- Email: [email protected]
Coastal circulation in the southwestern Yellow Sea in the summers of 2008 and 2009
Author’s Accepted Manuscript Coastal Circulation in the Southwestern Yellow Sea in the Summers of 2008 and 2009 Dongliang Yuan, Yao Li, Bin Wang, Lei He, Naoki Hirose www.elsevier.com/locate/csr
PII: DOI: Reference:
S0278-4343(17)30070-5 http://dx.doi.org/10.1016/j.csr.2017.01.022 CSR3553
To appear in: Continental Shelf Research Received date: 29 August 2015 Revised date: 4 November 2016 Accepted date: 10 January 2017 Cite this article as: Dongliang Yuan, Yao Li, Bin Wang, Lei He and Naoki Hirose, Coastal Circulation in the Southwestern Yellow Sea in the Summers of 2008 and 2009, Continental Shelf Research, http://dx.doi.org/10.1016/j.csr.2017.01.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Coastal Circulation in the Southwestern Yellow Sea in the Summers of 2008 and 2009
Dongliang Yuan1,2*, Yao Li1,2, Bin Wang3,4, Lei He5, and Naoki Hirose4 1
Key Laboratory of Ocean Circulation and Wave, Institute of Oceanology, Chinese
Academy of Sciences, Qingdao, China 2
Qingdao Collaborative Innovation Center of Marine Science and Technology,
Qingdao, China 3
College of oceanography, Hohai University, Nanjing, China
4
Research Institute for Applied Mechanics, Kyushu University, Kasuga, Japan
5
School of Marine Sciences, Sun Yat-Sen University, Guangzhou, China
*
Corresponding author at: Institute of Oceanology, Chinese Academy of Sciences, 7
Nanhai Road, Qingdao, China 266071. Email: [email protected]
Abstract The coastal ocean circulation off the Chinese Subei (north Jiangsu Province) coasts in the southwestern Yellow Sea in summer is studied using hydrographic data from four research cruises, moored current meter data in the Lunan (south Shandong Province) trough, and Lagrangian trajectories of satellite-tracked ARGOS surface drifters in the summers of 2008 and 2009.
The hydrographic data indicate a fresh
water plume extending offshore in the northeastward direction from the Sheyang River mouth and a belt of relatively cold surface temperature from the surrounding areas located offshore of the Subei coasts to the south of the plume in summer.
The
drifter trajectories show that the Subei coastal current, which is traditionally thought to flow southward year round, flows northward in June through July under the forcing of the southerly monsoon.
A northward intrusion along the submarine valley off the
1
Changjiang mouth is indicated by the movement of the drifters, which is persistently strong and independent of the winds.
Associated with the northward flow, coastal
upwelling is generated offshore of the Subei coast to the south of the Sheyang River plume, below which upwelling is generated but is shut down by the buoyancy of the fresh water at the surface.
Coastal upwelling is also observed off the coast of
Qingdao city in middle July through early August of 2008, which is forced by the along-shore component of the southerly and southwesterly winds.
The Qingdao
upwelling in summer is reported for the first time in history.
Keywords Subei coastal current, Lunan coastal current, Southwestern Yellow Sea, Sheyang River plume, Coastal upwelling, Buoyancy shut-down
1. Introduction The southwestern Yellow Sea is the part of the Yellow Sea south of the Shandong Peninsula and north of the Changjiang (aka the Yangtze River) mouth between Mainland China and the center of the Yellow Sea trough (Figure 1). The topography of this area features a very shallow bank called the Subei Bank and a submarine trough called the Lunan (south Shandong) trough extending from the center of the Yellow Sea trough westward into the Haizhou Bay between the Shandong Peninsula and the Subei Bank.
Offshore of the Changjiang mouth in the southern Yellow Sea,
a large area of shallow water called the Changjiang Bank forms an obstacle between the currents in the Yellow Sea center and in the western East China Sea.
South of
the Changjiang mouth off the Hangzhou Bay estuary, a submarine valley is located between the nearshore shallow estuary and the offshore Changjiang Bank, forming an
2
important conduit of deep-sea water intrusion onto the shallow shelf off the coast of the Chinese Jiangsu Province. The topographic features suggest that the circulation in the southwestern Yellow Sea is subject to complicated topographic steering. Historical understanding of the circulation in this area has been vague, except that the currents in this area are speculated to flow southward year round (Zhao et al., 1991; Guan, 1994; Hu, 1994; Su, 1998).
In summer, the southward currents are argued to be associated with the
strong temperature and salinity fronts along roughly the 40-50 m isobaths in the western Yellow Sea (Hu, 1994).
Recently, Yuan et al. (2008) have suggested that the
Subei coastal current might flow to the north in summer in response to the forcing of the southwesterly monsoon.
However, due to lack of in situ measurements of the
currents, the path and the dynamics of the Subei coastal current in summer remain unclear. Lately, the circulation in the southwestern Yellow Sea has received international attention due to repeated outbursts of large-scale Ulva Prolifera blooms off the coasts of the south Shandong Peninsula in summer since 2006. The green tide events have generated significant impact on regional environment and economy, the largest event of which took place in the summer of 2008 at the dawn of the 29th summer Olympic Game in the Olympic sailing competition site offshore of the city of Qingdao (Hu and He, 2008; Qiao et al., 2008).
Significant green tide events of similar scope and
amplitude recurred in the summers of 2009 and the years to follow.
It is, therefore,
of great interests to study the ocean circulation associated with these events. In the past, the study of the vertical water motion in the Yellow Sea in summer has focused mostly on the circulation associated with the Yellow Sea Bottom Cold Water. Cold surface temperature patches have been identified in the southwestern
3
Yellow Sea in summer and have been attributed to frontal upwelling induced by tidal mixing and rectification (Lü et al., 2010). Historical direct observations of the cold SST off the Qingdao coast in summer were rare (see the review of the regional observational and modeling studies in Lü et al. 2010).
No previous studies,
including the Lü et al. (2010) study, have conducted focused investigation of the upwelling off the Qingdao coast.
Because the Subei coastal current was thought to
flow southward all year round, the upwelling associated with the wind-driven circulation has not been studied before. In this paper, we use temperature and salinity measurements collected during four research cruises in the summers of 2008 and 2009 to study the ocean circulation in the southwestern Yellow Sea in summer.
These hydrographic data are used in
combination with trajectories of satellite-tracked surface drifters to study the path, direction, and dynamics of the Subei coastal current in summer.
A numerical model
is then used to simulate the shelf circulation in the summer of 2008, based on which the dynamics of the coastal upwelling currents are studied. The data used in this study are described in the next section. The temperature and salinity distributions during the 2008 and 2009 research cruises are presented and discussed in Sections 3 and 4, respectively. The structure and dynamics of the Subei coastal current in the southwestern Yellow Sea are discussed using drifter trajectories in Section 5.
In Section 6, a numerical simulation of the shelf circulation in the
southwestern Yellow Sea in summer is carried out, based on which the dynamics of the coastal upwelling are investigated. Section 7 summarizes the conclusions of this study.
4
2. Data The data used in this study include the hydrography data, moored current meter data, satellite-tracked surface drifter trajectories, and a numerical simulation using the Princeton Ocean Model. 2.1 Hydrography and moored current meter data The data used in this study are collected during three cruises in the summer of 2008 and during one cruise in the summer of 2009 in the southwestern Yellow Sea. The major surveys of the regional circulation took place during July 22-August 3, 2008 and May 28-June 21, 2009 in the western Yellow Sea from the Changjiang mouth to the northern Yellow Sea, with a spatial resolution of about 0.5 degree in the zonal direction and 2 degrees in the meridional direction (Figure 2).
We will focus
on the shelf circulation in the western Yellow Sea between the Changjiang mouth and the Shandong Peninsula, which is called the southwestern Yellow Sea in this study. Besides the two major surveys, two complementary cruises are carried out on July 9-13 and August 9-11 of 2008, respectively, to sample smaller areas, between 34ºN and 36ºN in the Lunan trough, with higher spatial resolutions, where a large amount of the floating algal patches are located and tracked (Figures 2b and 2c).
The
temperature and salinity data were collected using Conductivity-Temperature-Depth (CTD) instruments of the Seabird Electronics, Inc. of the USA. During the cruise of July 22-August 3, 2008, a bottom-mounted mooring is deployed at station B (34°59.77'N, 120°30.52'E) with a water depth of about 30 m (Figure 1).
An Acoustic Doppler Current Profile (ADCP) current meter with a 250
kHz band Sontek ADP sensor capable of measuring full-column current profiles in the shallow water was tethered at about 3 m above the bottom of the ocean to return
5
continuous current measurements during July 22-28, 2008. The ADCP profiles of east-west and north-south components of the current at station B were recorded at every 10 minutes.
The sub-tidal currents are obtained by removing high-frequency
signals using a 5th order low-pass Butterworth filter with a cut-off period at 40 hours. Due to intensive fishing activities in the area, long-term mooring measurements of the shelf currents are rare and precious. The moored current measurements thus provide us with important observations to validate and calibrate the model simulation, based on which the regional circulation and dynamics of the currents are studied. 2.2 Ocean model The Princeton Ocean Model (POM) with a nested grid is applied to study the structure and dynamics of the ocean circulation in the southwestern Yellow Sea. The model domain includes a coarse-grid region of (15°N-45°N, 105°E-135°E) with a horizontal resolution of 1/6° longitude by 1/6° latitude and a nested-grid smaller domain of the southwestern Yellow Sea (32°N-37°N, 119°E-124°E) with a horizontal resolution of 1/24° longitude by 1/24° latitude.
The topography of the large domain
is based on the 5′×5′ global ETOPO5 elevation data whereas the topography of the small domain is based on the local navigational chart.
Both the large and the small
domain models have 30 sigma levels in the vertical. We will focus on the wind-driven circulation in this study.
The effect of tidal
mixing is parameterized by the Munk–Anderson scheme as described in Lee et al. (2006).
The large-domain model is forced by the global 0.5°×0.5° model simulation
of Xia et al. (2004a, 2004b) at the open boundaries, and by monthly climatological wind stress and the Haney-type heat flux of the interpolated COADS (Comprehensive Ocean-Atmosphere Data Set) data (da Silva et al., 1994) at the sea surface.
The
surface salinity of the model is relaxed to the summer monthly climatology at a time
6
scale of 30 days.
The large-domain model has been integrated for 6 years from the
Levitus (1982) annual mean climatological temperature and salinity and geostrophic currents of zero vertically integrated transport.
The kinetic energy of the model is
found to reach a quasi-steady state in about 4 years of integration. The small-domain and the large-domain models are then spun up for 6 years. The surface wind stress and heat/salt flux conditions are the same as in the large-domain model.
The Sheyang River runoff in the small-domain model is set at
1000 m3 s-1. Due to lack of accurate measurements of the river discharge in this area, the specification of the river runoff in the small-domain model is based on testing and order-of-magnitude estimate.
Some of the latest studies based on numerical
modeling (e.g. Wu et al. 2014) suggest that the Changjiang fresh water is transported to the area by a nearshore tidal residual current, which is then transported offshore at the Sheyang River mouth. We do not distinguish between these two sources of the fresh water. Here, the model fresh water specification at the Sheyang River mouth represents the total freshwater to be transport offshore of the Subei coasts, which are supposed to be much larger than the Sheyang River discharge based on numerical experiments.
Finally, the NCEP-One 6-hourly re-analysis wind stress (Kalnay et al.,
1996) is used to force the model for three months from May 1 through July 31 of 2008.
The details of the model configuration and parameterization are described in
the master thesis of Wang (2010) and the Ph. D. dissertation of Li (2010). 2.3 ARGOS trajectory data In the summer of 2009, six satellite-tracked ARGOS floats were deployed in the southwestern Yellow Sea to measure the currents off the coast of Jiangsu Province. Two of the floats were released in the submarine valley area offshore of the Changjiang mouth to measure the movement of the waters in the East China Sea.
7
The remaining four floats were released in the 34ºN section to track the movement of the Subei coastal current.
All of the drifters have their drogues dragged at the 15 m
depth except the two released at about the 12 m isobath at 34ºN, whose drogue cables are set at 6 m.
The drogues of the ARGOS floats are nylon tubes of about 6.44 m
long and about 0.6 m in diameter. The diameter of the ARGOS float is 38 cm.
3. Hydrography in the summer of 2008 The surface and subsurface distributions of temperature and salinity are presented in this section, which form the basis of observational and numerical modeling studies of the dynamics of the coastal upwelling. 3.1 Temperature and salinity distributions The distributions of surface salinity and temperature in the southwestern Yellow Sea during July 9-13, 2008 show that the surface water off the coast of the Jiangsu Province is dominated by warm (> 25.5 ℃) and fresh (<30.5 psu) water from the rivers near 34ºN (the waters from multiple rivers together are called the Sheyang River plume in the following text) and by two patches of cold water: one off the coast of Subei to the south of the plume and one offshore of Qingdao, which indicate coastal upwelling (Figure 3).
The warm and extremely fresh (< 28 psu salinity)
Changjiang Diluted Water is only seen near the southern boundary.
The distribution
of the Changjiang Diluted Water south of 32ºN was not cover by the survey in the summer of 2008.
The Changjiang Dilute Water, as suggested by the salinity
concentration below 30 psu, is evidently confined south of the 33ºN latitude, and, therefore, is of limited influence to the circulation over the Subei Bank and in the Lunan trough in the north.
The cold sea surface temperature center off the coast of
the south Shandong Peninsula has not been reported before. We call this center the 8
Qingdao upwelling in the following text. At the 40 m depth, the distributions of salinity and temperature suggest that the western Yellow Sea trough is occupied by relatively warm (> 10℃) and saline (>33 psu) bottom water in the south (lower panels of Figure 3), with a salinity and temperature front at about 34.5ºN. This area is suggested to be the confluence of the residual Yellow Sea Warm Current from the south and the cold fresh water from north, which is seen every year in the 35ºN and 36ºN vertical sections in historical data and is evidenced by the southward movement of satellite-tracked drifters from the northern Yellow Sea into this area (Yuan et al., 2012). The temperature and salinity distributions in the vertical section of 32ºN show that the subsurface water at the southern boundary is warm (>22℃) and saline (>32 psu) compared to the waters in the 33ºN and 34ºN sections (Figure 4).
The warm
and fresh Changjiang Diluted Water is observed in a very thin layer of about 10 m offshore at the surface in this section. Further north at the 33ºN and 34ºN sections, the warm and saline water is absent, suggesting that it either exits the survey area through the eastern boundary, as evidenced by the salinity distribution at the 40 m depth, or mixes with the coastal fresh waters to lose its salinity identity.
The detailed
movement and evolution of this water mass is not resolved by our survey. A water mass of high salinity (> 33 psu) and low temperature (<14 ℃) is identified between the 35 m and 80 m isobaths in the 33ºN and 34ºN sections,.
The
temperature of this water mass is higher than the coldest water (< 10℃) of the Yellow Sea Bottom Cold Water at the central bottom of the Yellow Sea trough.
At the 35ºN
and 36ºN sections, the saline water are flanked by two even colder water masses: one at the western flank, believed to be connected with the cold-core water in the 34ºN 9
section, and one at the central bottom of the Yellow Sea trough, which is the traditional Yellow Sea Bottom Cold Water, respectively.
The cold-core water mass
over the western flank has been shown to be associated with the southward movement of the northern Yellow Sea water in summer (Yuan et al., 2012; Wang et al., 2014). 3.2 Upwelling fronts During the summer of 2008, northward drifting of large patches of the green algae was tracked by satellite observations (Hu and He, 2008).
Associated with the
northward coastal currents, onshore bottom Ekman transport and coastal upwelling are indicated by the temperature and salinity distributions in the vertical sections in Figure 4. The upwelling front is located above roughly the 20 m isobath in the 33ºN section, more than 50 km from the coast. Further north at 34ºN and 35ºN, the upwelling is shut down by the buoyancy of the warm and fresh water at the surface. Similar cases have been reported off the southeast coast of the U.S. in summer (Yuan, 2006). Further north at 36ºN, the nearshore surface temperature at Qingdao is also lower, evidently due to upwelling of bottom cold water. The evolution of the upwelling in the southwestern Yellow Sea in the summer of 2008 is further disclosed by the two supplementary surveys in the smaller area between 34ºN and 36ºN on July 9 and August 9, respectively.
During July 9-13,
2008, the surface temperature and salinity distributions off the Subei coast are similar to those during July 22-August 3 2008 survey, except that the cold center off Qingdao is absent (Figure 5).
The Subei coastal area is still dominated by the Sheyang River
plume, which extends northeastward into the central Yellow Sea.
The upwelling
associated with the northward coastal currents is evidenced by a belt of cold water between 120ºE and 120.5ºE at the 20 m depth produced by the onshore bottom Ekman transport (bottom panels of Figure 5 and Figure 6). 10
This upwelling is
evidently capped by the fresh water from the Sheyang River. By August 9-11, 2008, the surface salinity over the southwestern Yellow Sea is still characterized by the Sheyang River plume, although the scale and the pattern are different (Figure 7). The surface temperature distribution shows cold surface water entering the survey area from the south along an offshore path east of 122ºE, which is likely the upwelled water in the south.
The cold center off Qingdao is clearly
identified, indicating that the Qingdao upwelling has persisted through late July. The upwelling is evidently associated with the onshore bottom Ekman transport, as suggested by the temperature and salinity distribution in the vertical sections in Figure 8. Notice that the surface cold centers of the upwelling are located further nearshore instead of directly above the tidal fronts, suggesting these upwelling currents are associated with the winds instead of the tides.
4. Hydrography in the summer of 2009 The hydragraphy in the summer of 2009 are presented in this section to be compared with that in the summer of 2008. Drifter trajectories in the summer 2009 provide direct current measurements to supplement the hydrography surveys in the summers of the two years. 4.1 Temperature and salinity distributions The horizontal distributions of temperature and salinity at the surface and at the 30 m depth during May 28-June 21, 2009 are plotted in Figure 9, showing the dominance of fresh water west of 123ºE in the southwestern Yellow Sea.
Offshore
of the Changjiang mouth, a warm water tongue extending northward is in contract to the surface temperature distributions in the summer of 2008, which show cold offshore water due to upwelling at 33ºN.
This is because sections 32ºN and 33ºN
11
are not measured by the cruise.
During the summer 2009 survey, the sections south
and along of 35ºN were surveyed before May 31 and the sections north of 35ºN were measured after June 15 (Figure 9 upper panels).
The coastal upwelling is indicated
by the center of cold surface water in the Haizhou Bay and east of the tip of the Shandong Peninsula. In the subsurface, the northward intrusion of salty and warm water along the submarine canyon offshore of the Changjiang mouth is suggested by the salinity and temperature contours at the 30 m depth.
The salty water in the western Yellow Sea
trough and the southward spread of the northern Yellow Sea water around the tip of the Shandong Peninsula are also indicated by the subsurface temperature and salinity distributions. During the summer 2009, the vertical section of 29ºN is characterized by the extremely fresh water from the Changjiang in a thin surface layer (Figure 10 bottom panels).
A surface temperature inversion occurs offshore at about 123.5ºE, with a
warm-core but slightly high salinity at the subsurface. warm-core water is not clear.
The reason of the subsurface
However, it is unlikely that this water mass has impact
the circulation in the north significantly, because the temperature and salinity distributions in the 34ºN sections do not show the presence of this mass. 4.2 Direct current measurements using ARGOS drifter trajectories Due to lack of direct current measurements off the Subei coast in history, the direction and the path of the Subei coastal current have been in debate. To study this current, six satellite-tracked ARGOS surface drifters are released at 34ºN, 32ºN, and 31ºN. The drifter released at 31ºN in the submarine valley south of the Changjiang mouth is seen to move northward steadily and has repeated the path of the drifter
12
released at 32ºN precisely, suggesting the quasi-steady northward intrusion of the current in the area (Figure 11).
The southern drifter (Drifter 90) later got stuck over
the shallow shelf for nearly a month only to revive later in middle July to drift eastward into the central Yellow Sea. The northern drifter (Drifter 76) released at 32 ºN has been moving northward continuously and eventually enters the central Yellow Sea. Under very weak wind conditions, the four ARGOS drifters released at 34ºN initially made circular movement around their releasing positions within a range of about 10-20 km (~20 km for floats 77 and 78, ~10 km for floats 88 and 91) during the period between May 30 through June 3. The periodic movement is obviously forced by the tidal currents and has suggested weak tidal residual currents over the Subei Bank. Since June 4, after the southerly winds pick up, all four drifters started to move north- and northeastward, suggesting the dominance of the wind-driven currents over the tidal residual currents. The trajectories of the drifters clearly suggest that the Subei coastal current flows northward in summer, against the traditional understanding of the Yellow Sea Coast Current in summer. The northeastward drifting trajectories and the final landing positions at the coast of the southern Shandong Peninsula of the three drifters (Drifters 91,78, and 88) initially released at the 34ºN are in agreement with the drifting and reported landing of the algae patches in late July, 2009. The drifter released at 122ºE, 34ºN (Drifter 77) and the two southern drifters eventually drift into the central Yellow Sea. Interestingly, the three drifters all have crossed the 50 m isobaths into the center Yellow Sea, suggesting no significant tidal residual currents or frontal circulation in the southeastward direction along roughly the 50 m isobaths as in the traditional understanding of the regional circulation (Hu, 1994; Guan, 1994) or in numerical 13
simulations (e.g. Xia et al., 2006).
The circulation indicated by the drifter
trajectories seems to be in agreement with the wind-driven circulation in Xia et al. (2006) simulation without the tidal forcing. The three Drifters of 76, 77, and 90 have survived for a longer time (Figure 12). The trajectories of 76 and 90 show that the surface circulation over the western flank of the Yellow Sea changed directions in August 2009 to flow to the west and south. Again, no significant southeastward current along the 50 m isobath as simulated by Xia et al. (2006) is indicated by the drifter trajectories. The movement of the drifters in this period is forced by the easterly and northerly winds, as seem in the integrated progressive vector plot of the wind stress in Figure 13.
The drifter trajectories thus
suggest that the surface circulation in the western Yellow Sea is different in June-July from that in August-September.
Interestingly, Drifter 77, which was located in a
position very close to that of Drifter 76 on August 1, 2008, took a drifting course significantly different from that of Drifter 76.
The difference suggests chaotic
movement and mixing in the horizontal direction in the central Yellow Sea in summer. The patterns of wind stress variations in the summers of 2008 and 2009 are similar to each other (Figure 13).
In both years, southerly winds started to blow in
late May to early June and persisted through June and July, which drove the northward movement of the surface drifters. The southerly winds subsided in early August and were overwhelmed by northerly winds in early September, which marked the beginning of the northeasterly monsoon.
The surface circulation in the western
and central Yellow Sea evidently follows this pattern of wind stress variations. The direction of the wind stress in the summers of 2008 suggests that the winds off the south Shandong coast were onshore on July 9-13, 2008 (Figure 13), which explains the absence of the Qingdao upwelling on this day.
14
Upwelling favorable
winds were present along the south Shandong coast during late July through early August of 2008 and during May 31 through mid-July of 2009, which explain the observed upwelling during these periods (see numerical study results in the next section).
5. Dynamics of the circulation in the southwestern Yellow Sea The trajectories of the satellite-tracked surface drifters have shown that the currents over the western Yellow Sea flow northward in summer, in an opposite direction to the traditional understanding of the regional circulation.
The
measurements are in glaring contract to the existing numerical modeling results showing south- and southeast-ward flows over the western flank of the Yellow Sea trough forced by tides in summer.
The direction and variations of the drifter
movements suggest that the surface currents are forced by the winds instead of tides. In this section, the time series of the sub-tidal currents are analyzed and correlated with the wind stress variations to examine the wind-driven hypothesis.
The time
series of the drifter velocities are obtained from the trajectories, which are low-pass filtered with the Butterworth filter at the cutoff period of 40 hours. Daily-averaged time series are then calculated and are compared with the daily wind stress of the NCEP/NCAR reanalysis at (35ºN, 121ºE) in Figure 14. The time series of the meridional wind stress component show that the winds over the southwestern Yellow Sea are weak or to the south before June 1, after which a series of southerly wind bursts took place in the southwestern Yellow Sea. At the end of July, the southerly winds subside until northerly wind bursts pick up in September.
Similar wind stress variations are also observed in the summer of 2008
(Figure 13).
15
The meridional velocity of Drifter 90 shows that the drifter moves to the north in late May (bottom panel), against the weak northerly winds in May 21 but mostly during calm days, suggesting that this intrusion is forced remotely.
The small
intrusion on May 22 suggests wind-forced elements in the intrusion variations (Yuan, 2002).
After July 11, the drifter is offshore of the Jiangsu coast. Two peaks of
northward movement on July 15, and 19 are clearly in agreement with and lag behind the southerly wind bursts on July 13 and 17 by one or two days. The time lags are consistent with the spin-up of the coastal currents in the presence of bottom friction (Clarke and Brink, 1985). August 20 and 27.
Similar wind-forced currents are also observed around
The southward movement of this drifter on August 13 and 29,
and on September 9 and 14 are evidently forced by the northerly wind bursts. However, the currents in the Yellow Sea trough are obviously influenced by forces other than the winds. For example, the northward flow during August 1-10 of this drifter (Figure 12) cannot be explained by the local wind forcing, and is possibly associated with the northward intrusion of the Yellow Sea Warm Current (Zhao et al., 1986). The northward movement of Drifters 91, 78, and 88 during June 8-13, June 17-21, and June 25-July 1 is evidently forced by southerly wind bursts.
The drifters move
slowly to the south under weak wind forcing before June 1, which may suggest the effects of tides or some northerly wind forcing.
From the drifter trajectories, it is
clearly identified that the movement of the drifters in the absence of the winds are very slow compared with the wind-driven movement, suggesting that the tidal residual currents over the Subei Bank should be negligibly small.
By middle July,
the drifters had moved close to the south Shandong coasts. The correlation of the meridional currents and the wind are low because the coastal currents there are
16
dictated by the along-shore winds. Similar responses are also observed in the movement of Drifters 76 and 77, except that these drifters are mostly in the central Yellow Sea far away from the Subei coast. In summary, the analyses of the time series of the winds and the drifter movement suggest that the Subei coastal current is forced primarily by the winds to move northward in summer and southward in fall.
A northward intrusion along the
submarine valley offshore of the Changjiang mouth in summer has been observed by the drifter trajectories for the first time in history, the dynamics of which are not clear at present.
The drifer movement suggests that the coastal upwelling of the Subei
coast should be wind-driven instead of tide-driven, the dynamics of which are investigated in the next section.
6. Numerical modeling The above observations have disclosed important oceanographic phenomena, some of which are uncovered for the first time in history.
In particular, the
northward flow of the Subei coastal current in July and the associated upwelling along the Subei and the south Shandong (Lunan) coasts are subject to dynamics explanation. In the above, the drifter trajectories have evidenced the wind-driven dynamics of the Subei coastal current.
The dynamics of the upwelling are investigated in this section
using a numerical model. 6.1 Model validation The model has reproduced the surface drifter trajectories (Figure 15) and the sub-tidal currents at station B in the summer 2009 well (Wang et al., 2013).
The
correlation coefficients between the simulated and the observed currents for the east-west and north-south components at the middle depth (17 m) are 0.64 and 0.49,
17
respectively, both of which have passed the 95% significant level of 0.19.
The first
EOF modes of the observed east-west and north-south velocity time series, explain about 68% and 80% of the total variances, respectively, are reproduced by the numerical model to explaining 54% and 82% of the total simulated variance, respectively.
The correlations between the simulated and observed PC1 modes for
the east-west and north-south velocity components are as high as 0.89 and 0.95, above the 95% significant level, respectively.
In particular, the model has reproduced the
lack of movement of the surface drifters in late May through early June, when the winds were weak, and the northeastward movement of the wind-driven currents during middle June through early July of 2009 well (Figure 15).
Details of the
model-data inter-comparisons are reported in Wang et al. (2013), Li (2010), and He (2011) The comparisons suggest that the nested model can be used to study the dynamics of the regional circulation. The coastal upwelling off the Subei coast in the summer of 2008 has been simulated well by the model, as suggested by the agreement between the simulated and the observed temperature distribution in the vertical section along 33ºN (see Figure 7 of Wang et al., 2013). The thermocline in the model is a little diffused than in the observations and the near shore surface temperature is much higher than in the observations, due to inaccurate vertical diffusion coefficients and surface heat flux of the model.
However, the horizontal distribution patterns of the simulated surface
temperature and salinity (Figure 16) clearly resemble those of the observed (Figure 3), with the centers of upwelling located off the Subei and Lunan coasts. The observed offshore extension in the northeastward direction of the Sheyang River plume at the surface is also identified in the simulation. These comparisons suggest that the model can be used to study the dynamics of the upwelling currents in this area.
18
6.2 Effects of winds on the coastal upwelling According to the wind-driven coastal upwelling theory, the strength of the coastal upwelling is determined primarily by the along-shore component of the wind stress.
Therefore, the strengths of the Subei and Lunan coastal upwelling are
sensitive to the direction of the southerly winds in summer.
In two sensitivity
experiments, the model is forced by winds of 5 m s-1 at 10 m above the sea surface from the southwest and southeast, respectively, corresponding to about 0.3 dyn cm-2 of wind stress.
The results suggest that the southeasterly winds favor the Subei
coastal upwelling whereas the southwesterly winds favor the Lunan coastal upwelling (Figure 17).
The heat balance analysis of the surface mixed layer in the cold center
off the Lunan coast in the steady southwesterly experiment suggests the surface heating of 0.5×10-4 ℃ s-1 balanced by the advection (primarily the upwelling advection) of -0.3×10-4 ℃ s-1 and the vertical diffusivity of -0.2×10-4 ℃ s-1. These wind-forced experiments explain the dynamics of the Subei and Lunan coastal upwelling events in the observations.
On July 9-13, 2008, the Lunan coastal
upwelling is absent due to the southeasterly wind forcing, which is onshore at the Lunan coast and is not favorable for coastal upwelling there.
During late July
through early August of 2008 and during late May through middle June of 2009, the winds are dominated by southerly and southwesterly so that upwelling is generated along the Lunan coasts. Along the Subei coasts, upwelling is observed in all of the cruises, which can be explained by the strong and persistent forcing of the southerly winds over the shelf. In the southeasterly wind experiment above, the Subei coastal upwelling stays south of the Sheyang River plume, suggesting buoyancy shut-down by the fresh water plume.
This dynamics are investigated in the next subsection.
19
6.3 Effects of Sheyang River plume on the Subei coastal upwelling A controlled numerical experiment is conducted to repeat the base-line simulation in Section 6.1, except that the Sheyang River runoff is turned off (Figure 18). The resulting distributions of average surface temperature and salinity in July indicate that the Subei coastal upwelling extends much more northward without the Sheyang River plume.
The comparison with and without the river runoff clearly demonstrates the
shut-down effects by the buoyancy of the Sheyang River plume on the Subei coastal upwelling.
7. Summary The circulation in the southwestern Yellow Sea in summer is studied using hydrographic data of four research cruises, moored curremeter data, and trajectories of satellite-tracked ARGOS surface drifters in the summers of 2008 and 2009.
The
hydrographic data suggest coastal upwelling off the Subei and Lunan coasts and significant northeastward extension of fresh water from the Sheyang River mouth. Movement of surface drifters have suggested that the Subei coastal current flows northward in summer under the southerly wind forcing, opposite to the traditional understanding of the regional circulation in existing publications. The northward currents explain the drifting of the green algal patches from offshore of Subei to the Lunan coast in the summers of 2008 and 2009.
The drifter trajectories have also
shown that, during periods of no winds, the residual currents of the tide are small in this area and that a significant northward intrusion along the submarine valley off the Changjiang mouth persist, suggesting remote forcing dynamics of the current. Results of numerical modeling have shown that, associated with the northward wind-driven currents, onshore bottom Ekman transport and coastal upwelling is
20
generated along the Subei coast in summer.
At about 34ºN, the coastal upwelling is
shut-down by the buoyancy from the solar heating and fresh water input from the Sheyang River.
Coastal upwelling is observed to develop off the coast of Qingdao
city sometimes in summer, due to forcing of the along-shore component of the southerly and southwesterly winds. The fresh water input from the Sheyang River is found to have significant impact on the circulation in the southwestern Yellow Sea.
The low salinity water from the
Sheyang River is observed to extend northeastward over a large distance into the Lunan trough.
It is hypothesized that the wind-driven upwelling and the fresh water
plume of the Sheyang River play an important role in the nutrient dynamics and ecological blooms of Ulva Prolifera in summer in this area, which lead to the green tide events in July through August off the southern coast of the Shandong Peninsula.
Acknowledgment The authors thank the open cruises of the IOCAS and the hard work of the crews and scientists onboard of “KeXue-3”, “KeXue-1”, and “Beidou” research vessels. This work is support by NSFC (41176019, 41421005, U1406401), the National Basic Research Program of China (Project 2012CB956001), CMA (GYHY201306018), CAS (XDA11010301), and SOA (GASI-03-01-01-05).
References 1.
Clarke, A. J., and K. H. Brink, 1985. The response of stratified, frictional flow of shelf and slope waters to fluctuating large-scale, low-frequency wind forcing. J. Phys. Oceanogr., 15, 439–453.
2.
da Silva, A., C.Young and S. Levitus, ,1994. Atlas of Surface Marine Data 1994,
21
vol. 1, Algorithms and Procedures, NOAA Atlas NESDIS 6, U.S.Dep. of Commer., Washington, D.C. 3.
Guan, B., 1994. Patterns and structures of the currents in Bohai, Huanghai and East China Seas. Oceanology of China Seas. 11:1126–1134.
4.
He, L., 2011. Ocean circulation in the western Yellow and East China Seas and its impact on phytoplankton ecosystem dynamics. Postdoc Res. Rep., Institute of Oceanology, CAS, pp94.
5.
Hu, C., and M. He, 2008. Origin and offshore extent of floating algae in Olympic sailing area. EOS, 89(33):302–303.
6. Hu, D., 1994. Some striking features of circulation in Huanghai Sea and East China Sea. Oceanology of China Seas, 1. Netherlands: Kluwer Acadamic Publishers, 27–28. 7.
Kalnay et al., 1996. The NCEP/NCAR 40-year reanalysis project, Bull. Amer. Meteor. Soc., 77, 437-470.
8.
Lee, C., Rosati A., Spelman, J, 2006. Barotropic tidal mixing effects in a coupled climate model: Oceanic conditions in the Northern Atlantic, Ocean Modelling,11:464–477.
9.
Levitus, S., 1982. Climatological Atlas of the World Ocean, NOAA Prof. Pap. 13, 173 pp. plus 17 microfiche, U.S. Govt. Print. Off., Washington,D. C.
10. Li, Y., 2010. Structure and Dynamics of the Subtidal Circulation in the Southwestern Yellow Sea, PhD Thesis, Institute of Oceanology, Chinese Academy of Sciences. 11. Lü, X., F. Qiao, C. Xia, G. Wang, and Y. Yuan, 2010. Upwelling and surface cold patches in the Yellow Sea in summer: Effects of tidal mixing on the vertical circulation, Cont. Shelf Res., 30, 620-632, doi:10.1016/j.csr.2009.09.002.
22
12. Qiao, F. L., D. Y. Ma, M. Y. Zhu, R. X. Li, J. Y. Zang, and H. J. Yu, 2008. The Background of the green tide events (Ulva Prolifero bloom) off Qingdao 2008 and scientific response. Advances in Marine Science, 3, 409–410. (in Chinese with English abstract) 13. Su, J. L., 1998. Circulation dynamics of the China Seas: north of 18°N. In: "The Sea", Vol. 11, The Global coastal Ocean: Regional Studies and Syntheses, eds. A. R. Robinson and K. Brink, John Wiley, 483–506. 14. Wang B., 2010. The Influence of Topography on the Circulation in the Southwestern Yellow Sea, Master Thesis, Institute of Oceanology, Chinese Academy of Sciences. 15. Wang B., Li Y., Yuan D. L., 2013. Effects of topography on the sub-tidal circulation in the southwestern Huanghai Sea (Yellow Sea) in summer. Acta Oceanologica Sinica, 32(3): 1–9. 16. Wang B., N. Hirose, B. S. Kang, and K. Takayama, 2014. Seasonal migration of the Yellow Sea Bottom Cold Water. J. Geophys. Res. Oceans, 119, 4430–4443, doi:10.1002/2014JC009873. 17. Wu, H., J. Shen, J. Zhu, J. Zhang, and L. Li, 2014. Characteristics of the Changjiang plume and its extension along the Jiangsu Coast.
Cont. Shelf. Res.,
76, 108-123. 18. Xia, C., F. Qiao, M. Zhang, Y. Yang, and Y. Yuan, 2004a. Simulation of double cold cores of the 35ºN section in the Yellow Sea with a wave-tide-circulation coupled model, Chin. J. Oceanol. Limnol, 22, 292–298. 19. Xia, C., F. Qiao, Q. Zhang, and Y. Yuan, 2004b. Numerical modeling of the quasi-global ocean circulation based on POM, J. Hydrodyn., Ser. B,16:537–543. 20. Xia, C. S., F. L. Qiao, Y. Z. Yang, J. Ma, and Y. L. Yuan, 2006. Three
23
dimensional structure of the summertime circulation in the Yellow Sea from a wave-tide-circulation coupled model. J. Geophys. Res., 111, C11S03. Doi: 10.1029/2005JC003218. 21. Yuan, D., 2002. A numerical study of barotropicly forced intrusion in DeSoto Canyon. J. Geophys. Res. 107 (C2), doi:10.1029/2001JC000793. 22. Yuan, D., 2006. Dynamics of the Cold Water Events off the Southeast Coast of the United States in the summer of 2003. J. Phys. Oceanogr. , 36, 19121927. 23. Yuan, D., J. Zhu, C. Li, and D. Hu, 2008. Cross-shelf circulation in the Yellow and East China Seas indicated by MODIS satellite observations. J. Mar. Sys., 70 (1–2), 134–149. 24. Yuan, D., Y. Li, F. Qiao, and W. Zhao, 2012. Temperature inversion in the Yellow Sea Bottom Cold Water in summer. 25. Zhao, B. R., Q. Xiong, and F. G. Zhang, 1986. The internal hydrographic structure of the Huanghai Cold Water Mass (HCWM) in summer. Chin. J. Oceanol. Limnol., 4(1), 27–40. 26. Zhao, B. R., R. Limeberner, and D. Hu, 1991. Oceanographic Characteristics of the Southern Yellow Sea and the Northern East China Sea in summer. Oceanologia Et Limnologia Sinica, 22(2), 132–139. (in Chinese with English Abstract)
24
Figure Caption Figure 1.
Topography of the southwestern Yellow Sea. Unit of the depth is meter.
Dot marks the mooring station. Rectangle delineates the domain of the nested model. Figure 2.
Hydrographic survey stations in the southwestern Yellow Sea on July
22-August 3, 2008 (a), July 9-13, 2008 (b), August 9-11, 2008 (c), and May 28-June 21, 2009 (d), respectively. Figure 3.
Horizontal distribution of temperature (a, c) and salinity (b, d) at the
surface and 40 m depth, respectively, during July 22-August 3, 2008. Temperature unit is °C, salinity unit is psu. Figure 4.
Temperature (left column) and salinity (right column) distributions in
vertical sections along 32ºN, 33ºN, 34ºN, 35ºN, and 36ºN, respectively, in the summer of 2008. Figure 5.
Temperature unit is °C, salinity unit is psu.
Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20
m depth, respectively, in the southwestern Yellow Sea during the July 9-13, 2008 survey. Temperature unit is °C, salinity unit is psu. Figure 6.
Temperature (a) and salinity (b) distributions in vertical sections along 35º
N during July 9-13, 2008. Temperature unit is °C, salinity unit is psu. Figure 7.
Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20
m depth, respectively, in the southwestern Yellow Sea during the August 9-11, 2008 survey. Temperature unit is °C, salinity unit is psu. Figure 8.
Temperature (a) and salinity (b) distributions in the 35.5ºN vertical
sections on August 9-11, 2008. Temperature unit is °C, salinity unit is psu. Figure 9.
Temperature (a, c) and salinity (b, d) distributions at the sea surface and 30
m depth, respectively, in the western Yellow Sea during May 28-June 21, 2009 25
survey. Temperature unit is °C, salinity unit is psu. Figure 10. Temperature (a, c) and salinity (b, d) distributions along the vertical sections of 29ºN and 34ºN during May 28-June 21, 2009 survey. Temperature unit is °C, salinity unit is psu. Figure 11. Trajectories of satellite-tracked drifters in the southwestern Yellow Sea during May 28 through July31 of 2009. Figure 12. Trajectories of satellite-tracked surface drifters in the southwestern Yellow Sea after August 1, 2009. Figure 13. Integrated progressive vector plot of the NCEP wind stress at (35ºN, 121º E) in the summers of 2008 (a) and 2009 (b). Unit is Pa s. Figure 14. Comparison of daily averaged meridional wind stress from NCAR/NCEP reanalysis (top panel) and the meridional velocity of the ocean currents measured by the satellite-tracked surface drifters (low panels). Figure 15. Simulated surface drifter trajectories during May 1 through July 10 of 2009 based on the currents at the 5 m depth. Figure 16. Simulated average surface temperature (a) and salinity (b) distributions in July 2008. Temperature unit is °C, salinity unit is psu. Figure 17. Simulated surface temperature distribution forced by steady southwest (a) and southeast (b) winds. Temperature unit is °C. Figure 18. Simulated surface temperature distribution with (a) and without (b) the river runoff. Temperature unit is °C.
26
Highlights
This study provide solid evidence that the Subei coastal currents flows northward in summer, using the Lagrangian trajectories of satellite-tracked ARGOS surface drifters in the summers of 2008 and 2009. The wind-driven upwelling off the Subei coasts and the buoyancy shut-down by the Sheyang River fresh water plume is revealed for the first time in history. Coastal upwelling is also observed off the coast of Qingdao city in middle July through early August of 2008, which is forced by the along-shore component of the southerly and southwesterly winds. The Qingdao upwelling in summer is reported for the first time in history. A northward intrusion along the submarine valley off the Changjiang mouth, which is persistently strong and independent of the winds in summer, is indicated by the movement of the drifters for the first time in history.
27
Figure 1. Topography of the southwestern Yellow Sea. Unit of the depth is meter. Dot marks the mooring station. Rectangle delineates the domain of the nested model.
Figure 2. Hydrographic survey stations in the southwestern Yellow Sea on Julyne
22-August 328,2008(a), July 9-13, 2008(b), August 9-117, 2008(c), and May 28-June 21, 2009(d), respectively.
Figure 3. Horizontal distribution of temperature (a, c) and salinity (b, d) at the surface and 40 m depth, respectively, during July 22-28, 2008. Temperature unit is °C, salinity unit is psu.゜C
Figure 4. Temperature (left column) and salinity (right column) distributions in vertical sections along 32ºN, 33ºN, 34ºN, 35ºN, and 36ºN, respectively, in the summer of 2008. Temperature unit is °C, salinity unit is psu.
Figure 5. Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20 m depth, respectively, in the southwestern Yellow Sea during the July 9, 2008 survey. Temperature unit is °C, salinity unit is psu.
Figure 6. Temperature (a) and salinity (b) distributions in vertical sections along 35º N on July 9, 2008. Temperature unit is °C, salinity unit is psu.
Figure 7. Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20 m depth, respectively, in the southwestern Yellow Sea during the August 9, 2008 survey. Temperature unit is °C, salinity unit is psu.
Figure 8. Temperature (a) and salinity (b) distributions in the 35.5ºN vertical sections on August 9, 2008. Temperature unit is °C, salinity unit is psu.
Figure 9. Temperature (a, c) and salinity (b, d) distributions at the sea surface and 30 m depth, respectively, in the western Yellow Sea during May 28-June 21, 2009 survey. Temperature unit is °C, salinity unit is psu.
Figure 10. Temperature (a, c) and salinity (b, d) distributions along the vertical sections of 29ºN and 34ºN during May 28-June 21, 2009 survey. Temperature unit is °C, salinity unit is psu.
Figure 11. Trajectories of satellite-tracked surface drifters in the southwestern Yellow Sea during May 28 through July31 of 2009.゜
Figure 12. Trajectories of satellite-tracked surface drifters in the southwestern Yellow Sea after August 1, 2009.
Figure 13. Integrated progressive vector plot of the NCEP wind stress at (35ºN, 121º E) in the summers of 2008 (a) and 2009 (b). Unit is Pa s.
Figure 14. Comparison of daily averaged meridional wind stress from NCAR/NCEP reanalysis (top panel) and the meridional velocity of the ocean currents measured by the satellite-tracked surface drifters (low panels).
Figure 15. Simulated surface drifter trajectories during May 1 through July 10 of 2009 based on the currents at the 5 m depth.
Figure 16. Simulated average surface temperature (a) and salinity (b) distributions in July 2008. Temperature unit is °C, salinity unit is psu.
Figure 17. Simulated surface temperature distribution forced by steady southwestly (a) and southeastly (b) winds. Temperature unit is °C.
Figure 18. Simulated surface temperature distribution with (a) and without (b) the river runoff. Temperature unit is °C.
PII: DOI: Reference:
S0278-4343(17)30070-5 http://dx.doi.org/10.1016/j.csr.2017.01.022 CSR3553
To appear in: Continental Shelf Research Received date: 29 August 2015 Revised date: 4 November 2016 Accepted date: 10 January 2017 Cite this article as: Dongliang Yuan, Yao Li, Bin Wang, Lei He and Naoki Hirose, Coastal Circulation in the Southwestern Yellow Sea in the Summers of 2008 and 2009, Continental Shelf Research, http://dx.doi.org/10.1016/j.csr.2017.01.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Coastal Circulation in the Southwestern Yellow Sea in the Summers of 2008 and 2009
Dongliang Yuan1,2*, Yao Li1,2, Bin Wang3,4, Lei He5, and Naoki Hirose4 1
Key Laboratory of Ocean Circulation and Wave, Institute of Oceanology, Chinese
Academy of Sciences, Qingdao, China 2
Qingdao Collaborative Innovation Center of Marine Science and Technology,
Qingdao, China 3
College of oceanography, Hohai University, Nanjing, China
4
Research Institute for Applied Mechanics, Kyushu University, Kasuga, Japan
5
School of Marine Sciences, Sun Yat-Sen University, Guangzhou, China
*
Corresponding author at: Institute of Oceanology, Chinese Academy of Sciences, 7
Nanhai Road, Qingdao, China 266071. Email: [email protected]
Abstract The coastal ocean circulation off the Chinese Subei (north Jiangsu Province) coasts in the southwestern Yellow Sea in summer is studied using hydrographic data from four research cruises, moored current meter data in the Lunan (south Shandong Province) trough, and Lagrangian trajectories of satellite-tracked ARGOS surface drifters in the summers of 2008 and 2009.
The hydrographic data indicate a fresh
water plume extending offshore in the northeastward direction from the Sheyang River mouth and a belt of relatively cold surface temperature from the surrounding areas located offshore of the Subei coasts to the south of the plume in summer.
The
drifter trajectories show that the Subei coastal current, which is traditionally thought to flow southward year round, flows northward in June through July under the forcing of the southerly monsoon.
A northward intrusion along the submarine valley off the
1
Changjiang mouth is indicated by the movement of the drifters, which is persistently strong and independent of the winds.
Associated with the northward flow, coastal
upwelling is generated offshore of the Subei coast to the south of the Sheyang River plume, below which upwelling is generated but is shut down by the buoyancy of the fresh water at the surface.
Coastal upwelling is also observed off the coast of
Qingdao city in middle July through early August of 2008, which is forced by the along-shore component of the southerly and southwesterly winds.
The Qingdao
upwelling in summer is reported for the first time in history.
Keywords Subei coastal current, Lunan coastal current, Southwestern Yellow Sea, Sheyang River plume, Coastal upwelling, Buoyancy shut-down
1. Introduction The southwestern Yellow Sea is the part of the Yellow Sea south of the Shandong Peninsula and north of the Changjiang (aka the Yangtze River) mouth between Mainland China and the center of the Yellow Sea trough (Figure 1). The topography of this area features a very shallow bank called the Subei Bank and a submarine trough called the Lunan (south Shandong) trough extending from the center of the Yellow Sea trough westward into the Haizhou Bay between the Shandong Peninsula and the Subei Bank.
Offshore of the Changjiang mouth in the southern Yellow Sea,
a large area of shallow water called the Changjiang Bank forms an obstacle between the currents in the Yellow Sea center and in the western East China Sea.
South of
the Changjiang mouth off the Hangzhou Bay estuary, a submarine valley is located between the nearshore shallow estuary and the offshore Changjiang Bank, forming an
2
important conduit of deep-sea water intrusion onto the shallow shelf off the coast of the Chinese Jiangsu Province. The topographic features suggest that the circulation in the southwestern Yellow Sea is subject to complicated topographic steering. Historical understanding of the circulation in this area has been vague, except that the currents in this area are speculated to flow southward year round (Zhao et al., 1991; Guan, 1994; Hu, 1994; Su, 1998).
In summer, the southward currents are argued to be associated with the
strong temperature and salinity fronts along roughly the 40-50 m isobaths in the western Yellow Sea (Hu, 1994).
Recently, Yuan et al. (2008) have suggested that the
Subei coastal current might flow to the north in summer in response to the forcing of the southwesterly monsoon.
However, due to lack of in situ measurements of the
currents, the path and the dynamics of the Subei coastal current in summer remain unclear. Lately, the circulation in the southwestern Yellow Sea has received international attention due to repeated outbursts of large-scale Ulva Prolifera blooms off the coasts of the south Shandong Peninsula in summer since 2006. The green tide events have generated significant impact on regional environment and economy, the largest event of which took place in the summer of 2008 at the dawn of the 29th summer Olympic Game in the Olympic sailing competition site offshore of the city of Qingdao (Hu and He, 2008; Qiao et al., 2008).
Significant green tide events of similar scope and
amplitude recurred in the summers of 2009 and the years to follow.
It is, therefore,
of great interests to study the ocean circulation associated with these events. In the past, the study of the vertical water motion in the Yellow Sea in summer has focused mostly on the circulation associated with the Yellow Sea Bottom Cold Water. Cold surface temperature patches have been identified in the southwestern
3
Yellow Sea in summer and have been attributed to frontal upwelling induced by tidal mixing and rectification (Lü et al., 2010). Historical direct observations of the cold SST off the Qingdao coast in summer were rare (see the review of the regional observational and modeling studies in Lü et al. 2010).
No previous studies,
including the Lü et al. (2010) study, have conducted focused investigation of the upwelling off the Qingdao coast.
Because the Subei coastal current was thought to
flow southward all year round, the upwelling associated with the wind-driven circulation has not been studied before. In this paper, we use temperature and salinity measurements collected during four research cruises in the summers of 2008 and 2009 to study the ocean circulation in the southwestern Yellow Sea in summer.
These hydrographic data are used in
combination with trajectories of satellite-tracked surface drifters to study the path, direction, and dynamics of the Subei coastal current in summer.
A numerical model
is then used to simulate the shelf circulation in the summer of 2008, based on which the dynamics of the coastal upwelling currents are studied. The data used in this study are described in the next section. The temperature and salinity distributions during the 2008 and 2009 research cruises are presented and discussed in Sections 3 and 4, respectively. The structure and dynamics of the Subei coastal current in the southwestern Yellow Sea are discussed using drifter trajectories in Section 5.
In Section 6, a numerical simulation of the shelf circulation in the
southwestern Yellow Sea in summer is carried out, based on which the dynamics of the coastal upwelling are investigated. Section 7 summarizes the conclusions of this study.
4
2. Data The data used in this study include the hydrography data, moored current meter data, satellite-tracked surface drifter trajectories, and a numerical simulation using the Princeton Ocean Model. 2.1 Hydrography and moored current meter data The data used in this study are collected during three cruises in the summer of 2008 and during one cruise in the summer of 2009 in the southwestern Yellow Sea. The major surveys of the regional circulation took place during July 22-August 3, 2008 and May 28-June 21, 2009 in the western Yellow Sea from the Changjiang mouth to the northern Yellow Sea, with a spatial resolution of about 0.5 degree in the zonal direction and 2 degrees in the meridional direction (Figure 2).
We will focus
on the shelf circulation in the western Yellow Sea between the Changjiang mouth and the Shandong Peninsula, which is called the southwestern Yellow Sea in this study. Besides the two major surveys, two complementary cruises are carried out on July 9-13 and August 9-11 of 2008, respectively, to sample smaller areas, between 34ºN and 36ºN in the Lunan trough, with higher spatial resolutions, where a large amount of the floating algal patches are located and tracked (Figures 2b and 2c).
The
temperature and salinity data were collected using Conductivity-Temperature-Depth (CTD) instruments of the Seabird Electronics, Inc. of the USA. During the cruise of July 22-August 3, 2008, a bottom-mounted mooring is deployed at station B (34°59.77'N, 120°30.52'E) with a water depth of about 30 m (Figure 1).
An Acoustic Doppler Current Profile (ADCP) current meter with a 250
kHz band Sontek ADP sensor capable of measuring full-column current profiles in the shallow water was tethered at about 3 m above the bottom of the ocean to return
5
continuous current measurements during July 22-28, 2008. The ADCP profiles of east-west and north-south components of the current at station B were recorded at every 10 minutes.
The sub-tidal currents are obtained by removing high-frequency
signals using a 5th order low-pass Butterworth filter with a cut-off period at 40 hours. Due to intensive fishing activities in the area, long-term mooring measurements of the shelf currents are rare and precious. The moored current measurements thus provide us with important observations to validate and calibrate the model simulation, based on which the regional circulation and dynamics of the currents are studied. 2.2 Ocean model The Princeton Ocean Model (POM) with a nested grid is applied to study the structure and dynamics of the ocean circulation in the southwestern Yellow Sea. The model domain includes a coarse-grid region of (15°N-45°N, 105°E-135°E) with a horizontal resolution of 1/6° longitude by 1/6° latitude and a nested-grid smaller domain of the southwestern Yellow Sea (32°N-37°N, 119°E-124°E) with a horizontal resolution of 1/24° longitude by 1/24° latitude.
The topography of the large domain
is based on the 5′×5′ global ETOPO5 elevation data whereas the topography of the small domain is based on the local navigational chart.
Both the large and the small
domain models have 30 sigma levels in the vertical. We will focus on the wind-driven circulation in this study.
The effect of tidal
mixing is parameterized by the Munk–Anderson scheme as described in Lee et al. (2006).
The large-domain model is forced by the global 0.5°×0.5° model simulation
of Xia et al. (2004a, 2004b) at the open boundaries, and by monthly climatological wind stress and the Haney-type heat flux of the interpolated COADS (Comprehensive Ocean-Atmosphere Data Set) data (da Silva et al., 1994) at the sea surface.
The
surface salinity of the model is relaxed to the summer monthly climatology at a time
6
scale of 30 days.
The large-domain model has been integrated for 6 years from the
Levitus (1982) annual mean climatological temperature and salinity and geostrophic currents of zero vertically integrated transport.
The kinetic energy of the model is
found to reach a quasi-steady state in about 4 years of integration. The small-domain and the large-domain models are then spun up for 6 years. The surface wind stress and heat/salt flux conditions are the same as in the large-domain model.
The Sheyang River runoff in the small-domain model is set at
1000 m3 s-1. Due to lack of accurate measurements of the river discharge in this area, the specification of the river runoff in the small-domain model is based on testing and order-of-magnitude estimate.
Some of the latest studies based on numerical
modeling (e.g. Wu et al. 2014) suggest that the Changjiang fresh water is transported to the area by a nearshore tidal residual current, which is then transported offshore at the Sheyang River mouth. We do not distinguish between these two sources of the fresh water. Here, the model fresh water specification at the Sheyang River mouth represents the total freshwater to be transport offshore of the Subei coasts, which are supposed to be much larger than the Sheyang River discharge based on numerical experiments.
Finally, the NCEP-One 6-hourly re-analysis wind stress (Kalnay et al.,
1996) is used to force the model for three months from May 1 through July 31 of 2008.
The details of the model configuration and parameterization are described in
the master thesis of Wang (2010) and the Ph. D. dissertation of Li (2010). 2.3 ARGOS trajectory data In the summer of 2009, six satellite-tracked ARGOS floats were deployed in the southwestern Yellow Sea to measure the currents off the coast of Jiangsu Province. Two of the floats were released in the submarine valley area offshore of the Changjiang mouth to measure the movement of the waters in the East China Sea.
7
The remaining four floats were released in the 34ºN section to track the movement of the Subei coastal current.
All of the drifters have their drogues dragged at the 15 m
depth except the two released at about the 12 m isobath at 34ºN, whose drogue cables are set at 6 m.
The drogues of the ARGOS floats are nylon tubes of about 6.44 m
long and about 0.6 m in diameter. The diameter of the ARGOS float is 38 cm.
3. Hydrography in the summer of 2008 The surface and subsurface distributions of temperature and salinity are presented in this section, which form the basis of observational and numerical modeling studies of the dynamics of the coastal upwelling. 3.1 Temperature and salinity distributions The distributions of surface salinity and temperature in the southwestern Yellow Sea during July 9-13, 2008 show that the surface water off the coast of the Jiangsu Province is dominated by warm (> 25.5 ℃) and fresh (<30.5 psu) water from the rivers near 34ºN (the waters from multiple rivers together are called the Sheyang River plume in the following text) and by two patches of cold water: one off the coast of Subei to the south of the plume and one offshore of Qingdao, which indicate coastal upwelling (Figure 3).
The warm and extremely fresh (< 28 psu salinity)
Changjiang Diluted Water is only seen near the southern boundary.
The distribution
of the Changjiang Diluted Water south of 32ºN was not cover by the survey in the summer of 2008.
The Changjiang Dilute Water, as suggested by the salinity
concentration below 30 psu, is evidently confined south of the 33ºN latitude, and, therefore, is of limited influence to the circulation over the Subei Bank and in the Lunan trough in the north.
The cold sea surface temperature center off the coast of
the south Shandong Peninsula has not been reported before. We call this center the 8
Qingdao upwelling in the following text. At the 40 m depth, the distributions of salinity and temperature suggest that the western Yellow Sea trough is occupied by relatively warm (> 10℃) and saline (>33 psu) bottom water in the south (lower panels of Figure 3), with a salinity and temperature front at about 34.5ºN. This area is suggested to be the confluence of the residual Yellow Sea Warm Current from the south and the cold fresh water from north, which is seen every year in the 35ºN and 36ºN vertical sections in historical data and is evidenced by the southward movement of satellite-tracked drifters from the northern Yellow Sea into this area (Yuan et al., 2012). The temperature and salinity distributions in the vertical section of 32ºN show that the subsurface water at the southern boundary is warm (>22℃) and saline (>32 psu) compared to the waters in the 33ºN and 34ºN sections (Figure 4).
The warm
and fresh Changjiang Diluted Water is observed in a very thin layer of about 10 m offshore at the surface in this section. Further north at the 33ºN and 34ºN sections, the warm and saline water is absent, suggesting that it either exits the survey area through the eastern boundary, as evidenced by the salinity distribution at the 40 m depth, or mixes with the coastal fresh waters to lose its salinity identity.
The detailed
movement and evolution of this water mass is not resolved by our survey. A water mass of high salinity (> 33 psu) and low temperature (<14 ℃) is identified between the 35 m and 80 m isobaths in the 33ºN and 34ºN sections,.
The
temperature of this water mass is higher than the coldest water (< 10℃) of the Yellow Sea Bottom Cold Water at the central bottom of the Yellow Sea trough.
At the 35ºN
and 36ºN sections, the saline water are flanked by two even colder water masses: one at the western flank, believed to be connected with the cold-core water in the 34ºN 9
section, and one at the central bottom of the Yellow Sea trough, which is the traditional Yellow Sea Bottom Cold Water, respectively.
The cold-core water mass
over the western flank has been shown to be associated with the southward movement of the northern Yellow Sea water in summer (Yuan et al., 2012; Wang et al., 2014). 3.2 Upwelling fronts During the summer of 2008, northward drifting of large patches of the green algae was tracked by satellite observations (Hu and He, 2008).
Associated with the
northward coastal currents, onshore bottom Ekman transport and coastal upwelling are indicated by the temperature and salinity distributions in the vertical sections in Figure 4. The upwelling front is located above roughly the 20 m isobath in the 33ºN section, more than 50 km from the coast. Further north at 34ºN and 35ºN, the upwelling is shut down by the buoyancy of the warm and fresh water at the surface. Similar cases have been reported off the southeast coast of the U.S. in summer (Yuan, 2006). Further north at 36ºN, the nearshore surface temperature at Qingdao is also lower, evidently due to upwelling of bottom cold water. The evolution of the upwelling in the southwestern Yellow Sea in the summer of 2008 is further disclosed by the two supplementary surveys in the smaller area between 34ºN and 36ºN on July 9 and August 9, respectively.
During July 9-13,
2008, the surface temperature and salinity distributions off the Subei coast are similar to those during July 22-August 3 2008 survey, except that the cold center off Qingdao is absent (Figure 5).
The Subei coastal area is still dominated by the Sheyang River
plume, which extends northeastward into the central Yellow Sea.
The upwelling
associated with the northward coastal currents is evidenced by a belt of cold water between 120ºE and 120.5ºE at the 20 m depth produced by the onshore bottom Ekman transport (bottom panels of Figure 5 and Figure 6). 10
This upwelling is
evidently capped by the fresh water from the Sheyang River. By August 9-11, 2008, the surface salinity over the southwestern Yellow Sea is still characterized by the Sheyang River plume, although the scale and the pattern are different (Figure 7). The surface temperature distribution shows cold surface water entering the survey area from the south along an offshore path east of 122ºE, which is likely the upwelled water in the south.
The cold center off Qingdao is clearly
identified, indicating that the Qingdao upwelling has persisted through late July. The upwelling is evidently associated with the onshore bottom Ekman transport, as suggested by the temperature and salinity distribution in the vertical sections in Figure 8. Notice that the surface cold centers of the upwelling are located further nearshore instead of directly above the tidal fronts, suggesting these upwelling currents are associated with the winds instead of the tides.
4. Hydrography in the summer of 2009 The hydragraphy in the summer of 2009 are presented in this section to be compared with that in the summer of 2008. Drifter trajectories in the summer 2009 provide direct current measurements to supplement the hydrography surveys in the summers of the two years. 4.1 Temperature and salinity distributions The horizontal distributions of temperature and salinity at the surface and at the 30 m depth during May 28-June 21, 2009 are plotted in Figure 9, showing the dominance of fresh water west of 123ºE in the southwestern Yellow Sea.
Offshore
of the Changjiang mouth, a warm water tongue extending northward is in contract to the surface temperature distributions in the summer of 2008, which show cold offshore water due to upwelling at 33ºN.
This is because sections 32ºN and 33ºN
11
are not measured by the cruise.
During the summer 2009 survey, the sections south
and along of 35ºN were surveyed before May 31 and the sections north of 35ºN were measured after June 15 (Figure 9 upper panels).
The coastal upwelling is indicated
by the center of cold surface water in the Haizhou Bay and east of the tip of the Shandong Peninsula. In the subsurface, the northward intrusion of salty and warm water along the submarine canyon offshore of the Changjiang mouth is suggested by the salinity and temperature contours at the 30 m depth.
The salty water in the western Yellow Sea
trough and the southward spread of the northern Yellow Sea water around the tip of the Shandong Peninsula are also indicated by the subsurface temperature and salinity distributions. During the summer 2009, the vertical section of 29ºN is characterized by the extremely fresh water from the Changjiang in a thin surface layer (Figure 10 bottom panels).
A surface temperature inversion occurs offshore at about 123.5ºE, with a
warm-core but slightly high salinity at the subsurface. warm-core water is not clear.
The reason of the subsurface
However, it is unlikely that this water mass has impact
the circulation in the north significantly, because the temperature and salinity distributions in the 34ºN sections do not show the presence of this mass. 4.2 Direct current measurements using ARGOS drifter trajectories Due to lack of direct current measurements off the Subei coast in history, the direction and the path of the Subei coastal current have been in debate. To study this current, six satellite-tracked ARGOS surface drifters are released at 34ºN, 32ºN, and 31ºN. The drifter released at 31ºN in the submarine valley south of the Changjiang mouth is seen to move northward steadily and has repeated the path of the drifter
12
released at 32ºN precisely, suggesting the quasi-steady northward intrusion of the current in the area (Figure 11).
The southern drifter (Drifter 90) later got stuck over
the shallow shelf for nearly a month only to revive later in middle July to drift eastward into the central Yellow Sea. The northern drifter (Drifter 76) released at 32 ºN has been moving northward continuously and eventually enters the central Yellow Sea. Under very weak wind conditions, the four ARGOS drifters released at 34ºN initially made circular movement around their releasing positions within a range of about 10-20 km (~20 km for floats 77 and 78, ~10 km for floats 88 and 91) during the period between May 30 through June 3. The periodic movement is obviously forced by the tidal currents and has suggested weak tidal residual currents over the Subei Bank. Since June 4, after the southerly winds pick up, all four drifters started to move north- and northeastward, suggesting the dominance of the wind-driven currents over the tidal residual currents. The trajectories of the drifters clearly suggest that the Subei coastal current flows northward in summer, against the traditional understanding of the Yellow Sea Coast Current in summer. The northeastward drifting trajectories and the final landing positions at the coast of the southern Shandong Peninsula of the three drifters (Drifters 91,78, and 88) initially released at the 34ºN are in agreement with the drifting and reported landing of the algae patches in late July, 2009. The drifter released at 122ºE, 34ºN (Drifter 77) and the two southern drifters eventually drift into the central Yellow Sea. Interestingly, the three drifters all have crossed the 50 m isobaths into the center Yellow Sea, suggesting no significant tidal residual currents or frontal circulation in the southeastward direction along roughly the 50 m isobaths as in the traditional understanding of the regional circulation (Hu, 1994; Guan, 1994) or in numerical 13
simulations (e.g. Xia et al., 2006).
The circulation indicated by the drifter
trajectories seems to be in agreement with the wind-driven circulation in Xia et al. (2006) simulation without the tidal forcing. The three Drifters of 76, 77, and 90 have survived for a longer time (Figure 12). The trajectories of 76 and 90 show that the surface circulation over the western flank of the Yellow Sea changed directions in August 2009 to flow to the west and south. Again, no significant southeastward current along the 50 m isobath as simulated by Xia et al. (2006) is indicated by the drifter trajectories. The movement of the drifters in this period is forced by the easterly and northerly winds, as seem in the integrated progressive vector plot of the wind stress in Figure 13.
The drifter trajectories thus
suggest that the surface circulation in the western Yellow Sea is different in June-July from that in August-September.
Interestingly, Drifter 77, which was located in a
position very close to that of Drifter 76 on August 1, 2008, took a drifting course significantly different from that of Drifter 76.
The difference suggests chaotic
movement and mixing in the horizontal direction in the central Yellow Sea in summer. The patterns of wind stress variations in the summers of 2008 and 2009 are similar to each other (Figure 13).
In both years, southerly winds started to blow in
late May to early June and persisted through June and July, which drove the northward movement of the surface drifters. The southerly winds subsided in early August and were overwhelmed by northerly winds in early September, which marked the beginning of the northeasterly monsoon.
The surface circulation in the western
and central Yellow Sea evidently follows this pattern of wind stress variations. The direction of the wind stress in the summers of 2008 suggests that the winds off the south Shandong coast were onshore on July 9-13, 2008 (Figure 13), which explains the absence of the Qingdao upwelling on this day.
14
Upwelling favorable
winds were present along the south Shandong coast during late July through early August of 2008 and during May 31 through mid-July of 2009, which explain the observed upwelling during these periods (see numerical study results in the next section).
5. Dynamics of the circulation in the southwestern Yellow Sea The trajectories of the satellite-tracked surface drifters have shown that the currents over the western Yellow Sea flow northward in summer, in an opposite direction to the traditional understanding of the regional circulation.
The
measurements are in glaring contract to the existing numerical modeling results showing south- and southeast-ward flows over the western flank of the Yellow Sea trough forced by tides in summer.
The direction and variations of the drifter
movements suggest that the surface currents are forced by the winds instead of tides. In this section, the time series of the sub-tidal currents are analyzed and correlated with the wind stress variations to examine the wind-driven hypothesis.
The time
series of the drifter velocities are obtained from the trajectories, which are low-pass filtered with the Butterworth filter at the cutoff period of 40 hours. Daily-averaged time series are then calculated and are compared with the daily wind stress of the NCEP/NCAR reanalysis at (35ºN, 121ºE) in Figure 14. The time series of the meridional wind stress component show that the winds over the southwestern Yellow Sea are weak or to the south before June 1, after which a series of southerly wind bursts took place in the southwestern Yellow Sea. At the end of July, the southerly winds subside until northerly wind bursts pick up in September.
Similar wind stress variations are also observed in the summer of 2008
(Figure 13).
15
The meridional velocity of Drifter 90 shows that the drifter moves to the north in late May (bottom panel), against the weak northerly winds in May 21 but mostly during calm days, suggesting that this intrusion is forced remotely.
The small
intrusion on May 22 suggests wind-forced elements in the intrusion variations (Yuan, 2002).
After July 11, the drifter is offshore of the Jiangsu coast. Two peaks of
northward movement on July 15, and 19 are clearly in agreement with and lag behind the southerly wind bursts on July 13 and 17 by one or two days. The time lags are consistent with the spin-up of the coastal currents in the presence of bottom friction (Clarke and Brink, 1985). August 20 and 27.
Similar wind-forced currents are also observed around
The southward movement of this drifter on August 13 and 29,
and on September 9 and 14 are evidently forced by the northerly wind bursts. However, the currents in the Yellow Sea trough are obviously influenced by forces other than the winds. For example, the northward flow during August 1-10 of this drifter (Figure 12) cannot be explained by the local wind forcing, and is possibly associated with the northward intrusion of the Yellow Sea Warm Current (Zhao et al., 1986). The northward movement of Drifters 91, 78, and 88 during June 8-13, June 17-21, and June 25-July 1 is evidently forced by southerly wind bursts.
The drifters move
slowly to the south under weak wind forcing before June 1, which may suggest the effects of tides or some northerly wind forcing.
From the drifter trajectories, it is
clearly identified that the movement of the drifters in the absence of the winds are very slow compared with the wind-driven movement, suggesting that the tidal residual currents over the Subei Bank should be negligibly small.
By middle July,
the drifters had moved close to the south Shandong coasts. The correlation of the meridional currents and the wind are low because the coastal currents there are
16
dictated by the along-shore winds. Similar responses are also observed in the movement of Drifters 76 and 77, except that these drifters are mostly in the central Yellow Sea far away from the Subei coast. In summary, the analyses of the time series of the winds and the drifter movement suggest that the Subei coastal current is forced primarily by the winds to move northward in summer and southward in fall.
A northward intrusion along the
submarine valley offshore of the Changjiang mouth in summer has been observed by the drifter trajectories for the first time in history, the dynamics of which are not clear at present.
The drifer movement suggests that the coastal upwelling of the Subei
coast should be wind-driven instead of tide-driven, the dynamics of which are investigated in the next section.
6. Numerical modeling The above observations have disclosed important oceanographic phenomena, some of which are uncovered for the first time in history.
In particular, the
northward flow of the Subei coastal current in July and the associated upwelling along the Subei and the south Shandong (Lunan) coasts are subject to dynamics explanation. In the above, the drifter trajectories have evidenced the wind-driven dynamics of the Subei coastal current.
The dynamics of the upwelling are investigated in this section
using a numerical model. 6.1 Model validation The model has reproduced the surface drifter trajectories (Figure 15) and the sub-tidal currents at station B in the summer 2009 well (Wang et al., 2013).
The
correlation coefficients between the simulated and the observed currents for the east-west and north-south components at the middle depth (17 m) are 0.64 and 0.49,
17
respectively, both of which have passed the 95% significant level of 0.19.
The first
EOF modes of the observed east-west and north-south velocity time series, explain about 68% and 80% of the total variances, respectively, are reproduced by the numerical model to explaining 54% and 82% of the total simulated variance, respectively.
The correlations between the simulated and observed PC1 modes for
the east-west and north-south velocity components are as high as 0.89 and 0.95, above the 95% significant level, respectively.
In particular, the model has reproduced the
lack of movement of the surface drifters in late May through early June, when the winds were weak, and the northeastward movement of the wind-driven currents during middle June through early July of 2009 well (Figure 15).
Details of the
model-data inter-comparisons are reported in Wang et al. (2013), Li (2010), and He (2011) The comparisons suggest that the nested model can be used to study the dynamics of the regional circulation. The coastal upwelling off the Subei coast in the summer of 2008 has been simulated well by the model, as suggested by the agreement between the simulated and the observed temperature distribution in the vertical section along 33ºN (see Figure 7 of Wang et al., 2013). The thermocline in the model is a little diffused than in the observations and the near shore surface temperature is much higher than in the observations, due to inaccurate vertical diffusion coefficients and surface heat flux of the model.
However, the horizontal distribution patterns of the simulated surface
temperature and salinity (Figure 16) clearly resemble those of the observed (Figure 3), with the centers of upwelling located off the Subei and Lunan coasts. The observed offshore extension in the northeastward direction of the Sheyang River plume at the surface is also identified in the simulation. These comparisons suggest that the model can be used to study the dynamics of the upwelling currents in this area.
18
6.2 Effects of winds on the coastal upwelling According to the wind-driven coastal upwelling theory, the strength of the coastal upwelling is determined primarily by the along-shore component of the wind stress.
Therefore, the strengths of the Subei and Lunan coastal upwelling are
sensitive to the direction of the southerly winds in summer.
In two sensitivity
experiments, the model is forced by winds of 5 m s-1 at 10 m above the sea surface from the southwest and southeast, respectively, corresponding to about 0.3 dyn cm-2 of wind stress.
The results suggest that the southeasterly winds favor the Subei
coastal upwelling whereas the southwesterly winds favor the Lunan coastal upwelling (Figure 17).
The heat balance analysis of the surface mixed layer in the cold center
off the Lunan coast in the steady southwesterly experiment suggests the surface heating of 0.5×10-4 ℃ s-1 balanced by the advection (primarily the upwelling advection) of -0.3×10-4 ℃ s-1 and the vertical diffusivity of -0.2×10-4 ℃ s-1. These wind-forced experiments explain the dynamics of the Subei and Lunan coastal upwelling events in the observations.
On July 9-13, 2008, the Lunan coastal
upwelling is absent due to the southeasterly wind forcing, which is onshore at the Lunan coast and is not favorable for coastal upwelling there.
During late July
through early August of 2008 and during late May through middle June of 2009, the winds are dominated by southerly and southwesterly so that upwelling is generated along the Lunan coasts. Along the Subei coasts, upwelling is observed in all of the cruises, which can be explained by the strong and persistent forcing of the southerly winds over the shelf. In the southeasterly wind experiment above, the Subei coastal upwelling stays south of the Sheyang River plume, suggesting buoyancy shut-down by the fresh water plume.
This dynamics are investigated in the next subsection.
19
6.3 Effects of Sheyang River plume on the Subei coastal upwelling A controlled numerical experiment is conducted to repeat the base-line simulation in Section 6.1, except that the Sheyang River runoff is turned off (Figure 18). The resulting distributions of average surface temperature and salinity in July indicate that the Subei coastal upwelling extends much more northward without the Sheyang River plume.
The comparison with and without the river runoff clearly demonstrates the
shut-down effects by the buoyancy of the Sheyang River plume on the Subei coastal upwelling.
7. Summary The circulation in the southwestern Yellow Sea in summer is studied using hydrographic data of four research cruises, moored curremeter data, and trajectories of satellite-tracked ARGOS surface drifters in the summers of 2008 and 2009.
The
hydrographic data suggest coastal upwelling off the Subei and Lunan coasts and significant northeastward extension of fresh water from the Sheyang River mouth. Movement of surface drifters have suggested that the Subei coastal current flows northward in summer under the southerly wind forcing, opposite to the traditional understanding of the regional circulation in existing publications. The northward currents explain the drifting of the green algal patches from offshore of Subei to the Lunan coast in the summers of 2008 and 2009.
The drifter trajectories have also
shown that, during periods of no winds, the residual currents of the tide are small in this area and that a significant northward intrusion along the submarine valley off the Changjiang mouth persist, suggesting remote forcing dynamics of the current. Results of numerical modeling have shown that, associated with the northward wind-driven currents, onshore bottom Ekman transport and coastal upwelling is
20
generated along the Subei coast in summer.
At about 34ºN, the coastal upwelling is
shut-down by the buoyancy from the solar heating and fresh water input from the Sheyang River.
Coastal upwelling is observed to develop off the coast of Qingdao
city sometimes in summer, due to forcing of the along-shore component of the southerly and southwesterly winds. The fresh water input from the Sheyang River is found to have significant impact on the circulation in the southwestern Yellow Sea.
The low salinity water from the
Sheyang River is observed to extend northeastward over a large distance into the Lunan trough.
It is hypothesized that the wind-driven upwelling and the fresh water
plume of the Sheyang River play an important role in the nutrient dynamics and ecological blooms of Ulva Prolifera in summer in this area, which lead to the green tide events in July through August off the southern coast of the Shandong Peninsula.
Acknowledgment The authors thank the open cruises of the IOCAS and the hard work of the crews and scientists onboard of “KeXue-3”, “KeXue-1”, and “Beidou” research vessels. This work is support by NSFC (41176019, 41421005, U1406401), the National Basic Research Program of China (Project 2012CB956001), CMA (GYHY201306018), CAS (XDA11010301), and SOA (GASI-03-01-01-05).
References 1.
Clarke, A. J., and K. H. Brink, 1985. The response of stratified, frictional flow of shelf and slope waters to fluctuating large-scale, low-frequency wind forcing. J. Phys. Oceanogr., 15, 439–453.
2.
da Silva, A., C.Young and S. Levitus, ,1994. Atlas of Surface Marine Data 1994,
21
vol. 1, Algorithms and Procedures, NOAA Atlas NESDIS 6, U.S.Dep. of Commer., Washington, D.C. 3.
Guan, B., 1994. Patterns and structures of the currents in Bohai, Huanghai and East China Seas. Oceanology of China Seas. 11:1126–1134.
4.
He, L., 2011. Ocean circulation in the western Yellow and East China Seas and its impact on phytoplankton ecosystem dynamics. Postdoc Res. Rep., Institute of Oceanology, CAS, pp94.
5.
Hu, C., and M. He, 2008. Origin and offshore extent of floating algae in Olympic sailing area. EOS, 89(33):302–303.
6. Hu, D., 1994. Some striking features of circulation in Huanghai Sea and East China Sea. Oceanology of China Seas, 1. Netherlands: Kluwer Acadamic Publishers, 27–28. 7.
Kalnay et al., 1996. The NCEP/NCAR 40-year reanalysis project, Bull. Amer. Meteor. Soc., 77, 437-470.
8.
Lee, C., Rosati A., Spelman, J, 2006. Barotropic tidal mixing effects in a coupled climate model: Oceanic conditions in the Northern Atlantic, Ocean Modelling,11:464–477.
9.
Levitus, S., 1982. Climatological Atlas of the World Ocean, NOAA Prof. Pap. 13, 173 pp. plus 17 microfiche, U.S. Govt. Print. Off., Washington,D. C.
10. Li, Y., 2010. Structure and Dynamics of the Subtidal Circulation in the Southwestern Yellow Sea, PhD Thesis, Institute of Oceanology, Chinese Academy of Sciences. 11. Lü, X., F. Qiao, C. Xia, G. Wang, and Y. Yuan, 2010. Upwelling and surface cold patches in the Yellow Sea in summer: Effects of tidal mixing on the vertical circulation, Cont. Shelf Res., 30, 620-632, doi:10.1016/j.csr.2009.09.002.
22
12. Qiao, F. L., D. Y. Ma, M. Y. Zhu, R. X. Li, J. Y. Zang, and H. J. Yu, 2008. The Background of the green tide events (Ulva Prolifero bloom) off Qingdao 2008 and scientific response. Advances in Marine Science, 3, 409–410. (in Chinese with English abstract) 13. Su, J. L., 1998. Circulation dynamics of the China Seas: north of 18°N. In: "The Sea", Vol. 11, The Global coastal Ocean: Regional Studies and Syntheses, eds. A. R. Robinson and K. Brink, John Wiley, 483–506. 14. Wang B., 2010. The Influence of Topography on the Circulation in the Southwestern Yellow Sea, Master Thesis, Institute of Oceanology, Chinese Academy of Sciences. 15. Wang B., Li Y., Yuan D. L., 2013. Effects of topography on the sub-tidal circulation in the southwestern Huanghai Sea (Yellow Sea) in summer. Acta Oceanologica Sinica, 32(3): 1–9. 16. Wang B., N. Hirose, B. S. Kang, and K. Takayama, 2014. Seasonal migration of the Yellow Sea Bottom Cold Water. J. Geophys. Res. Oceans, 119, 4430–4443, doi:10.1002/2014JC009873. 17. Wu, H., J. Shen, J. Zhu, J. Zhang, and L. Li, 2014. Characteristics of the Changjiang plume and its extension along the Jiangsu Coast.
Cont. Shelf. Res.,
76, 108-123. 18. Xia, C., F. Qiao, M. Zhang, Y. Yang, and Y. Yuan, 2004a. Simulation of double cold cores of the 35ºN section in the Yellow Sea with a wave-tide-circulation coupled model, Chin. J. Oceanol. Limnol, 22, 292–298. 19. Xia, C., F. Qiao, Q. Zhang, and Y. Yuan, 2004b. Numerical modeling of the quasi-global ocean circulation based on POM, J. Hydrodyn., Ser. B,16:537–543. 20. Xia, C. S., F. L. Qiao, Y. Z. Yang, J. Ma, and Y. L. Yuan, 2006. Three
23
dimensional structure of the summertime circulation in the Yellow Sea from a wave-tide-circulation coupled model. J. Geophys. Res., 111, C11S03. Doi: 10.1029/2005JC003218. 21. Yuan, D., 2002. A numerical study of barotropicly forced intrusion in DeSoto Canyon. J. Geophys. Res. 107 (C2), doi:10.1029/2001JC000793. 22. Yuan, D., 2006. Dynamics of the Cold Water Events off the Southeast Coast of the United States in the summer of 2003. J. Phys. Oceanogr. , 36, 19121927. 23. Yuan, D., J. Zhu, C. Li, and D. Hu, 2008. Cross-shelf circulation in the Yellow and East China Seas indicated by MODIS satellite observations. J. Mar. Sys., 70 (1–2), 134–149. 24. Yuan, D., Y. Li, F. Qiao, and W. Zhao, 2012. Temperature inversion in the Yellow Sea Bottom Cold Water in summer. 25. Zhao, B. R., Q. Xiong, and F. G. Zhang, 1986. The internal hydrographic structure of the Huanghai Cold Water Mass (HCWM) in summer. Chin. J. Oceanol. Limnol., 4(1), 27–40. 26. Zhao, B. R., R. Limeberner, and D. Hu, 1991. Oceanographic Characteristics of the Southern Yellow Sea and the Northern East China Sea in summer. Oceanologia Et Limnologia Sinica, 22(2), 132–139. (in Chinese with English Abstract)
24
Figure Caption Figure 1.
Topography of the southwestern Yellow Sea. Unit of the depth is meter.
Dot marks the mooring station. Rectangle delineates the domain of the nested model. Figure 2.
Hydrographic survey stations in the southwestern Yellow Sea on July
22-August 3, 2008 (a), July 9-13, 2008 (b), August 9-11, 2008 (c), and May 28-June 21, 2009 (d), respectively. Figure 3.
Horizontal distribution of temperature (a, c) and salinity (b, d) at the
surface and 40 m depth, respectively, during July 22-August 3, 2008. Temperature unit is °C, salinity unit is psu. Figure 4.
Temperature (left column) and salinity (right column) distributions in
vertical sections along 32ºN, 33ºN, 34ºN, 35ºN, and 36ºN, respectively, in the summer of 2008. Figure 5.
Temperature unit is °C, salinity unit is psu.
Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20
m depth, respectively, in the southwestern Yellow Sea during the July 9-13, 2008 survey. Temperature unit is °C, salinity unit is psu. Figure 6.
Temperature (a) and salinity (b) distributions in vertical sections along 35º
N during July 9-13, 2008. Temperature unit is °C, salinity unit is psu. Figure 7.
Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20
m depth, respectively, in the southwestern Yellow Sea during the August 9-11, 2008 survey. Temperature unit is °C, salinity unit is psu. Figure 8.
Temperature (a) and salinity (b) distributions in the 35.5ºN vertical
sections on August 9-11, 2008. Temperature unit is °C, salinity unit is psu. Figure 9.
Temperature (a, c) and salinity (b, d) distributions at the sea surface and 30
m depth, respectively, in the western Yellow Sea during May 28-June 21, 2009 25
survey. Temperature unit is °C, salinity unit is psu. Figure 10. Temperature (a, c) and salinity (b, d) distributions along the vertical sections of 29ºN and 34ºN during May 28-June 21, 2009 survey. Temperature unit is °C, salinity unit is psu. Figure 11. Trajectories of satellite-tracked drifters in the southwestern Yellow Sea during May 28 through July31 of 2009. Figure 12. Trajectories of satellite-tracked surface drifters in the southwestern Yellow Sea after August 1, 2009. Figure 13. Integrated progressive vector plot of the NCEP wind stress at (35ºN, 121º E) in the summers of 2008 (a) and 2009 (b). Unit is Pa s. Figure 14. Comparison of daily averaged meridional wind stress from NCAR/NCEP reanalysis (top panel) and the meridional velocity of the ocean currents measured by the satellite-tracked surface drifters (low panels). Figure 15. Simulated surface drifter trajectories during May 1 through July 10 of 2009 based on the currents at the 5 m depth. Figure 16. Simulated average surface temperature (a) and salinity (b) distributions in July 2008. Temperature unit is °C, salinity unit is psu. Figure 17. Simulated surface temperature distribution forced by steady southwest (a) and southeast (b) winds. Temperature unit is °C. Figure 18. Simulated surface temperature distribution with (a) and without (b) the river runoff. Temperature unit is °C.
26
Highlights
This study provide solid evidence that the Subei coastal currents flows northward in summer, using the Lagrangian trajectories of satellite-tracked ARGOS surface drifters in the summers of 2008 and 2009. The wind-driven upwelling off the Subei coasts and the buoyancy shut-down by the Sheyang River fresh water plume is revealed for the first time in history. Coastal upwelling is also observed off the coast of Qingdao city in middle July through early August of 2008, which is forced by the along-shore component of the southerly and southwesterly winds. The Qingdao upwelling in summer is reported for the first time in history. A northward intrusion along the submarine valley off the Changjiang mouth, which is persistently strong and independent of the winds in summer, is indicated by the movement of the drifters for the first time in history.
27
Figure 1. Topography of the southwestern Yellow Sea. Unit of the depth is meter. Dot marks the mooring station. Rectangle delineates the domain of the nested model.
Figure 2. Hydrographic survey stations in the southwestern Yellow Sea on Julyne
22-August 328,2008(a), July 9-13, 2008(b), August 9-117, 2008(c), and May 28-June 21, 2009(d), respectively.
Figure 3. Horizontal distribution of temperature (a, c) and salinity (b, d) at the surface and 40 m depth, respectively, during July 22-28, 2008. Temperature unit is °C, salinity unit is psu.゜C
Figure 4. Temperature (left column) and salinity (right column) distributions in vertical sections along 32ºN, 33ºN, 34ºN, 35ºN, and 36ºN, respectively, in the summer of 2008. Temperature unit is °C, salinity unit is psu.
Figure 5. Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20 m depth, respectively, in the southwestern Yellow Sea during the July 9, 2008 survey. Temperature unit is °C, salinity unit is psu.
Figure 6. Temperature (a) and salinity (b) distributions in vertical sections along 35º N on July 9, 2008. Temperature unit is °C, salinity unit is psu.
Figure 7. Temperature (a, c) and salinity (b, d) distributions at the sea surface and 20 m depth, respectively, in the southwestern Yellow Sea during the August 9, 2008 survey. Temperature unit is °C, salinity unit is psu.
Figure 8. Temperature (a) and salinity (b) distributions in the 35.5ºN vertical sections on August 9, 2008. Temperature unit is °C, salinity unit is psu.
Figure 9. Temperature (a, c) and salinity (b, d) distributions at the sea surface and 30 m depth, respectively, in the western Yellow Sea during May 28-June 21, 2009 survey. Temperature unit is °C, salinity unit is psu.
Figure 10. Temperature (a, c) and salinity (b, d) distributions along the vertical sections of 29ºN and 34ºN during May 28-June 21, 2009 survey. Temperature unit is °C, salinity unit is psu.
Figure 11. Trajectories of satellite-tracked surface drifters in the southwestern Yellow Sea during May 28 through July31 of 2009.゜
Figure 12. Trajectories of satellite-tracked surface drifters in the southwestern Yellow Sea after August 1, 2009.
Figure 13. Integrated progressive vector plot of the NCEP wind stress at (35ºN, 121º E) in the summers of 2008 (a) and 2009 (b). Unit is Pa s.
Figure 14. Comparison of daily averaged meridional wind stress from NCAR/NCEP reanalysis (top panel) and the meridional velocity of the ocean currents measured by the satellite-tracked surface drifters (low panels).
Figure 15. Simulated surface drifter trajectories during May 1 through July 10 of 2009 based on the currents at the 5 m depth.
Figure 16. Simulated average surface temperature (a) and salinity (b) distributions in July 2008. Temperature unit is °C, salinity unit is psu.
Figure 17. Simulated surface temperature distribution forced by steady southwestly (a) and southeastly (b) winds. Temperature unit is °C.
Figure 18. Simulated surface temperature distribution with (a) and without (b) the river runoff. Temperature unit is °C.