Duan Hongtao

Spatio-temporal distribution of cyanobacteria blooms based on satellite imageries in Lake Taihu, China

Of all the cloudless 340 satellite imageries 11 were shot by Landsat MSS/TM/ETM sensor since 1979, 216 by EOS MODIS sensor since 2002, 10 by CEBERS CCD sensor since 2005 and 3 by IRS P6 LISS-3 sensor in 2007. On the basis of the spectral analysis, an applicable model was developed to extract the cyanobacteria bloom information from multi-source remote sensing images. And then, the model was used to acquire cyanobacteria bloom-covered area and its spatial distribution. The result showed that the initial time of cyanobacteria bloom was moved from June or July to March or April at present, three-four months advanced than the normal. The statistics showed that the cyanobacteria bloom occurs most frequently in June or July and the second was in October or November. Additionally, the duration of cyanobacteria bloom is becoming longer and longer and almost over time of March to December at present. The most intensity of cyanobacteria bloom occurred in September and the second was in June. The north of Lake Taihu, including Meiliang Bay and Zhushan Bay, was the initial location of cyanobacteria bloom and also the heavy disaster area, where the cyanobacteria bloom occurred every year since its first time. However, cyanobacteria bloom also occurred along the south bank of Lake Taihu almost every year since 2001, where the cyanobacteria bloom-covered area was becoming larger and larger and the duration time was becoming longer and longer, and which was becoming the earliest location of cyanobacteria bloom. It was noticeable that the cyanobacteria bloom was pervading gradually from the north, the west and the south to the center since 2003, and it sometimes covered almost the whole non-vegetation area. Additionally, there were cyanobacteria blooms occurring at Gonghu Bay since 2005, and it occurred more frequently in 2007.

Cyanobacteria bloom monitoring with remote sensing in Lake Taihu

It is significant that remote sensing methods is used for monitoring cyanobacteria bloom in Lake Taihu,since it breaks out frequently each year.Based on spectral characters of cyanobacteria bloom,different algorithm including single band,band subtraction and band ratio,were used for bloom mapping,with different instruments such as the MODIS/Terra,CBERS-2 CCD,ETM and IRS-P6.They noted that all these sensors were able to detect cyanobacteria bloom,while the algorithm of band ratio between infrared and red band has a stable correlation with blooms,and it can be developed into a universal pattern.Except that,spatial cyanobacteria bloom concentrations were separated into five classes based on digital number values(DNs)in ETM and IRS-P6 Band 4.This study showed that satellite observations was effectively applied to cyanobacteria bloom monitoring and early-warning for Lake Taihu.

Progress in lake water color remote sensing

We analyzed in detail the status quo of lake water color remote sensing from some aspects of satellite sensor, atmospheric correction, optical properties measurement, bio-optical model, radiative transfer model for the waters, and water quality parameter retrieval approach. It has a great difficulty to have practical application of water color remote sensing at a regional scale at present, depending on the complex components of lake water and on the inconsistency between satellite sensor and its actual demand from water quality monitoring. The progress in some key problems for lake water color remote sensing is still small, and there is a long way to go in applications of lake water color remote sensing. However, to be greatly pleasure, the satellite sensor and water color parameter retrieval approach are developing and making progresses, and the application in the future is hopeful.

Scale analysis of cyanobacteria bloom in Lake Taihu from MODIS observations

In order to study the spatio-temporal patterns of cyanobacteria blooms,it is important to detect and monitor them effectively by satellite observations.However,such detection and monitoring by low resolution data induces a scaling bias.Based on MODIS(250m and 500m) data at Oct.17,2005 and Dec.3,2010,the area of cyanobacteria bloom in Lake Taihu was derived by a approach named Floating Algae Index(FAI).The low resolution FAI was then achieved in two ways: FAI500,was directly calculated from MODIS(500m);and the FAImean was the mean of FAI250 which was directly calculated from MODIS(250m).Results reveal a serious overestimation of FAI500 and the area of cyanobacteria bloom due to the scale error all over Lake Taihu.The causes for scaling error are discussed and it is found that the spatial heterogeneous is the key factor which may lead to the error in the detection and monitoring on cyanobacteria bloom.

Cyanobacteria bloom detection and monitoring from satellite observations in the coastal region of Finland

This study investigates and analyses the possibility of cyanobacteria bloom detection as a case study in the Gulf of Finland and southern Finish lakes. The results show that MODIS data are difficult to be used in cyanobacteria bloom early warning. The main reason is that MODIS data are more suitable for CASE I waters as there are five bands in the spectral region of 400-550nm. However, for CASE II waters (estuary, coastal, and lakes), the maxima of reflectance spectra are shifting towards longer wavelengths with increasing turbidity or increasing amount of CDOM. This suggests that the reflectance maximum in coastal waters often occurs in the spectral range of 550-670nm where MODIS data do not have any spectral bands. In contrast, MERIS data have three bands of 560, 620, and 665nm at the spectral region of 550-670nm. The two bands at 620nm and 665nm are potentially important bands of detecting cyanobacteria bloom of the coastal waters (and/or lakes). The reason is that there are the phycocyanin absorption near 630nm and a reflectance peak near 650nm. In fact, MODIS and MERIS data are unlikely applicable to warn potentially cyanobacteria bloom in its early stages, but able to monitor the already formed blooms. This methodology of satellite-based observations can be applied to detection and monitoring of cyanobacteria bloom in Lake Taihu and coastal regions of China.

The neural network model for estimation of chlorophyll-a with water temperature in Lake Taihu

The advantage of neural network method for estimating water quality parameters of complex water body has been approved. Using in-situ measurement data of chlorophyll-a concentration, imageries of MODIS 250m and retrieval model of water temperature, we develop two single-hidden-layer BP neural network models for estimating chlorophyll-a in Lake Taihu: Model NN1 without temperature input and Model NN2 with temperature input. The training method is used by Levenberg-Marquardt algorithm, and the early-stage determinationin the modeling is used to improve generalization. The results show that: the estimation precision of the two models is high, in which the estimation precision of neural network input with temperature has been improved although the test is not significant.

Comparison of the extraction methods of phycocyanin pigments in eutrophic lake waters

The paper compares the effects of three methods of phycocyanin extraction buffer and tries to find out which method is the best. Cultured Microcystis aeruginosa and cyanobacteria blooms water samples of Lake Chaohu were used as extract objects. By repeated freezing and thawing method,cyanobacterial cells were broken,and then applied with Asolctin-CHAPS buffer,phosphate buffer and Tris-HCl buffer solution as extractant agent for extracting phycocyanin. Finally,spectrophotometry was used to detect the content of phycocyanin. We analyzed the absorption spectrum of the phycocyanin extract,the concentration of phycocyanin and the relevance between phycocyanin concentration and chlorophyll-a concentration to compare the quality of three methods. Absorption spectra show that all the characteristic absorption peaks of phycocyanin are emerged at 620 nm. Experimental results show that AC buffer and PBS buffer are better than the Tris-HCl buffer in extraction efficiency. In comparison,AC buffer is more expensive and difficult to obtain and save than the other two buffers,which makes it not suitable for large-scale use. So we recommend PBS buffer as the normal buffer which can meet the requirements of large-scale water quality monitoring.

Estimate of gas transfer velocity between water-air interface in a large lake based on different models: A case study of Lake Taihu

Gas transfer velocity ( k) is a factor driving Greenhouse gases exchange between water-air interface. But the estimate of k showed great uncertainly. To explore the control factors and variation of k,and select suitable k estimate model in large lake,three different parameterized equations were chosen for estimating k in Lake Taihu. Results indicated that velocity scales for wind shear and convection ( u* and w* ,respectively) showed diurnal cycle with high values during nighttime and low values during day, which were two important parameters in k estimate. As a result,k showed similar variation. In the large lake,the k was mainly dominated by wind shear,and wind speed was useful for estimating k. The annual mean value of k ranged from 1.27 to 1.46 m /d in Lake Taihu. Because of the spatial heterogeneity,the parameterized models in single site maybe not suitable for estimating k in other lakes. However,the lake area may be useful predictor for k.

The absorption spectral decomposition of water in Lake Taihu, China (II):the decomposition of absorption due to phytoplankton pigments

Concentrations of pigments could reflect the dominant phyla of algae.Investigations of algae in eutrophic lakes revealed that Chl.b,Chl.c and phycobilin(PC) are the critical pigments of chlorophyta,bacilliariphyta and cyanophyta,respectively.In this study,partial least square(PLS) regress was used on the retrieval of Chl.a,Chl.b,Chl.c and PC through absorption spectral of phytoplankton.Retrieve of Chl.a was based on the data collected in situ during 2011.Retrieve of Chl.b and Chl.c was based on the data collected in situ in March,since there is no significant dominant phyla of algae in spring.Result shows that PLS is a more effective method than the original least square regress and could be of help for remote sensing of multi-pigments and the distribution of main algae in eutrophic lakes.Concentrations of pigments could reflect the dominant phyla of algae.Investigations of algae in eutrophic lakes revealed that Chl.b,Chl.c and phycobilin(PC) are the critical pigments of chlorophyta,bacilliariphyta and cyanophyta,respectively.In this study,partial least square(PLS) regress was used on the retrieval of Chl.a,Chl.b,Chl.c and PC through absorption spectral of phytoplankton.Retrieve of Chl.a was based on the data collected in situ during 2011.Retrieve of Chl.b and Chl.c was based on the data collected in situ in March,since there is no significant dominant phyla of algae in spring.Result shows that PLS is a more effective method than the original least square regress and could be of help for remote sensing of multi-pigments and the distribution of main algae in eutrophic lakes.

Estimation of the contribution of chromophoric dissolved organic matter to total light absorption by remote sensing in Lake Taihu

Chromophoric dissolved organic matter(CDOM) mainly absorbs light in water which may influence the nature water color in lakes.Its absorption and photochemical degradation products play an important role in the primary productivity of water and carbon cycle.In Lake Taihu,a total of 333 sites were sampled in October 2004,October 2008,April 2010 and January and March 2011 to analyze the contribution of CDOM to total light absorption and estimate [aCDOM/at](412) from remote sensing.It was found that the average of [aCDOM/at](412) exhibited highly temporal variations during the five cruises.The maximum(0.369) was determined in 2011,comparing with all samples in Lake Taihu(0.295±0.139).The minimum average of [aCDOM/at](412) in the dataset 201004 was 0.236±0.108,varing from 0.046 to 0.455.No significant difference was observed in the dataset 200410 and 200810.The mean of [aCDOM/at](412) in Zhushan Bay was higher than that in both whole Lake Taihu and Meiliang Bay.For Meiliang Bay,it had almost the same value with the whole lake.A multi-band algorithm was adopted to estimate the [aCDOM/at](412) by remote sensing and acceptable results were detected(n=333,RMSE=34.60%).Suspended sediments and pigments had an important impact on determination of [aCDOM/at](412) from remote sensing.It was underestimated because of pigments and overestimated as the suspended sediments in water and the latter was worse.The results also showed that the CDOM and detritus optically dominate the water color in Lake Taihu.