-
Contents lists available at ScienceDirect
Continental Shelf Research
journal homepage: www.elsevier.com/locate/csr
Double fronts in the Yellow Sea in summertime identified using
sea surfacetemperature data of multi-scale ultra-high resolution
analysis
Lei Lina,⁎, Dongyan Liua, Chongxin Luob, Lian Xiec
a State Key Laboratory of Estuarine and Coastal Research, East
China Normal University, No. 500 Dongchuan Road, Shanghai 200241,
Chinab Key Laboratory of Marine Environment and Ecology, Ministry
of Education, Ocean University of China, Qingdao, Chinac Department
of Marine, Earth and Atmospheric Sciences, North Carolina State
University, Raleigh, NC, USA
A R T I C L E I N F O
Keywords:Yellow SeaThermal frontRemote sensingSSTTidal
mixingTopography
A B S T R A C T
Fronts are ubiquitous phenomena in oceans, and they play a
significant role in marine hydrodynamics andecology. During the
stratified season of a shelf sea, the coastal front is usually
considered as a single front, i.e.,the tidal mixing front. However,
using high resolution (~1 km) Multi-scale Ultra-high Resolution
(MUR) analysissea surface temperature (SST) data, this study
observes persistent double fronts along the Yellow Sea coast
insummertime. The double fronts comprise the well-known offshore
tidal mixing front and a nearshore front, andthe nearshore front
has not been previously reported. The climatological (2002–2017)
monthly mean resultshows that the double fronts with two SST
gradient peaks exceeding ~2 °C/100 km and opposite SST
gradientdirections basically remain unchanged from June to August,
whereas the frontal spacing decreases in September.Analyses based
on a two-layer concept model suggest that a topographic slope along
with tidal mixing couldinduce the pattern of double fronts. The
frontogenesis of nearshore thermal front could be associated with
thedifferent responses of the water column of different water
depths to insolation. The offshore movement of thenearshore front
in September could be related to the fast cooling of nearshore
water and intensified offshorewind, and the topographic slope is
important for determining the pattern of double fronts (loose or
tight). Thisstudy shows a new pattern of coastal fronts in the
stratified season, and indicates the significance of high
re-solution satellite data. The discovery of the double front
pattern implies the influence of coastal fronts during
thestratified season on marine ecology and environment in a shelf
sea might be underestimated.
1. Introduction
Oceanic fronts defined as narrow zones of intensified
horizontalgradients of water properties are common and important
marine phe-nomena (Joyce, 1983; Belkin et al., 2009; Acha et al.,
2015). Frontscould act as dynamic barriers to momentum and offshore
water trans-port (Belkin et al., 2009; Chen et al., 2002). Fronts
could increaseprimary production by fertilizing the surface
euphotic zone (Chen et al.,2003), could induce phytoplankton bloom,
and may develop into redtides (Pingree et al., 1975). A front as a
convergence zone could alsoaffect the distribution of zooplankton,
larva, fish, and even sea birds(Durazo et al., 1998; Lee et al.,
2005; Lough and Manning, 2001;Schultes et al., 2013; Woodson and
Litvin, 2015). To explore thestructure and variation of fronts is
significant for marine hydro-dynamics and ecology.
The Yellow Sea, which has a mean depth of ~44m, an area
of~380,000 km2 and is located at the north of the East China Sea
and
surrounded by China and the Korean Peninsula, is part of the
westernPacific marginal sea (Fig. 1). The Yellow Sea, which has
high pro-ductivity, has important fishing grounds for surrounding
countries (Jinet al., 2003). Coastal fronts play an important role
in Yellow Seaecology and fisheries (e.g., Huang et al., 2010; Wei
et al., 2003; Liuet al., 2003).
With the advances in satellites and numerical models, patterns
ofYellow Sea fronts that were difficult to observe from in situ
data havebeen investigated and gradually recognized. In wintertime,
prevailingsouthern cool currents forced by the northerly East Asian
monsoonwind occur along the coasts of both China and the Korean
Peninsula,while a compensated warm current (i.e., the Yellow Sea
Warm Current)flows north basically along the western flank of the
central trough inthe Yellow Sea (Su, 2001; Huang et al., 2005).
Strong sea surfacecooling and the topography function together with
the shear betweenthe cool coastal current and Yellow Sea warm
current induced strongSST coastal fronts (SST gradient was ~3–8
°C/100 km) along the
https://doi.org/10.1016/j.csr.2019.02.004Received 19 June 2018;
Received in revised form 24 December 2018; Accepted 10 February
2019
⁎ Corresponding author.E-mail address: [email protected]
(L. Lin).
Continental Shelf Research 175 (2019) 76–86
Available online 11 February 20190278-4343/ © 2019 Elsevier Ltd.
All rights reserved.
T
http://www.sciencedirect.com/science/journal/02784343https://www.elsevier.com/locate/csrhttps://doi.org/10.1016/j.csr.2019.02.004https://doi.org/10.1016/j.csr.2019.02.004mailto:[email protected]://doi.org/10.1016/j.csr.2019.02.004http://crossmark.crossref.org/dialog/?doi=10.1016/j.csr.2019.02.004&domain=pdf
-
Shandong, Jiangsu, and Korean coasts (Hickox et al., 2000;
Huanget al., 2010; Wang et al., 2012). From spring to summer, the
Yellow SeaCool Water Mass at the bottom of the central Yellow Sea
graduallyforms, and the intensified sea surface heating induces the
strong stra-tification in summertime (Zhang et al., 2008). Due to
the high tidaldissipation, a tidal mixing front that separates the
onshore well mixedwater and offshore stratified water occurs along
the Yellow Sea coast.Zhao (1985, 1987) investigated the location of
the tidal front in theYellow Sea by using the Simpson–Hunter index
k (Simpson Amp andHunter, 1974), and their results suggested that
the tidal front occurredbasically at k=1.8–2.0 and roughly along
the 40m isobath. Lie (1989)confirmed the tidal front in the
southeastern Yellow Sea via an in situCTD survey. Hickox et al.
(2000) used 12 years of Pathfinder SST datawith ~9 km resolution to
detect the climatology and seasonal varia-bility of Yellow Sea
fronts and found that fronts were weak in summercompared to winter,
which was due to the surface solar heatingwarming the sea surface
and smoothing out the SST filed in summer(Hickox et al., 2000;
Huang et al., 2010). The results from an oceannumerical model
suggested that the weak surface thermal front insummer corresponded
to a strong subsurface thermal front (Park andChu, 2006). Based on
simulations of a numerical model with horizontal1/12° resolution
(~9 km), Liu et al. (2003) and Lü et al. (2010) in-vestigated the
current structure and the coastal cold water belts inducedby Yellow
Sea tidal mixing fronts. By using eight-years cloud-pene-trating
microwave imager data (6–22 km resolution) from the TropicalRain
Measuring Mission (TRMM) satellite, Huang et al. (2010) identi-fied
seasonal SST fronts in the Yellow Sea and indicated that the
frontsin the stratified months were expressed as tidal fronts
(monthly meanSST gradient was ~2–4 °C/100 km). However, there was a
blind zone of~25 km at the coast in TRMM results due to the
inherent defect ofmicrowave measurements.
Previous studies using satellite or model data of relatively
coarseresolution (≥9 km) have provided a general pattern of the
tidal mixingfront in the Yellow Sea during summertime. However, the
finer frontal
structure in the Yellow Sea, especially near the coast, is not
clear due tothe limitation of data spatial resolution. In this
study, using high re-solution (~1 km) Multi-scale Ultra-high
Resolution SST analysis data(Chin et al., 2017), we explored the
finer features of fronts in the YellowSea. Interestingly, we found
a distinct but not previously reportednearshore front in summertime
that occurred along with the well-known tidal mixing front, forming
a double front pattern in the YellowSea summertime. The discovery
of the double fronts could imply adouble or intensified impact of
fronts on coastal hydrodynamics andecology during the stratified
season.
This paper is organized as follows. Section 2 introduces the
datasource and the frontal-detection method. The characteristics of
thedouble fronts are presented in Section 3. The mechanism of the
doublefronts formation and variation are discussed in Section 4.
Finally, abrief conclusion is given in Section 5.
2. Data and method
2.1. SST data source
In this study, Multi-scale Ultra-high Resolution (MUR) analysis
SSTproducts from the Jet Propulsion Laboratory (JPL) were used to
detectoceanic fronts. The MUR analysis SST data merged from
multiple sa-tellites and in situ data are a global field produced
daily on a0.01°× 0.01° grid with a resolution of ~1 km from 2002 to
2017 (Chinet al., 2013, 2017). According to Chin et al. (2017), the
global rootmean squares difference between MUR and the
Multi-product Ensembleanalysis data is about 0.36 °C, which is
comparable to other SST pro-ducts, e.g., Optimum Interpolation SST
analysis from NOAA NationalCenter for Environmental Information,
Operational Sea Surface Tem-perature and Sea Ice Analysis from the
UK Met Office, and CMC SSTanalysis from the Canadian Meteorological
Center. MUR SST data havebeen successfully used to diagnose oceanic
fronts (e.g., Vazquez-Cuervoet al., 2013, 2017) and used in
research on coastal jets, coastal up-welling and the warm SST
anomaly (e.g., Gentemann et al., 2017;Patricola and Chang, 2016;
Wiafe and Nyadjro, 2015). More informa-tion about MUR SST data can
be found at
http://podaac.jpl.nasa.gov/Multi-scale_Ultra-high_Resolution_MUR-SST.
In this study, the climato-logical (2002–2017) monthly mean SST
data created from the dailydata by NOAA ERDDAP
(http://coastwatch.pfeg.noaa.gov/erddap)were used for the frontal
detection.
In addition, to confirm the presence of the double fronts, the
cli-matological monthly mean SST of the Moderate Resolution
ImagingSpectroradiometer (MODIS, 2003–2017) and Advanced Very
HighResolution Radiometer (AVHRR, 1982–2012) data were used.
TheMODIS and AVHRR SST data were also download from NOAA ERDDAP.For
the analysis on the formation and variation of the double
fronts,climatological monthly mean QuickScat sea surface wind
data(2000–2009) provided by NOAA ERDDAP and net heat flux
data(1981–2010) provided by NOAA NCEP-NCAR Reanalysis 1
data(https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.derived.html)
were used.
2.2. Frontal detection
The SST front is usually characterized by a band with higher
SSTgradient magnitude than background. The gradient algorithm
devel-oped by Belkin and O'Reilly (2009) is used to diagnose the
positions andintensities of SST fronts. Based on the contextual
feature-preservingfilter and traditional gradient method, the
algorithm can efficientlyremove data noise and have a good
shape-preservation (Belkin andO'Reilly, 2009). This algorithm has
been widely used in studies ofoceanic fronts (e.g., Karimova, 2014;
Liu and Hou, 2012; Zeng et al.,2014). The front detection algorithm
is presented as below briefly.
1) Pre-processing the data using contextual median filter (MF)
until
Fig. 1. The topography of the Yellow Sea. The grey lines are the
water depthisobaths. The red dashed lines denote the four transects
for the double frontsshown in Fig. 4 (For interpretation of the
references to color in this figure le-gend, the reader is referred
to the web version of this article).
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
77
http://podaac.jpl.nasa.gov/Multi-scale_Ultra-high_Resolution_MUR-SSThttp://podaac.jpl.nasa.gov/Multi-scale_Ultra-high_Resolution_MUR-SSThttp://coastwatch.pfeg.noaa.gov/erddaphttps://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.derived.htmlhttps://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.derived.html
-
iterative MF convergence. In the process of filtering, the data
keepunchanged if the window center is a significant 5-point
extremum in5× 5 windows, otherwise if the window center is a spike
(ex-tremum in 3× 3 windows) the data will be filtered by a 2D
3×3median filter.
2) Calculating the gradient magnitude for fronts. The gradient
iscomputed by using the Sobel operator consisting of 3×3
kernels:
= ⎡
⎣⎢
− +− +− +
⎤
⎦⎥ =
⎡
⎣⎢
+ + +
− − −
⎤
⎦⎥G A G A
1 0 12 0 21 0 1
* ,1 2 1
0 0 01 2 1
*x y(1)
where A is the data from step 1), Gx, Gy are two images which
containapproximations for derivate in horizontal and vertical
directions, and *is the convolution sign. Then, the SST gradient
magnitude (GM) can becalculated as:
= +GM G Δx G Δy( / ) ( / )x y2 2 (2)
The above processes and calculations were all carried out by
usingMatlab in this study. Finally, the SST gradient in units of
°C/100 km canbe derived. The more detailed description of the
algorithm can be foundin Belkin and O'Reilly (2009).
To compare the cross-shore variation of SST in different
months,SST deviation was derived using the SST minus the transect
mean ofSST. And the SST data pre-processed by using contextual
median filterwas used to calculate the SST deviation.
3. Results
3.1. Location and monthly variation of the double front
The climatological monthly mean SST gradients indicate that
frontsin the Yellow Sea are obvious throughout the year and
relatively strongin wintertime (results not shown). The fronts
occur along the ShandongPeninsula, Jiangsu coast and Korean
Peninsula. The general pattern andseasonal variation of the fronts
are basically consistent with previousstudies (Zhao, 1987; Hickox
et al., 2000; Huang et al., 2010; Lie, 1989;Lü et al., 2010; Wang
et al., 2012). However, the results from June toSeptember show that
distinct and basically parallel double fronts occuralong the Yellow
Sea east coast (Fig. 2). The double fronts consist ofnearshore
fronts and offshore fronts, which are respectively indicatedby blue
and black arrows in Fig. 2. The nearshore fronts and the off-shore
fronts are roughly along the 10m and 40m isobaths, respectively.The
location of the offshore fronts agrees with the well-known
tidalmixing front in the Yellow Sea from previous studies (e.g.,
Zhao, 1987;Huang et al., 2010; Lü et al., 2010). The SST gradient
magnitudes ofboth the nearshore and offshore fronts are in
basically the same orderand are greater than 2 °C/100 km. The
double fronts in Seohan Bay (SB)and Kyunggi Bay (KB) are clearer
due to their large frontal spacing,whereas the double fronts along
the southern Liaodong Peninsula (SLP)and the southwestern Korean
Peninsula (SKP) have two relatively closefronts. The location of
the double fronts from June to August basicallyremains unchanged,
while in September, the nearshore fronts in SB andKB move offshore,
and the spacing of the two double fronts apparentlydecreases. The
SST gradient directions of the nearshore and offshorefronts are
basically opposite, i.e., landwards and seawards, respectively(Fig.
3). A double front is also present at the Jiangsu coast, though it
isnot very obvious due to the weak offshore SST front
(basically< 2 °C/100 km).
3.2. Loose and tight double fronts
To further analyze the characteristics of double fronts, four
transectsacross the clear double fronts at SB, KB, SLP and SKP are
selected(Transects I–IV in Fig. 1). The transects are selected
basically along theSST gradient direction and perpendicular to
isobaths. The variations inSST gradient magnitude, SST deviation,
together with water depth
along the transects are shown in Fig. 4. The detailed data on
themaximum SST gradients of the double fronts and corresponding
waterdepths and locations are presented in Table 1.
In the transect SST gradients shown in Fig. 4a-d, two
apparentgradient peaks greater than 2 °C/100 km indicate a double
front patternalong each transect. In general, the double front
patterns along Trans-ects I and II are similar, and both have large
frontal spacings, whereasthe frontal spacings along Transects III
and IV are relatively small.Depending on the frontal spacing, we
classify the double fronts as twotypes, namely, loose double fronts
(Transects I and II) and tight doublefronts (Transects III and
IV).
For the two loose double fronts from June to August, two SST
gra-dient peaks occur offshore at 30–40 km and 120–130 km,
respectively(Fig. 4a and b). The average spacing of the two
gradient peaks is~90 km. The water depths at the SST gradient peaks
of the nearshoreand offshore fronts are approximately 10m and 40m,
respectively(Table 1). In addition, the SST gradient magnitudes of
the fronts areslightly larger in August than in the other three
months. In September,the nearshore front has an obvious offshore
movement of ~30 km. Thewater depth at the maximum SST gradient of
the nearshore frontsdeepens to ~20m. However, the offshore tidal
mixing fronts basicallyremain unchanged. The spacing of two double
fronts declines by~30 km. The maximum SST gradient of both double
fronts in Sep-tember decrease by ~20% compared to August.
For the two tight double fronts, two SST gradient peaks with
smallerspacing (~40 km) than that of the loose double fronts occur
offshore at10–20 km and 50–60 km, respectively (Fig. 4c and d). For
Transect III,the location of the double fronts basically remains
unchanged fromJune to September, whereas the SST gradient of double
fronts is rela-tively strong in August. In September, the SST
gradient of the nearshorefront decreases, while the double front
spacing experiences no obviouschange. For Transect IV, the pattern
of the double fronts basically re-mains unchanged from June to
August. In September, the double frontsmove offshore obviously,
especially for the offshore front. This may bedue to Transect IV
being close to the East China sea and the influence ofthe shelf
circulation from the south.
To clearly show the frontal SST gradient formation, the SST
devia-tion along the transects was calculated from the SST minus
the transectmean of the SST. As shown in Fig. 4e-h, the SST
deviations are basicallythe same from June to August. Offshore SST
cold anomalies exist be-tween the nearshore and the offshore
fronts. In addition, the landwardand seaward increases in SST
correspond to the nearshore and offshorefronts, respectively, which
could induce the opposite directions of SSTgradient of the double
fronts. In September, the SST cold anomalies ofTransects I, II, and
IV move offshore, which is consistent with the off-shore movement
of fronts.
Both transects with loose double fronts show a gentle
topographicslope with a slope ratio of ~0.35‰ (Fig. 4e and f),
while the topo-graphy is much steeper for the tight double front
transects (Fig. 4g andh). This indicates that the topographic slope
may play an important rolein the double frontal spacing.
4. Discussion
4.1. Providing confidence in the presence of the double
fronts
The double fronts in the Yellow Sea are observed in the
monthlymean MUR satellite analysis data. The double or multiple
fronts patternhas been found in the Zhe-Min coastal front zone
(induced by themeeting between the China eastern coastal cold
current and Taiwanwarm current) in winter (He et al., 2015) and the
Patagonian shelfbreak front zone (Franco et al., 2008). However, to
our knowledge, thedouble front pattern occurring along the tidal
mixing front during thestratified season in a shelf sea has not
been reported.
Although the MUR data accuracy in the Yellow Sea has not
beenevaluated by in situ data, several aspects could provide
confidence in
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
78
-
the presence of the double fronts. First, some studies have
suggestedfronts can be effectively detected based on gradient
algorithms whenapplied on monthly mean satellite SSTs (e.g., Ullman
and Cornillon,1999; Franco et al., 2008). Second, offshore cold SST
anomalies withinthe double fronts induced the landwards and
seawards SST gradients ofthe double fronts. The locations and
timings of cold SST belts from MUR(Fig. 5) are consistent with the
in situ observations in the Ocean atlas(Chen et al., 1992) and
model results (Lü et al., 2010). Third, the doublefronts are also
detected by other satellite data sets, including for ex-ample,
those of the MODIS and AVHRR (see Fig. 6). The MODIS andAVHRR
results basically agree with the occurrences, locations andtimings
of the double fronts in the Yellow Sea, although the intensityhave
slight differences and the tight double fronts are not very clear
dueto the relatively coarse spatial resolution (~4 km) of MODIS
andAVHRR data.
The data with relatively coarse resolution (≥9 km) used by
previousstudies on Yellow Sea fronts missed the nearshore front in
summertime,thus could not detect the double fronts. The impact of
the data re-solution on the frontal detection is investigated by
averaging the ~1 kmresolution MUR data into the coarse resolutions
of 9 km and 25 km. Asshown in Fig. 7, the coarse resolution data
can detect the offshore tidalmixing front but not the nearshore
front. Therefore, the fine spatialresolution (at least 4 km) data
is necessary for detecting the doublefront.
4.2. The formation and monthly variation of the double
fronts
The offshore SST cold anomaly, which induced the SST gradient
ofboth sides and the opposite gradient direction, was the key
factor in theformation of the double fronts. Lü et al. (2010)
noticed the SST anomalybeside the tidal mixing front in the Yellow
Sea and suggested it wasinduced by the upwelling from the tidal
mixing front. However, thethermal balance analysis from an ocean
model by Ren et al. (2014)showed that the vertical diffusion and
net radiation flux were dominantterms and were roughly balanced
near the tidal mixing front in theYellow Sea. This suggests that
the tidal mixing could predominatelyinduce the offshore SST cold
anomaly, while the upwelling could besecondary. Therefore, in this
section, we offer an explanation of theformation of the offshore
cold SST anomaly based on a two-layer con-cept model with only
slope topography and vertical mixing, and thendiscuss the formation
and variation of the double fronts. According tothe model results
of Lü et al. (2010), tide plays a dominant role invertical mixing
in the nearshore water. Thus, it is assumed that thevertical mixing
is induced only by tide motion in the concept model.And the concept
model did not contain the horizontal diffusion whichmight be
induced by some hydrodynamic processes, for instance,
tidalhorizontal dispersion, horizontal current shears, frontal
instabilities,and intra-tidal and fortnight tidal variations.
In summertime, the strong stratification occurs in the Yellow
Sea
Fig. 2. Climatological monthly means of SST gradient magnitude
(°C/100 km) in the Yellow Sea from June to September. The blue and
black arrows indicate thenearshore and offshore fronts of double
fronts, respectively. The nearshore and offshore black lines are
the 10m and 40m isobaths, respectively (For interpretation ofthe
references to color in this figure legend, the reader is referred
to the web version of this article).
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
79
-
due to the strong surface heating (Fig. 8). Assuming uniform
surfaceheating and no tidal mixing, the summer Yellow Sea could be
idealizedas two layers, i.e., the warm upper mixed layer and the
cold bottomlayer (Fig. 9a). The temperatures of the upper and
bottom layers are T1and T2, respectively. In the Yellow Sea east,
the water depth variesbasically linearly along the double front
transect (see Fig. 4e-h).Therefore, an idealized linear topography
is used in the concept model.In addition, SST in nearshore waters
(offshore distance< 20 km) wasbasically uniform according to the
observation (Fig. 4e-f), which mightbe induced by the strong tidal
dispersion in the shallow waters.Therefore, the temperature in
nearshore waters shallower than themixed layer depth (i.e., x <
X0) is assumed to be uniform and un-changed during the tidal
mixing.
The offshore tidal mixing front separating the well-mixed
waterwith stratified water is basically along the 40m isobaths in
the YellowSea (Fig. 2). Therefore, the water shallower than 40m
becomes verti-cally well-mixed after adding the tidal mixing as
shown in Fig. 9b-d. Invertically well-mixed water, the thermal
inertia of a water column isproportional to the bottom depth, which
determines the temperaturechange rate of the water column (e.g.,
Xie et al., 2002). Deep waterwarms more slowly than nearshore
shallow water. Due to the differentresponses of the water column to
insolation at different depths, ap-parent offshore SST cold
anomalies occur between the nearshore andoffshore warm waters after
tidal mixing, which is consistent with thetransect SST pattern in
Fig. 4e-h. Then, two strong SST gradients nearX0 and X1 correspond
to the nearshore front and the offshore tidalmixing front,
respectively, whose location be basically consistent withthe
satellite result.
The temperature structure near X1 is a typical and well-known
tidalmixing front (e.g., Heijst, 1985; Acha et al., 2015; Krause et
al., 1986).
Therefore, we only focus on the nearshore front. As shown in
Fig. 9aand b, the temperature of the well-mixed water becomes
verticallyuniform and equals the vertical temperature average, as
below.
=+ ⋅ −+ ⋅ −
∈T x T D T SR xD SR x
x( ) ( X )( X )
, [X , X ]MLML
1 2 0
00 1
(3)
where x is the offshore distance, DML denotes the upper mixed
layerdepth, SR denotes the slope ratio, X0 is the offshore distance
of theintersection of the interface and the sea floor, X1 is the
offshore distanceof the 40m water depth, and T1 and T2 are the
temperatures in theupper layer and lower layer, respectively. Then,
the horizontal gradientof the temperature (gT) can be derived by
the partial derivative of T(x)with respect to x as below.
=∂
∂=
−+ ⋅ −
∈gT x T xx
D SR T TD SR x
x( ) ( ) ( )[ ( X )]
, (X , X )MLML
2 1
02 0 1 (4)
According to Expression (4), |gT| increases landwards and
deceasesseawards. The total heat content of the bottom water is
basically linearincrease with increasing distance from the coast.
Due to mixing with awater body of increasing depth, the effect of
the cold bottom water onthe heat content of the entire water column
is decreasing seawards.Therefore, the temperature gradient is
greater nearshore than offshore.When →x X0 (i.e., the location
where the strong nearshore front oc-curs), the magnitude of SST
gradient reached a maximum and is ex-pressed as below.
=−
+gTSR T T
D(X ) ( )
ML0
2 1
(5)
This expression suggests that the intensity of the nearshore
front isrelated to mixed layer depth, the slope ratio, and the
temperature
Fig. 3. Climatological monthly mean of SST gradient vectors in
the Yellow sea from June to September. Only vectors with gradient
magnitudes greater than 1 °C/100 km are shown.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
80
-
difference between the upper and lower layers. According to the
ob-servation by Zhao (1989), −T T2 1 is approximately 10 °C in
summer-time. The SR is ~0.35‰ according to the topography of the
loosedouble fronts (Fig. 4e and f). For the case from June to
August, DML is~10m (Fig. 8). Then, gT (X )0 is ~35 °C/100 km for
the nearshore front,
according to Expression (5). This indicates that the SST
gradient in-duced by the slope topography and tidal vertical mixing
is enough forthe formation of the nearshore front of double fronts
(3–4 °C/100 km).Therefore, results of the concept model suggest
that a slope topographyalong with tidal mixing could induce the SST
double front pattern. In
Fig. 4. The climatological monthly (June to September) mean SST
gradient magnitudes, temperature deviations, and water depths along
Transects I–IV in Fig. 1(a-d)show the SST gradients for the four
months along Transects I–IV, respectively. The blue and black
arrows roughly indicate the nearshore and the offshore fronts.
(e-h)show the SST deviations for the four months along Transects I-
IV, respectively. Black dotted lines denote the variations in water
depth along transects (Forinterpretation of the references to color
in this figure legend, the reader is referred to the web version of
this article).
Table 1The maximum SST gradients of the double fronts and
corresponding water depths and locations for the four
transects.
Transect Month Maximum SST gradient (°C/100 km) Water depth (m)
Location (offshore distance, km) Frontal spacing (km)
Front a Front b Front a Front b Front a Front b
I Jun. 3.6 2.3 5 46 31 128 97Jul. 3.8 2.8 6 45 37 125 88Aug. 4.3
3.3 10 44 42 121 79Sep. 3.5 2.8 23 44 70 124 54
II Jun. 2.8 2.6 11 48 40 144 104Jul. 2.1 2.3 11 43 41 126 85Aug.
2.8 2.7 11 46 41 114 73Sep. 2.1 2.4 22 43 78 122 44
III Jun. 1.7 1.8 11 44 9 60 51Jul. 2.4 2.8 11 31 10 51 41Aug.
3.8 4.8 12 32 7 48 41Sep. 2.0 5.2 11 36 9 54 45
IV Jun. 2.8 2.2 7 55 26 70 44Jul. 3.3 2.1 7 20 24 49 25Aug. 3.7
2.4 6 20 20 48 28Sep. 3.4 2.3 17 60 34 85 51
Note. Fronts a and b denote the nearshore and the offshore
fronts of the double fronts, respectively.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
81
-
reality, the monthly mean SST gradient of the nearshore front
fromMUR data is much less than that of the concept model, which may
besmoothed by strong surface heating and the ignored horizontal
diffu-sion. In addition, −T T2 1 has a maximum in August (Fig. 8),
which couldexplain the relatively strong double fronts in
August.
For the case in September, due to DML deepening and −T T2 1
de-creasing (Fig. 8), the SST gradient of double fronts will
diminish basedon Expression (5). The offshore movement of the
nearshore front inSeptember may be related to the drastically
reduced surface heat fluxand the intensified offshore wind. The
monthly mean of net surface heatflux in September is almost zero
and much less than that in June-August(Fig. 10a). There must be
heat loss for the Yellow Sea in the second halfof September. The
temperature at nearshore waters should go downfaster due to less
water depth (i.e., less heat capacity) and higher SST atnearshore
waters. The faster reduction in temperature of nearshorewaters
could make nearshore temperature approaching that of the ad-jacent
offshore water. In addition, the wind from June to August isweak
and is southeasterly or easterly basically. In September, the
windspeed increases, and the wind turns to be northeasterly (Fig.
10b),which could induce an intensified offshore transport along the
YellowSea east coast. The offshore transport could also be helpful
for the ex-change of the nearshore waters. As shown in Fig. 11,
both of the fastcooling of the nearshore water and the intensified
offshore wind couldfacilitate the offshore movement of X0, i.e.,
the offshore movement ofthe nearshore front in September.
For the tight double fronts case, the topographic slope is
obviouslysteeper than that of the loose double fronts case (Fig.
4g-h). Therefore,SR is set to 0.7‰ in the concept model (Fig. 9d).
As indicated by pre-vious studies (e.g., Zhao, 1987), the energy of
tidal mixing can basicallymaintain the well-mixing of water
depth< 40m. That is consistentwith the observation in this study
that the offshore fronts basically
occurred along the 40m isobath (Fig. 2). Therefore, X1 (i.e.,
the off-shore front) was fixed at the point of water depth =40m in
the conceptmodel. The offshore front for the transect with a steep
slope is closer toshore than that with a gentle slope, and thus is
closer to the nearshorefront (Fig. 9d). Therefore, the different
topographic slope could be ac-count for the different spacings
between the tight and loose doublefronts. According to Expression
(5), the SST gradient should have beenstronger than that of the
loose tight double fronts case. However, thereis no significantly
frontal intensification in satellite results for tightdouble fronts
(Fig. 4a-d). This could be due to the strong tidal disper-sion in
the nearshore region limiting the SST gradient, which was ig-nored
in the concept model.
4.3. Implications from the double fronts
During the stratified season, the coastal front in a shelf sea
is usuallyconsidered as a single front, i.e., the tidal mixing
front which is theboundary between stratified and well-mixed waters
(e.g., Holt andUmlauf, 2008; Lwiza et al., 1991; Pingree and
Griffiths, 1978; Schulteset al., 2013). Analyses suggest that the
strong tidal mixing together witha slope topography could induce
the double front pattern during thestratified season in the Yellow
Sea. Therefore, the double front may alsooccur at other seas with
the condition of the occurrence of a tidal frontand a suitable
slope topography. Whereas, the observation of thedouble fronts
needs using high spatial resolution data as the front isusually
close to the coastline. With the development of satellite
mea-surement, more double front cases and much finer structure of
oceanicfront would be detected.
Fronts play a significant role in marine ecology. They could
con-verge and enhance the marine biomass (Acha et al., 2015;
Woodson andLitvin, 2015; Choi et al., 2017), and could also act as
barriers to block
Fig. 5. The climatological monthly (June to September) mean SSTs
(°C) from MUR data.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
82
-
Fig. 6. The climatological monthly mean SST gradients from June
to September as in Fig. 2 but from MODIS (a-d) and AVHRR (e-h)
satellite data with ~4 kmresolution. The 2003–2017 and 1982–2012
SST data were used for the MODIS and AVHRR results,
respectively.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
83
-
Fig. 7. The climatological SST gradients in August and September
resulting from the spatial averaging of MUR SST data. (a-b) 9 km.
(c-d) 25 km.
Fig. 8. The observed water temperature profiles in the Yellow
sea summertime, redrawn after Zhao (1989). (a) shows the locations
of observed sites. (b-e) are theobserved results for sites B-E,
respectively.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
84
-
the material and organism at the coast (Woodson et al., 2012).
Thedouble fronts mean an augment of the frontal convergence zone
nearcoast, which may enhance the probability of coastal
phytoplanktonbloom and the risk of harmful algal blooms (e.g.,
Pingree et al., 1975).The double fronts may also induce double
fencing for the offshoretransport of coastal high nutrient or
pollutant in summertime, whichcould intensify nearshore
eutrophication and environmental pollutionand influence the
connectivity and the recruitment of coastal species(e.g., Woodson
et al., 2012). Therefore, the discovery of the doublefronts implies
the impact of fronts on marine ecology in the Yellow Seaduring the
stratification season might be greatly underestimated be-cause it
did not consider the existence of the nearshore front.
5. Conclusions
Double fronts are found in the Yellow Sea in summertime from
theclimatological monthly mean MUR satellite SST data with high
re-solution (~1 km). The double fronts comprise an offshore tidal
mixingfront and a nearshore front with the opposite SST gradient
direction.The double fronts are of loose frontal spacing (~90 km)
in SB and KB,whereas they are of tight frontal spacing (~40 km) in
SLP and SKP. Thedouble fronts basically remain stable in
summertime, while their
intensities and spacings decrease in SB and KB in September.
Thedouble fronts are also found from the MODIS and AVHRR SST
satellitedata. A two-layer concept model including tidal mixing and
slope to-pography could basically explain the formation and the
general patternof double fronts. The concept model suggests that
the increase in uppermixed layer depth in September could reduce
the intensity. The off-shore movement of the nearshore front in
September may be related tothe fast cooling of nearshore water and
intensified offshore wind, andthe slope is a key factor for the
pattern of double fronts (loose or tight).This study indicates the
necessity of using high resolution satellite datato reveal the fine
structure of oceanic fronts.
The double fronts in the Yellow Sea are observed from satellite
datain this study. However, field observations and modeling studies
areneeded to further confirm the underlying dynamic processes of
thedouble fronts and investigate their hydrodynamic
characteristics. Theecological and environmental effects of the
double fronts are also an-ticipated from future studies.
Fig. 9. Two-layer concept model for double fronts. (a) shows the
initial watertemperature distribution of idealized two-layer modes,
in which the tempera-tures in the upper layer and lower layer are
T1 and T2, respectively. The in-tersection of the interface and the
sea floor is at x= X0, and the 40m waterdepth is at x=X1. (b-d)
show the temperature distributions after tidal verticalmixing
derived from Eq. (3) for different cases. Only the water shallower
than40m is vertically mixed. The blue and black arrows indicate the
nearshore andthe offshore fronts of double fronts, respectively.
(b) and (c) are the cases withthe same slope ratio (SR) of 0.35‰
for idealizing Transects I or II and withupper mixed layer depths
(DML) of 10m and 20m, respectively. (d) is the casewith a steeper
slope (SR=0.70‰) for idealizing Transects III or IV. The intervalof
contours in (b-d) is (T1-T2)/10 (For interpretation of the
references to color inthis figure legend, the reader is referred to
the web version of this article).
Fig. 10. Climatological monthly mean net heat fluxes (a) and
mean sea surfacewinds (b) of the Yellow Sea from QuickScat
satellite data (2000–2009) andNOAA NCEP-NCAR reanalysis 1 data
(1981–2010), respectively.
Fig. 11. The schematic diagram for the offshore movement of the
nearshorefront in September. The darker color denote the higher
water temperature alongthe transect. The upper panel shows the
initial pattern of temperature, and thelower panel shows the
temperature pattern after heat loss and intensification ofoffshore
wind.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
85
-
Acknowledgments
This work was supported by the National Key FundamentalResearch
and Development Plan of China (Grant No. 2016YFC1402106-06) and the
National Natural Science Foundation of China (No.41706011). Lei Lin
was also supported by the China PostdoctoralScience Foundation
funded project and Key Laboratory of CoastalEnvironmental Processes
and Ecological Remediation, YICCAS (GrantNo.: 2016KFJJ02). The MUR
data were provided by JPL under supportby the NASA MEaSUREs
program. The authors thank NOAA ERDDAPfor the MODIS and AVHRR SST
data support (http://coastwatch.pfeg.noaa.gov/erddap). The authors
declare that the research was conductedin the absence of any
commercial or financial relationships that couldbe construed as a
potential conflict of interest. The data used can bedownloaded from
the given websites.
References
Acha, E.M., Piola, A., Iribarne, O., Mianzan, H., 2015.
Ecological Processes at MarineFronts: Oases in the Ocean. Springer,
Cham.
Belkin, I.M., Cornillon, P.C., Sherman, K., 2009. Fronts in
Large Marine Ecosystems. Prog.Oceanogr. 81 (1–4), 223–236.
Belkin, I.M., O'Reilly, J.E., 2009. An algorithm for oceanic
front detection in chlorophylland SST satellite imagery. J. Mar.
Syst. 78 (3), 319–326.
Chen, C., Zhu, J., Kang, K.R., Liu, H., Ralph, E., Green, S.A.,
Budd, J.W., 2002. Cross-frontal transport along the Keweenaw coast
in Lake Superior: a Lagrangian modelstudy. Dyn. Atmos. Oceans 36
(1), 83–102.
Chen, D., Liu, W.T., Tang, W., Wang, Z., 2003. Air‐sea
interaction at an oceanic front:implications for frontogenesis and
primary production. Geophys. Res. Lett. 30 (14),253–266.
Chen, et al., 1992. Marine Atlas of Bohai Sea, Yellow Sea, East
China Sea: Hydrology.China Ocean Press, Beijing.
Chin, T.M., Vazquez, J., Armstrong, E., 2013. A Multi-scale,
High-resolution Analysis ofGlobal Sea Surface Temperature. In:
Algorithm Theoretical Basis Document, Version1.3. Jet Propulsion
Laboratory, Pasadena, CA, pp. 13.
Chin, T.M., Vazquez-Cuervo, J., Armstrong, E.M., 2017. A
multi-scale high-resolutionanalysis of global sea surface
temperature ☆. Remote Sens. Environ. 200, 154–169.
Choi, B.J., Lee, J.A., Choi, J.S., Park, J.G., Lee, S.H., Yih,
W., 2017. Influence of the tidalfront on the three-dimensional
distribution of spring phytoplankton community inthe eastern Yellow
Sea. Chemosphere 173, 299–306.
Durazo, R., Harrison, N.M., Hill, A.E., 1998. Seabird
observations at a tidal mixing frontin the Irish Sea. Estuar.
Coast. Shelf Sci. 47 (2), 153–164.
Franco, B.C., Piola, A.R., Rivas, A.L., Baldoni, A., Pisoni,
J.P., 2008. Multiple thermalfronts near the Patagonian shelf break.
Geophys. Res. Lett. 35 (2), 344–360.
Gentemann, C.L., Fewings, M.R., García Reyes, M., 2017.
Satellite sea surface tempera-tures along the West Coast of the
United States during the 2014–2016 northeastPacific marine heat
wave. Geophys. Res. Lett. 44, 1.
He, S., Huang, D., Zeng, D., 2015. Double SST fronts observed
from MODIS data in theEast China Sea off the Zhejiang–Fujian coast,
China. J. Mar. Syst. 154.
Heijst, G.J.F.V., 1985. A geostrophic Adjustment model of a
tidal mixing front. J. Phys.Oceanogr. 15 (9), 1182–1190.
Hickox, R., Belkin, I., Cornillon, P., Shan, Z., 2000.
Climatology and seasonal variabilityof ocean fronts in the East
China, Yellow and Bohai seas from satellite SST data.Geophys. Res.
Lett. 27 (18), 2945–2948.
Holt, J., Umlauf, L., 2008. Modelling the tidal mixing fronts
and seasonal stratification ofthe Northwest European Continental
shelf. Cont. Shelf Res. 28 (7), 887–903.
Huang, D., Fan, X., Xu, D., Tong, Y., Su, J., 2005. Westward
shift of the Yellow Sea warmsalty tongue. Geophys. Res. Lett. 32
(24), 348–362.
Huang, D., Zhang, T., Zhou, F., 2010. Sea-surface temperature
fronts in the Yellow andEast China Seas from TRMM microwave imager
data. Deep Sea Res. Part II Top. Stud.Oceanogr. 57 (11–12),
1017–1024.
Jin, X., Xu, B., Tang, Q., 2003. Fish assemblage structure in
the East China Sea andsouthern Yellow Sea during autumn and spring.
J. Fish. Biol. 62 (5), 1194–1205.
Joyce, T.M., 1983. Varieties of ocean fronts (Technical Report.
No. 83-41) In: Stern, M.E.,Mellor, F.K. (Eds.), Baroclinic
Instability and Ocean Fronts. Woods HoleOceanographic
Institution.
Karimova, S., 2014. Hydrological fronts seen in visible and
infrared MODIS imagery of theBlack Sea. Int. J. Remote Sens. 35
(16), 6113–6134.
Krause, G., Budeus, G., Gerdes, D., Schaumann, K., Hesse, K.,
1986. Frontal systems in theGerman Bight and their physical and
biological effects. Elsevier Oceanogr. Ser. 119,
20–29.Lee, O., Nash, R.D.M., Danilowicz, B.S., 2005. Small-scale
spatio-temporal variability in
ichthyoplankton and zooplankton distribution in relation to a
tidal-mixing front inthe Irish Sea. Ices J. Mar. Sci. 62 (6),
1021–1036.
Lie, H.J., 1989. Tidal fronts in the southeastern Hwanghae
(Yellow Sea). Cont. Shelf Res.9 (6), 527–546.
Liu, G.M., Sun, S., Wang, H., Zhang, Y., Yang, B., Ji, P., 2003.
Abundance of Calanussinicus across the tidal front in the Yellow
Sea, China. Fish. Oceanogr. 12 (4–5),291–298.
Liu, G., Wang, H., Sun, S., Han, B., 2003. Numerical study on
the velocity structurearound tidal fronts in the Yellow Sea. Adv.
Atmos. Sci. 20 (3), 453–460.
Liu, Z., Hou, Y., 2012. Kuroshio front in the East China Sea
from satellite SST and remotesensing data. IEEE Geosci. Remote
Sens. Lett. 9 (3), 517–520.
Lough, R.G., Manning, J.P., 2001. Tidal-front entrainment and
retention of fish larvae onthe southern flank of Georges Bank ☆.
Deep-Sea Res. Part II 48 (1), 631–644.
Lü, X., Qiao, F., Xia, C., Wang, G., Yuan, Y., 2010. Upwelling
and surface cold patches inthe Yellow Sea in summer: effects of
tidal mixing on the vertical circulation. Cont.Shelf Res. 30 (6),
620–632.
Lwiza, K.M.M., Bowers, D.G., Simpson, J.H., 1991. Residual and
tidal flow at a tidalmixing front in the North Sea. Cont. Shelf
Res. 11 (11), 1379–1395.
Park, S., Chu, P.C., 2006. Thermal and haline fronts in the
Yellow/East China Seas:Surface and subsurface seasonality
comparison. J. Oceanogr. 62 (5), 617–638.
Patricola, C.M., Chang, P., 2016. Structure and dynamics of the
Benguela low-levelcoastal jet. Clim. Dyn. 1–24.
Pingree, R.D., Griffiths, D.K., 1978. Tidal fronts on the shelf
seas around the British Isles.J. Geophys. Res. Oceans 83 (C9),
4615–4622.
Pingree, R.D., Pugh, P.R., Holligan, P.M., Forster, G.R., 1975.
Summer phytoplanktonblooms and red tides along tidal fronts in the
approaches to the English Channel.Nature 258 (5537), 672–677.
Ren, S., Xie, J., Zhu, J., 2014. The roles of different
mechanisms related to the tide-induced fronts in the Yellow Sea in
summer. Adv. AtmIos. Sci. 31 (5), 1079–1089.
Schultes, S., Sourisseau, M., Masson, E.L., Lunven, M., Mari
Eacute, L., 2013. Influence ofphysical forcing on mesozooplankton
communities at the Ushant tidal front. J. Mar.Syst. 109–110 (1),
S191–S202.
Simpson Amp, J.H., Hunter, J.R., 1974. Fronts in the Irish Sea.
Nature 250 (5465),404–406.
Su, J., 2001. A review of circulation dynamics of the coastal
oceans near China. ActaOceanol. Sin. 23 (4), 1–16 (in Chinese with
English abstract).
Ullman, D.S., Cornillon, P.C., 1999. Satellite-derived sea
surface temperature fronts onthe continental shelf off the
northeast U.S. coast. J. Geophys. Res. 104 (C10), 23459.
Vazquez-Cuervo, J., Dewitte, B., Chin, T.M., Armstrong, E.M.,
Purca, S., Alburqueque, E.,2013. An analysis of SST gradients off
the Peruvian coast: the impact of going tohigher resolution. Remote
Sens. Environ. 131 (8), 76–84.
Vazquez-Cuervo, J., Torres, H.S., Menemenlis, D., Chin, T.,
Armstrong, E.M., 2017.Relationship between SST gradients and
upwelling off Peru and Chile: model/satellitedata analysis. Int. J.
Remote Sens. 38 (23), 6599–6622.
Wang, F., Liu, C., Meng, Q., 2012. Effect of the Yellow Sea warm
current fronts on thewestward shift of the Yellow Sea warm tongue
in winter. Cont. Shelf Res. 45 (3),98–107.
Wei, H., Su, J., Wan, R., Lei, Wang, Lin, Y., 2003. Tidal front
and the convergence ofanchovy (Engraulis japonicus) eggs in the
Yellow Sea. Fish. Oceanogr. 12 (4–5),434–442.
Wiafe, G., Nyadjro, E.S., 2015. Satellite observations of
upwelling in the Gulf of Guinea.IEEE Geosci. Remote Sens. Lett. 12
(5), 1066–1070.
Woodson, C.B., McManus, M.A., Tyburczy, J.A., Barth, J.A.,
Washburn, L., Caselle, J.E.,Palumbi, S.R., 2012. Coastal fronts set
recruitment and connectivity patterns acrossmultiple taxa. Limnol.
Oceanogr. 57 (2), 582–596.
Woodson, C.B., Litvin, S.Y., 2015. Ocean fronts drive marine
fishery production andbiogeochemical cycling. Proc. Natl. Acad.
Sci. USA 112 (6), 1710–1715.
Xie, S.P., Jan, H., Youichi, T., Liu, W.T., Hiroki, T., Xu, H.,
2002. Bathymetric effect on thewinter sea surface temperature and
climate of the Yellow and East China Seas.Geophys. Res. Lett. 29
(24), 81.
Zeng, X., Belkin, I.M., Peng, S., Li, Y., 2014. East Hainan
upwelling fronts detected byremote sensing and modelled in summer.
Int. J. Remote Sens. 35 (11–12),4441–4451.
Zhang, S.W., Wang, Q.Y., Lü, Y., Cui, H., Yuan, Y.L., 2008.
Observation of the seasonalevolution of the Yellow Sea cold water
mass in 1996–1998. Cont. Shelf Res. 28 (3),442–457.
Zhao, B.R., 1985. The fronts of the Huanghai Sea cold water mass
induced by tidalmixing. Oceanol. Limnol. Sin. 16, 451–459 (in
Chinese with English abstract).
Zhao, B.R., 1987. The continental shelf fronts induced by tidal
mixing in the Huanghaisea. J. Oceanogr. Huanghai Bohai sea 5, 16–23
(in Chinese with English abstract).
Zhao, B.R., 1989. The basic characteristics and formation
mechanism of the strongthermocline in Bohai sea, Yellow Sea and
East China Sea North. Acta Oceanol. Sin. 11(4), 401–410.
L. Lin, et al. Continental Shelf Research 175 (2019) 76–86
86
http://coastwatch.pfeg.noaa.gov/erddaphttp://coastwatch.pfeg.noaa.gov/erddaphttp://refhub.elsevier.com/S0278-4343(18)30280-2/sbref1http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref1http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref2http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref2http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref3http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref3http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref4http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref4http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref4http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref5http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref5http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref5http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref6http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref6http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref7http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref7http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref7http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref8http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref8http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref9http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref9http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref9http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref10http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref10http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref11http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref11http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref12http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref12http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref12http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref13http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref13http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref14http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref14http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref15http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref15http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref15http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref16http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref16http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref17http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref17http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref18http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref18http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref18http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref19http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref19http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref20http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref20http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref20http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref21http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref21http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref22http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref22http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref22http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref23http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref23http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref23http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref24http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref24http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref25http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref25http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref25http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref26http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref26http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref27http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref27http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref28http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref28http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref29http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref29http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref29http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref30http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref30http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref31http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref31http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref32http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref32http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref33http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref33http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref34http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref34http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref34http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref35http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref35http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref36http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref36http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref36http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref37http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref37http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref38http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref38http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref39http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref39http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref40http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref40http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref40http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref41http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref41http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref41http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref42http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref42http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref42http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref43http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref43http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref43http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref44http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref44http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref45http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref45http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref45http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref46http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref46http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref47http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref47http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref47http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref48http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref48http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref48http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref49http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref49http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref49http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref50http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref50http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref51http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref51http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref52http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref52http://refhub.elsevier.com/S0278-4343(18)30280-2/sbref52
Double fronts in the Yellow Sea in summertime identified using
sea surface temperature data of multi-scale ultra-high resolution
analysisIntroductionData and methodSST data sourceFrontal
detection
ResultsLocation and monthly variation of the double frontLoose
and tight double fronts
DiscussionProviding confidence in the presence of the double
frontsThe formation and monthly variation of the double
frontsImplications from the double fronts
ConclusionsAcknowledgmentsReferences