Search

Remote sensing has become an essential technique in water environment monitoring. However, the cloud contamination of remote sensing images breaks its spatial and temporal continuity in practical application. The current cloud removal algorithms always based on the hypothesis that the ground surface has no significant changes temporally, which is not proper for optically complex water body. Therefore, to overcome the weakness of current cloud removal algorithms on temporally fast-changing landscapes, a temporal unmixing-based cloud removal (TUBCR) algorithm is proposed to estimate the under-cloud information of optically complex water images. In this algorithm, the spatial information and temporal information were separated into temporal abundance images and temporal endmembers. By extracting the high temporal resolution endmembers, the fast-changing water information were estimated. The performance of the proposed algorithm was compared with those of two widely used cloud removal algorithms: the modified neighbourhood similar pixel interpolator (mNSPI) algorithm and the weighted linear regression (WLR) algorithm. The results of four simulated datasets (500 images in total) and two real Geostationary Ocean Colour Imager (GOCI) datasets (8 images in total) show that the proposed algorithm performed better in both quantitative indexes and visual effect compared with the mNSPI algorithm and the WLR algorithm. In satellite image time series, the proposed algorithm could accurately estimate under-cloud water information and correct the offsets of the statistical temporal trends that caused by cloud contamination. Hence, theoretically, the proposed TUBCR algorithm altered the similar pixel-based strategy of classic methods and got a satisfied performance in fast-changing water images. In applications, the TUBCR algorithm has great potential to improve the spatiotemporal coverage of optically complex water remote sensing. [ABSTRACT FROM AUTHOR]

A spectral–temporal-unmixing-based multitemporal image fusion (MTIF) algorithm is proposed to fuse Geostationary Ocean Color Imager (GOCI) and Himawari images. The algorithm was applied to two data sets. The fusions are quantitatively and qualitatively compared with four widely used algorithms. The results show that the MTIF algorithm performs better using both evaluation indexes and visual comparisons. For optical complex water monitoring, the MTIF-derived chlorophyll-a concentration (${C} _{{\text {chla}}}$) map has better spatial detail and temporal trends compared with the other algorithms. For cloudy images, the MTIF algorithm can estimate part of the undercloud water reflectance information: when there are more than 32.6 cloud-free pixels in the current study area, the MTIF algorithm can successfully recover the under cloud information. The MTIF algorithm has great potential to advance the monitoring of optical complex inland water. [ABSTRACT FROM AUTHOR]

Sediment resuspension is closely related to the endogenetic release of nutrients in Lake Taihu. Thus, understanding the factors associated with sediment resuspension is important. In this study, a new technology, which integrates a hydrodynamic model and remote sensing techniques, was applied to derive the distribution of the erosion flux and obtain the spatially variable critical shear stress. Then, the spatially variable critical shear stress was used in the sediment simulations at Lake Taihu. Compared to the traditional model, based on uniform values of critical shear stress, the new method, using variable values of critical shear stress calibrated from the Geostationary Ocean Color Imager (GOCI) data, significantly improved the sediment simulations at Lake Taihu. Based on the erosion flux from August 6-8, 2013, the correlations between erosion and wind speed, wind fetch, mud depth, and water depth were analyzed for different subsections and spots in Lake Taihu. The potential sources of error were also addressed. Further improvement of the model is necessary. [ABSTRACT FROM AUTHOR]

Although remote sensing technology has been widely used to monitor inland water bodies, the lack of suitable data with high spatial and spectral resolutions has seriously obstructed the practical development of inland water color remote sensing. An inversion-based fusion (IBF) algorithm is proposed to fuse water color and high-spatial resolution images. The algorithm was applied to two datasets: The Hyperion simulated dataset (dataset #1) and a pair of Environmental Satellite 1 (HJ1, launched by China in 2009) and medium resolution imaging spectrometer images (dataset #2). The fusions are quantitatively and qualitatively compared with three widely used algorithms. The results show that the IBF algorithm performs better using both evaluation indexes and visual comparisons. A discussion of the free parameter window size n and the contribution of the low-resolution image (LRI) to the spatial distribution ($w_{{{\rm LRI}}}$) shows that a larger n will result in both greater model errors and better control of geometric errors, whereas $w_{{{\rm LRI}}}$ helps stabilize the algorithm. Finally, chlorophyll-a concentration maps are developed from the fusions. The significant advantage of the IBF-derived chlorophyll-a concentration map indicates that the IBF algorithm has the potential to advance the monitoring of optical complex inland water. [ABSTRACT FROM PUBLISHER]

Land target-based atmospheric correction methods over inland waters are an effective alternative method for when the standard or improved atmospheric correction method that is based on water is invalid. The type of aerosol model that is selected and land dark target that is distinguished are two key factors that could determine the atmospheric correction’s accuracy. We present an improved land target-based iterative method [denoted as dense dark vegetation (DDV)-WC] for the atmospheric correction of Moderate Resolution Imaging Spectroradiometer (MODIS) over Lake Taihu. The improvements include the use of aerosol models in Optical Properties of Aerosols and Clouds (OPAC), which fully consider the impact of relative humidity (RH) on aerosol properties, and a rigorous land dark target selection rule (DDV-WC-Selection). The DDV-WC algorithm includes three steps: the selection of DDV pixels from lakeshore areas, the retrieval of the aerosol optical thickness (AOT) at 550 nm (AOT550) for the selected DDV pixels, and the retrieval of the remote sensing reflectance (Rrs) over water. The AOT550 inversion accuracy of different aerosol models in OPAC is tested in Lake Taihu. The “continental clean model” has the best performance compared with five other aerosol models for every month of the year. Therefore, the aerosol model of the DDV-WC algorithm is fixed to the “continental clean model” in Lake Taihu because of its stable and excellent performance. When using the “continental clean model” to perform these retrievals, the DDV pixels selection (DDV-WC-Selection) in the DDV-WC method (\text MRE(\text Mean\;\text Relative\;\text Error) = 14.63\% ) is more reliable than the DDV pixels selection (DDV-Selection) in the MODIS V5.2 method (\text MRE = 17.24\% ). The atmospheric correction of MODIS indicates that this approach produces superior performance in the green and red bands but poor results in the near-infrared (NIR) and blue bands. The MRE is only 13.15% at 645 nm. These findings illustrate that the DDV-WC algorithm is a practical MODIS atmospheric correction method for inland turbid waters. [ABSTRACT FROM PUBLISHER]

Understanding changes in climate and environment on a regional scale can provide useful guidance for regional socio-economic development. The present study characterizes changes in the environment, climate, land use and cover types via in situ observed, statistical data and remote sensing images for Jiangsu Province, China, during the period 1980–2012. Statistical and spatial analyses indicate that the pace of urbanization in southern Jiangsu is more rapid than that in northern Jiangsu. Urbanization (92.7%) results primarily from the loss of farmland. While emissions of pollutants from industrial sources were well controlled, and wastewater, which more frequently derives from urban domestic sources, was found to be increasing. The rates of wastewater to population increased from 0.17 ± 0.017 to 0.32 ± 0.090 (billion ton/million persons) during the two periods of 1980–2000 and 2000–2012. However, the rates of wastewater to Gross Domestic Product (GDP) decreased from 0.26 ± 0.20 to 0.014 ± 0.009 (billion ton/billion Yuan), respectively. The significant increase in scattering radiance and Earth's albedo caused by the urbanization and its process (Pearson correlation coefficient (r) between urban land and scattering radiance = 0.86, p < 0.0001; r between farmland and scattering radiance = − 0.92, p < 0.0001) aggravates the warming in the regional scale. This correlation analysis indicates that temperature will decrease with the increase of woodland, grassland and farmland, and will increase with the increase of urbanized and unexploited lands. Added to warming caused by an increase in CO 2 , land use/cover change and human activities may be the primary reason for the rising temperatures in Jiangsu Province. The change in regional thermal conditions reduces both local humidity and land atmosphere flux exchange. The low atmosphere flux exhange contributes to the spread of atmospheric pollutants and the deposition of atmospheric particles. [ABSTRACT FROM AUTHOR]

Treatment of eutrophication and algal blooms in large shallow lakes has been constrained by the endogenetic release of nutrients from sediment resuspension despite the well-controlled input of nutrients from surrounding land in Taihu Lake, China. In this study, we propose a method to simplify regionally empirical algorithms for total suspended matter (${{C}_{{\rm{TSM}}}}$), chlorophyll-a (${{C}_{{\rm{Chl\hbox{-}a}}}}$), and phosphorus concentration (${{C}_{\rm{p}}}$) for the Geostationary Ocean Color Imager (GOCI) for Taihu Lake. The results show that the GOCI-derived ${{C}_{{\rm{TSM}}}}$, ${{C}_{{\rm{Chl\hbox{-}a}}}}$, and ${{C}_{\rm{p}}}$ match well with the in situ measured data. Based on the validated synchronous and timely observation data (${{C}_{{\rm{TSM}}}}$, ${{C}_{{\rm{Chl\hbox{-}a}}}}$ and ${{C}_{\rm{p}}}$) from GOCI level-1b data, the interaction of phosphorus with algae and suspended sediment were analyzed from August 6 to August 9, 2013. The results indicate that phosphorus released from sediment resuspension is significant in Taihu Lake, with a high spatial–temporal variation. The effect of the restriction of nutrients (nitrogen and phosphorus) on algal growth varied temporally and spatially with the hydrodynamic characteristics of the lake. In most cases, restricted phosphorus in the lake is the primary reason for the algal growth, especially in open regions such as Center Lake (CL), Northwestern Lake (NWL), and Southwestern Lake (SWL). Both nitrogen and phosphorus are the restrictive factors to algal growth in Meiliang and Gonghu Bays. More attention should thus be focused on the endogenetic release of phosphorus in Taihu Lake when treating eutrophication and algal blooms in addition to controlling the input of nitrogen and phosphorus from surrounding land. [ABSTRACT FROM PUBLISHER]

Humans have had a significant impact on the terrestrial pedosphere through activities such as agriculture and urbanization. The effects of human activities on land use and the related environmental changes were investigated through point and areal studies surrounding Meiliang Bay, which is an open area of extreme eutrophication in Taihu Lake, China. This study used remote sensing and environmental-tracer profiles [total nitrogen (TN), total phosphorus (TP), total organic carbon (TOC), grain size, and geochemical parameters] to determine the causes of changes in land use and the associated environmental parameters. The results of LUCCs (Land use/cover changes) indicate that over the past three decades, total farmland decreased by 862.49 km2, with an annual decrement rate of 28.75 km2/year, and total urbanized land increased by 859.71 km2, with an annual growth rate of 28.66 km2/year. The geochemical results indicate that the trophic state of Taihu Lake was persistently intensifying and that the TN, TP, and TOC concentrations increased twofold, threefold, and twofold, respectively, from 1949 to 2010. The sources of TN, TP, and TOC were highly similar after 1975. However, before 1974, TN and TP originated from different sources than TOC. The grassland and woodland around the lake retain nutrients and sand from the land of study area. The increase in urbanized land and tertiary industries significantly increased the sediment concentrations of TN, TP, and TOC after 1980. [ABSTRACT FROM AUTHOR]

Assessments of specific inherent optical properties (SIOPs) and their variability in highly turbid and productive inland waters are essential for the accurate estimation of water quality. A new optical classification method including two classification criteria [i.e., normalized remote sensing reflectance slope (NS), and normalized remote sensing reflectance depth (ND)] was developed to divide remote sensing reflectance into four classes, i.e., class 1 (NS < −0.0017 and ND < 0.21) is low turbid and productive water; class 2 (NS < −0.0017 and ND > 0.21) is low turbid and high productive water; class 3 (NS > −0.0017 and ND < 0.09) is high turbid and low productive water; and class 4 (NS > −0.0017 and ND > 0.009) is high turbid and high productive water. The relationships between phytoplankton absorption at 440 nm [ a(440)] and chlorophyll- a concentration [ C] as well as between particle backscattering coefficient at 440 nm [ b(440)] and total suspended matter concentration ( C) after classification were obtained from a large number of in situ data in Lake Taihu. The measured specific phytoplankton absorption $$[ {a_{\text{ph}}^{*} \left( \lambda \right)} ]$$ and particle backscattering coefficient $$[ {b_{\text{bp}}^{*} \left( \lambda \right)} ]$$ show significant variations even within the same class. The mean values of $$a_{\text{ph}}^{*} \left( \lambda \right)$$ at 440 nm $$[ {a_{\text{ph}}^{*} \left( {440} \right)} ]$$ for each class are 0.048 ± 0.013, 0.060 ± 0.012, 0.083 ± 0.021, and 0.056 ± 0.017 m/mg, respectively. The mean values of $$b_{\text{bp}}^{*} \left( \lambda \right)$$ at 440 nm $$[ {b_{\text{bp}}^{*} \left( {440} \right)} ]$$ for each class are 0.035 ± 0.01, 0.024 ± 0.004, 0.041 ± 0.009, and 0.038 ± 0.009 m/g, respectively. The power functions of SIOPs and water constituents' concentration indicate that $$a_{\text{ph}}^{*} \left( {440} \right)$$ and $$b_{\text{bp}}^{*} \left( {440} \right)$$ vary with C and C. The validation results show that our proposed values for $$a_{\text{ph}}^{*} \left( {440} \right)$$ and $$b_{\text{bp}}^{*} \left( {440} \right)$$ cover a very wide range of water optical properties, which are characterized from clear water to highly turbid productive water. The validation results also suggest that the retrieval accuracy of C and C bio-optical model was improved after classification. The root mean square error (RMSE) of C was improved from 14.18 to 7.43 μg/L (mean value of all classes) and RMSE of C was improved from 32.98 to 26.10 mg/L (mean value of all classes). Thus, the temporal and spatial variation of $$a_{\text{ph}}^{*} \left( {440} \right)$$ and $$b_{\text{bp}}^{*} \left( {440} \right)$$ should be considered in the bio-optical retrieval model of water quality. Graphical Abstract: In complex optical properties of inland water, retrieving the water constituents with high accuracy needs to classify the water optical properties from the remote sensing spectrum by optical classification method. The figure shows the water color examples of each class[Figure not available: see fulltext.]. [ABSTRACT FROM AUTHOR]

Abstract: Human activities contribute highly to water eutrophication. In this study, the relationship between human activities and water eutrophication in Dianchi Lake in China was characterized using a combination of satellite imaging, sedimentary physicochemical and meteorological data analyses. The heavy eutrophic status and algal bloom in Dianchi Lake were first observed by satellite in 1977 and 1989, respectively. The C/N ratio, an indicator of organic sources in sediments, also showed that the planktonic organic source in the sediment significantly increased beginning in 1989. The land use cover in the Dianchi basin showed that both farm lands and forests, but particularly farmlands, were reduced during the period from 1974 to 2009. The urbanized land area increased from 1974 to 2009, particularly after 2000. The effects of human activities on water eutrophication were expressed by land use cover, population, gross domestic product (GDP; separated into primary, secondary and tertiary industries) and wastewater discharge. For land use cover, farm and urbanized lands were the main sources of water nutrients; forest contributed slightly to these nutrients. For GDP, primary (correlation coefficient=0.94, P<0.001) and tertiary (correlation coefficient=0.95, P<0.001) industries were highly correlated with total nitrogen. Secondary (correlation coefficient=0.95, P<0.001) and tertiary (correlation coefficient=0.96, P<0.001) industries were highly correlated with total phosphorus. The algal bloom area was significantly correlated with wastewater discharge (correlation coefficient=0.78, P<0.005) (except industrial wastewater), which was primarily led by the non-agricultural population, from 2000 to 2009. This study suggests that the protection of water environments requires a comprehensive protection policy in addition to a unilateral protection policy. [Copyright &y& Elsevier]

Pigment C-phycocyanin (C-PC) is a useful indicator for the presence of cyanobacteria in inland waters, which has been well known as a phytoplankton group with many negative effects on human, animal, and aquatic ecosystem health. In recent years, the remote detection of the C-PC concentrations for inland waters has received much attention. However, their accurate quantification by means of remote sensing is still a challenge due to the significant bio-optical complexity of turbid inland waters. In this paper, three typical turbid inland lakes in China were investigated through in situ observed data sets containing optical and water quality parameters. By using a recently proposed TD680 optical classification method, all collected samples were first classified into three types. For each type of water, we determined specific spectral sensitive regions for the pigment C-PC. Then, we developed three type-specific support vector regression (SVR) algorithms and an aggregated SVR algorithm. The performances of these algorithms were evaluated through the validation data sets. The results show that the type-specific algorithms generally have significantly improved performance over the aggregated SVR algorithm. Their assessment errors [mean absolute percentage error (MAPE) and root-mean-square error ( rmse)] were as follows: 1) MAPE = 15.6\% and rmse =\break 30.6\ \mg\cdot \m^-3 for Type 1 water; 2) MAPE = 47.1\% and rmse = 61.5\ \mg\cdot\m^-3 for Type 2 water; and 3) MAPE =\break 26.4\% and rmse = 19.1\ \mg\cdot\ m^-3 for Type 3 water. The findings in this paper demonstrate that a prior water classification is needed for the development of accurate C-PC retrieval algorithms. This paper provides a valid strategy for improving C-PC estimation accuracy and enhancing algorithm commonality for optically complex turbid waters. [ABSTRACT FROM AUTHOR]

[Display omitted] • Combining R rs (λ) and δ13 C POC can track and monitor the source of POC. • The source of POC in surface water of Taihu Lake in summer is mainly endogenous POC. • Endogenous POC has a certain physiological relationship with water temperature and nutrient (TN, TP), while terrestrial POC was closely related to wind speed. • The a ph (λ) and a nph (λ) can effectively reflect the source of POC. The sources of particulate organic carbon (POC) determine its conversion, thereby playing an important role in the carbon cycle of lakes. Accurate estimation of the sources and dynamic characteristics of POC is important for understanding the migration and transformation of organic carbon. However, the synchronous observation of POC sources with large areas through remote sensing is still challenging because of the complex composition of POC and the optical conditions of inland lakes. In this study, a three-band (1/ R rs (689)–1/ R rs (717)) × R rs (697)) empirical algorithm of POC sources was constructed based on remote sensing reflectance (R rs (λ)) and the proportion of endogenous POC estimated from the field-measured stable isotope (δ13 C POC) values. The validation and calibration results of the three-band algorithm showed robust performance, with MAPE and RMSE of estimated values and measured values of 10% and 0.07, respectively. The three-band algorithm had good simulation results for the Ocean and Land Color Instrument (OLCI), Moderate Imaging Spectroradiometer (MODIS), Geostationary Ocean Color Imager (GOCI), and Medium Resolution Imaging Spectrometer (MERIS) spectra. The POC sources estimated by the three-band algorithm suggest that the endogenous POC of Taihu Lake in August showed a decreasing trend from 2006 to 2019. The variation in terrestrial POC was slow and stable for both annual and monthly variations. The analysis of POC sources with total phosphorus (TP), total nitrogen (TN), water temperature, and wind speed indicated that terrestrial POC was closely related to wind speed (r = 0.33, P < 0.001), while endogenous POC was significantly associated with TP (r = 0.6, P < 0.001), TN (r = 0.56, P < 0.001), and water temperature (r = 0.49, P < 0.001). The use of remote sensing algorithms to evaluate POC from different sources is convenient and effective; furthermore, it helps to better understand the carbon cycle in lacustrine ecosystems. [ABSTRACT FROM AUTHOR]

The estimation of particulate organic carbon (POC) concentrations from satellite images can provide crucial spatiotemporal continuous observation data for the carbon cycle and ecological environmental governance. Here, we developed a novel inversion algorithm for deriving POC in inland water based on remote sensing and geochemical isotopes, which is summarized as follows. First, we developed empirical relationships between the phytoplankton absorption coefficient and endogenous POC concentration (C endPOC) and between the nonalgal particulate absorption coefficient and terrestrial POC concentration (C terPOC). Second, based on the valid relationships, semianalytical retrieval models were established to estimate C endPOC and C terPOC. Third, the proportions of endogenous POC (R endPOC) and terrestrial POC (R terPOC) to the total POC concentration (C POC) were derived using a three-band empirical model. Finally, C POC was obtained by dividing C endPOC by R endPOC (R endPOC ≥ 0.5) or dividing C terPOC by R terPOC (R endPOC < 0.5). Validation with field data shows that our proposed algorithm can accurately derive C POC (0–20 mg/L), with a root mean square deviation (RMSD), median bias (MB), median absolute percent difference (MAPD), and median ratio (MR) of 1.15 mg/L, −0.05 mg/L, 24%, and 0.98, respectively. Synchronous validation based on Sentinel-3/OLCI images confirmed the accuracy, with RMSD, MB, MAPD, and MR values of 0.41 mg/L, −0.16 mg/L, 28%, and 0.91, respectively. The algorithm was applied to ocean and land color sensor (OLCI) images to reveal the temporal and spatial variations in POC in Lake Taihu. • A novel bio-optical based POC semi-analytical inversion algorithm for inland water. • We improved POC estimation strategy by estimating POC terminal members. • a ph (674) and a nap (443) can be used to drive C endPOC and C terPOC , respectively. • Endogenous POC dominance in Lake Taihu increased significantly in summer. [ABSTRACT FROM AUTHOR]

The Secchi disk depth (Z SD ) plays a critical role in describing water clarity. Several studies have shown linkages between Three Gorges Dam (TGD) and the downstream lacustrine ecosystem in the middle and lower Yangtze River basin. However, the potential influence on the Z SD fluctuation in the entire anthropogenic reservoirs of Three Gorges (ER) and Dongting Lake (DTL) has not been reported, possibly due to technical obstacles in obtaining statistically significant spatial and temporal results. We addressed this challenge by using remote sensing technology: the Landsat 8 Operational Land Imager (OLI). We proposed a new, robust remote-sensing algorithm to estimate Z SD from OLI imagery using red and green band-ratio, leading to MAPE of 21.68% and RMSE of 0.076 m for Z SD ranging from 0.1 m to 1.05 m. After satisfactory image-based validation, the algorithm was implemented on OLI data to derive Z SD patterns over ER and DTL from 2013 to 2017. Several crucial findings can be drawn: 1) Spatial-temporal patterns of Z SD exhibited notable fluctuations over both ER and DTL, and they also demonstrated a significant correlation with each other because of the opposite temporal cycle of Z SD fluctuations between ER and DTL; 2) Temporally, monthly fluctuations of Z SD between ER and DTL had opposite temporal cycles, which was mainly attributed to the surface runoff and sediment discharge driven by the outbound runoff variations of TGD. Spatially, the heterogeneity of the Z SD pattern in ER might have resulted from the different geographical regions being divided by large anthropologic hydrological facilities, such as TGD; 3) The relationship between Z SD and total suspended matter (TSM) showed a significant negative correlation, as did the relationship between Z SD and K d (490). These findings demonstrate that TSM often plays a principal role in light attenuation of extremely turbid inland waters; 4) An inversed phenomenon of water clarity was observed at the intersection of DTL and the Yangtze River around Chenglingji site (YRAC), which was due to the opposite temporal cycle of Z SD fluctuations between DTL and ER after the impoundment of TGD; and 5) Owing to the analysis of noise-equivalent Z SD , OLI data can be used to derive Z SD, since the imagery uncertainty is 0.07 m by means of our band-ratio algorithm, which demonstrates similar results to MODIS. The proposed Z SD -derived algorithm in this study could be suitable for other turbid lakes or reservoirs to formulate related strategies of water quality management in the middle and lower Yangtze River basin, and the unveiled findings here improve our understanding of Z SD spatiotemporal fluctuations in large river-connected lakes, such as Poyang Lake. [ABSTRACT FROM AUTHOR]

Abstract: Accurate assessment of phytoplankton chlorophyll a (Chla) concentration in turbid waters by means of remote sensing is challenging due to optically complexity and significant variability of case 2 waters, especially in inland waters with multiple optical types. In this study, a water optical classification algorithm is developed, and two semi-analytical algorithms (three- and four-band algorithm) for estimating Chla are calibrated and validated using four independent datasets collected from Taihu Lake, Chaohu Lake, and Three Gorges Reservoir. The optical classification algorithm is developed using the dataset collected in Taihu Lake from 2006 to 2009. This dataset is also used to calibrate the three- and four-band Chla estimation algorithms. The optical classification technique uses remote sensing reflectance at three bands: Rrs(G), Rrs(650), and Rrs(NIR), where G indicates the location of reflectance peak in the green region (around 560nm), and NIR is the location of reflectance peak in the near-infrared region (around 700nm). Optimal reference wavelengths of the three- and four-band algorithm are located through model tuning and accuracy optimization. The three- and four-band algorithm accuracy is further evaluated using other three independent datasets. The improvement of optical classification in Chla estimation is revealed by comparing the performance of the two algorithms for non-classified and classified waters. Using the slopes of the three reflectance bands, the 138 reflectance spectra samples in the calibration dataset are classified into three classes, each with a specific spectral shape character. The three- and four-band algorithm performs well for both non-classified and classified waters in estimating Chla. For non-classified waters, strong relationships are yielded between measured and predicted Chla, but the performance of the two algorithms is not satisfactory in low Chla conditions, especially for samples with Chla below 30mg m−3. For classified waters, the class-specific algorithms perform better than for non-classified waters. Class-specific algorithms reduce considerable mean relative error from algorithms for non-classified waters in Chla predicting. Optical classification makes that there is no need to adjust the optimal position to estimate Chla for other waters using the class-specific algorithms. The findings in this study demonstrate that optical classification can greatly improve the accuracy of Chla estimation in optically complex waters. [Copyright &y& Elsevier]

Accurate estimation of phytoplankton chlorophyll a (Chla) concentration from remotely sensed data is particularly challenging in turbid, productive waters. The objectives of this study are to validate the applicability of a semi-analytical three-band algorithm in estimating Chla concentration in the highly turbid, widely variable waters of Taihu Lake, China, and to improve the algorithm using a proposed four-band algorithm. The improved algorithm is expressed as [Rrs(λ 1)−1 −Rrs(λ 2)−1][Rrs(λ 4)−1 −Rrs(λ 3)−1]−1. The two semi-analytical algorithms are calibrated and evaluated against two independent datasets collected from 2007 and 2005 in Taihu Lake. Strong linear relationships were established between measured Chla concentration and that derived from the three-band algorithm of [Rrs−1(660)−Rrs−1(692)]Rrs(740) and the four-band algorithm of [Rrs−1(662)−Rrs−1(693)][Rrs−1(740)−Rrs−1(705)]−1. The first algorithm accounts for 87% and 80% variation in Chla concentration in the 2007 and 2005 datasets, respectively. The second algorithm accounts for 97% of variability in Chla concentration for the 2007 dataset and 87% of variation in the 2005 dataset. The three-band algorithm has a mean relative error (MRE) of 43.9% and 34.7% for the 2007 and 2005 datasets. The corresponding figures for the four-band algorithm are 26.7% and 28.4%. This study demonstrates the potential of the four-band model in estimating Chla even in highly turbid case 2 waters. [Copyright &y& Elsevier]

Accurate remote assessment of phytoplankton chlorophyll- a (Chl a) concentration in turbid case-2 waters is a challenge, owing largely to terrestrial substances (such as minerals and humus) that are optically significant but do not co-vary with phytoplankton. Here, we propose an improved Quasi-Analytical Algorithm (QAA) (denoted as TC2) for retrieving Chl a concentrations from remote sensing reflectance (R rs (λ)) which can be applied to Sentinel-3 Ocean and Land Colour Instrument (OLCI) images in turbid case-2 waters. TC2 has two main extensions when compared with QAA. First, TC2 makes an additional assumption to separate the total non-water absorption at 665 nm (a nw (665)) into phytoplankton absorption (a ph (665)) and yellow matter (a ym (665)), which is the sum of colored dissolved matter (CDOM) and detritus. Second, for selecting the position of the near-infrared (NIR) band which is used to estimate the signal of total backscattering coefficient (b b (λ 0)) at QAA reference band (λ 0), we take into account the assumption that the absorption of pure water should be dominant at this band, as well as the impact of the signal-to-noise ratio (SNR) in the NIR band on the Chl a concentration estimating model. When applied to in situ R rs (λ) and OLCI match-up R rs (λ) data in this study, TC2 provided more accurate Chl a estimation than previous Ch a concentration retrieval algorithms for turbid case-2 waters. TC2 has the potential for use as a simple and effective algorithm for monitoring Chl a concentrations in the turbid case-2 waters at a global scale from space. • A new algorithm for estimating Chl a concentrations for turbid case-2 waters. • It makes an additional assumption to isolate Chl a absorption at the phytoplankton absorption peak of the red band. • The algorithm considers the impact of the SNR in the NIR band on the water total backscattering coefficient estimation. • It outperformed previous Chl a concentration retrieval algorithms using OLCI remote sensing reflectance. [ABSTRACT FROM AUTHOR]