A four-band semi-analytical model for estimating chlorophyll a in highly turbid lakes: The case of Taihu Lake, China

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(λ₁)^− 1 − Rrs(λ₂)^− 1][Rrs(λ₄)^− 1 − Rrs(λ₃)^− 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.     

Influence of phytoplankton pigment composition on remote sensing of cyanobacterial biomass

An extensive field campaign was carried out for the validation of a previously published reflectance ratio-based algorithm for quantification of the cyanobacterial pigment phycocyanin (PC). The algorithm uses band settings of the Medium Resolution Imaging Spectrometer (MERIS) onboard ENVISAT, and should accurately retrieve the PC concentration in turbid, cyanobacteria-dominated waters. As algae and cyanobacteria often co-occur, the algorithm response to varying phytoplankton composition was explored. Remote sensing reflectance and reference pigment measurements were obtained in the period 2001-2005 in Spain and the Netherlands using field spectroradiometry and various pigment extraction methods. Additional field data was collected in Spain in May 2005 to allow intercalibration of spectroradiometry and pigment assessment methods. Two methods for extraction of PC from concentrated water samples, and in situ measured PC fluorescence, compared well. Reflectance measurements with different field spectroradiometers used in Spain and the Netherlands also gave similar results. Residual analysis of PC predicted by the algorithm showed that overestimation of PC mainly occurred in the presence of chlorophylls b and c, and phaeophytin. The errors were strongest at low PC relative to Chl a concentrations. A correction applied for absorption by Chl b markedly improved the prediction. Without such a correction, the quality of the PC prediction still increased markedly with estimates > 50 mg PC m^− 3, allowing monitoring of the cyanobacterial status of eutrophic waters. The threshold concentration may be lowered when a high intracellular PC:Chl a ratio or cyanobacterial dominance is expected. Below the limit, predicted PC concentrations should be considered as the highest estimate. We evaluated that remote sensing of both PC and Chl a would allow assessment of cyanobacterial risk to water quality and public health in over 70% of our cases.     

Remote estimation of chlorophyll a in optically complex waters based on optical classification

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 560 nm), and NIR is the location of reflectance peak in the near-infrared region (around 700 nm). 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 30 mg 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.     

Hyperspectral remote sensing of cyanobacterial pigments as indicators for cell populations and toxins in eutrophic lakes

The growth of mass populations of toxin-producing cyanobacteria is a serious concern for the ecological status of inland waterbodies and for human and animal health. In this study we examined the performance of four semi-analytical algorithms for the retrieval of chlorophyll a (Chl a) and phycocyanin (C-PC) from data acquired by the Compact Airborne Spectrographic Imager-2 (CASI-2) and the Airborne Imaging Spectrometer for Applications (AISA) Eagle sensor. The retrieval accuracies of the semi-analytical models were compared to those returned by optimally calibrated empirical band-ratio algorithms. The best-performing algorithm for the retrieval of Chl a was an empirical band-ratio model based on a quadratic function of the ratio of reflectance at 710 and 670 nm (R² = 0.832; RMSE = 29.8%). However, this model only provided a marginally better retrieval than the best semi-analytical algorithm. The best-performing model for the retrieval of C-PC was a semi-analytical nested band-ratio model (R² = 0.984; RMSE = 3.98 mg m⁻³). The concentrations of C-PC retrieved using the semi-analytical model were correlated with cyanobacterial cell numbers (R² = 0.380) and the particulate and total (particulate plus dissolved) pools of microcystins (R² = 0.858 and 0.896 respectively). Importantly, both the empirical and semi-analytical algorithms were able to retrieve the concentration of C-PC at cyanobacterial cell concentrations below current warning thresholds for cyanobacteria in waterbodies. This demonstrates the potential of remote sensing to contribute to early-warning detection and monitoring of cyanobacterial blooms for human health protection at regional and global scales.     

An evaluation of algorithms for the remote sensing of cyanobacterial biomass

Most remote sensing algorithms for phytoplankton in inland waters aim at the retrieval of the pigment chlorophyll a (Chl a), as this pigment is a useful proxy for phytoplankton biomass. More recently, algorithms have been developed to quantify the pigment phycocyanin (PC), which is characteristic of cyanobacteria, a phytoplankton group of relative importance to inland water management due to their negative impact on water quality in response to eutrophication.We evaluated the accuracy of three published algorithms for the remote sensing of PC in inland waters, using an extensive database of field radiometric and pigment data obtained in the Netherlands and Spain in the period 2001–2005. The three algorithms (a baseline, single band ratio, and a nested band ratio approach) all target the PC absorption effect observed in reflectance spectra in the 620 nm region. We evaluated the sensitivity of the algorithms to errors in reflectance measurements and investigated their performance in cyanobacteria-dominated water bodies as well as in the presence of other phytoplankton pigments.All algorithms performed best in moderate to high PC concentrations (50–200 mg m^? 3) and showed the most linear response to increasing PC in cyanobacteria-dominated waters. The highest errors showed at PC < 50 mg m^? 3. In eutrophic waters, the presence of other pigments explained a tendency to overestimate the PC concentration. In oligotrophic waters, negative PC predictions were observed. At very high concentrations (PC > 200 mg m^? 3), PC underestimations by the baseline and single band ratio algorithms were attributed to a non-linear relationship between PC and absorption in the 620 nm region. The nested band ratio gave the overall best fit between predicted and measured PC. For the Spanish dataset, a stable ratio of PC over cyanobacterial Chl a was observed, suggesting that PC is indeed a good proxy for cyanobacterial biomass. The single reflectance ratio was the only algorithm insensitive to changes in the amplitude of reflectance spectra, which were observed as a result of different measurement methodologies.     

持续有界扰动系统的嵌套不变椭圆集鲁棒控制

为了抑制外部持续有界扰动和模型不确定性对系统稳定性控制的影响,通过不变集理论,采用嵌套不变椭圆集鲁棒控制算法实现系统的快速稳定控制。控制算法分为离线算法和在线算法两部分。离线时根据公式得到一维状态变量序列,通过线性矩阵不等式方法优化得到嵌套不变椭圆集。在线时,根据系统状态变量在嵌套不变椭圆集的位置,构建新的不变椭圆集并计算得到系统的控制律。给出新的不变椭圆集满足系统控制要求的理论证明。通过与不变单椭圆集控制算法进行仿真比较,结果验证了上述算法的有效性,为持续有界扰动下模型不确定性系统的稳定控制,提供一种有效的控制方法。     

Intercomparison of ocean colour band-ratio algorithms for chlorophyll concentration in the Subtropical Front east of New Zealand

This study intercompared the performance of eight band-ratio chlorophyll-a algorithms which together can be used to process measurements from the ocean colour satellite sensors CZCS, OCTS, SeaWiFS, MODIS, MERIS, and GLI. The study area included Subtropical, Subtropical Front and Subantarctic waters east of New Zealand, and Case 1 waters of the New Zealand northeast continental shelf. Over 170 co-incident measurements of spectral normalised water-leaving radiance and near-surface concentration of chlorophyll-a were made on nine research voyages between 1998 and 2000. The studentised bootstrap method was used to identify statistically significant bias in algorithm products relative to in situ measurements. The band-ratio algorithms used by CZCS, OCTS and SeaWiFS missions systematically underestimated chlorophyll-a concentration in the offshore regions by between 21% and 45%, but showed no systematic bias in the continental shelf waters. The band-ratio algorithms applicable to the MODIS and MERIS sensors had no clear bias with respect to in situ measurements in offshore waters, but had a positive bias of 20% over the continental shelf. The proposed GLI band-ratio algorithm led to estimates that were negatively biased with respect to in situ measurement offshore (− 30%), and positively biased over the continental shelf (20%). The results were consistent with unusually high values of absorption in the blue part of the spectrum (443-490 nm) compared to the green part (∼ 550 nm) by phytoplankton pigments in the offshore waters, and high chlorophyll-specific absorption over the continental shelf.     

Multispectral derivation of bathymetry in Singapore's shallow,turbid waters

In very shallow waters, active sensing determinations of bathymetry are often expensive and unwieldy. Sea depth estimation using passive remote-sensing methods is an attractive alternative, especially using cheap multispectral imagery with high spatial resolution. Three models for the determination of bathymetry from multispectral imagery were utilized with new eight-band images from DigitalGlobe's Worldview-2 satellite platform. All three were trained with electronic navigational chart data and evaluated for accuracy in Singapore's turbid shallow coastal waters. These waters are characterized by high turbidity, suspended sediment, and vehicle traffic. Of the three models, a linear band algorithm performed best, with a root-mean-square error (RMSE) of 0.48 m. A look-up table classification provided a precision of 0.64 m, but was limited by a training set that did not fully represent variance in water column and benthic properties. Possibly owing to the domination of particle backscatter over pigment absorption in these turbid waters, a linear ratio algorithm did not perform as well as the linear band algorithm, achieving an RMSE of only 0.56 m. Analysis found that the usual relationship between ratios of low-absorption to high-absorption bands and depth does not hold as well for these waters, likely due to backscatter dominating leaving-water signals, masking relative absorption effects. High turbidity, with a Secchi disk depth of 1.9 m, limited analysis to shallow reefs and coastline and likely impacted the sensitivity of the bathymetric algorithms. A larger validation data set containing water quality and benthic data is required for further investigation to determine specific sources of error.     

Hyperspectral remote sensing of cyanobacteria in turbid productive water using optically active pigments,chlorophyll a and phycocyanin

Nuisance blue-green algal blooms contribute to aesthetic degradation of water resources by means of accelerated eutrophication, taste and odor problems, and the production of toxins that can have serious adverse human health effects. Current field-based methods for detecting blooms are costly and time consuming, delaying management decisions. Methods have been developed for estimating phycocyanin concentration, the accessory pigment unique to freshwater blue-green algae, in productive inland water. By employing the known optical properties of phycocyanin, researchers have evaluated the utility of field-collected spectral response patterns for determining concentrations of phycocyanin pigments and ultimately blue-green algal abundance. The purpose of this research was to evaluate field spectroscopy as a rapid cyanobacteria bloom assessment method. In-situ field reflectance spectra were collected at 54 sampling sites on two turbid reservoirs on September 6th and 7th in Indianapolis, Indiana using ASD Fieldspec (UV/VNIR) spectroradiometers. Surface water samples were analyzed for in-vitro pigment concentrations and other physical and chemical water quality parameters. Semi-empirical algorithms by Simis et al. [Simis, S., Peters, S., Gons, H. (2005). Remote sensing of the cyanobacterial pigment phycocyanin in turbid inland water. American Society of Limnology and Oceanography 50(11): 237–245] were applied to the field spectra to predict chlorophyll a and phycocyanin absorption at 665 nm and 620 nm, respectively. For estimation of phycocyanin concentration, a specific absorption coefficient of 0.0070 m² mg PC^-1 for phycocyanin at 620 nm, a_PC^?(620), was employed, yielding an r² value of 0.85 (n = 48, p < 0.0001), mean relative residual value of 0.51 (σ = 1.41) and root mean square error (RMSE) of 19.54 ppb. Results suggest this algorithm could be a robust model for estimating phycocyanin. Error is highest in water with phycocyanin concentrations of less than 10 ppb and where phycocyanin abundance is low relative to chlorophyll a. A strong correlation between measured phycocyanin concentrations and biovolume measurements of cyanobacteria was also observed (r = 0.89), while a weaker relationship (r = 0.66) resulted between chlorophyll a concentration and cyanobacterial biovolume.     

Phycocyanin detection from LANDSAT TM data for mapping cyanobacterial blooms in Lake Erie

Algorithms were developed from LANDSAT 7 ETM+ data for the July 1, 2000 overpass and LANDSAT 5 Thematic Mapper (TM) data for the September 27, 2000 overpass for Path 20 Row 31 (including Toledo, OH) to measure relative phycocyanin content (PC) and turbidity in the western basin of Lake Erie. Water samples were collected from discrete hydrographic stations arranged in a 20×4 km grid adjacent to the Ohio shoreline during a 6-h period spanning each of the two LANDSAT overpasses. The samples were analyzed for chlorophyll (chl) a content and turbidity. In addition, the concentration of phycocyanin, a light-harvesting pigment associated with cyanobacteria, was estimated from the ratio of phycocyanin/chl a in vivo fluorescence (IVPF/IVCF). A dark-object-subtracted, spectral ratio model derived from the July 1, 2000 data was found to be the most robust, when applied to the September 27, 2000 data. The same July 1, 2000 model (or algorithm) for PC was then applied to LANDSAT 7 ETM+ frames for July 16 and August 1, 2002 of the Path 19 Row 31 frame (including Cleveland, OH) and to the August 8, 2002 frame of Path 20 Row 31. Moderate, very low, and high PC values were detected in the western basin of Lake Erie on July 16, August, 1, and August 8, 2002, respectively. On September 17, 2002, local media reported a large Microcystis bloom in the western basin. The high PC values on August 8, 2002 may have represented early stage detection of the large Microcystis bloom that was reported 5 weeks later. The PC algorithm derived in this study will improve our understanding of the temporal and spatial dynamics of cyanobacterial bloom formation in Lake Erie and other systems. It may also serve to alert municipalities to the presence of potentially toxic bloom events.     

Towards a practical remote-sensing model of suspended sediment concentrations in turbid waters using MERIS measurements

A new empirical index, termed the normalized suspended sediment index (NSSI), is proposed to predict total suspended sediment (TSS) concentrations in inland turbid waters using Medium Resolution Imaging Spectrometer (MERIS) full-resolution (FR) 300 m data. The algorithm is based on the normalized difference between two MERIS spectral bands, 560 and 760 nm. NSSI shows its potential in application to our study region – Poyang Lake – the largest freshwater lake in China. An exponential function (R² = 0.90, p < 0.01) accurately explained the variance in the in situ data and showed better performance for the TSS range 10–524 mg l^?1. The algorithm was then validated with TSS estimates using an atmospheric-corrected MERIS FR image. The validation showed that the NSSI algorithm was a more robust TSS algorithm than the band-ratio algorithms. Findings of this research imply that NSSI can be successfully used on MERIS images to obtain TSS in Poyang Lake. This work provided a practical remote-sensing approach to estimate TSS in the optically and hydrologically complex Poyang Lake and the method can be easily extended to other similar waters.     

Water property monitoring and assessment for China's inland Lake Taihu from MODIS-Aqua measurements

We provide results of quantitative measurements and characterization for inland freshwater Lake Taihu from the Moderate Resolution Imaging Spectroradiometer (MODIS) on the satellite Aqua. China's Lake Taihu, which is located in the Yangtze River delta in one of the world's most urbanized and heavily populated areas, contains consistently highly turbid waters in addition to frequent large seasonal algae blooms in various lake regions. Thus, satellite data processing requires use of the shortwave infrared (SWIR) atmospheric correction algorithm. Specifically for Lake Taihu, an iterative SWIR-based atmospheric correction algorithm has been developed and proven to provide reasonably accurate water-leaving radiance spectra data. Using MODIS-Aqua measurements, the blue-green algae bloom in Lake Taihu in 2007 has been studied in detail, demonstrating the importance and usefulness of satellite water color remote sensing for effectively monitoring and managing a bloom event.Seasonal and interannual variability, as well as spatial distributions, of lake water properties were studied and assessed using the MODIS-Aqua measurements from 2002 to 2008. Results show that overall waters in Lake Taihu are consistently highly turbid all year round, with the winter and summer as the most and least turbid seasons in the lake, respectively. Extremely turbid waters in the winter are primarily attributed to strong winter winds that lead to significant amounts of total suspended sediment (TSS) in the water column. In addition, MODIS-Aqua-measured water-leaving radiance at the blue band is consistently low in various bay regions in Lake Taihu, indicating high algae concentration in these regions. Climatological water property maps, including normalized water-leaving radiance spectra nL_w(λ), chlorophyll-a concentration, and water diffuse attenuation coefficient at the wavelength of 490 nm (K_d(490)), are derived from all MODIS-Aqua data from 2002 to 2008 for Lake Taihu, showing overall spatial distribution features for the lake water property.     

Spectral variations in the near-infrared ocean reflectance

The optical properties of natural waters beyond the visible range, in the near-infrared (NIR, 700-900 nm), have received little attention because they are often assumed to be mostly determined by the large absorption coefficient of pure water, and because of methodological difficulties. It is now growingly admitted that the NIR represents a potential optical source of unambiguous information about suspended sediments in turbid waters, thence the need for better understanding the NIR optical behaviour of such waters. It has recently been proposed (Ruddick et al., Limnology and Oceanography. 51, 1167-1179, 2006) that the variability in the shape of the surface ocean reflectance spectrum in the NIR is negligible in turbid waters. In the present study, we show, based on both in situ and remote sensing data, that the shape of the ocean reflectance spectrum in the NIR does vary in turbid to extremely turbid waters. Space-borne ocean reflectance data were collected using 3 different sensors (SeaWiFS, MODIS/Aqua and MERIS) over the Amazon, Mackenzie and Rio de la Plata turbid river plumes during extremely clear atmospheric conditions so that reliable removal of gas and aerosol effects on reflectance could be achieved. In situ NIR reflectance data were collected in different European estuaries where extremely turbid waters were found. In both data sets, a flattening of the NIR reflectance spectrum with increasing turbidity was observed. The ratio of reflectances at 765 nm and 865 nm, for instance, varied from ca. 2 down to 1 in our in situ data set, while a constant value of 1.61 had been proposed based on theory in a previous study. Radiative transfer calculations were performed using a range of realistic values for the seawater inherent optical properties, to determine the possible causes of variations in the shape of the NIR reflectance spectrum. Based on these simulations, we found that the most significant one was the gradual increase in the contribution of suspended sediments to the color of surface waters, which often leads to the flattening of the reflectance spectrum. Changes in the scattering and absorption properties of particles also contributed to variations in the shape of the NIR surface ocean reflectance spectrum. The impact of such variations on the interpretation of ocean color data is discussed.     

Assessing the potential of SeaWiFS and MODIS for estimating chlorophyll concentration in turbid productive waters using red and near-infrared bands

Bio-optical algorithms for remote estimation of chlorophyll-a concentration (Chl) in case-1 waters exploit the upwelling radiation in the blue and green spectral regions. In turbid productive waters other constituents, that vary independently of Chl, absorb and scatter light in these spectral regions. As a consequence, the accurate estimation of Chl in turbid productive waters has so far not been feasible from satellite sensors. The main purpose of this study was to evaluate the extent to which near-infrared (NIR) to red reflectance ratios could be applied to the Sea Wide Field-of-View Sensor (SeaWiFS) and the Moderate Imaging Spectrometer (MODIS) to estimate Chl in productive turbid waters. To achieve this objective, remote-sensing reflectance spectra and relevant water constituents were collected in 251 stations over lakes and reservoirs with a wide variability in optical parameters (i.e. 4 ≤ Chl ≤ 240 mg m^− 3; 18 ≤ Secchi disk depth ≤ 308 cm). SeaWiFS and MODIS NIR and red reflectances were simulated by using the in-situ hyperspectral data. The proposed algorithms predicted Chl with a relative random uncertainty of approximately 28% (average bias between − 1% and − 4%). The effects of reflectance uncertainties on the predicted Chl were also analyzed. It was found that, for realistic ranges of R_rs uncertainties, Chl could be estimated with a precision better than 40% and an accuracy better than ± 35%. These findings imply that, provided that an atmospheric correction scheme specific for the red-NIR spectral region is available, the extensive database of SeaWiFS and MODIS images could be used to quantitatively monitor Chl in turbid productive waters.     

The potential use of geostationary MTG/FCI to retrieve chlorophyll-a concentration at high temporal resolution for the open oceans

In a few years, the Flexible Combined Imager (FCI) on-board Meteosat Third Generation will provide images of European Seas, the Atlantic Ocean, and the Mediterranean Sea every 2.5 min (regions above 30° N) or 10 min (full disk). Although dedicated to meteorological applications, this sensor has blue, green, and red spectral bands allowing to consider the adaptation of a band-ratio algorithm to retrieve chlorophyll-a concentration (chl-a). However, the radiometric specification of the FCI sensor is far from the minimum requirement recommended for ocean colour sensors and the validity of FCI data for oceanic applications is not clear. This present article aims to determine if, and under which conditions, chl-a could be estimated from FCI data. From the National Aeronautics and Space Administration bio-Optical Marine Algorithm data set in situ data set, a blue green band-ratio algorithm adapted to FCI spectral characteristics is proposed. Then, the impact of FCI radiometric noise on chl-a estimations is investigated in detail. Results show that noise-induced chl-a error increases with chl-a and solar zenith angle. For a chl-a estimation based on a unique pixel, this error ranges between 20% and 100% which prevents any direct utilisation and suggests that it is necessary to degrade the spatio-temporal resolution to obtain an acceptable noise-related uncertainty on chl-a. With a spatial (9 pixels) and temporal (1 h) averaging process, chl-a can be estimated with a noise-induced error less than 10% for chl-a up to 5 mg m^?3 and solar zenith angle lower than 60°. Our analysis also showed that the noise-related error associated to the atmospheric correction process can be neglected compared to the radiometric noise of the visible bands themselves if it is assumed that aerosol type is uniform over large areas (9 km × 9 km boxes).     

动态输出反馈鲁棒模型预测控制离线算法

研究具有多包不确定性和有界噪声系统的动态输出反馈鲁棒模型预测控制(Robust model predictive control, RMPC)的离线方法. 先前的在线方法中, 在估计状态和估计误差集合已知的情况下, 在每一采样时刻通过近似最优算法求解控制器参数. 本文采用先前的方法计算离线控制器参数和吸引域. 首先, 选定一系列估计状态, 其中,每个估计状态对应同样一组嵌套的估计误差集合. 然后,针对每一估计状态和每一估计误差集合的组合,离线计算唯一的控制器参数和对应的吸引域. 这些控制器参数和对应的吸引域存储在表中. 如果离线确定的吸引域包含实时的扩展状态, 则该离线控制器参数是实时可行的. 在线时, 根据实时估计状态和选取实时估计误差集合, 在表中搜索包含实时扩展状态且优化性能指标最小的吸引域所对应的控制器参数. 通过连续搅拌釜式反应器控制系统验证了该方法的有效性.     

Measurements and analysis of in situ multi-angle reflectance of turbid inland water: a case study in Meiliang Bay,Taihu Lake,China

The bidirectional reflectance properties of the anisotropic light field above the water surface are important for a range of applications. The bidirectional reflectance distribution function of oceanic waters has been well characterized but there is a lack of information for turbid inland waters. In addition, there is a lack of bidirectional reflectance data measured in turbid inland waters partially due to the difficulty in collecting in situ water-surface multi-angle remote-sensing reflectance data. To facilitate bidirectional reflectance studies of turbid inland waters using in situ multi-angular reflectance data, we have designed and developed a simple hand-held 3D positioning pole to position the spectrometer optical fibre probe and a specific method to collect the multi-angular reflectance data above the water surface with this pole. Using this device, we collected multi-angular reflectance data in Meiliang Bay, Taihu Lake, China, and analysed the uncertainties in this method. We analysed the bidirectional distribution characteristics of the data, and compared the findings to those in the literature. Both uncertainty analysis and bidirectional distribution characteristics analysis showed that our method is effective in collecting multi-angular reflectance above the water surface and can be applied to validate bidirectional correction models in the future.     

Estimation of cyanobacterial pigments in a freshwater lake using OCM satellite data

Cyanobacteria represent a major harmful algal group in fresh to brackish water environments. Lac des Allemands, a freshwater lake of 49 km² southwest of New Orleans, Louisiana on the upper end of the Barataria Estuary, provides a natural laboratory for remote characterization of cyanobacterial blooms because of their seasonal occurrence. The Oceansat-1 satellite Ocean Colour Monitor (OCM) provides measurements similar to SeaWiFS but with higher spatial resolution, and this work is the first attempt to use OCM measurements to quantify cyanobacterial pigments. The satellite signal was first vicariously calibrated using SeaWiFS as a reference, and then corrected to remove the atmospheric effects using a customized atmospheric correction procedure. Then, empirical inversion algorithms were developed to convert the OCM remote sensing reflectance (R_rs) at bands 4 and 5 (centered at 510.6 and 556.4 nm, respectively) to concentrations of phycocyanin (PC), the primary cyanobacterial pigment. A holistic approach was used to minimize the influence of other optically active constituents on the PC algorithm. Similarly, empirical algorithms to estimate chlorophyll a (Chl a) concentrations were developed using OCM bands 5 and 6 (centered at 556.4 and 669 nm, respectively). The best PC algorithm (R² = 0.7450, p < 0.0001, n = 72) yielded a root mean square error (RMSE) of 36.92 μg/L with a relative RMSE of 10.27% (PC from 2.75 to 363.50 μg/L, n = 48). The best algorithm for Chl a (R² = 0.7510, p < 0.0001, n = 72) produced an RMSE of 31.19 μg/L with a relative RMSE of 16.56% (Chl a from 9.46 to 212.76 μg/L, n = 48). While more field data are required to further validate the long-term performance of these algorithms, currently they represent the best protocol for establishing a long time-series of cyanobacterial blooms in the Lac des Allemands using OCM data.     

基于生态地理分区的5套土地利用覆盖数据的不确定性研究

基于生态地理分区从面积精度和位置精度两个方面定量探讨了5套全球土地利用/覆盖(LULC)数据产品在实际应用中的不确定性,为基于生态地理分区的相关研究选择合理数据集提供参考依据,同时为中国生产全球LULC产品提供有关信息。选择中国地区20个典型生态地理分区为研究对象,采用最小误差频率法,分析5套数据集对各个类型面积估计的不确定性大小及原因;采用混淆矩阵法,基于位置分析5套数据集在类型混分方面的不确定性大小,原因及空间分布规律。结果表明:在生态地理分区尺度,MODIS,Meris300以及Glc2000这3套数据集明显优于Umd和Usgs这两套数据集,并且随着生态地理分区自南向北\,自东向西的空间分布,这3套数据集的不确定性呈减小趋势。对所有生态地理分区而言,Meris300数据集整体估计的稳定性最高,但是估计精度不是最高,并且它对建设用地和水域的估计最有优势。MODIS数据集整体估计精度和稳定性次之,对耕地的估计最有优势。 Glc2000数据集更适用于土地利用/覆盖简单的生态地理分区。研究还发现地形和土地利用/覆盖的复杂程度是引起数据集不确定性的两个重要因素。     

Evaluation of four MERIS atmospheric correction algorithms in Lake Kasumigaura,Japan

Accurate atmospheric correction for turbid inland waters remains a significant challenge. Several atmospheric correction algorithms have been proposed to address this issue, but their performance is unclear in regard to Asian lakes, some of which have extremely high turbidity and different inherent optical properties from lakes in other continents. Here, four existing atmospheric correction algorithms were tested in Lake Kasumigaura, Japan (an extremely turbid inland lake), using in situ water-leaving reflectance and concurrently acquired medium resolution imaging spectrometer (MERIS) images. The four algorithms are (1) GWI (the standard Gordon and Wang algorithm with an iterative process and a bio-optical model) (2) MUMM (Management Unit of the North Sea Mathematical Models); (3) SCAPE-M (Self-Contained Atmospheric Parameters Estimation for MERIS Data) and (4) C2WP (Case-2 Water Processor). The results show that all four atmospheric correction algorithms have limitations in Lake Kasumigaura, even though SCAPE-M and MUMM gave acceptable accuracy for atmospheric correction in several cases (relative errors less than 30% for the 2006 and 2008 images). The poor performance occurred because the conditions in Lake Kasumigaura (i.e. the atmospheric state and/or turbidity) did not always meet the assumptions in each atmospheric correction algorithm (e.g. in 2010, the relative errors ranged from 42% to 83%). These results indicate that further improvements are necessary to address the issue of atmospheric correction for turbid inland waters such as Lake Kasumigaura, Japan.