alexa Water Quality Seasonal Variability (2000 to 2015) in Yangtze River Estuary and Its Adjacent Coastal Area

ISSN: 2469-4134

Journal of Remote Sensing & GIS

  • Research Article   
  • J Remote Sensing & GIS 6: 216, Vol 6(4)
  • DOI: 10.4172/2469-4134.1000216

Water Quality Seasonal Variability (2000 to 2015) in Yangtze River Estuary and Its Adjacent Coastal Area

Yang Xianping, Leonid Sokoletsky* and Hui Wu
State Key Laboratory of Estuarine and Coastal Research, East China Normal University, Shanghai, China
*Corresponding Author: Leonid Sokoletsky, State Key Laboratory of Estuarine and Coastal Research, East China Normal University, Shanghai, China, Tel: 086-021-52134943, Email: [email protected]

Received Date: Sep 22, 2017 / Accepted Date: Oct 26, 2017 / Published Date: Oct 30, 2017


Three separate algorithms: 1) Modified NIR-SWIR atmospheric correction, 2) Suspended sediment concentration (SSC), and 3) Diffuse attenuation coefficient at 490 nm, Kd (490) were developed and used for mapping SSC and Kd (490) for the East China natural waters. A geographic area located between 27-35°N and 119-125°N was selected to analyze wet (flood) and dry seasons from 2000 to 2015. Remote sensing acquisition has been realized using the MODIS/Terra and MODIS/Aqua satellite sensors. Results showed large differences between these seasons in terms of spatial pattern of SSC and Kd (490) levels: SSC and Kd (490) values were higher during the dry season than during the flood season in the most part of area for most of the area. The area with high SSC>80 gm-3 [or, correspondingly, Kd (490)>2.3 m-1] within the Subei Bank of the Yellow Sea and the Zhejiang coastal area was estimated to be almost twice as large during the dry season when compared to the wet season. Results also revealed an impact of the Three Gorges Dam power station on the water quality in the Yangtze River Estuary.

Keywords: Turbid estuarine and coastal waters; Atmospheric correction; Remote-sensing reflectance; Suspended sediment concentration; Diffuse attenuation coefficient


An increasing population and the accompanied rising anthropogenic pressure in the East China coastal region, comprising the Yangtze River Estuary (YRE) and its adjacent coastal area (ACA), dictate the necessity for improved management and monitoring of these waters. From optical point of view, the part of ACA far away from the coastal line (open oceanic waters) may be referred to the Case 1 waters, i.e., such waters for which phytoplankton and their immediate detrital material, along with the water itself, do control the optical properties of the medium [1-3]. Oppositely, the YRE and the nearest part of ACA, for which the suspended sediment concentration (SSC) is the central factor influencing the optical properties, are the typical Case 2 waters [1-3].

Thus, SSC, one of the most important water quality components, especially in inland, estuarine, and coastal turbid areas [4-6], directly affects optical properties such as color, light transmittance, absorption, backscattering, scattering angular distribution, and reflectance [7-9], and it is strongly responsible for the abundance and the transport of carbon, phytoplankton, nutrients, metals, and pollutants in aquatic environments [10-12].

In sediment-dominated Case 2 waters, the SSC is closely related with the diffuse attenuation coefficient (Kd), which in turn has been used for different systems of water quality classification [13-20] and also in models of light penetration, for example, to compute a primary production as a function of light available at depth [21-23]. It is important that Kd is one of the geophysical parameters that can be derived from ocean color images [24-26]. Kd together with SSC, are important water quality parameters (WQPs), and they were the focus of many in situ measurements and remote sensing investigations performed in East China inland, estuarine and coastal waters [19,20,26-46].

In many of these publications, the seasonal variations in SSC and Kd (λ) were noted during both field and remote-sensing [26,32-34,36,38,40-47] observations. These publications give a somewhat contradictory picture of the seasonal variations of SSC and Kd (490) in East China waters. For example, Shi and Wang [26], based on results obtained during 2002-2008 from the Moderate Resolution Imaging Spectroradiometer (MODIS) onboard the satellite Aqua, concluded that Kd (490) values derived for Subei Bank (SB) of the Yellow Sea, Taihu Lake (TL), and Hangzhou Bay (HB) were substantially larger in winter than in summer, while for YRE the averaged (within the 2002-2008) seasonal values of Kd(λ) were approximately the same for all seasons.

However, He et al. [39] wrote, "Numerous factors can affect temporal variations of the TSM in HZB (‘total suspended matter in Hangzhou Bay’-our comm.). In general, the TSM concentrations inputted by the Chang jiang and Qiantang Rivers are much higher in the wet season with large discharge from May through October, than in the dry season with low discharge from November through April [28,37].

Oppositely, results found by Ref. [26-28,33,35,36,38,39,42] show that generally SSC and Kd (490) in YRE, HB, SB, and TL are significantly higher in the dry season than in the wet (flood) season. Moreover, from Ref. [26,36,38] follows even more detailed rule for all four seasonal average values of SSC and Kd (490), which may be written, in short form, as: {Summer}<{Autumn}<{Spring}<{Winter}.

This obvious contradiction between the above statement in Ref. [39] and the conclusions from the other publications (and from Ref. [39] itself) could be caused by the separate deviations from the general picture, observed in the two stations in the inner part of YRE by Chen et al. [28] and at the Nanjing station (Lower Yangtze River) by Mao et al. [48]. This is in accordance with conclusions made by Chen [26] that "in the inner estuary predominantly affected by river water discharge, the SSC is higher during the summer than during the winter. However, the SSC out the mouth of Yangtze Estuary is higher during the winter than during the summer, hence the sediment discharge carried by the runoff does not directly affect the SSC of the study area."

Another important task is a monitoring of the multi-year SSC and Kd (490) dynamics, strongly related with activity of the Three Gorges Dam (TGD), the world's largest power station, spanning the Yangtze River [32,35,48]. Three main periods may be highlighted for the purposes of the current study: before 2006 (when the dam body was completed), after 2012 (when the dam project was completed and become fully functional) and between 2006 and 2012 (transition period).

Therefore, our aim is to study the seasonal and multi-year water quality dynamics for East China natural waters based on the last achievements in atmospheric and ocean optics. With this purpose, we have performed remote sensing observations of SSC and Kd (490) via MODIS, which is onboard two NASA spacecrafts (Terra and Aqua). We have used images obtained by this satellite sensor (with a spatial resolution of 1 km) over the East China area, during 2000-2015 and grouped all images for flood (May to October) and dry (November to the next April) seasons. This allowed us to check, extend, and improve current knowledge of water quality in this region. Thus, these new results span much longer time period and simultaneously they better seasonally grouped than it was in previous studies of the area.

Geographical region and observations

The coastal region along the YRE and ACA (Figure 1) is one of the most active economic belts, and it is impacted by huge stresses from anthropogenic activities and population growth. The Yangtze River is the longest river (6300 km) in Asia, carrying abundant sediment (470 Mt/yr averagely) to its estuary [48].


Figure 1: Map of the eastern Chinese waters. Numbers 1-5 correspond to Subei Bank of the Yellow Sea, Yangtze River Estuary, Taihu Lake, Hangzhou Bay, and Zhejiang coastal area, respectively.

The appropriate literature shows that the SSC varies from 1~70 gm-3 in East China Sea (ECS), up to ~150 gm-3 in Taihu Lake, up to ~2000 gm-3 in YRE, up to ~2500 gm-3 in the Subei Bank of the Yellow Sea, and up to ~8000 gm-3 in the HB [31,32,34,36,37,39,40,43-47]. These values indicate that the suspended sediment ranges in the YRE and its ACA differ from most worldwide estuaries and coastal zones. Similarly, the Kd (490) values found in the literature [26,34,36,37] demonstrate huge variability in this parameter, which strongly exceed the limits (0.02-1.1 m-1) provided by the classical Jerlov's classification for Kd (490) [13].

Three databases were used in the present work for model development and validation purposes: 1) SSC and laboratory tank radiometric measurements performed in July 2006 (118 samples), when sediment samples used in tank experiments were collected in the Yangtze River and Yellow River estuaries [31]; 2) in situ measurements of SSC, radiometry, and inherent optical properties (IOPs), including absorption a, attenuation c, and back scattering bb coefficients, performed on May 4-12, 2011, in the YRE and its ACA (19 samples); and 3) in situ measurements of the chlorophyll concentration (Chl), SSC, and IOPs, including a, c, bb, and colored dissolved organic matter (CDOM) absorption, performed on February 21-March 10, 2014, in the YRE and its ACA (94 samples) [46].

The images of MODIS/Terra from March 2000 to December 2015 were downloaded from the Level 1 Atmosphere Archive and Distribution System (LAADS), MODIS/Aqua images from January 2010 to December 2015 were also downloaded from the LAADS system as a supplement. In total, 908 MODIS/Terra images and 192 MODIS/Aqua images were processed during the 16-year time period.

Development of a New (Modified Nir-Swir) Atmospheric Correction Approach

Introduction to a new approach

Aerosol type and load can be observed by ground stations, however, generally there is a problem with the simultaneous collection of aerosol data and satellite images, especially for the sea and ocean where there are no meteorological stations. Nevertheless, numerous atmospheric correction approaches were developed during last decades. Atmospheric and ocean optics pioneers, such as Howard Gordon, assumed that the ocean is totally absorbing in the band centered at λ=750 nm, i.e., Lw (750)=0 [49]. Moreover, worldwide in situ observations [50] have shown that in very clear waters with the pigment concentration C<0.25 mg m-3, even Lw (λ ≥ 670 nm) can be assumed to be zero. Similar assumptions were widely used in numerous studies (e.g., [51-53]. The further development of such approach was suggested by Wang and Shi (2005) [54], who assumed that the ocean is black (i.e., bb(λ)=Lw(λ)=Rrs(λ)=0) at the short wave infrared (SWIR) band to solve an atmospheric correction problem for the turbid Case 2 natural waters.

On the other hand, another strategy, so called "the best fit," is popular, and it is commonly used for the remote sensing retrieval of aerosol properties for land [55,56] and ocean [57,58]. Lookup tables (LUTs) generating preliminary, are used for comparison of the measured water-leaving radiance with the LUTs values for determination of aerosol types or/and other atmospheric parameters. The spectral remote-sensing reflectance Rrs (or normalized water- leaving radiance, Lwn) values derived from LUTs then may be used as inputs for SSC and Kd products.

Some serious problems may be noted, however, under exploration of the LUTs: 1) Not all LUTs are achievable for customary researchers; 2) A lot of satellite images do not have SWIR bands, which are used often by LUTs; and 3) Our experience shows that results derived from the NIR-SWIR based on the using LUTs are not reasonable enough.

Therefore, we suggest the compromise variant of the atmospheric correction algorithm by combining the “best fit” and NIR-SWIR approaches as follows:

• For Case 1 waters, the black ocean assumption is accepted for all NIR and SWIR bands, while for Case 2 waters, it is only accepted for SWIR bands.

• A boundary between Case 1 and Case 2 waters has been determined by the turbid water index, Tindi, λj), which may be expressed approximately [59,60] (under condition that the aerosol reflectance at wavelength λj, ρwj)=0) as Tindi, λj) ≈ 1+ρw, TOA(λi)/ρai)=1+Lw, TOA(λi)/La(λi), where index "TOA" means top-of- atmosphere level; we accepted Tind(746.4,1241.6)=1.3 (here 746.4 and 1241.6 nm are the central wavelengths of 15th and 5th MODIS bands, respectively) as a boundary value following Ref. [59,60].

• For MODIS-derived images with NIR and SWIR bands, the NIRSWIR atmospheric correction was applied.

• For satellite images without SWIR band, the NIR bands were exploited first for the Case 1 waters, and then the same aerosol type as was derived for these waters was applied to the Case 2 waters as well.

Another problem with the NIR-SWIR atmospheric correction that not all satellite sensors have SWIR bands. The Geostationary Ocean Color Imager (GOCI) launched on June 2010 by South Korea [39,43] is an example of such sensor. The sensor has a spatial resolution of 500 m, temporal resolution of 1 hour, and two NIR bands centered about 745 and 865 nm, but does not has SWIR bands. Thus, we suggest to use these two bands for an atmospheric correction in Case 1 waters (i.e., for ACA), and then to do atmospheric correction for the remaining Case 2 waters with preliminary derived (from LUTs) aerosol type.

Generation of LUTs

A preliminary lookup table was generated for four small modes ("S") and five large modes ("L") [56,58] of aerosols shown in Table 1, assuming bimodal lognormal aerosol size distribution (ASD) [56]. Details of such ASD may be found in Ref. [56]. Table 1 contains a full set of the complex refractive indices and ASD parameters for all nine aerosols modes included in this LUT.

Mode Refractive indices ASD* parameters (mm)  
  λ=470-860 nm λ=1240 nm λ=1640 nm λ=2130 nm rg s ref Aerosol type*
S1 1.45-0.0035i 1.45-0.0035i 1.43-0.01i 1.40-0.005i 0.07 0.4 0.1 WWS
S2 1.45-0.0035i 1.45-0.0035i 1.43-0.01i 1.40-0.005i 0.06 0.6 0.15 WWS
S3 1.40-0.002i 1.40-0.002i 1.39-0.005i 1.36-0.003i 0.08 0.6 0.2 WSWH
S4 1.40-0.002i 1.40-0.002i 1.39-0.005i 1.36-0.003i 0.1 0.6 0.25 WSWH
L1 1.45-0.0035i 1.45-0.0035i 1.43-0.0035i 1.43-0.0035i 0.4 0.6 0.98 WSS
L2 1.45-0.0035i 1.45-0.0035i 1.43-0.0035i 1.43-0.0035i 0.6 0.6 1.48 WSS
L3 1.45-0.0035i 1.45-0.0035i 1.43-0.0035i 1.43-0.0035i 0.8 0.6 1.98 WSS
L4 1.53-0.003i (470)
1.53-0.001i (550)
1.53-0.000i (660)
1.53-0.000i (860)
1.46-0.000i 1.46-0.000i 1.46-0.000i 0.6 0.6 1.48 DL
L5 1.53-0.003i (470)
1.53-0.001i (550)
1.53-0.000i (660)
1.53-0.000i (860)
1.46-0.000i 1.46-0.000i 1.46-0.000i 0.5 0.8 2.5 DL

Table 1: Complex refractive indices, geometric mean radii (rg), standard deviations of radii (sr) and effective radii (ref) for small (S1 to S4) and large modes (L1 to L5) used in the MODIS LUT. ASD (aerosol size distribution); WSWH (water soluble with humidity), WWS (wet water soluble), WSS (wet sea salt), DL (dust-like).

The output [Rrs(λ)] LUT was calculated by using 6S code with different values of aerosol total loading expressed through the aerosol optical thickness τa at 550 nm. We have used five different values of τa (550) for calculations: 0 (a pure molecular atmosphere), 2 (a very turbid atmosphere observed, for example, during dust or biomass burning events [61-63]), and three intermediate values: 0.2, 0.5, and 1.0. Thus, the final LUT contains 9 × 5=45 rows (it is not shown here).

Validation of the Atmospheric Correction Algorithm

Standard (MOD09) and modified NIR-SWIR algorithms vs. in situ values

The data obtained from the extensive field campaigns along YRE and its ACA were used for validation of the atmospheric correction algorithm. Details for the field and satellite pixels, which were found appropriate for validation purposes (due to their spatial and temporal closeness), are given in Table 2.

Fig. ID St. ID Date Site Water Type Time
(deg N)
(deg E)
a A7-1 3/9/2014 HB Case 2 6:18
b B7 2/21/2014 YRE Case 2 5:47
c D2 5/7/2011 YRE Case 2 5:01
d D1 5/7/2011 YRE Case 2 7:00

Table 2: Match-up stations selected for validation of the modified NIR-SWIR algorithm. All UTC times referred to are morning hours. Times (UTC) and geographical coordinates belonging to each station are shown for in situ and remote sensing observations in the upper and bottom rows, respectively.

This table demonstrate that the maximal temporal difference between corresponding pixels does not exceed 1.4 hours, and distances between them are shorter than 700 m. Below (Figure 2), we show examples of the spectral Rrs (λ) obtained from the field observations and satellite monitoring in the area. In addition to our (modified NIRSWIR algorithm), we show here the results derived from the standard MOD09 atmospheric correction algorithm, created by the MODIS Land Surface Reflectance Science Computing Facility [64-66]. The output spectral surface reflectance ρs (λ) for this algorithm may be presented with good accuracy (assuming the upwelling light as completely diffuse and unpolarized) as ρs (λ)=πRrs (λ) [13,20,67,68].


Figure 2: MODIS/Aqua-derived remote sensing reflectance spectra (open squares and closed triangles are for the standard MOD09 and modified NIR-SWIR algorithms, respectively) compared with in situ measurements (curves) acquired at A7-1 (a), B-7 (b), D2 (c), and D1 (d). Time and geographical locations of field stations and corresponding MODIS targets are listed in Table 2.

The results undoubtedly confirm a significant contribution of waterleaving radiance into the sensor level signal for the NIR bands. They also show that the modified NIR-SWIR atmospheric correction algorithm performed generally better than the standard MOD09 algorithm for the highly variable waters under consideration. Besides, it is obvious that there is a closeness between measured in situ and predicted from the satellite imagery values of Rrs (λ), especially in the green-NIR spectral range. However, at λ<550 nm, the modified NIR-SWIR approach may slightly underestimate (Figure 2a-2c) or overestimate (Figure 2d) in situ Rrs (λ) values.

Nevertheless, as numerous studies show (see, e.g., reviews in Ref. [46,64]), wavelengths λ>550 nm (where not strong impact of chlorophyll and dissolved organic matter absorption) should be used for SSC and Kd (490) retrievals in turbid waters. Moreover, our previous search of the best MODIS bands for SSC retrieval led to using the fourth band centered at 553.6 nm and the first band centered at 645.5 nm, for in situ [64] and remote sensing [46] algorithms.

At present, we have found that these two wavelengths yield the minimum of the normalized (to the mean measured in situ values) root-mean-squares error (NRMSE) for the remote sensing Rrs (λ) values relative to in situ values; NRMSE=17.7% and 11.7%, respectively (Figure 3), which confirms our previous choice of these wavelengths.


Figure 3: Normalized root-mean-squares error (NRMSE) for the Rrs(λ) values derived from the modified NIR-SWIR atmospheric correction algorithm relative to in situ values.

Modified NIR-SWIR vs. standard (MOD09) algorithm

Figure 4 demonstrates comparison results between the modified NIR-SWIR and MOD09 algorithms for April 5, 2011. It is obvious that both approaches yield close results for the turbid waters, with Rrs (645.5) approximately more than 0.02 sr-1 [that corresponds to SSC>24 gm-3, Rrs (553.6)>0.026 sr-1, Rrs (645.5)/Rrs (553.6)>0.76, and Kd (0°, 490)>0.70 m-1, according to our algorithm]; however, there are sometimes large discrepancies for the more clear waters.


Figure 4: Scattering plots of Rrs (λ) derived from the modified NIRSWIR algorithm versus Rrs (λ) derived from the standard MODIS algorithm (MOD09) for 645.5 nm (a), 553.6 nm (b), and Rrs (645.5)/Rrs (553.6) (c). Note that color palettes (right) yield color scales for the number of points per pixel.

Cloud Mask

Empirical cloud masking was based on testing two quantities: the atmospheric-corrected Rrs (λ) at the bands centered at 856.6 nm and 1241.6 nm (the second and fifth MODIS bands, respectively). The pixel under investigation has considered as a cloud pixel, when Rrs (856.6)>0.08 or Rrs (1241.6)>0.006.

SSC Retrieval

Retrieval of SSC (in gm-3) for the subsurface layer (accepted as the layer lying between 0- and 1 m below the air-sea surface) was realized by the model suggested by [47,65] as follows:


where X=Rrs (645.5)/Rrs (553.6).

Kd (0°, 490 nm) Retrieval

An estimation of diffuse attenuation coefficient at zenith angle (i.e., at solar zenith angle θ0=0°) and near surface (layer<1 m) was performed as follows:

Kd(0°, 490) = 0.03012(SSC)0.9895 (2)

Eq. (2) was derived from our three databases and following Gordon's (quasi-single-scattering) approximation, QSSA [69-71]:

Kd(0°, 490) = a(490) + bb(490)(3)

It was shown in Ref. [71] that QSSA yields very accurate solutions for the near surface layers of natural waters with widely varied optical conditions.

Results and Discussion

Remote-sensing seasonal analysis of water quality parameters was performed for the East China geographical area during 2000-2015 (Figure 5) based on the sub-algorithms described above, namely, the modified NIR-SWIR atmospheric correction (with cloud masking), SSC, and Kd (0°, 490). We proceeded with all achievable (mostly cloud free) images for this time period and averaged results for each year and for two seasons: the flood (May to October) and the dry (November to the next April) seasons, according to [28,29,36,41,48]. All images were decomposed into 0.05° × 0.05° boxes in the 27°N-35°N and 119°E-125°E geographical area.


Figure 5: The flood season and dry season average of SSC and Kd (0°, 490 nm) derived from 2000 to 2015 MODIS images.

Our results actually correspond to numerous in situ and remote observations [26-28,33,36,38,39,42], that reported seasonal features in the region, namely, generally higher values of SSC and Kd (0°, 490) during the dry season when compared to the wet season. However, this seasonal feature may be different for different years and sub-areas. The most obvious and multi-year stable situation was observed in the East China Sea, Subei Bank, Taihu Lake, Hangzhou Bay, and Zhejiang coastal area. For example, in the ECS a multi-year water transparency is the rather high, with SSC<20 gm-3 (that corresponds the Jerlov 7C water type with Kd (0°, 490)<0.58 m-1; see algorithm above and Ref. [13,72] for details) regardless of the year or season of the year. However, four other areas -SB, TL, HB, and ZCA-have shown clear seasonal features during the whole time period (2000-2015).

The most variable multi-year results were found in the South Branch of YRE, namely, very high SSC values (~1000 gm-3 and more) observed there in 2000, 2001, 2003, 2006, and 2007 flood seasons; however, it was not the case in 2008-2015. Obviously, the YRE accepted a substantial amount of sediments originating from the freshwater discharge-carried sediments of the Yangtze River before 2008. In these years, SSC and Kd (0°, 490) stayed approximately the same during the whole year, i.e., without significant seasonal differences. It is worth to note that such situation may be explained by different balancing factors, namely, large concentrations of suspended matter discharged by Yangtze River during wet seasons, from one side and stronger wind wave and weaker thermal stratification in winter seasons, from another side.

After completion in 2006 of the TGD, significant amounts of sediments that used to be flushed away were captured by the dam, decreasing sediment load and SSC downstream of the river and the YRE [33,36,49]. Moreover, a field study by Li et al. [36] has shown that the most dramatic decline of sediment discharge in the Lower Yangtze River (at Datong) -more than twice comparing 2003-2009 vs. 1950-2002- happened from April to November, i.e., mostly in the flood season. Our results (Figure 5) completely confirm these field observations, showing significant decrease of SSC in the Lower Yangtze River since 2008, during the flood months. Interesting to note that this dramatic decline of sediment discharge occurs on the background of relatively weak decline (less than 20%) of water discharge in the Lower Yangtze River during the wet season (see Figure 2 in Ref. [36]).

For numerical presentation of results, we have calculated total turbid area for YRE and ACA with SSC>80 gm-3 [this corresponds Kd (0°, 490)>2.3 m-1 according to our optical model] for each false color image and then summarized results for each year in Table 3 and Figure 6.


Figure 6: Dynamics of turbid area in the East China coastal area during 2000-2015. Dynamics for the flood and dry seasons are shown on the main ordinate axis, while the flood to dry area ratio is shown on the secondary ordinate axis.

Year Flood season area (km2) Dry season area (km2)
2000 40400 59900
2001 35300 75600
2002 29700 81600
2003 45700 66800
2004 37000 64000
2005 36900 59400
2006 39500 70400
2007 24500 70500
2008 32300 67100
2009 36000 68400
2010 30900 73100
2011 34700 79500
2012 28100 59400
2013 40800 68000
2014 51500 79600
2015 31600 72000

Table 3: The flood and dry seasonal areas in with SSC>80 gm-3.

These numerical results confirm a general trend observed in the East China waters with the prevailing SSC during the dry season. However, total results are not sensitive enough to reveal an impact of the Yangtze River dams (especially, the TGD) on the total turbid area in YRE and ACA. This fact may be explained by the relatively small area of the YRE relative to whole area under consideration.

The similar analysis performed for separate regions shows that Subei Bank and Zhejiang coastal area are among the most dramatically variable regions. For example, during the wet seasons, the Subei Bank high turbid area (with SSC>80 gm-3) has varied from 4600 km2 in 2007 to 23600 km2 in 2014, at an average area of 12300 km2. The high turbid area for the same region, but during the dry season, is significantly larger, even though variations are less dramatic: from 24600 km2 in 2005 up to 42000 km2 in 2002, with the average area of 32300 km2.

Another important finding is decreasing SSC with increasing distance from the coastal line during whole year and along the whole coastal line. However, huge tongue-shaped "clouds" of more turbid waters (plumes) extend from the coast to the open parts of the ECS between approximately 31.5°N and 33°N, sometimes achieving a longitude of 126°E (see Figure 7 as example). These plumes are more expressed during dry seasons that was noted also by Ref. [26,36-38,43-46], but there is no consensus on the mechanisms of this phenomenon. For example, Shi and Wang [25] found that the main reason of formation of the plume is the strong mixing in the water column that, in turn, explained by shallow water depths, strong surface cooling, and high wind conditions. However, Chen et al. [45] suggested other reasons for plume formation in the Bohai, Yellow, and East China Seas, namely, river discharge, tidal and oceanic currents. Thus, this phenomenon obviously deserves further research.


Figure 7: Example of SSC mapping of the East China waters on March 24, 2012.


Three separate algorithms: 1) Modified NIR-SWIR atmospheric correction, 2) Suspended sediment concentration (SSC), and 3) Diffuse attenuation coefficient at 490 nm, Kd (490) were developed and used for mapping SSC and Kd (490) for the East China natural waters. Based on these findings, we have presented satellite-based maps of SSC and Kd (490) during 2000-2015 time period. Three main issues have been addressed in the study:

• Impact of dry and wet seasons on the water quality [expressed via SSC, Kd (490), and high turbid area] in the region;

• Impact of the Three Gorges Dam (TGD) power station on the water quality;

• Plume formation in the region.

We have found, in particular, that Hangzhou Bay, Yangtze River Estuary and Subei Bank are the most turbid areas in the region (and, probably, in the world). Especially high levels of SSC and high turbid areas were observed during dry seasons. Concerning wet seasons, turbidity in the South Branch of YRE was significantly reduced after the TGD completion. Huge plumes of more turbid waters extending toward East China Sea were also observed on satellite-based images.


The research leading to these results has received funding from the National Science Foundation of China (grants number 41271375 and 41371346). We would like also to express our gratitude to the two anonymous reviewers for their constructive and helpful feedback on earlier drafts of this paper.


Citation: Yang X, Sokoletsky L, Wu H (2017) Water Quality Seasonal Variability (2000 to 2015) in Yangtze River Estuary and Its Adjacent Coastal Area. J Remote Sensing & GIS 6:216. Doi: 10.4172/2469-4134.1000216

Copyright: © 2017 Xianping Y, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Select your language of interest to view the total content in your interested language

Post Your Comment Citation
Share This Article
Recommended Conferences
  • International Conference on Geological and Environmental Sustainability

    August 13-14, 2018, Bali, Indonesia

  • International Conference on GIS and Remote Sensing

    September 27-28, 2018 Berlin, Germany

  • International Convention on Geochemistry and Environmental Chemistry

    October 19-20, 2018 Ottawa, Canada

Article Usage
  • Total views: 532
  • [From(publication date): 0-2017 - Jun 24, 2018]
  • Breakdown by view type
  • HTML page views: 460
  • PDF downloads: 72

Post your comment

captcha   Reload  Can't read the image? click here to refresh
Leave Your Message 24x7