Deyong Sun

Estimating phycocyanin pigment concentration in productive inland waters using Landsat measurements: A case study in Lake Dianchi

Using remote sensing reflectance (R(rs)(λ), sr(-1)) and phycocyanin (PC, mg m(-3)) pigment data as well as other bio-optical data collected from two cruises in September and December 2009 in Lake Dianchi (a typical plateau lake of China), we developed a practical approach to estimate PC concentrations that could be applied directly to Landsat measurements. The visible and near-IR bands as well as their band ratios of simulated Landsat data were used as inputs to the algorithms, where the algorithm coefficients for each Landsat sensor were determined through multivariate regressions. The coefficients of determination (R(2)) between the R(rs)-modeled and measured PC were all > 0.97 for the spectral bands corresponding to Landsat 8 OLI, Landsat 7 ETM + , Landsat 5 TM, and Landsat 4 TM, with mean absolute percentage errors (MAPE) < 10% for PC ranging between ~80 and 700 mg m(-3) (n = 14). The algorithms were further evaluated using an independent data set (n = 14), yielding larger but still acceptable MAPE (~30%) for PC ranging between ~80 and 500 mg m(-3). Application of the approach to Landsat 8 measurements over Lake Dianchi suggests potential use of the approach for periodical assessment of the lakes bloom conditions, yet its empirical nature together with the lack of specific narrow bands on Landsat sensors to explicitly account for the PC absorption around 625 nm calls for extra caution when applied to other eutrophic lakes.

Reconstruction of hyperspectral reflectance for optically complex turbid inland lakes: test of a new scheme and implications for inversion algorithms.

A new scheme has been proposed by Lee et al. (2014) to reconstruct hyperspectral (400 - 700 nm, 5 nm resolution) remote sensing reflectance (R_rs(λ), sr^-1) of representative global waters using measurements at 15 spectral bands. This study tested its applicability to optically complex turbid inland waters in China, where R_rs(λ) are typically much higher than those used in Lee et al. (2014). Strong interdependence of R_rs(λ) between neighboring bands (≤ 10 nm interval) was confirmed, with Pearson correlation coefficient (PCC) mostly above 0.98. The scheme of Lee et al. (2014) for R_rs(λ) re-construction with its original global parameterization worked well with this data set, while new parameterization showed improvement in reducing uncertainties in the reconstructed R_rs(λ). Mean absolute error (MAE_Rrs(λ_i)) in the reconstructed R_rs(λ) was mostly < 0.0002 sr^-1 between 400 and 700nm, and mean relative error (MRE_Rrs(λ_i)) was < 1% when the comparison was made between reconstructed and measured R_rs(λ) spectra. When R_rs(λ) at the MODIS bands were used to reconstruct the hyperspectral R_rs(λ), MAE_Rrs(λ_i) was < 0.001 sr^-1 and MRE_Rrs(λ_i) was < 3%. When R_rs(λ) at the MERIS bands were used, MAE_Rrs(λ_i) in the reconstructed hyperspectral R_rs(λ) was < 0.0004 sr^-1 and MRE_Rrs(λ_i) was < 1%. These results have significant implications for inversion algorithms to retrieve concentrations of phytoplankton pigments (e.g., chlorophyll-a or Chla, and phycocyanin or PC) and total suspended materials (TSM) as well as absorption coefficient of colored dissolved organic matter (CDOM), as some of the algorithms were developed from in situ R_rs(λ) data using spectral bands that may not exist on satellite sensors.

Daytime sea fog retrieval based on GOCI data: a case study over the Yellow Sea

In this paper, a new daytime sea fog detection algorithm has been developed by using Geostationary Ocean Color Imager (GOCI) data. Based on spectral analysis, differences in spectral characteristics were found over different underlying surfaces, which include land, sea, middle/high level clouds, stratus clouds and sea fog. Statistical analysis showed that the Rrc (412 nm) (Rayleigh Corrected Reflectance) of sea fog pixels is approximately 0.1-0.6. Similarly, various band combinations could be used to separate different surfaces. Therefore, three indices (SLDI, MCDI and BSI) were set to discern land/sea, middle/high level clouds and fog/stratus clouds, respectively, from which it was generally easy to extract fog pixels. The remote sensing algorithm was verified using coastal sounding data, which demonstrated that the algorithm had the ability to detect sea fog. The algorithm was then used to monitor an 8-hour sea fog event and the results were consistent with observational data from buoys data deployed near the Sheyang coast (121°E, 34°N). The goal of this study was to establish a daytime sea fog detection algorithm based on GOCI data, which shows promise for detecting fog separately from stratus.

Innovative GOCI algorithm to derive turbidity in highly turbid waters: a case study in the Zhejiang coastal area.

An innovative algorithm is developed and validated to estimate the turbidity in Zhejiang coastal area (highly turbid waters) using data from the Geostationary Ocean Color Imager (GOCI). First, satellite-ground synchronous data (n = 850) was collected from 2014 to 2015 using 11 buoys equipped with a Yellow Spring Instrument (YSI) multi-parameter sonde capable of taking hourly turbidity measurements. The GOCI data-derived Rayleigh-corrected reflectance (R(rc)) was used in place of the widely used remote sensing reflectance (R(rs)) to model turbidity. Various band characteristics, including single band, band ratio, band subtraction, and selected band combinations, were analyzed to identify correlations with turbidity. The results indicated that band 6 had the closest relationship to turbidity; however, the combined bands 3 and 6 model simulated turbidity most accurately (R(2) = 0.821, p<0.0001), while the model based on band 6 alone performed almost as well (R(2) = 0.749, p<0.0001). An independent validation data set was used to evaluate the performances of both models, and the mean relative error values of 42.5% and 51.2% were obtained for the combined model and the band 6 model, respectively. The accurate performances of the proposed models indicated that the use of R(rc) to model turbidity in highly turbid coastal waters is feasible. As an example, the developed model was applied to 8 hourly GOCI images on 30 December 2014. Three cross sections were selected to identify the spatiotemporal variation of turbidity in the study area. Turbidity generally decreased from near-shore to offshore and from morning to afternoon. Overall, the findings of this study provide a simple and practical method, based on GOCI data, to estimate turbidity in highly turbid coastal waters at high temporal resolutions.

Seasonal and Interannual Variability of Satellite-Derived Chlorophyll-a (2000–2012) in the Bohai Sea, China

Knowledge of the chlorophyll-a dynamics and their long-term changes is important for assessing marine ecosystems, especially for coastal waters. In this study, the spatial and temporal variability of sea surface chlorophyll-a concentration (Chl-a) in the Bohai Sea were investigated using 13-year (2000–2012) satellite-derived products from MODIS and SeaWiFS observations. Based on linear regression analysis, the results showed that the entire Bohai Sea experienced an increase in Chl-a on a long-term scale, with the largest increase in the central Bohai Sea and the smallest increase in the Bohai strait. Distinct seasonal patterns of Chl-a existed in different sub-regions of the Bohai Sea. A long-lasting Chl-a peak was observed from May to September in coastal waters (Liaodong bay, Qinhuangdao coast, and Bohai bay) and the central Bohai Sea, whereas Laizhou bay had relatively low Chl-a in early summer. In the Bohai strait, two pronounced Chl-a peaks occurred in March and September, but the lowest Chl-a was in summer. This pattern was quite different from those in other regions of the Bohai Sea. The water column condition (stratified or mixed) was likely an important physical factor that affects the seasonal pattern of Chl-a in the Bohai Sea. Meanwhile, increased human activity (e.g., river discharge) played a significant role in changing the Chl-a distribution in both coastal waters and the central Bohai Sea, especially in summer. The increasing trend of Chl-a in the Bohai Sea might be attributed to the increase in nutrient contents from riverine inputs. The Chl-a dynamics documented in this study provide basic knowledge for the future exploration of marine biogeochemical processes and ecosystem evolution in the Bohai Sea.

Light beam attenuation and backscattering properties of particles in the Bohai Sea and Yellow Sea with relation to biogeochemical properties

This study reports the first results of the variability in light beam attenuation and the backscattering properties of particles and their controlling factors during the summer in the Bohai Sea (BS) and Yellow Sea (YS), which are two typical shallow and semienclosed seas. We observe large variations in the particulate beam attenuation (cp) and backscattering coefficients (bbp); such variations are mainly attributed to changes in the total suspended matter, while the cross-sectional area concentration shows tighter relationships with both cp and bbp. The mass-specific beam attenuation ( cp*) and backscattering coefficients ( bbp*) vary more widely over about two orders of magnitude. The attenuation (Qce) and backscattering efficiencies (Qbbe) are important factors that control cp* and bbp*, which clearly separate all the samples into two types. Type 1 samples show low Qce and Qbbe and contain relatively high proportions of organic or large particles, while type 2 samples have high Qce and Qbbe and mainly contain relatively small mineral particles. The majority of the variability in cp* and bbp* within each type is related to the inverse of the product of particle apparent density (ρa) and mean diameter (DA); ρa plays a major role, while DA exerts only a slight impact. Overall, this study provides general knowledge of particulate beam attenuation and the backscattering properties in the BS and YS, which may improve our understanding of underwater radiative transfer processes, marine biogeochemical processes and ocean color algorithms.

A hybrid method to estimate suspended particle sizes from satellite measurements over Bohai Sea and Yellow Sea

Particle size distribution (PSD), a measure of particle concentrations at different sizes, is of great importance to the understanding of many biogeochemical processes in coastal marine ecosystems. Here, a hybrid method, including analytical, semi-analytical, and empirical steps, is developed to estimate PSD through the median diameter of suspended particles (Dv50). Four cruise surveys were conducted to measure optical scattering properties, particle concentrations, spectral reflectance, and particle size distributions (obtained with a LISST instrument covering a size range of 2.5-500 μm) in coastal waters of Bohai Sea, Yellow Sea, and Jiangsu coastal region. Based on the Mie scattering theory, Dv50 is closely related to mass-specific backscattering coefficient of suspended particles (bbp*), and their relationship is calibrated through a power model (R2=0.796, n=67, p<0.001) for the Dv50 range of 23.5-379.8 μm. The model is shown to perform better than the previously used inverse-proportion model. The retrieval of bbp* is through a bio-optical model that links remote sensing reflectance just beneath the surface to inherent optical properties, where a close empirical relationship is established between particulate backscattering and particle concentration. The hybrid method shows high degree of fitting (R2=0.875, n=46, p<0.001) between the measured and estimated Dv50 for the size range of 17.2-325.2 μm used in the model calibration, while validation using two independent datasets shows mean absolute percentage errors of 46.0% and 64.7%, respectively. Application of the hybrid method to MODIS (Moderate Resolution Imaging Spectroradiometer) data results in spatial distributions of Dv50 that are generally consistent with those from in situ observations, suggesting potential use of the method in studying particle dynamics through time-series of remote sensing observations. However, its general applicability to other regions still requires further research. This article is protected by copyright. All rights reserved.

Remote Sensing of Particle Cross-Sectional Area in the Bohai Sea and Yellow Sea: Algorithm Development and Application Implications

Suspended particles in waters play an important role in determination of optical properties and ocean color remote sensing. To link suspended particles to their optical properties and thereby remote sensing reflectance (Rrs(λ)), cross-sectional area is a key factor. Till now, there is still a lack of methodologies for derivation of the particle cross-sectional area concentration (AC) from satellite measurements, which consequently limits potential applications of AC. In this study, we investigated the relationship between AC and Rrs(λ) based on field measurements in the Bohai Sea (BS) and Yellow Sea (YS). Our analysis confirmed the strong dependence of Rrs(λ) on AC and that such dependence is stronger than on mass concentration. Subsequently, a remote sensing algorithm that uses the slope of Rrs(λ) between 490 and 555 nm was developed for retrieval of AC from satellite measurements of the Geostationary Ocean Color Imager (GOCI). In situ evaluations show that the algorithm displays good performance for deriving AC and is robust to uncertainties in Rrs(λ). When the algorithm was applied to satellite data, it performed well, with a coefficient of determination of 0.700, a root mean squared error of 2.126 m−1 and a mean absolute percentage error of 40.7%, and it yielded generally reasonable spatial and temporal distributions of AC in the BS and YS. The satellite-derived AC using our algorithm may offer useful information for modeling the inherent optical properties of suspended particles, deriving the water transparency, estimating the particle composition and possibly improving particle mass concentration estimations in future.

Using Landsat 8 data to estimate suspended particulate matter in the Yellow River estuary

The distribution of suspended particulate matter (SPM) and its variations in estuary regions are key to promoting carbon, oxygen, and nutrient cycling in coastal regions and nearby seas. This study presents SPM estimations for the Yellow River estuary from Landsat 8 Operational Land Imager (L8/OLI) data from 2013 to 2016. L8/OLI-measured remote sensing reflectance (Rrs) was cross-validated with Moderate Resolution Imaging Spectroradiometer (MODIS) measurements, and SPM concentrations calculated from the tuned retrieval model were validated with in situ observations. The validation shows that L8/OLI can provide reasonably Rrs, which can be used to quantify SPM distributions and variations in the Yellow River estuary. Three year averaged SPM maps show that highly turbid waters are mostly found in an ovate area surrounding the mouth of the Yellow River. The corresponding area proportion is less than 30%, with SPM concentrations greater than 100 g m−3. High variations of SPM distributions are consistent with high SPM concentrations, and vice versa. Significant difference is observed between dry and wet seasons. Higher SPM in the dry season is observed both in range and intensity compared to those of the wet season. Furthermore, multiyear averaged SPM distributions with high concentrations are mainly attributable to currents. Significant seasonal variations are mainly controlled by sediment resuspension processes driven by wind-wave forces. Due to human interventions, seasonal variability in river runoff and sediment discharge from the Yellow River has decreased in recent years. Accordingly, seasonal variability in SPM distributions in the Yellow River estuary due to sediment discharge has decreased.

Remote‐Sensing Estimation of Phytoplankton Size Classes From GOCI Satellite Measurements in Bohai Sea and Yellow Sea

Phytoplankton size class (PSC), a measure of different phytoplankton functional and structural groups, is a key parameter to the understanding of many marine ecological and biogeochemical processes. In turbid waters where optical properties may be influenced by terrigenous discharge and nonphytoplankton water constituents, remote estimation of PSC is still a challenging task. Here based on measurements of phytoplankton diagnostic pigments, total chlorophyll a, and spectral reflectance in turbid waters of Bohai Sea and Yellow Sea during summer 2015, a customized model is developed and validated to estimate PSC in the two semienclosed seas. Five diagnostic pigments determined through high-performance liquid chromatography (HPLC) measurements are first used to produce weighting factors to model phytoplankton biomass (using total chlorophyll a as a surrogate) with relatively high accuracies. Then, a common method used to calculate contributions of microphytoplankton, nanophytoplankton, and picophytoplankton to the phytoplankton assemblage (i.e., Fm, Fn, and Fp) is customized using local HPLC and other data. Exponential functions are tuned to model the size-specific chlorophyll a concentrations (Cm, Cn, and Cp for microphytoplankton, nanophytoplankton, and picophytoplankton, respectively) with remote-sensing reflectance (Rrs) and total chlorophyll a as the model inputs. Such a PSC model shows two improvements over previous models: (1) a practical strategy (i.e., model Cp and Cn first, and then derive Cm as C-Cp-Cn) with an optimized spectral band (680 nm) for Rrs as the model input; (2) local parameterization, including a local chlorophyll a algorithm. The performance of the PSC model is validated using in situ data that were not used in the model development. Application of the PSC model to GOCI (Geostationary Ocean Color Imager) data leads to spatial and temporal distribution patterns of phytoplankton size classes (PSCs) that are consistent with results reported from field measurements by other researchers. While the applicability of the PSC model together with its parameterization to other optically complex regions and to other seasons is unknown, the findings of this study suggest that the approach to develop such a model may be extendable to other cases as long as local data are used to select the optimal band and to determine the model coefficients.