Niiler (1998) showed that the vertical movement of water by wind-eddy interactions depends on the relative angle between the wind and the current.In a mesoscale eddy, the maximum upwelling occurs in the area where the wind direction is the same as the current direction and the mean secondary circulation generated by the wind-eddy interaction has a magnitude comparable to that of the Ekman current.However, their model has not been tested in field observations.Here, we confirm the validity of Lee and Niiler's model using empirical measurements in the East/Japan Sea using sea-level anomalies and a moored bottom-tethered buoy equipped with wind and chlorophyll a sensors.

Materials and methods
The buoy system was deployed at a water depth of 139 m at 3.2 km east of the coast of the Dok Islets, Korea (35 • 20 11.5 N, 131 • 54 29 E).Sensors attached to the buoy system are listed in the Table 1.Real-time in vivo fluorescence was measured using an ECO fluorometer at the surface and was converted to chlorophyll a concentrations using an embedded algorithm.Wind speed and direction were measured at the surface using RM Young 05106 equipment.An OTRONIX Tiny Wave was used to measure wave height and period at the surface.Conductivity-temperature-depth (CTD) sensors were mounted at 0, 20, 40, 60, 80, 100, and 120 m depths.The observed data were transmitted to the Uljin shore laboratory every 10 min through a CDMA-based Introduction

Conclusions References
Tables Figures

Back Close
Full Sea-level anomalies were calculated following Lee and Niiler (2010a).Weekly gridded (1/3-degree resolution) Merged Sea Level Anomaly (MSLA) fields from all available altimeter satellites (TOPEX/Poseidon, GFO, ERS-1, ERS-2, Jason-1, Jason-2, and Envisat) were obtained from Archiving, Validation, and Interpretation of Satellite Oceanographic data (AVISO) from October 1992 to May 2012.Pascual et al. (2006) have described the processing procedures for the global MSLA datasets that were used in this study.The error in the MSLA in the Ulleung Basin (UB) due to tide and aliasing of high-frequency fluctuations was estimated to be about 2-3 cm (Choi et al., 2004) and the Root Mean Square (RMS) value of the difference between the sea-level anomaly observed at Ulleung Island and MSLA after removing seasonal variations is 2.7 cm with a correlation coefficient of 0.93 (Choi et al., 2004).The surface temperature was > 23

Daily
• C from 8 August to 30 September and decreased thereafter.Seasonal temperature variations were not apparent at or below depths of 40 m.The time-series evolution of the pycnocline structure can also be seen in Fig. 1 as density is largely determined by temperature in the East/Japan Sea (Lim et al., 2012).Above the pycnocline, centered at 80 m, a mixed layer developed with an average depth of 60 m.Temperatures measured at 40, 60, 80, 100, and 120 m exhibited distinct semidiurnal variations (data not shown).

Salinity
Salinity varied along with water temperature at 40, 60, and 80 m depths, with low salinity always related to low temperature (Fig. 2).From 11 to 30 April, salinity decreased gradually from 34.436 psu to 34.211 psu.Salinity increased thereafter until it reached 34.382 psu on 16 May.A similar salinity decrease was also observed between 16 and 31 August 2009.Seasonal salinity variations were not apparent at or below depths of 40 m, although they were somewhat lower in summer than in winter.

Chlorophyll a
The daily mean chlorophyll a concentration varied seasonally with two maxima in spring (March-April) and fall (October-December) and with two minima in summer and winter (Fig. 3), as observed previously by satellite measurements (Yamada et al., 2004) and in other temperate seas (Parsons et al., 1984).The instantaneous chlorophyll a concentration varied from less than 0.01 to 8.9 mg m −3 (data not shown).Introduction

Conclusions References
Tables Figures

Back Close
Full

Diurnal variation of chlorophyll a
A persistent diel rhythm in the chlorophyll a concentration was measured using in vivo fluorescence, with daytime values about 2-6-fold lower than nighttime values in April 2009 (Fig. 4), and this diel feature was apparent year-round at the DETS buoy.A diel fluorescence cycle arises mainly from inhibition by light because fluorescence begins to decrease at dawn and increases again after dark.However, the subsurface chlorophyll a maximum was present at 35-45 m all day long at the buoy site (Fig. 5).

Daily mean wind speed and daily mean wave height at DETS buoy site
The daily mean sea surface wind speed was highly variable throughout the year with relatively strong winds in winter and weaker winds in summer.It varied from 1.4 to 14.3 m s −1 with a mean of 3.9 m s −1 (Fig. 6).
Daily mean wave height was also low in summer and high in winter with large variations.It varied from 0.2 to 4.0 m with a mean of 1.1 m (Fig. 7).Introduction

Conclusions References
Tables Figures

Back Close
Full The East/Japan Sea is moderately productive, with carbon fixation of as much as 161-240 g C m −2 yr −1 based on in situ carbon-uptake measurements and satellite measurements from the Coastal Zone Color Scanner (CZCS) and Sea-viewing Wide Field-ofview Sensor (SeaWiFS) (Yoo and Park, 2009).Therefore, the sea as a whole draws down 0.16 to 0.22 Gt C yr −1 through photosynthesis.Even if community respiration is more than half of the gross production (Waku and Furuya, 1998), this amount of gross primary productivity is an order of magnitude greater than the net annual CO 2 flux at the air-sea interface (0.032 ± 0.012 Gt C) based on the fugacity of CO 2 in the surface water and air, and based on the pH and alkalinity of the surface water (Oh et al., 1999).The settling flux of organic carbon in the southwestern part of the Japan Basin at a depth of 2.8 km was 2.9 g C m −2 yr −1 , which falls in the upper range of the world ocean (Hong et al., 1977).This preliminary comparison indicates that the biological pathway for the drawing down of atmospheric CO 2 in the East/Japan Sea is very significant.Moreover, organic matter produced through phytoplankton photosynthesis sinks to the bottom of the sea to enhance the accumulation of CO 2 in the deep sea.One of the current estimates of the accumulation of anthropogenic CO 2 in the East/Japan Sea (80 mol C m −2 as of 1999) is much greater than that in the North Pacific and comparable to that of the North Atlantic (Park et al., 2006).Consequently, a proper understanding of the environmental factors controlling phytoplankton primary productivity is of great importance in terms of quantifying the carbon dioxide budget in the East/Japan Sea.
In the sea, mesoscale eddies occur where major currents and oceanic fronts are present.The mesoscale eddies separate from the respective current or front and meander into water masses with different physical, chemical, and biological characteristics.The East/Japan Sea is one of the most eddy-rich fields in the world ocean (Ichiye and Takano, 1988;Isoda, 1996).The warm and cold eddies appear periodically in the southern part of the UB with lifetimes of about 1 month or longer (Morimoto et al., Introduction Conclusions References Tables Figures

Back Close
Full ).However, using lag correlation analysis, Morimoto et al. (2000) determined an eastward propagation of about 1.8 cm s −1 , suggesting eastward advection by the Offshore Branch of the Tsushima Warm Current, a branch of the Kuroshio Current.There have been many descriptions of meandering currents in the UB (Naganuma, 1985;Katoh, 1994;Morimoto and Yanagi, 2001), and they include both warm-and cold-water intrusions (Ichiye and Takano, 1988) and warm and cold eddies (Ichiye and Takano, 1988;An et al., 1994;Isoda, 1996;Lee and Niiler, 2010b).
Five eddy circulation patterns have been identified in the UB based on acoustic travel time from an array of inverted echo sounders (Mitchell et al., 2005).An anticyclonic eddy called the Ulleung Warm Eddy (UWE) is a major hydrographic feature of the UB.
One very significant cold cyclonic eddy was located south of the Dok Islets throughout much of the year and was named the Dok Cold Eddy (DCE) by Mitchell et al. (2005).Previously, there were not enough systematic hydrographic observations in the eastern UB to detect the DCE, as the weak SST signal of the cold eddy precluded its detection based on an analysis of water mass properties.The DCE is typically ∼ 60 km in diameter and originates from the pinching off of a Subarctic Front meander between Ulleung Island and Dok Islets (Mitchell et al., 2005).After its formation, the DCE remains southwest of the Dok Islets for 1-6 months before propagating westward toward Korea, where it deflects the path of the East Korean Warm Current (EKWC).
It has been observed that eddies displace isopycnals, which are uplifted at the center of cyclonic eddies.Up-sloping of the isopycnals due to eddy pumping is known to enrich oligotrophic surface layers (Falkowski et al., 1991;Olaisala et al., 1993;McGillicuddy et al., 1998;Seki et al., 2001).Phytoplankton biomass is determined by the availability of light and nutrients.These growth-limiting factors are in turn regulated by the physical processes of ocean circulation, mixed-layer dynamics, atmospheric-dust deposition, and the solar cycle.The East/Japan Sea experiences bimodal phytoplankton blooms in spring and fall, as do other temperate zones in the global oceans (Nagata, 1994;Ishizaka et al., 1997;Nagata and Ogawa, 1997;Kim et al., 2000;Yamada et al., 2004Yamada et al., , 2005)).Satellite-derived

Conclusions References
Tables Figures

Back Close
Full ocean color imagery has allowed large-spatial-scale observations of the seasonal development of phytoplankton biomass.Kim et al. (2000) demonstrated that phytoplankton biomass observed in CZCS images exhibited a bimodal seasonal pattern of spring and fall blooms in the East/Japan Sea.They attributed the seasonal patterns in phytoplankton populations to seasonal density stratification.In the East/Japan Sea, the depth of the seasonal thermocline increases to 100-150 m in winter and decreases to 10-20 m in summer.During the summer, the water is stratified and the upper layer is nutrient limited.The surface chlorophyll a concentration falls below 0.3 mg m −3 at this time, which is the lowest value of the year.In winter, deep mixing over the permanent thermocline replenishes nutrients in the upper mixed layer, while preventing phytoplankton from absorbing enough light for growth.Stratification in spring briefly satisfies the light-nutrient requirements for phytoplankton growth, inducing a temporary increase in phytoplankton abundance in the upper layer.In summer, stratification inhibits nutrient input from below, suppressing phytoplankton growth in the surface mixed layer.A second seasonal bloom occurs in the fall as weakening of the stratification allows nutrient input into the upper mixed layer.Phytoplankton blooms also occur periodically when physical phenomena of various scales raise the concentrations of growth-limiting nutrients in the euphotic layer of the East/Japan Sea waters between late spring (after spring bloom) and early fall (before fall bloom).Examples of such events include mixing and decay of stratification due to passage of warm and cold currents and the formation of mesoscale eddy features (Falkowski et al., 1991).These phenomena are likely to be significant transport mechanisms of large quantities of nutrients to the euphotic zone, resulting in the enhanced phytoplankton populations seen in satellite ocean color observations in the East/Japan Sea (Ahn et al., 2005).Other than a few preliminary investigations (e.g., Kim et al., 2012), the relationship between these physical mechanisms and phytoplankton blooms in the region is not been yet clearly understood.Introduction

Conclusions References
Tables Figures

Back Close
Full

Cross-correlation analysis
Interpretation of time-series observations at fixed locations in the ocean is often complicated by spatial heterogeneity, which varies over time.Thus, it is important to differentiate spatial heterogeneity from temporal variation (McGillicuddy et al., 1999).In order to understand the variability of the surface chlorophyll a concentration in terms of the physical environment, a cross-correlation analysis was performed.As discussed above, relatively high chlorophyll a concentrations were observed in spring and fall.
Apart from this bimodal feature of phytoplankton blooms, small-scale variations in chlorophyll a concentrations were observed during the year.Increases in chlorophyll a concentrations were often related to decreases in subsurface temperatures with a short lag time of a few days.
In the UB, subsurface chlorophyll a concentration maxima usually occurred at around 30-40 m owing to the diffusion of nutrients across the seasonal thermocline (Hong et al., 1997).However, this type of nutrient diffusion cannot support phytoplankton growth in surface waters.In most cases, small amounts of nutrients that diffuse through the strong thermocline are readily absorbed by phytoplankton living in subsurface layers, thus producing subsurface chlorophyll maxima (Fig. 5).An active mechanism is needed to transport nutrients upward to enhance phytoplankton growth in surface waters.Synchronous water temperature oscillations have been observed at 40-80 m, which seem to be coupled with the barotropic tide or internal waves (Park and Watts, 2006) as they travel relatively rapidly (Sangrà et al., 2005).Enhancement of biological productivity in shelf-break regions has also been attributed to internal tidal mixing, which induces upward transport of nutrient-rich cold water (Pingree et al., 1986;Mann and Lazier, 1991).
We postulated that mesoscale physical processes occurring in subsurface layers increases nutrients and influences chlorophyll a concentrations in the surface waters near the Dok Islets.In this context, we statistically analyzed the relationships between surface chlorophyll a concentrations and the subsurface temperature.The period for Introduction

Conclusions References
Tables Figures

Back Close
Full the cross-correlation analysis was chosen to be from 1 May to 30 June when significant correlations were found between the chlorophyll a concentration and subsurface temperature.Daily mean values were exclusively used in this cross-correlation analysis.This highly correlated period occurred when the DETS buoy was located near the boundary of the DCE after the spring bloom (Fig. 8).When chlorophyll a concentration at the surface and the temperature at a depth of 40 m were cross-correlated, the highest correlation coefficient was found for a 6 day lag (Fig. 9a; r = −0.47,significant at the 95 % confidence level).This indicated that the highest chlorophyll a concentration typically occurred 6 days after a decrease in temperature at 40 m, which was an indication of upwelling, and thus, of transport of nutrients from the deep layer to the surface layer.
In October, the cross-correlation values for the temperature at a depth of 40 m versus the chlorophyll concentration reached their most significant (negative) value for a maximum lag time of 5 days (r = −0.418),implying that 5 days were required for the chlorophyll a values to change following a decrease in temperature at 40 m.
The mean surface chlorophyll a concentration was not correlated with the subsurface temperature from July to September.During the summer, the phytoplankton biomass in the upper layer is suppressed by low nutrient concentrations (typically < 1 µM of nitrate) caused by the stratification of the water column (Rho et al., 2010;Kim et al., 2010).The nitrate concentration increased sharply at depths below 30 m and was > 10 µM below 50 m (Fig. 5).Wind-induced upwelling of subsurface waters with surface water is considered to be one of the primary driving forces of the phytoplankton growth in late spring and early fall.The daily mean surface chlorophyll a concentration was not significantly correlated with wind speed or wave height.

Eddy and wind interactions
The DCE was observed near to the DETS mooring site year-round in 2009 (Fig. 8) using the eddy detection method of Lee and Niiler (2010b) and satellite measurements of the sea level anomaly.The five-day mean wind (red arrows) measured from Introduction

Conclusions References
Tables Figures

Back Close
Full DETS and the geostrophic current (blue arrows) from the sea-level anomaly (Fig. 10a) clearly show two unusually high phytoplankton blooms (shaded periods) related to the upwelling events (Fig. 10b).The maximum chlorophyll a concentration was found in the area near the eddy rim, similar to the chlorophyll a concentration distribution reported by Yoder et al. (2010), and the mooring site was located in the eddy rim area (Fig. 8a and b) where the wind and current had the same direction over several weeks (Fig. 10a).These observations are the first direct observations that the upwelling induced by wind-eddy interactions (Lee and Niiler, 1998) is related to nutrient input to the euphotic zone in the cyclonic eddy.During the summer months, a strong and shallow thermocline with weak wind may hinder nutrient supply through upwelling driven by the wind-eddy interaction.These variations in subsurface temperature, particularly at a depth of 20 m (Fig. 10b), appear to be indicative of deep-water upwelling or turbulent mixing due to eddy pumping in May, June, and October 2009.Episodic nutrient inputs into the euphotic zone associated with mesoscale eddies have previously been proposed to explain variations in chlorophyll a concentrations at the sea surface.We conclude that the nutrient supply from upsloping of the isopycnals due to eddy pumping may result in mesoscale changes in phytoplankton in the UB of the southwestern East/Japan Sea.We further suggest that the nutrients supplied through eddy pumping during seasonal bloom periods enhance phytoplankton bloom further (Fig. 10b).MODIS chlorophyll a images showed that cyclonic eddies often occurred and appeared to be responsible for scattered patches of phytoplankton in the East/Japan Sea (Lee and Niiler, 2010a).

Conclusions
Significant energy is introduced into the East/Japan Sea by inflow conditions through the Korea Strait, in particular to the Ulleung Basin in the southwestern part of the sea, and the intrusion of the Tsushima Warm Current into the sea and the bottom geomorphology of the sea favor eddy formation.Moreover, the sea is subject to winds that Introduction

Conclusions References
Tables Figures

Back Close
Full vary over 3 to 4 days, as it is located to the eastern side of the Eurasian Continent (e.g., Hong et al., 2012).Therefore, the interaction between eddies and the wind plays a critical role in sustaining the high primary productivity in the sea during summer when the surface water becomes otherwise depleted of nutrients.A bottom-tethered surface buoy moored in the middle of the East/Japan Sea to observe metrological and oceanic variables provided a unique opportunity to test the hypothesis of Lee et al. (1994) and Lee and Niiler (1998).When the wind blows in the direction of the current (against the current), a narrow upwelling (downwelling) area and broad downwelling (upwelling) area are formed.This secondary circulation pattern extends well below the mixed layer.
When the wind blows perpendicular to the jet, the secondary circulation does not extend below the mixed layer.Therefore, the unusually high primary productivity in the East/Japan Sea, particularly during the late spring -after the spring bloom -as observed in this study, appears to be sustained by the interaction between eddies and the wind, as well as by other factors such as coastal upwelling and terrestrial nutrient input during the rainy season.

BGD Introduction
Full with optical and electrical sensors (time-series stations) attached to mooring arrays at Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | communication network.The instrumentation on the DETS Buoy included a Sea-Bird temperature sensor, a beam transmissometer, which measures the beam attenuation coefficient (beam c) at 660 nm, and a stimulated fluorometer.Buoy data acquired from 30 March 2009 to 7 February 2010 were used in this study.In the data analysis, temperature at 40 m depth was cross-correlated with chlorophyll a concentrations to determine the relationship between the physical variables and phytoplankton response.Statistical analyses were conducted using the Statistical Package for the Social Sciences (SPSS 13.0 for Windows).
mean temperature, salinity, chlorophyll a concentrations, wind speed, and wave height obtained from the DETS Buoy for the period 30 March 2009 to 7 February 2010 are shown in Figs.1-3 and 5-6, respectively.Chlorophyll a concentrations and hydrographic variables were averaged on a daily basis because the magnitudes of diurnal variations in fluorescence and semi-diurnal variations in temperature at depth were comparable to those of seasonal variations.The surface water temperature increased gradually from 10.6 • C to 14.4 • C during 1-11 April 2009, and then decreased until it reached 12.0• C on 28 April (Fig.1).From the 11 April to 17 May 2009, relatively low temperatures were measured at all depths.
Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Aside from seasonal variability, small-scale variations in chlorophyll a concentrations were seen in the DETS buoy data.During 10-13 May 2009, chlorophyll a concentrations in the surface waters were > 1.5 mg m −3 .Prior to this period, the temperature gradually decreased from 28 April to 6 May at depths of 40, 60, and 80 m.Similarly, a decrease in temperature at depths of 40, 60, and 80 m during 27-31 May preceded relatively high chlorophyll a concentrations observed on 3-7 June.The temperature at 40 m was < 16 • C during 13-15 October.After this temperature decrease at 40 m, a small peak in the chlorophyll a concentration (> 0.5 mg m −3 ) was observed during 17-20 October 2009.
Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | 2000 Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Fig. 10 .
Fig. 10.(a) Wind (red arrows) measured by DETS and the geostrophic current (blue arrows) from the sea-level anomaly at the mooring site.(b) Chlorophyll a concentration (blue line) and temperature (red line) at a 20 m depth.Gray shaded areas represent blooms resulting from the wind-eddy interaction induced upwelling.