Fangfang Zhang

Algorithms and Schemes for Chlorophyll a Estimation by Remote Sensing and Optical Classification for Turbid Lake Taihu, China

Monitoring chlorophyll a (CHLA) by remote sensing is particularly challenging for turbid productive waters. Although several empirical and semianalytical algorithms have been developed for such waters, their accuracy varies significantly due to variability in optical properties. In this paper, we evaluated the performance of six CHLA concentration (Cchla) estimation algorithms [e.g., two-band ratio algorithm (TBR), normalized difference chlorophyll index (NDCI), synthetic chlorophyll index (SCI), three-band algorithm (TBS), four-band algorithm (FBS), and improved four-band algorithm (IOC3M)] for a highly turbid lake based on remote sensing reflectance classification. Remote sensing reflectance was classified using the iterative k-mean clustering method. We also developed four estimation schemes (S1-S4) for the six algorithms to assess the effect of the estimation scheme on the accuracy of the algorithms. The estimation schemes were developed based on classification methods (no, soft, or hard classification) and the optimization bands used. The six algorithms performed differently for different remote sensing reflectance classes and different estimation schemes. The optimal algorithms for Classes 1, 2, and 3 were TBS, NDCI, and TBR, respectively. For the four estimation schemes, TBS and NDCI outperformed the other four algorithms. The accuracy of TBS and NDCI was higher than FBS, IOC3M, TBR, and SCI. The accuracy of all six algorithms was improved by remote sensing reflectance classification, particularly for Classes 2 and 3. Soft classification with recalibration of the bands for each class outperformed hard classification for all the three classes.

MODIS-Based Radiometric Color Extraction and Classification of Inland Water With the Forel-Ule Scale: A Case Study of Lake Taihu

Serious difficulties are present in the application of remote sensing techniques for optically complex waters, as retrieval of water quality parameters is generally based on detailed local knowledge of optical properties of water bodies for specific areas and at specific times. Water color is measured in traditional water quality observations and characterized by the Forel-Ule scale, as it is intimately related to water compositions. In this paper, a Moderate Resolution Imaging Spectroradiometer (MODIS) based water color extraction and classification approach is developed and applied to Lake Taihu. By using MODIS data together with field data, we attempted to 1) retrieve the dominant wavelength of water color and classify water color into FU-classes; 2) analyze the relationship between water color dominant wavelength and the abundance of optically active component (OACs) in water; and 3) discover seasonal variations of water color based on Lake Taihu. Our results show that the dominant wavelength exhibits some relationship with the three types of OAC concentrations under certain conditions, particularly TSM and Chl-a; inorganic suspended matter (ISM) can be retrieved by using MODIS derived dominant wavelength in appropriate water body. Moreover, differences in water quality for different seasons can be detected by dominant wavelength and FU-class with some prior knowledge of the studied water. Therefore, dominant wavelength may be used as a comprehensive and promising indicator of water quality situation even though much work has to be done in the future to optimize the analyses and verify it on diverse sites.

Synthetic aperture radar detection and characteristic analysis of cyanobacterial scum in Lake Taihu

Abstract. To compensate for the limitations of optical remote sensing when restricted by cloud cover, it is worth exploring how to detect cyanobacterial blooms using synthetic aperture radar (SAR), which can penetrate clouds. A satellite–ground synchronous experiment was conducted in Lake Taihu, the third largest freshwater lake in China. A lipopeptide biosurfactant was detected in the algal scum layer, with an average content of 1.8  g/L. The viscosity (1.41 to 332 mPa.s) of the scum was significantly higher than that of scum-free water. The surface tension of the algal scum decreased by 12.5%, and the SAR microwave backscatter was reduced by 7.3 dB. This indicated that the cyanobacterial scum could effectively attenuate capillary waves and appear as dark patches in SAR images. SAR has the potential to be developed as a tool for the remote sensing of algal scum in lake waters.

MODIS observations of water color of the largest 10 lakes in China between 2000 and 2012

ABSTRACT Forel-Ule (FU) index of water color is an important parameter in traditional water quality investigations. We retrieved the FU index of the largest 10 lakes in China during 2000-2012 from MODerate-resolution Imaging Spectroradiometer surface reflectance product (MOD09) images. Since FU index is an optical parameter, it can be derived from optical remote sensing data by direct formulas, which is invariant with region and season. Based on validation by in situ measured reflectance data, the FU index products are reliable, with average relative error of 7.7%. FU index can be used to roughly assess water clarity: the clearer a water body is, and the bluer it is in color, the smaller its FU index is. FU index can also be used to roughly classify trophic state into three classes: oligotrophic, mesotrophic, and eutrophic. We analyzed the spatial, interannual, and seasonal variations of the FU index and its implications for water clarity and trophic state, and the findings are mostly consistent with the results from related literature. All in all, it might be a feasible way to roughly assess inland water quality by FU index in large region and over long time period.

A simple automated dynamic threshold extraction method for the classification of large water bodies from landsat-8 OLI water index images

ABSTRACT Traditional manual methods of extracting water bodies from remote sensing images cannot satisfy the requirements for mass processing of remote sensing data, and new automated methods are complicated and require a large amount of auxiliary data. The histogram bimodal method is a frequently used objective tool for threshold selection in image segmentation. However, automatically calculating the threshold is difficult because of complex surfaces and image noise, which lead to imperfect twin peaks. To overcome these difficulties, we developed an operational automated water extraction method. This method does not require the identification of twin histogram peaks but instead seeks minimum values in the threshold range to achieve an automated dynamic threshold. We calibrated the method for 18 lakes in China using Landsat 8 Operational Land Imager images, for which the relative error (RE) and coefficient of determination (R2) for threshold accuracy were 2.1% and 0.96, respectively. The RE of area accuracy was 0.59%. The advantages of the method lie in its simplicity and minimal requirements for auxiliary data while still achieving an accuracy comparable to that of other automatic water extraction methods. It can be applied to mass remote sensing data to calculate water thresholds and automatically extract large water bodies.

Monitoring cyanobacteria-dominant algal blooms in eutrophicated Taihu Lake in China with synthetic aperture radar images

Monitoring algal blooms by optical remote sensing is limited by cloud cover. In this study, synthetic aperture radar (SAR) was deployed with the aim of monitoring cyanobacteria-dominant algal blooms in Taihu Lake in cloudy weather. The study shows that dark regions in the SAR images caused by cyanobacterial blooms damped the microwave backscatter of the lake surface and were consistent with the regions of algal blooms in quasi-synchronous optical images, confirming the applicability of SAR for detection of surface blooms. Low backscatter may also be associated with other factors such as low wind speeds, resulting in interference when monitoring algal blooms using SAR data alone. After feature extraction and selection, the dark regions were classified by the support vector machine method with an overall accuracy of 67.74%. SAR can provide a reference point for monitoring cyanobacterial blooms in the lake, particularly when weather is not suitable for optical remote sensing. Multi-polarization and multi-band SAR can be considered for use in the future to obtain more accurate information regarding algal blooms from SAR data.

MODIS surface reflectance product (MOD09) validation for typical inland waters in China

This paper is aiming at the problem that the MODIS surface reflectance product (MOD09) does not offer an accurate aerosol correction for inland water, for the constraints of MODIS atmospheric correction algorithm. In-situ data collected in Taihu Lake and Yuqiao Reservoir were used to validate and assess the quasi-synchronous MOD09 product. The results showed that there is linear relationship on the whole between MOD09 bands and in-situ data in inland water with acceptable deviation level. The reason for the deviations is analyzed primarily and a simple correction for MOD09 product in Lake Taihu is introduced based on bands calculation. The results also illustrated that it is possible to monitor inland water quality globally with MOD09 product by providing validation evidence in typical inland waters. And it would be most accurate by using bands ratio algorithm for the water quality retrieval using MOD09. The validation is also important to improve atmospheric algorithms of MODIS.

Validation of a synthetic chlorophyll index for remote estimates of chlorophyll-a in a turbid hypereutrophic lake

Remote sensing techniques can offer powerful tools for measuring concentrations of chlorophyll-a (chl-a), which is an important proxy for water quality. However, remote estimates of chl-a can be difficult in water bodies that have high levels of total suspended matter (TSM). In this study, we examined the applicability of the synthetic chlorophyll index (SCI) and a parameter relevant to chlorophyll pigments (Hchl) used in conjunction with remote-sensing data to predict chl-a concentrations (Cchl-a) in Taihu Lake, a highly turbid hypereutrophic lake in eastern China. We sampled water quality and surface spectral properties at 250 field stations throughout the lake over five sampling periods spanning 2 years. Because data acquired at 31 stations could not be used due to equipment failure or blue-green algal blooms, we used data acquired at the remaining 219 stations. We then randomly selected parts of the spectral properties data (N = 164) to calibrate bands used in the SCI algorithm and established cubic polynomial models to estimate Cchl-a with SCI and Hchl as the independent variables. We evaluated the accuracy of these models using data from the remaining 55 stations that were not used for calibration. Our results showed the following trends: (1) the parameter of Hchl performed better than SCI in estimating Cchla in Taihu Lake; (2) Hchl showed optimal performance in winter, average performance in spring, and poor performance in summer and autumn; (3) Hchl was appropriate for the NAP-dominant waters with high CTSM and low Cchl-a, but was not suitable for organism-dominant waters with low CTSM; and (4) in short, Hchl had limited usability in turbid and eutrophic waters.

Modification of 6SV to remove skylight reflected at the air-water interface: Application to atmospheric correction of Landsat 8 OLI imagery in inland waters

During the atmospheric correction of remote sensing data in inland waters, the original Second Simulation of the Satellite Signal in the Solar Spectrum-Vector version (6SV) model does not eliminate the specular reflection of downward skylight radiance at the air-water interface. Thus, we propose a modified version of the 6SV model (M6SV) that does remove reflected skylight at the air-water interface. We apply the new model to the atmospheric correction of a Landsat 8 Operational Land Imager (OLI) image over Taihu Lake, China, where the aerosol optical depth is known. In situ reflectance measurements acquired concurrently with the L8/OLI image are used to validate the performance of the new M6SV algorithm. To further analyze the merits and demerits of M6SV, the model is compared with two short-wave infrared (SWIR)-based atmospheric correction models: the Sea-Viewing Wide Field-of-View Sensor Data Analysis System short-wave infrared (SD-SWIR) model and the Vanhellemont & Ruddick short-wave infrared with a per scene fixed aerosol type (VR-SWIR-F) model. Comparisons of results from all three L8/OLI image atmospheric corrections with the in situ remote sensing reflectance data show that M6SV produces reliable atmospheric corrections in the green and red spectral bands and is an effective alternative for Landsat 8 OLI atmospheric correction in inland waters.

Landsat-satellite-based analysis of spatial–temporal dynamics and drivers of CyanoHABs in the plateau Lake Dianchi

ABSTRACT Dianchi Lake, located in southwest China’s Yungui plateau, is facing severe eutrophication and frequent outbreaks of harmful cyanobacteria blooms (CyanoHABs). It is of great significance to monitor the occurrence and development of CyanoHABs in Dianchi Lake over a long period and analyse the main influences. Based on Landsat Thematic Mapper/Enhanced Thematic Mapper Plus/Operational Land Imager 1986–2016 data, we derived the distribution of the CyanoHABs in Dianchi Lake, and analysed spatial–temporal dynamics of the CyanoHABs by correlation with nutrition, meteorological, and humanities data. The results showed that the first outbreak of CyanoHABs in Dianchi Lake occurred in 1987, which is likely to be influenced by a rapid increase of nutrients in the lake, while the weather conditions also have some impact on the CyanoHABs occurrence. After 1990, the frequency of CyanoHABs is relatively high in the water near Longmen village, Fubao Bay, Hui Bay, and the lake inlet of the Panlong River to the north of Waihai in Dianchi Lake from June to November every year. Moreover, the CyanoHABs increased year by year until 2000. This is closely related to population growth and economic development. Furthermore, a large amount of precipitation and small wind speeds can also promote the occurrence of CyanoHABs. After 2000, the frequency of CyanoHABs decreased, as the large-scale management of water pollution in Dianchi Lake achieved certain effects. The area and frequency of CyanoHABs from 2011 to 2014 are the smallest in the last 20 years, which may be related to the large-scale planting of Eichhornia crassipes in the north of Dianchi Lake.