Suspended solids concentration is an important parameter to
evaluate the quality and environment of water. It has very
important significance to get the time-spatial distribution
information of suspended solids concentration in lakes for
their environmental management based on the remote sensing
technology. Poyang Lake is the largest freshwater lake in
China, located south of the Yangtze River in northern Jiangxi
Province. One of the most notable features of Poyang Lake is
the large difference in water level between the wet and dry
seasons; during the wet season, the reservoir stretches hundreds
of miles, while shallower wetlands and reed-dominated areas
are widespread throughout the dry season.
The measured suspended solid concentrations and chlorophyll-a
concentration data consisted of 16 data sets in January, April,
July, and October from 2009 to 2012, obtained from the Poyang
Lake Wetland Observation Station, Chinese Academy of Science.
MOD09 data products were used for suspended solid concentration
retrieval. As the second-level standard products of MODIS, they
are obtained after atmospheric and aerosol corrections. MOD09
data products have higher temporal resolution and 250 and 500 m
MODIS data products were obtained for the 5th–15th of January,
April, July, and October from 2000 to 2013 and preprocessed by
transferring the image projection to the Albers Conical Equal
Area project. To extract the water boundary for Poyang Lake, a
water boundary extraction method called the Floating Algae Index
(FAI, Hu 2009) was used.
This method is based on an effective combination of the red,
nearinfrared, and shortwave-infrared bands of MODIS imagery.
Based on the semi-empirical method, sensitive bands with higher
R2 values were selected through regression analysis and curve
fitting by combining the values between in situ measured suspended
solid concentrations and each band of the MODIS imagery. Selected
sensitive bands were the modeling input factors used to obtain the
nonlinear response relationship between the remote-sensing reflectance
values and observed suspended solid concentrations.
The remaining measured suspended solid concentrations data (that were
not used for modeling) and the spectral reflectance of the MODIS bands
were then used to evaluate the accuracy of the seasonal retrieval models,
Because of the clear seasonal changes in the water boundary of Poyang Lake,
the standard deviation was used to indicate the fluctuation of suspended
solid concentrations in the same month across different years, calculated
where σ is the standard deviation of suspended solid concentrations,
N is years, xi is the suspended solid concentration value of each year,
and μ is the average value of suspended solid concentrations.
The four seasons’ inversion models were built according to the suspended
solids concentration data of Poyang Lake continuously measured in spring,
summer, autumn and winter season of 2009 and 2012 year and combining with
the regression analysis of the lake’s multi-band MODIS images in the same
period. Then they are used to obtain the suspended solids concentration of
the lake in annual spring, summer, autumn and winter seasons from year 2000
The results show that the suspended solids concentration of Poyang Lake is
higher in the center of the lake and the main channel while is relatively
lower in coastal waters, what is more the concentration was increasing year
by year from 2000 to 2013 with high value in spring and autumn but low in
summer and winter. Human activities have a great influence on the suspended
solids concentration changes, too.