[go: up one dir, main page]
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
8.3
4.2
Journals
Remote Sensing
Special Issues
Satellite Monitoring of Water Quality and Water Environment
remotesensing-logo
Submit to Remote Sensing Review for Remote Sensing Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Satellite Monitoring of Water Quality and Water Environment

A special issue of Remote Sensing (ISSN 2072-4292).

Deadline for manuscript submissions: closed (30 June 2019) | Viewed by 66779

Special Issue Editors

Dr. Wei Yang

E-Mail Website
Guest Editor
Center for Environmental Remote Sensing, Chiba University, Chiba 263-8522, Japan
Interests: quantitative remote sensing; algorithm development; environmental modeling; phenology
Special Issues, Collections and Topics in MDPI journals
Dr. Bunkei Matsushita

E-Mail Website
Guest Editor
Faculty of Life and Environmental Science, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8572, Japan
Interests: remote sensing of case-2 waters; water quality; water environment; land cover/use changes; estimation of impervious surface area of watersheds
Prof. Dr. Ronghua Ma

E-Mail Website
Guest Editor
Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China
Interests: remote sensing of inland lakes; water quality; water environment; aquatic ecology; machine learning; GIS
Special Issues, Collections and Topics in MDPI journals
Dr. Steven Loiselle

E-Mail Website
Guest Editor
University of Siena, Via Aldo Moro 2, 53100 Siena SI, Italy
Interests: environmental spectroscopy; optical analysis and modelling of aquatic ecosystems; analysis of organic matter in aquatic ecosystems
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

Variety of water bodies plays important roles in human societies, in terms of providing multiple ecosystem services. However, in recent decades, they have encountered intensive pollution problems, all over the world. For example, the water quality of inland waters is strongly influenced by land use/cover changes (LUCC) in their corresponding watersheds. Understanding the interactions between water quality and their watersheds is, therefore, crucial for sustainable management of water resources.

Remote sensing is an important technique for monitoring water environments and watershed LUCC, being able to cover large spatial areas at frequent intervals. Mapping of regional LUCC, based on satellite observations, is well established. In contrast, satellite remote sensing of water quality, especially for inland and coastal water areas is still very challenging, due mainly to complex interactions among optically active substances, diversity of specific inherent optical properties, as well as difficulties in atmospheric correction above many inland and coastal waterbodies. The application of satellite monitoring to inland waters has been far less successful than those in the oceanic areas. In recent years, field surveys for inland waterbody remote sensing have increased. Meanwhile, the rapid development of mathematic techniques (e.g., machine learning) and cloud computation platforms (e.g., Google Earth Engine) provide new opportunities to improve the utility of satellite remote sensing for inland water monitoring. However, there is a clear need to share approaches and new ideas that can be used to expand and strengthen an integrated approach to catchment and waterbody management.

To meet this urgent need, a Special Issue on “Satellite Monitoring of Water Quality and Water Environment” is being planned by the international journal, Remote Sensing, to address the technical challenges for satellite monitoring of inland and coastal waters and demonstrate successful applications of remote sensing on the links between water quality/resource and watershed LUCC.

We solicit your contributions in this field to our Remote Sensing Special Issue. Research or review articles with respect to the following topics are welcome:

  • Remote estimation of water quality parameters
  • Remote estimation of water optical properties
  • Optical classification of specific inherent optical properties of world-wide inland and coastal waters
  • Application of machine learning algorithms to remote sensing of water environment
  • Application of Google Earth Engine to remote sensing of water environment
  • Validation and development of atmospheric correction algorithms for inland and coastal waters
  • Satellite mapping of macrophytes in inland waters
  • Satellite monitoring of inland water resources
  • Long-term satellite monitoring of watershed LUCC
  • Linkage between water qualities/resources and watershed LUCC
  • Environmental modeling of inland waters and its watershed based on application of remote sensing

Dr. Wei Yang
Dr. Bunkei Matsushita
Dr. Ronghua Ma
Dr. Steven Loiselle
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Remote Sensing is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2700 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • water quality
  • water environment
  • watersheds
  • land use/land cover
  • remote sensing
  • atmospheric correction
  • algorithm development
  • environmental monitoring
  • machine learning
  • Google Earth Engine

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

Further information on MDPI's Special Issue policies can be found here.

Published Papers (12 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

19 pages, 4025 KiB  
Article
Remote Sensing Estimation of Lake Total Phosphorus Concentration Based on MODIS: A Case Study of Lake Hongze
by Junfeng Xiong, Chen Lin, Ronghua Ma and Zhigang Cao
Remote Sens. 2019, 11(17), 2068; https://doi.org/10.3390/rs11172068 - 3 Sep 2019
Cited by 49 | Viewed by 6119
Abstract
Phosphorus (P) is an important substance for the growth of phytoplankton and an efficient index to assess the water quality. However, estimation of the TP concentration in waters by remote sensing must be associated with optical substances such as the chlorophyll-a (Chla) and [...] Read more.
Phosphorus (P) is an important substance for the growth of phytoplankton and an efficient index to assess the water quality. However, estimation of the TP concentration in waters by remote sensing must be associated with optical substances such as the chlorophyll-a (Chla) and the suspended particulate matter (SPM). Based on the good correlation between the suspended inorganic matter (SPIM) and P in Lake Hongze, we used the direct and indirect derivation methods to develop algorithms for the total phosphorus (TP) estimation with the MODIS/Aqua data. Results demonstrate that the direct derivation algorithm based on 645 nm and 1240 nm of the MODIS/Aqua performs a satisfied accuracy (R2 = 0.75, RMSE = 0.029mg/L, MRE = 39% for the training dataset, R2 = 0.68, RMSE = 0.033mg/L, MRE = 47% for the validate dataset), which is better than that of the indirect derivation algorithm. The 645 nm and 1240 nm of MODIS are the main characteristic band of the SPM, so that algorithm can effectively reflect the P variations in Lake Hongze. Additionally, the ratio of the TP to the SPM is positively correlated with the accuracy of the algorithm as well. The proportion of the SPIM in the SPM has a complex effect on the accuracy of the algorithm. When the SPIM accounts for 78%, the algorithm achieves the highest accuracy. Furthermore, the performance of this direct derivation algorithm was examined in two inland lakes in China (Lake Nanyi and Lake Chaohu), it derived the expected P distribution in Lake Nanyi whereas the algorithm failed in Lake Chaohu. Different water properties influence significantly the accuracy of this direct derivation algorithm, while the TP, Chla, and suspended particular inorganic matter (SPOM) of Lake Chaohu are much higher than those of the other two lakes, thus it is difficult to estimate the TP concentration by a simple band combination in Lake Chaohu. Although the algorithm depends on the dataset used in the development, it usually presents a good estimation for those waters where the SPIM dominated, especially when the SPIM accounts for 60% to 80% of the SPM. This research proposed a direct derivation algorithm for the TP estimation for the turbid lake and will provide a theoretical and practical reference for extending the optical remote sensing application and the TP empirical algorithm of Lake Hongze’s help for the local government management water quality. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Location of the study site and sampling plots.</p> Full article ">Figure 2
<p>The optimal algorithms obtained by the direct derivation method: (<b>a</b>) The relationship between the in situ measured TP and (B2–B5)/(B2+B5); (<b>b</b>) the validation of the MODIS-estimated TP with the in situ measured TP based on an independent dataset.</p> Full article ">Figure 3
<p>The optimal algorithms obtained by the indirect derivation method: (<b>a</b>) The relationship between the in situ measured SPM and (B2−B5); (<b>b</b>) the validation of the MODIS-estimated SPM with the in situ measured SPM based on an independent dataset; (<b>c</b>) the relationship between the in situ measured TP and MODIS-estimated SPM; (<b>d</b>) the validation of the indirect derivation method-estimated TP with the in situ measured TP based on an independent dataset.</p> Full article ">Figure 4
<p>The seasonal average of the TP concentration from 2016 to 2018 in Lake Hongze in each season: (<b>a</b>) Spring; (<b>b</b>) summer; (<b>c</b>) autumn; (<b>d</b>) winter.</p> Full article ">Figure 5
<p>The correlation coefficients between the SPM, SPIM, SPOM, TP, and R<sub>rs</sub> measured in the sites.</p> Full article ">Figure 6
<p>Algorithm accuracy change with different substance changes. (<b>a</b>) Chlorophyll-a (Chla); (<b>b</b>) SPM; (<b>c</b>) TP/SPM; (<b>d</b>) SPIM/SPM.</p> Full article ">Figure 7
<p>Box-plot of the water quality attributes of the three lakes: (<b>a</b>) Chla; (<b>b</b>) TP; (<b>c</b>) SPM; (<b>d</b>) SPIM; (<b>e</b>) SPOM.</p> Full article ">Figure 8
<p>(<b>a</b>) Scatter plot of the (B2−B5)/(B2+B5) and TP concentration in Lake Chaohu; (<b>b</b>) scatter plot of the (B2−B5)/(B2+B5) and TP concentration in Lake Nanyi; (<b>c</b>) scatter plot of the B1*B5 and TP concentration in Lake Chaohu; (<b>d</b>) scatter plot of the (B1−B2)/(B1+B2) and TP concentration in Lake Nanyi.</p> Full article ">Figure 9
<p>(<b>a</b>) Scatter plot of the (B1–B2)/(B1+B2) and Chla concentration in Lake Nanyi; (<b>b</b>) scatter plot of the (B1–B2)/(B1+B2) and SPM concentration in Lake Nanyi; (<b>c</b>) scatter plot of the (B1–B2)/(B1+B2) and SPIM concentration in Lake Nanyi; (<b>d</b>) scatter plot of the (B1–B2)/(B1+B2) and SPOM concentration in Lake Nanyi.</p> Full article ">
19 pages, 8874 KiB  
Article
Removal of Chlorophyll-a Spectral Interference for Improved Phycocyanin Estimation from Remote Sensing Reflectance
by Igor Ogashawara and Lin Li
Remote Sens. 2019, 11(15), 1764; https://doi.org/10.3390/rs11151764 - 26 Jul 2019
Cited by 12 | Viewed by 4367
Abstract
Monitoring cyanobacteria is an essential step for the development of environmental and public health policies. While traditional monitoring methods rely on collection and analysis of water samples, remote sensing techniques have been used to capture their spatial and temporal dynamics. Remote detection of [...] Read more.
Monitoring cyanobacteria is an essential step for the development of environmental and public health policies. While traditional monitoring methods rely on collection and analysis of water samples, remote sensing techniques have been used to capture their spatial and temporal dynamics. Remote detection of cyanobacteria is commonly based on the absorption of phycocyanin (PC), a unique pigment of freshwater cyanobacteria, at 620 nm. However, other photosynthetic pigments can contribute to absorption at 620 nm, interfering with the remote estimation of PC. To surpass this issue, we present a remote sensing algorithm in which the contribution of chlorophyll-a (chl-a) absorption at 620 nm is removed. To do this, we determine the PC contribution to the absorption at 665 nm and chl-a contribution to the absorption at 620 nm based on empirical relationships established using chl-a and PC standards. The proposed algorithm was compared with semi-empirical and semi-analytical remote sensing algorithms for proximal and simulated satellite sensor datasets from three central Indiana reservoirs (total of 544 sampling points). The proposed algorithm outperformed semi-empirical algorithms with root mean square error (RMSE) lower than 25 µg/L for the three analyzed reservoirs and showed similar performance to a semi-analytical algorithm. However, the proposed remote sensing algorithm has a simple mathematical structure, it can be applied at ease and make it possible to improve spectral estimation of phycocyanin from space. Additionally, the proposed showed little influence from the package effect of cyanobacteria cells. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Figure 1

Figure 1
<p>Linear relationship between <span class="html-italic">a<sub>phy</sub></span> at 665 nm and R<sub>rs</sub> 709/665 (<b>A</b>), and between <span class="html-italic">a<sub>phy</sub></span> at 620 nm and R<sub>rs</sub> 709/620 (<b>B</b>).</p> Full article ">Figure 2
<p>Linear relationships between the absorption coefficients at 620 and 665 nm measured for Sigma Aldrich standards: Chl-<span class="html-italic">a</span> (<b>A</b>) and PC (<b>B</b>).</p> Full article ">Figure 3
<p>Average R<sub>rs</sub> (sr<sup>−1</sup>) spectra for each dataset.</p> Full article ">Figure 4
<p>Validation of the semi-empirical algorithms using 2010 datasets for ECR (<b>A</b>), GR (<b>B</b>), and MR (<b>C</b>).</p> Full article ">Figure 5
<p>Tuning results for SIM05 (<b>left</b>) and OGA19 (<b>right</b>) using 2005–2007 datasets for ECR (<b>A</b>,<b>B</b>), GR (<b>C</b>,<b>D</b>) and MR (<b>E</b>,<b>F</b>).</p> Full article ">Figure 6
<p>Validation results for SIM05 and OGA19 using the 2010 datasets of ECR (<b>A</b>), GR (<b>B</b>), and MR (<b>C</b>).</p> Full article ">Figure 7
<p>Sensitivity analysis of MIS14 (<b>A</b>) and OGA19 (<b>B</b>) when applied to the MR data collected in years 2005–2007 and 2010.</p> Full article ">Figure 8
<p>Validation results for the band ratio (BR) and OGA19 algorithms using the 2010 datasets of ECR (<b>A</b>), GR (<b>B</b>), and MR (<b>C</b>).</p> Full article ">
17 pages, 4407 KiB  
Article
Regional Vicarious Calibration of the SWIR-Based Atmospheric Correction Approach for MODIS-Aqua Measurements of Highly Turbid Inland Water
by Junsheng Li, Ziyao Yin, Zhaoyi Lu, Yuntao Ye, Fangfang Zhang, Qian Shen and Bing Zhang
Remote Sens. 2019, 11(14), 1670; https://doi.org/10.3390/rs11141670 - 13 Jul 2019
Cited by 5 | Viewed by 3068
Abstract
Water color remote sensing requires accurate atmospheric correction but this remains a significant challenge in highly turbid waters. In this respect, the shortwave infrared (SWIR) band-based atmospheric correction approach has proven advantageous when applied to the moderate resolution imaging spectroradiometer (MODIS) onboard the [...] Read more.
Water color remote sensing requires accurate atmospheric correction but this remains a significant challenge in highly turbid waters. In this respect, the shortwave infrared (SWIR) band-based atmospheric correction approach has proven advantageous when applied to the moderate resolution imaging spectroradiometer (MODIS) onboard the Aqua satellite. However, even so, uncertainties affect its accuracy. We performed a regional vicarious calibration of the MODIS-Aqua SWIR (1240, 2130)-based atmospheric correction using in situ water surface reflectance data measured during different seasons in Lake Taihu, a highly turbid lake. We then verified the accuracy of the (1240, 2130)-based atmospheric correction approach using these results; good results were obtained for the remote sensing reflectance retrievals at the 555, 645, and 859 nm, with average relative errors of 15%, 14%, and 22%, respectively, and no significant bias. Comparisons with the (1240, 2130)-based iterative approach and (1640, 2130)-based approach showed that the vicarious calibrated (1240, 2130)-based approach has the best accuracy and robustness. Thus, it is applicable to the highly turbid Lake Taihu. It may also be applicable to other highly turbid inland waters with similar optical and aerosol optical properties above water, but such applications will require further validation. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Figure 1

Figure 1
<p>Maps of the location of Lake Taihu, China, and the 377 in situ water sampling stations used during the 12 cruise surveys performed between 2005 and 2014.</p> Full article ">Figure 2
<p>The <span class="html-italic">R</span><sub>rs</sub> spectra measured in Lake Taihu from the 12 cruise surveys performed in (<b>a</b>) October 2005, (<b>b</b>) January 2006, (<b>c</b>) July to August 2006, (<b>d</b>) October 2006, (<b>e</b>) January 2007, (<b>f</b>) April 2007, (<b>g</b>) March 2009, (<b>h</b>) April 2009, (<b>i</b>) October 2012, (<b>j</b>) May 2013, (<b>k</b>) July 2014, and (<b>l</b>) October 2014.</p> Full article ">Figure 3
<p>A comparison between the moderate resolution imaging spectroradiometer (MODIS)-Aqua measured <span class="html-italic">L</span><sub>t</sub>(λ) in Lake Taihu with the computed <span class="html-italic">L</span><sub>t</sub>(λ) via the in situ-measured <span class="html-italic">R<sub>rs</sub></span>(λ) and atmospheric correction parameters calculated with the SeaDAS 7.2. The units for the MODIS-measured and computed <span class="html-italic">L</span><sub>t</sub>(λ) are W m<sup>−2</sup> μm<sup>−1</sup> sr<sup>−1</sup>.</p> Full article ">Figure 4
<p>The scatter plots of the matchups for the in situ <span class="html-italic">R<sub>rs</sub></span>(λ) measurements coincident with the atmospheric correction results of MODIS-Aqua based on the three methods, i.e., the (1640, 2130), (1240, 2130)-Ite, and (1240, 2130)-Cal. The points for the (1240, 2130)-Cal results were divided into calibration, referred to as (1240, 2130)-Cal (Cal), and validation points, referred to as (1240, 2130)-Cal (Val). <a href="#remotesensing-11-01670-t004" class="html-table">Table 4</a> lists the valid matchup numbers for each band based on each method.</p> Full article ">Figure 5
<p>The atmospheric-corrected <span class="html-italic">R<sub>rs</sub></span> images of a scene from a MODIS-Aqua image acquired on 26 April 2009, based on the (1640, 2130), (1240, 2130)-Ite, and (1240, 2130)-Cal methods. (<b>a</b>) is the <span class="html-italic">L</span><sub>t</sub>(645) image, which shows the locations of the five coincident in situ <span class="html-italic">R<sub>rs</sub></span> measurements as red dots. (<b>b</b>–<b>d</b>) are the atmospheric-corrected <span class="html-italic">R<sub>rs</sub></span>(645) images based on the (1640, 2130), (1240, 2130)-Ite, and (1240, 2130)-Cal methods, respectively. (<b>e</b>) is the <span class="html-italic">L</span><sub>t</sub>(859) image. (<b>f</b>–<b>h</b>) are the atmospheric-corrected <span class="html-italic">R<sub>rs</sub></span>(859) images based on the (1640, 2130), (1240, 2130)-Ite, and (1240, 2130)-Cal methods, respectively. The (1240, 2130)-Cal method retrieved an AOT(555) = 0.33 (Mean) ± 0.10 (Std).</p> Full article ">Figure 6
<p>A comparison of the five matchups for the in situ <span class="html-italic">R<sub>rs</sub></span> spectra with corresponding atmospheric-corrected <span class="html-italic">R<sub>rs</sub></span> spectra from the MODIS-Aqua image acquired on 26 April 2009, which was processed using the three different methods. The five subfigures show the comparison results on the five in situ locations, the red dots in <a href="#remotesensing-11-01670-f005" class="html-fig">Figure 5</a>a.</p> Full article ">Figure 7
<p>The atmospheric-corrected <span class="html-italic">R<sub>rs</sub></span>(645) images of a scene of a MODIS-Aqua image acquired on 9 January 2006, using the (1240, 2130)-Cal method. (<b>a</b>) is the <span class="html-italic">L</span><sub>t</sub>(645) image. (<b>b</b>,<b>c</b>) are the atmospheric-corrected results without BRDF correction and with the default BRDF correction, respectively. (<b>d</b>) is the BRDF correction factor computed during the atmospheric correction process with the default BRDF correction.</p> Full article ">
15 pages, 3940 KiB  
Article
Characteristics of Absorption Spectra of Chromophoric Dissolved Organic Matter in the Pearl River Estuary in Spring
by Xia Lei, Jiayi Pan and Adam T. Devlin
Remote Sens. 2019, 11(13), 1533; https://doi.org/10.3390/rs11131533 - 28 Jun 2019
Cited by 21 | Viewed by 4923
Abstract
In this study, absorption variation of chromophoric dissolved organic matter (CDOM) was investigated based on spectroscopic measurements of the water surface and bottom during a cruise survey on 2–12 May 2014 in the Pearl River Estuary (PRE). Multiple spectral signatures were utilized, including [...] Read more.
In this study, absorption variation of chromophoric dissolved organic matter (CDOM) was investigated based on spectroscopic measurements of the water surface and bottom during a cruise survey on 2–12 May 2014 in the Pearl River Estuary (PRE). Multiple spectral signatures were utilized, including the absorption ratios E2/E3 (a(250)/a(365)) and E2/E4 (a(254)/a(436))) as well as the spectral slopes over multiple wavelength ranges. The horizontal variations of a(300), E2/E3, spectral slope (S) of Ultraviolet C (SUVC, 250–280 nm), Ultraviolet B (SUVB, 280–315 nm), and S275–295 (275–295 nm) were highly correlated, revealing that CDOM of terrigenous origin in the upper estuary contained chromophores of larger molecular size and weight, while the marine CDOM in the lower estuary comprised organic compounds of smaller molecular size and weight; the molecular size of surface CDOM was generally larger than that at the bottom. Results of Gaussian decomposition methods showed that CDOM in the middle estuary of terrigenous origin produced more Gaussian components per spectrum than those of marine origin in the lower estuary and the adjacent Hong Kong waters. The surface CDOM composition was more diverse than at the bottom, inferred by the finding that the average number of Gaussian components yielded per surface sample (5.44) was more than that of the bottom sample (4.8). A majority of components was centered below 350 nm, indicating that organic compounds with relatively simple structures are ubiquitous in the estuary. Components centered above 350 nm only showed high peaks at the head of the estuary, suggesting that terrigenous CDOM with chromophores in complex structures rapidly lose visible light absorptivity during its transport in the PRE. The relatively low and homogenous peak heights of the components in Hong Kong waters imply higher light stability and composition consistency of the marine CDOM compared with the terrigenous CDOM. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>In situ absorption spectrum measured in the Pearl River Estuary in May 2014 (thick gray line) and the exponential curves (colored lines) fitted to Equation (1). The fits were obtained over the same wavelength range (250–700 nm) but with a different reference wavelength (<span class="html-italic">λ<sub>0</sub></span>), resulting in various spectral slope (<span class="html-italic">S</span>) values (see legend in (<b>a</b>)) and various residuals (see colored lines in (<b>b</b>)). The absorption spectrum sampled at the surface of Station A2 is used as an example (please refer to [<a href="#B29-remotesensing-11-01533" class="html-bibr">29</a>] for detailed information on Station A2).</p> Full article ">Figure 2
<p>Location of the sampling stations for the May 2014 cruise in the Pearl River Estuary.</p> Full article ">Figure 3
<p>Chromophoric dissolved organic matter (CDOM) absorption curves over the wavelength range of 250–700 nm, measured at the water surface and at the bottom in the Pearl River Estuary.</p> Full article ">Figure 4
<p>Scatter plot of surface and bottom spectral slope as a function of absorption ratio over a similar wavelength range: <span class="html-italic">S<sub>UVB</sub></span> (280–315 nm) versus <span class="html-italic">E2/E3</span> (<span class="html-italic">a</span>(250)/<span class="html-italic">a</span>(365)), and <span class="html-italic">S<sub>UVA</sub></span> (315–400 nm) versus <span class="html-italic">E2/E4</span> (<span class="html-italic">a</span>(254)/<span class="html-italic">a</span>(436)).</p> Full article ">Figure 5
<p>Surface and bottom variation of: (<b>a</b>) CDOM absorption <span class="html-italic">a</span>(300); (<b>b</b>) the spectral slope ratio <span class="html-italic">S<sub>R</sub></span> (<span class="html-italic">S<sub>275–295</sub></span>/<span class="html-italic">S<sub>350–400</sub></span>); (<b>c</b>–<b>h</b>) spectral slopes <span class="html-italic">S<sub>UVC</sub></span> (250–280 nm), <span class="html-italic">S<sub>UVB</sub></span> (280–315 nm), <span class="html-italic">S<sub>UVA</sub></span> (315–400 nm), <span class="html-italic">S<sub>VIS</sub></span> (400–700 nm), <span class="html-italic">S<sub>275–295</sub></span> (275–295 nm), and <span class="html-italic">S<sub>350–400</sub></span> (350–400 nm); and (<b>i</b>,<b>j</b>) absorption ratios <span class="html-italic">E2/E3</span> (<span class="html-italic">a</span>(250)/<span class="html-italic">a</span>(365)) and <span class="html-italic">E2/E4</span> (<span class="html-italic">a</span>(254)/<span class="html-italic">a</span>(436)) in the Pearl River Estuary. All figures were mapped by the Kriging interpolation method based on discrete sampling data.</p> Full article ">Figure 6
<p>The (<b>a</b>) baselines, (<b>b</b>) residuals, and (<b>c</b>) Gaussian components separated from CDOM absorption spectra collected in the Pearl River Estuary at the surface and bottom. The gray line in (<b>c</b>) is the residual curve of a surface sample (Station A3); the colored lines are the seven decomposed Gaussian components. The peak height (<span class="html-italic">φ</span>), peak position (<span class="html-italic">μ</span>), and curve width (<span class="html-italic">σ</span>) of the components are listed in the legend.</p> Full article ">Figure 7
<p>Surface and bottom variation of (<b>a</b>) the number of Gaussian components decomposed per spectrum and (<b>b</b>) the proportion of residuals from the original absorption at the peak position. Figures in (<b>b</b>) were mapped by the Kriging interpolation method based on discrete sampling data.</p> Full article ">Figure 8
<p>The frequency histogram of (<b>a</b>) peak height (<span class="html-italic">ϕ</span>), (<b>b</b>) peak position (<span class="html-italic">μ</span>), and (<b>c</b>) curve width (<span class="html-italic">σ</span>) of the 479 Gaussian components at the surface (left column) and 288 components at the bottom (right column). Eight groups were identified from the surface components and labeled in (<b>b1</b>); seven groups were identified from the bottom components and labeled in (<b>b2</b>).</p> Full article ">Figure 9
<p>Surface and bottom variations of peak height of Gaussian components in groups centered near wavelengths of (<b>a</b>) 260, (<b>b</b>) 275, (<b>c</b>) 300, (<b>d</b>) 350, and (<b>e</b>) 400 nm. The size of the dots denotes the peak height: the bigger the size, the higher the peak.</p> Full article ">
18 pages, 5396 KiB  
Article
Aquarius Sea Surface Salinity Gridding Method Based on Dual Quality–Distance Weighting
by Yanyan Li, Qing Dong and Yongzheng Ren
Remote Sens. 2019, 11(9), 1131; https://doi.org/10.3390/rs11091131 - 11 May 2019
Cited by 4 | Viewed by 3691
Abstract
A new method for improving the accuracy of gridded sea surface salinity (SSS) fields is proposed in this paper. The method mainly focuses on dual quality–distance weighting of the Aquarius level 2 along-track SSS data according to quality flags, which represent nonnominal data [...] Read more.
A new method for improving the accuracy of gridded sea surface salinity (SSS) fields is proposed in this paper. The method mainly focuses on dual quality–distance weighting of the Aquarius level 2 along-track SSS data according to quality flags, which represent nonnominal data conditions for measurements. In the weighting progress, 14 data conditions were considered, and their geospatial distributions and influences on the SSS were also visualized and evaluated. Three interpolation methods were employed, and weekly gridded SSS maps were produced for the period from September 2011 to May 2015. These maps were evaluated via comparisons with concurrent Argo buoy measurements. The results show that the proposed method improved the accuracy of the SSS fields by approximately 36% compared to the officially released weekly level 3 products and yielded root mean squared difference (RMSD), correlation and bias values of 0.19 psu, 0.98 and 0.01 psu, respectively. These findings indicate a significant improvement in the accuracy of the SSS fields and provide a better understanding of the influences of different conditions on salinity. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Mean spatial bias correction fields (psu) for Aquarius (left) ascending and (right) descending data: (top) beam 1, (middle) beam 2, and (bottom) beam 3.</p> Full article ">Figure 2
<p>Relationships between the RMSD and (<b>a</b>) <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">k</mi> <mn>1</mn> </msub> </mrow> </semantics></math>, (<b>b</b>) <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">k</mi> <mn>2</mn> </msub> </mrow> </semantics></math>, and (<b>c</b>) <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">k</mi> <mn>3</mn> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 3
<p>Weekly sea surface salinity (SSS) fields from Aquarius for 2 July–8 July 2013 constructed using different algorithms: (<b>a</b>) weighted average fitting (WAF), (<b>b</b>) weighted unary linear fitting (WULF), (<b>c</b>) weighted binary linear fitting (WBLF), and (<b>d</b>) the Aquarius SSS L3 product data provided by Aquarius Data Processing System (ADPS).</p> Full article ">Figure 4
<p>Time series of the weekly (<b>a</b>) root mean squared differences (RMSDs), (<b>b</b>) biases and (<b>c</b>) correlations between the Argo buoy data and the four Aquarius SSS analyses: WAF (magenta), weighted unary linear fitting (WULF) (green), WBLF (blue) and the official L3 SSS products provided by ADPS (red). The error statistics were calculated by comparing the Argo buoy measurements for a given week with the SSS values at the same locations obtained through interpolation of the corresponding SSS maps.</p> Full article ">Figure 5
<p>Statistics of the differences between the Argo buoy data and the results of the four Aquarius SSS analyses: (<b>a</b>) WAF, (<b>b</b>) WULF, (<b>c</b>) WBLF, and (<b>d</b>) the official standard L3 SSS products provided by ADPS. The error statistics were calculated by comparing the Argo buoy measurements for all weeks between September 2011 and May 2015 with the SSS values at the same locations obtained through interpolation of the corresponding Aquarius SSS maps.</p> Full article ">Figure 6
<p>Scatter plots of the results of the Aquarius weekly SSS analyses and the collocated Argo buoy data. The Aquarius SSS analyses are (<b>a</b>) WAF, (<b>b</b>) WULF, (<b>c</b>) WBLF, and (<b>d</b>) the official standard L3 SSS product provided by ADPS. The colors represent the number of points in each 0.1 psu bin. The error statistics were calculated by comparing the Argo buoy measurements with the SSS values at the same locations obtained via interpolation of the corresponding Aquarius SSS maps for all weeks between September 2011 and May 2015.</p> Full article ">Figure 7
<p>Distribution of each condition. The colors indicate the normalized frequency of each condition in each 2° × 2° bin. The conditions represented in panels (<b>a</b>) to (<b>n</b>) directly correspond to those listed in <a href="#remotesensing-11-01131-t002" class="html-table">Table 2</a>.</p> Full article ">Figure 8
<p>SSS difference map between the WAF and ADPS products (WAF - ADPS): (<b>a</b>) is the total difference, (<b>b</b>) is the part of the difference due to the large-scale bias adjustment, and (<b>c</b>) is the part of the difference due to the weighting process.</p> Full article ">
23 pages, 7862 KiB  
Article
Retrieving Phytoplankton Size Class from the Absorption Coefficient and Chlorophyll A Concentration Based on Support Vector Machine
by Lin Deng, Wen Zhou, Wenxi Cao, Wendi Zheng, Guifen Wang, Zhantang Xu, Cai Li, Yuezhong Yang, Shuibo Hu and Wenjing Zhao
Remote Sens. 2019, 11(9), 1054; https://doi.org/10.3390/rs11091054 - 4 May 2019
Cited by 22 | Viewed by 4220
Abstract
The phytoplankton size class (PSC) plays an important role in biogeochemical processes in the ocean. In this study, a regional model of PSCs is proposed to retrieve vertical PSCs from the total minus water absorption coefficient (at-w(λ)) and Chlorophyll a [...] Read more.
The phytoplankton size class (PSC) plays an important role in biogeochemical processes in the ocean. In this study, a regional model of PSCs is proposed to retrieve vertical PSCs from the total minus water absorption coefficient (at-w(λ)) and Chlorophyll a concentration (Chla). The PSC model is developed by first reconstructing phytoplankton absorption and Chla from at-w(λ), and then extracting PSC from them using the support vector machine (SVM). In situ bio-optical data collected in the South China Sea from 2006 to 2013 were used to train the SVM. The proposed PSC model was subsequently validated using an independent PSC dataset from the Northeast South China Sea Cruise in 2015. The results indicate that the PSC model performed better than the three components model, with a value of r2 between 0.35 and 0.66, and the absolute percentage difference between 56% and 181%. On the whole, our PSC model shows a remarkable utility in terms of inferring vertical PSCs from the South China Sea. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Locations of SCS dataset (purple square), NESCS dataset (red circle), WSCS dataset (yellow circle), the transect-A (green triangle), and Station 50 (cyan triangle). The upward direction is due north.</p> Full article ">Figure 2
<p>Schematic of regional PSC model building and steps of application.</p> Full article ">Figure 3
<p>Ternary plots showing the fm, fn, and fp of SCS and NESCS datasets.</p> Full article ">Figure 4
<p>Boxplot of retrieved PSCs for three inputs (red, green, and blue represent SVM-Type1, SVM-Type2, and SVM-Type3, respectively) for (<b>a</b>) the training dataset and (<b>b</b>) the test dataset.</p> Full article ">Figure 5
<p>Cross-validation of splitting the data into training and testing datasets. (<b>a</b>) Absolute percentage differences (in %) of model derived for PSCs with respect to ratio of training dataset. (<b>b</b>) Coefficient of determination of the derived PSCs with respect to ratio of training dataset. The broken lines indicate the test dataset. Red, green, and blue represent Cm, Cn, and Cp, respectively.</p> Full article ">Figure 6
<p>Cross-validation of random pick tests. (<b>a</b>) Absolute percentage differences (in %) of randomly picked training and test datasets in estimation of PSC with respect to statistic quartiles. (<b>b</b>) Coefficient of determination of randomly picked training and test datasets in the estimation of PSC with respect to statistic quartiles. The straight lines indicate the training datasets and broken lines indicate the test dataset. The dotted lines represent the statistical locations of the first and third quartiles. Red, green, and blue represent Cm, Cn, and Cp, respectively.</p> Full article ">Figure 7
<p>Scatter plots of the PSC derived from the model against in situ PSC. (<b>a</b>,<b>b</b>,<b>c</b>) The scatter plots of the training dataset. <b>(d</b>,<b>e</b>,<b>f)</b> The scatter plots of the test dataset. (<b>g</b>,<b>h</b>,<b>i</b>) The scatter plots of SVM-SCM applied to the NESCS dataset. (<b>j</b>,<b>k</b>,<b>l</b>) The scatter plots of SVM-Bricaud95 applied to the NESCS dataset. The black line represents the 1:1 line and dotted lines represent the 1:1 line ± 30% log<sub>10</sub> PSC.</p> Full article ">Figure 8
<p>The vertical distribution of PSC and Chla retrieved by SVM-Bricaud95 at Station 50. The solid circles represent the PSC measured using the HPLC method. The open circles represent the profile PSC derived from SVM-Bricaud95. The dotted lines represent the range within one-fold APD.</p> Full article ">Figure 9
<p>Vertical distribution along transect-A of the concentrations of total chlorophyll (<b>a</b>), and the proportions of micro- (<b>b</b>), nano- (<b>c</b>), and pico-phytoplankton (<b>d</b>), respectively.</p> Full article ">Figure 10
<p>(<b>a</b>) Fitted curve and scatter plots between a<sub>LH</sub>(676) and Chla of NESCS dataset. (<b>b</b>) Scatter plots of RChla<sub>LH</sub> and in situ Chla. The black line represents the 1:1 line and the dotted lines represent the 1:1 line ± 30% log<sub>10</sub> PSC.</p> Full article ">Figure 11
<p>Scatter plots of PSC retrieved using SVM-Bricaud95 and SVM-Bricaud95 (in situ Chla) for (<b>a</b>) Cm, (<b>b</b>) Cn, and (<b>c</b>) Cp. Scatter plots of SVM-Bricaud95 and SVM-Bricaud95 (in situ Chla) (open circle) versus (solid circle) (<b>d</b>). The black line represents the 1:1 line and the dotted lines represent the 1:1 line ± 30% log<sub>10</sub> PSC.</p> Full article ">Figure 12
<p>Scatter plots of PSC derived using SVM-SCM PSC versus SVM-Bricaud95 (<b>a</b>). <span class="html-italic">a<sub>ph</sub></span>(443)-specific absorption coefficients of phytoplankton reconstructed by Bricaud95 methods (<b>b</b>). Black line represents the 1:1 line and dotted lines represent the 1:1 line ± 30% log<sub>10</sub> PSC.</p> Full article ">
17 pages, 4733 KiB  
Article
Remote Sensing Estimation of Sea Surface Salinity from GOCI Measurements in the Southern Yellow Sea
by Deyong Sun, Xiaoping Su, Zhongfeng Qiu, Shengqiang Wang, Zhihua Mao and Yijun He
Remote Sens. 2019, 11(7), 775; https://doi.org/10.3390/rs11070775 - 31 Mar 2019
Cited by 24 | Viewed by 5416
Abstract
Knowledge about the spatiotemporal distribution of sea surface salinity (SSS) provides valuable and important information for understanding various marine biogeochemical processes and ecosystems, especially for those coastal waters significantly affected by human activities. Remote-sensing techniques have been used to monitor salinity in the [...] Read more.
Knowledge about the spatiotemporal distribution of sea surface salinity (SSS) provides valuable and important information for understanding various marine biogeochemical processes and ecosystems, especially for those coastal waters significantly affected by human activities. Remote-sensing techniques have been used to monitor salinity in the open ocean with their advantages of wide-area surveys and real-time monitoring. However, potential challenges remain when using satellite data with coarse spatiotemporal resolutions, leading to a loss of valuable information. In the current study, based on the local dataset collected over the southern Yellow Sea (SYS), a region-customized algorithm was developed to estimate SSS by using the remote sensing reflectance. The model evaluations indicated that our algorithm yielded good SSS estimation, with a root-mean-square error (RMSE) of 0.29 psu and a mean absolute percentage error (MAPE) of 0.75%. Satellite-derived SSS results compared well with those derived from in situ observations, further suggesting the good performance of our developed algorithm for the study regions. We applied this algorithm to Geostationary Ocean Color Imager (GOCI) data for the month of August from 2011 to 2018 in the SYS, and produced the spatial distribution patterns of the SSS for August of each year. The SSS values were high in offshore waters and lower in coastal waters, especially in the Yangtze River estuary. The negative correlation between the monthly Changjiang River discharge (CRD) and SSS (R = −0.71, p < 0.001) near the Yangtze River estuary was observed, suggesting that the SSS distribution in the Yangtze River estuary was potentially influenced by the CRD. In offshore waters, the correlation between SSS and CRD was weak (R < 0.2), suggesting that the riverine discharge’s effect might be weak. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Figure 1

Figure 1
<p>Locations of the sampling stations collected in the SYS during November 2014, August 2015, July 2016, and July 2018.</p> Full article ">Figure 2
<p>(<b>a</b>) Histogram of sea surface salinity (SSS) collected from the SYS. The black line represents a normal distribution curve; (<b>b</b>) Spectral shapes of the measured Rrs(λ) overlaid by the locations of 6 Geostationary Ocean Color Imager (GOCI) channels (blue bars). The green line represents the coefficient of variation of Rrs(λ). The solid black line represents the mean of Rrs(λ) spectra and the black dotted line is the mean plus or minus the standard deviation.</p> Full article ">Figure 3
<p>(<b>a</b>) Spectral distribution of the correlation coefficients between Rrs(λ) and SSS (black line) and log<sub>10</sub>SSS (red line) in the range of 400–700 nm; (<b>b</b>) Correlation coefficients between nine band combination forms derived from the 6 GOCI bands and log<sub>10</sub>SSS. The expressions of different band forms are listed in <a href="#remotesensing-11-00775-t001" class="html-table">Table 1</a>.</p> Full article ">Figure 4
<p>Scatter plots of (<b>a</b>) X5 and (<b>b</b>) X8 versus the log<sub>10</sub>SSS. The solid red lines are the fitted function curves.</p> Full article ">Figure 5
<p>Scatter plots of the in situ measured SSS versus the estimated SSS by using (<b>a</b>) the X5-based model and (<b>b</b>) the X8-based model. Note that both R values are significant with <span class="html-italic">p</span> &lt; 0.01.</p> Full article ">Figure 6
<p>Scatter plots of the measured SSS versus the satellite-derived SSS from GOCI data using (<b>a</b>) the X5-based model and (<b>b</b>) the X8-based model. Note that both R values are not very significant, with <span class="html-italic">p</span> = 0.42 and <span class="html-italic">p</span> = 0.25 for (<b>a</b>) and (<b>b</b>) figures, respectively. The shown red dotted lines are ±3% error lines.</p> Full article ">Figure 7
<p>Comparison of the measured SSS versus the derived SSS between (<b>a</b>) the retuned model of Song et al., (<b>b</b>) the retuned model of Yu, and (<b>c</b>) the proposed model in this paper. The solid red line refers to the 1:1 line. Note that the R values are significant with <span class="html-italic">p</span> &lt; 0.01.</p> Full article ">Figure 8
<p>Spatial distribution of SSS derived from GOCI satellite observations (<b>a</b>–<b>h</b>) during 2011–2018 in August and (<b>i</b>) eight-year SSS average in August.</p> Full article ">Figure 9
<p>Spatial distributions of sea surface salinity produced by SMAP satellite products for (<b>a</b>) August 2015 and (<b>b</b>) August 2016.</p> Full article ">Figure 10
<p>(<b>a</b>) Locations of three subareas of interest, including the central region of the SYS (CYS), the Yangtze River estuary (YRE), and the region near Jeju Island (JI); (<b>b</b>) Monthly mean time series of SSS for the three subareas.</p> Full article ">Figure 11
<p>(<b>a</b>) Monthly Changjiang River discharge (CRD) and the satellite-derived SSS of three subareas of interest in August during 2011 to 2016. Relationship between CRD and the satellite-derived SSS (<b>b</b>) in the YRE and (<b>c</b>) in the CYS (red points) and the JI (blue points). The solid lines are the fitted lines.</p> Full article ">
21 pages, 3985 KiB  
Article
Retrieving the Lake Trophic Level Index with Landsat-8 Image by Atmospheric Parameter and RBF: A Case Study of Lakes in Wuhan, China
by Yadong Zhou, Baoyin He, Fei Xiao, Qi Feng, Jiefeng Kou and Hui Liu
Remote Sens. 2019, 11(4), 457; https://doi.org/10.3390/rs11040457 - 22 Feb 2019
Cited by 17 | Viewed by 5027
Abstract
The importance of atmospheric correction is pronounced for retrieving physical parameters in aquatic systems. To improve the retrieval accuracy of trophic level index (TLI), we built eight models with 43 samples in Wuhan and proposed an improved method by taking atmospheric water vapor [...] Read more.
The importance of atmospheric correction is pronounced for retrieving physical parameters in aquatic systems. To improve the retrieval accuracy of trophic level index (TLI), we built eight models with 43 samples in Wuhan and proposed an improved method by taking atmospheric water vapor (AWV) information and Landsat-8 (L8) remote sensing image into the input layer of radical basis function (RBF) neural network. All image information taken in RBF have been radiometrically calibrated. Except model(a), image data used in the other seven models were not atmospherically corrected. The eight models have different inputs and the same output (TLI). The models are as follows: (1) model(a), the inputs are seven single bands; (2) model(c), besides seven single bands (b1, b2, b3, b4, b5, b6, b7), we added the AWV parameter k1 to the inputs; (3) model(c1), the inputs are AWV difference coefficient k2 and the seven bands; (4) model(c2), the input layers include seven single bands, k1 and k2; (5) model(b), seven band ratios (b3/b5, b1/b2, b3/b7, b2/b5, b2/b7, b3/b6, and b3/b4) were used as input parameters; (6) model(b1), the inputs are k1 and seven band ratios; (7) model(b2), the inputs are k2 and seven band ratios; (8) model(b3), the inputs are k1, k2, and seven band ratios. We estimated models with root mean squared error (RMSE), model(a) > model(b3) > model(b1) > model(c2) > model(c) > model(b) > model(c1) > model(b2). RMSE of the eight models are 12.762, 11.274, 10.577, 8.904, 8.361, 6.396, 5.389, and 5.104, respectively. Model b2 and c1 are two best models in these experiments, which confirms both the seven single bands and band ratios with k2 are superior to other models. Results also corroborate that most lakes in Wuhan urban area are in mesotrophic and light eutrophic states. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Study area and stations.</p> Full article ">Figure 2
<p>The workflow of this research.</p> Full article ">Figure 3
<p>Spectral of thermal infrared channels of Landsat8 and AVHRR made by ENVI5.3.</p> Full article ">Figure 4
<p>Scatter diagram of measured-retrieved TLI based on seven single bands and atmospheric parameters.</p> Full article ">Figure 5
<p>Scatter diagram of measured-retrieved TLI based on seven band ratios and atmospheric parameters.</p> Full article ">Figure 6
<p>Distribution of TLI-retrieved data based on single bands and atmospheric parameters.</p> Full article ">Figure 7
<p>Distribution of TLI-retrieved data based on band ratios and atmospheric parameters.</p> Full article ">
21 pages, 4830 KiB  
Article
Optical Classification of the Remote Sensing Reflectance and Its Application in Deriving the Specific Phytoplankton Absorption in Optically Complex Lakes
by Kun Xue, Ronghua Ma, Dian Wang and Ming Shen
Remote Sens. 2019, 11(2), 184; https://doi.org/10.3390/rs11020184 - 18 Jan 2019
Cited by 39 | Viewed by 5800
Abstract
Optical water types (OWTs) were identified from remote sensing reflectance (Rrs(λ)) values in a field-measured dataset of several large lakes in the lower reaches of the Yangtze and Huai River (LYHR) Basin. Four OWTs were determined from normalized remote sensing [...] Read more.
Optical water types (OWTs) were identified from remote sensing reflectance (Rrs(λ)) values in a field-measured dataset of several large lakes in the lower reaches of the Yangtze and Huai River (LYHR) Basin. Four OWTs were determined from normalized remote sensing reflectance spectra (NRrs(λ)) using the k-means clustering approach, and were identified in the Sentinel 3A OLCI (Ocean Land Color Instrument) image data over lakes in the LYHR Basin. The results showed that 1) Each OWT is associated with different bio-optical properties, such as the concentration of chlorophyll-a (Chla), suspended particulate matter (SPM), proportion of suspended particulate inorganic matter (SPIM), and absorption coefficient of each component. One optical water type showed an obvious characteristic with a high contribution of mineral particles, while one type was mostly determined by a high content of phytoplankton. The other types belonged to the optically mixed water types. 2) Class-specific Chla inversion algorithms performed better for all water types, except type 4, compared to the overall dataset. In addition, class-specific inversion algorithms for estimating the Chla-specific absorption coefficient of phytoplankton at 443 nm (a*ph(443)) were developed based on the relationship between a*ph(443) and Chla of each OWT. The spatial variations in the class-specific model-derived a*ph(443) values were illustrated for 2 March 2017, and 24 October 2017. 3) The dominant water type and the Shannon index (H) were used to characterize the optical variability or similarity of the lakes in the LYHR Basin using cloud-free OLCI images in 2017. A high optical variation was located in the western and southern parts of Lake Taihu, the southern part of Lake Hongze, Lake Chaohu, and several small lakes near the Yangtze River, while the northern part of Lake Hongze had a low optical diversity. This work demonstrates the potential and necessity of optical classification in estimating bio-optical parameters using class-specific inversion algorithms and monitoring of the optical variations in optically complex and dynamic lake waters. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Location of the lakes in the lower reaches of the Yangtze and Huai River (LYHR) Basin. The field samples of Lake Chaohu, Lake Taihu, and Lake Hongze were collected from 2011 to 2017. The validation data were match-up pairs of field data and Ocean Land Color Instrument (OLCI)-derived data.</p> Full article ">Figure 2
<p>Comparison of the field-measured <span class="html-italic">R<sub>rs</sub></span> and OLCI-derived <span class="html-italic">R<sub>rs</sub></span> using the (<b>a</b>) C2RCC, (<b>b</b>) POLYMER, and (<b>c</b>) 6SV atmospheric correction models for match-up pairs at different OLCI bands (<span class="html-italic">N</span> = 63). (<b>d</b>) MAPE of C2RCC, POLYMER, and 6SV at different OLCI bands, error bars represent one standard deviation of the absolute percentage error in the validation data.</p> Full article ">Figure 3
<p>Performance of the three unsupervised clustering methods: heritage clustering, fuzzy <span class="html-italic">c</span>-means (FCM), and <span class="html-italic">k</span>-means in clustering waters with different number of types: (<b>a</b>) silhouette coefficient, (<b>b</b>) SSE (sum of the squared errors), and (<b>c</b>) STD (standard deviation).</p> Full article ">Figure 4
<p>(<b>a</b>–<b>d</b>) <span class="html-italic">NR<sub>rs</sub></span>(λ) sorted into the four optical water types (OWTs) from the <span class="html-italic">k</span>-means cluster analysis (<span class="html-italic">N</span> = 535); blue lines: individual <span class="html-italic">NR<sub>rs</sub></span>(λ) values; red lines: mean <span class="html-italic">NR<sub>rs</sub></span>(λ) of each OWT. (<b>e</b>) The mean spectra of <span class="html-italic">NR<sub>rs</sub></span>(λ) of the four OWTs. The OWT means and covariance matrices are the basis for the membership function. Note that the optical classification was conducted using the <span class="html-italic">NR<sub>rs</sub></span>(λ) of the field data. (<b>f</b>) The mean spectra of <span class="html-italic">R<sub>rs</sub></span>(λ) of the four OWTs.</p> Full article ">Figure 5
<p>Mean spectrum of the absorption coefficients of phytoplankton (<span class="html-italic">a<sub>ph</sub></span>), NAP (<span class="html-italic">a<sub>d</sub></span>), and CDOM (<span class="html-italic">a<sub>g</sub></span>) in each OWT: (<b>a</b>) type 1, (<b>b</b>) type 2, (<b>c</b>) type 3, and (<b>d</b>) type 4.</p> Full article ">Figure 6
<p>(<b>a</b>) Mean spectra of the absorption coefficient of phytoplankton normalized to the Chl<span class="html-italic">a</span> concentration (<span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(λ)) of types 1–4. (<b>b</b>) Boxplots of <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443)/<span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(675) for each OWT in the field-measured data. (<b>c</b>) Mean spectra of the absorption coefficient of NAP normalized to the SPM concentration (<span class="html-italic">a<sup>*</sup><sub>d</sub></span>(λ)) of types 1–4. (<b>d</b>) Boxplots of <span class="html-italic">a<sup>*</sup><sub>d</sub></span>(443) for each OWT in the field-measured data. The sample median is indicated by a line within the box, the dots represent the mean value, and “x” represents data beyond the bounds of the error bars.</p> Full article ">Figure 7
<p>(<b>a</b>) Mer-3B versus field-measured Chl<span class="html-italic">a</span> content data for OLCI validation of each OWT and all data. (<b>b</b>) Chl<span class="html-italic">a</span> versus <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443) for OLCI validation of each OWT and all data. (<b>c</b>) Comparison of the field-measured Chl<span class="html-italic">a</span> and model-derived Chl<span class="html-italic">a</span> using unclassified models and classified models for each OWT and all data. (<b>d</b>) Comparison of the field-measured <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443) and model-derived <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443) using unclassified models and classified models for each OWT and all data. Note that the input Chl<span class="html-italic">a</span> data in calculating <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443) were the derived Chl<span class="html-italic">a</span> values using the class-specific model of each OWT. The number of samples (<span class="html-italic">N</span>) is 15, 15, 27, and 6, for type 1 to type 4, respectively.</p> Full article ">Figure 8
<p>(<b>a</b>) Optical water types, (<b>b</b>) Chl<span class="html-italic">a</span> derived using the unclassified Mer-3B Chl<span class="html-italic">a</span> model, (<b>c</b>) Chl<span class="html-italic">a</span> derived using the class-specific Mer-3B Chl<span class="html-italic">a</span> model, and (<b>d</b>) <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443) derived using the class-specific model on the 2 March 2017, OLCI image over the lakes in the LYHR Basin. (<b>e</b>) Optical water types, (<b>f</b>) Chl<span class="html-italic">a</span> derived using the unclassified Mer-3B Chl<span class="html-italic">a</span> model, (<b>g</b>) Chl<span class="html-italic">a</span> derived using the class-specific Mer-3B Chl<span class="html-italic">a</span> model, and (<b>h</b>) <span class="html-italic">a<sup>*</sup><sub>ph</sub></span>(443) derived using the class-specific model on the 24 October 2017, OLCI image over the lakes in the LYHR Basin.</p> Full article ">Figure 9
<p>The comparison of mean <span class="html-italic">R<sub>rs</sub></span>(λ) of the four optical water types with the optical water types in the previous studies [<a href="#B21-remotesensing-11-00184" class="html-bibr">21</a>,<a href="#B23-remotesensing-11-00184" class="html-bibr">23</a>]. The dashed lines represent mean <span class="html-italic">R<sub>rs</sub></span>(λ) of OWTs acquired from Table A1 in Moore et al. (2009) [<a href="#B21-remotesensing-11-00184" class="html-bibr">21</a>] and <a href="#remotesensing-11-00184-t002" class="html-table">Table 2</a> in Moore et al. (2014) [<a href="#B23-remotesensing-11-00184" class="html-bibr">23</a>].</p> Full article ">Figure 10
<p>(<b>a</b>) Dominant OWTs of the lakes in the LYHR Basin in 2017 (the class most frequently selected as the dominant class over the period); (<b>b</b>) Shannon index (<span class="html-italic">H</span>) computed from the frequency of the different OWTs of the lakes in the LYHR Basin in 2017. (<b>c</b>–<b>f</b>) The annual frequency of the different OWTs: (<b>c</b>) type 1, (<b>d</b>) type 2, (<b>e</b>) type 3, (<b>f</b>) type 4, associated with lakes in the LYHR basin in 2017.</p> Full article ">Figure 11
<p>Comparison of <span class="html-italic">R<sub>rs</sub></span>(λ) derived using ρ in Mobley (2015) [<a href="#B35-remotesensing-11-00184" class="html-bibr">35</a>] (<span class="html-italic">R<sub>rs</sub></span><sub>-M2015</sub>(λ)) and (<b>a</b>) <span class="html-italic">R<sub>rs</sub></span>(λ) derived using ρ in Mobley (1999) [<a href="#B34-remotesensing-11-00184" class="html-bibr">34</a>] (<span class="html-italic">R<sub>rs</sub></span><sub>-M1999</sub>(λ)), and (<b>b</b>) using ρ = 0.028 for match-up pairs (<span class="html-italic">N</span> = 63). (<b>c</b>) Comparisons between indexes (NR-2B, Mer-3B) derived using M2015 and M1999, 0.028, respectively. (<b>d</b>) Spectral RMSD of <span class="html-italic">R<sub>rs</sub></span>(λ) between ρ of M2015 and M1999 (blue line), 0.028 (red line), respectively.</p> Full article ">
23 pages, 7021 KiB  
Article
The Assessment of Landsat-8 OLI Atmospheric Correction Algorithms for Inland Waters
by Dian Wang, Ronghua Ma, Kun Xue and Steven Arthur Loiselle
Remote Sens. 2019, 11(2), 169; https://doi.org/10.3390/rs11020169 - 17 Jan 2019
Cited by 91 | Viewed by 7574
Abstract
The OLI (Operational Land Imager) sensor on Landsat-8 has the potential to meet the requirements of remote sensing of water color. However, the optical properties of inland waters are more complex than those of oceanic waters, and inland atmospheric correction presents additional challenges. [...] Read more.
The OLI (Operational Land Imager) sensor on Landsat-8 has the potential to meet the requirements of remote sensing of water color. However, the optical properties of inland waters are more complex than those of oceanic waters, and inland atmospheric correction presents additional challenges. We examined the performance of atmospheric correction (AC) methods for remote sensing over three highly turbid or hypereutrophic inland waters in China: Lake Hongze, Lake Chaohu, and Lake Taihu. Four water-AC algorithms (SWIR (Short Wave Infrared), EXP (Exponential Extrapolation), DSF (Dark Spectrum Fitting), and MUMM (Management Unit Mathematics Models)) and three land-AC algorithms (FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes), 6SV (a version of Second Simulation of the Satellite Signal in the Solar Spectrum), and QUAC (Quick Atmospheric Correction)) were assessed using Landsat-8 OLI data and concurrent in situ data. The results showed that the EXP (and DSF) together with 6SV algorithms provided the best estimates of the remote sensing reflectance (Rrs) and band ratios in water-AC algorithms and land-AC algorithms, respectively. AC algorithms showed a discriminating accuracy for different water types (turbid waters, in-water algae waters, and floating bloom waters). For turbid waters, EXP gave the best Rrs in visible bands. For the in-water algae and floating bloom waters, however, all water-algorithms failed due to an inappropriate aerosol model and non-zero reflectance at 1609 nm. The results of the study show the improvements that can be achieved considering SWIR bands and using band ratios, and the need for further development of AC algorithms for complex aquatic and atmospheric conditions, typical of inland waters. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Study lakes and in situ sampling points. (<b>a</b>) Lake Hongze, (<b>b</b>) Lake Chaohu, and (<b>c</b>) Lake Taihu. Dots, triangles, crosses, and forks represent the in situ sampling locations.</p> Full article ">Figure 2
<p>The remote sensing reflectance (<span class="html-italic">R<sub>rs</sub></span>) measured in (<b>a</b>) Lake Hongze on 24 October 2014, (<b>b</b>) Lake Chaohu on 11 October 2015 and 15 January 2016, and (<b>c</b>) Lake Taihu on 11 May 2017 and 27 May 2017. The color of lines correspond to the sampling date.</p> Full article ">Figure 3
<p>Average (solid line) and standard deviation (shadow) of <span class="html-italic">R<sub>rs</sub></span> for water types: turbid water (<b>a</b>), in-water algae (<b>b</b>), and floating bloom (<b>c</b>).</p> Full article ">Figure 4
<p>Scatter plots of <span class="html-italic">R<sub>rs</sub></span> retrieved by the water-AC algorithms (SWIR (<b>a</b>), EXP (<b>b</b>), DSF (<b>c</b>), and MUMM (<b>d</b>)) versus in situ <span class="html-italic">R<sub>rs</sub></span>.</p> Full article ">Figure 5
<p>Scatter plots of <span class="html-italic">R<sub>rs</sub></span> retrieved by land-AC algorithms (FLAASH (<b>a</b>), 6SV (<b>b</b>), and QUAC (<b>c</b>)) versus in situ <span class="html-italic">R<sub>rs</sub></span>.</p> Full article ">Figure 6
<p>Frequency of <span class="html-italic">R<sub>rs</sub></span>(λ) retrieved using EXP algorithm (blue line) and 6SV algorithm (red line) from OLI measurements in Lake Hongze on 24 October 2014, Lake Chaohu on 15 October 2015, January 11, 2016, and Lake Taihu on 11 May 2017 and 27 May 2017, respectively.</p> Full article ">Figure 7
<p>OLI true color image presenting, <span class="html-italic">R<sub>rs</sub></span>(λ) for Lake Hongze on 24 October 2014, Lake Chaohu on 15 October 2015, 11 January 2016, and Lake Taihu on 11 May 2017 and 27 May 2017, respectively. <span class="html-italic">R<sub>rs</sub></span>(λ) was derived by EXP and 6SV algorithm.</p> Full article ">Figure 8
<p>Scatter plots of SPM calibration between in situ measurement data and SPM model (<b>a</b>), validation between measured and simulated <span class="html-italic">R<sub>rs</sub></span>-based derived measurements (<b>b</b>).</p> Full article ">Figure 9
<p>OLI true color image, and distribution of estimated SPM patterns in Lake Chaohu on 15 October 2015, 11 January 2016. OLI-estimated SPM derived by EXP (<b>a</b>,<b>c</b>) and 6SV (<b>b</b>,<b>d</b>), and frequency of OLI-estimated SPM on 15 October 2015 (<b>f</b>) and 11 January 2016 (<b>g</b>), blue line was driven by EXP algorithm and red line was driven by 6SV algorithm. The comparison of in situ-measured SPM and OLI-estimated SPM derived by EXP and 6SV (<b>e</b>).</p> Full article ">Figure 10
<p>Distribution of RMSE (<b>right</b>) and MAPE (<b>left</b>) of bands on different water types. Blue is TW (turbid water), orange is IW (in-water algae) and black is FB (floating bloom). The different patterns of column represent different water-AC algorithms.</p> Full article ">Figure 11
<p>Distribution of RMSE (<b>right</b>) and MAPE (<b>left</b>) of bands on different water types used for land-AC algorithms. Blue is TW (turbid water), orange is IW (in-water algae), and black is FB (floating bloom). The different patterns of column represent different land-AC algorithms.</p> Full article ">Figure 12
<p>OLI-RGB images and Rayleigh-corrected reflectance (<span class="html-italic">ρ</span><sub>rc</sub>) images of two SWIR bands in Lake Taihu (<b>a</b>,<b>c</b>,<b>d</b>) and Lake Chaohu (<b>b</b>,<b>e</b>,<b>f</b>). The Rayleigh-corrected reflectance derived by SeaDAS processing. (<b>g</b>) and (<b>h</b>) were line graphs of <span class="html-italic">ρ</span><sub>rc</sub> value from figure (<b>a</b>) and (<b>b</b>) (100 × 100 pixel area), respectively.</p> Full article ">Figure 13
<p>The frequency of seasonal variation of aerosol types in Lake Taihu from 2005 to 2018. The HA, MA, SA, and HS represent highly absorbing, moderately absorbing, slightly absorbing, and highly scattering fine-mode aerosols, respectively.</p> Full article ">
16 pages, 3389 KiB  
Article
Satellite Retrieval of Surface Water Nutrients in the Coastal Regions of the East China Sea
by Difeng Wang, Qiyuan Cui, Fang Gong, Lifang Wang, Xianqiang He and Yan Bai
Remote Sens. 2018, 10(12), 1896; https://doi.org/10.3390/rs10121896 - 27 Nov 2018
Cited by 26 | Viewed by 7709
Abstract
Due to the tremendous flux of terrestrial nutrients from the Changjiang River, the waters in the coastal regions of the East China Sea (ECS) are exposed to heavy eutrophication. Satellite remote sensing was proven to be an ideal way of monitoring the spatiotemporal [...] Read more.
Due to the tremendous flux of terrestrial nutrients from the Changjiang River, the waters in the coastal regions of the East China Sea (ECS) are exposed to heavy eutrophication. Satellite remote sensing was proven to be an ideal way of monitoring the spatiotemporal variability of these nutrients. In this study, satellite retrieval models for nitrate and phosphate concentrations in the coastal regions of the ECS are proposed using the back-propagation neural network (BP-NN). Both the satellite-retrieved sea surface salinity (SSS) and remote-sensing reflectance (Rrs) were used as inputs in our model. Compared with models that only use Rrs or SSS, the newly proposed model performs much better in the study area, with determination coefficients (R2) of 0.98 and 0.83, and mean relative error (MRE) values of 18.2% and 17.2% for nitrate and phosphate concentrations, respectively. Based on the proposed model and satellite-retrieved Rrs and SSS datasets, monthly time-series maps of nitrate and phosphate concentrations in the coastal regions of the ECS for 2015–2017 were retrieved for the first time. The results show that the distribution of nutrients had a significant seasonal variation. Phosphate concentrations in the ECS were lower in spring and summer than those in autumn and winter, which was mainly due to phytoplankton uptake and utilization. However, nitrate still spread far out into the ocean in summer because the diluted Changjiang River water remained rich in nitrogen. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Figure 1

Figure 1
<p>Bathymetry and regional ocean circulations in the study area. The arrows indicate the Changjiang Diluted Water (CDW), Yellow Sea Warm Current (YSWC), SuBei Coastal Current (SBCC), Zhejiang–Fujian Coastal Current (ZFCC), and Taiwan Warm Current (TWC) [<a href="#B20-remotesensing-10-01896" class="html-bibr">20</a>,<a href="#B21-remotesensing-10-01896" class="html-bibr">21</a>]. The red dashed box is Zone A (122.5–125.5°E, 27.5–33.5°N), where the regional average was calculated for comparison with WOA13 data (detailed in <a href="#sec5dot2-remotesensing-10-01896" class="html-sec">Section 5.2</a>).</p> Full article ">Figure 2
<p>Locations of sampling sites. Two Changjiang Estuary cruises were conducted in August 2015 and March 2016, and two continental shelf cruises were conducted in December 2015 and August 2016. Black circles indicate the sites where the weather was clear during sampling and synchronous satellite observations were obtained.</p> Full article ">Figure 3
<p>The three-layer neural network used in this study. In this study, the nitrate and phosphate networks were trained separately using in situ data.</p> Full article ">Figure 4
<p>Relationships between in situ salinity and the concentration of nutrients. The color of the points represents chlorophyll concentration. Points circled in red indicate sites where nitrate and phosphate concentrations were significantly lower than conservative lines in high-salinity waters. Points circled in yellow indicate phosphate concentrations in the estuarine freshwater that are significantly lower than the conservative line.</p> Full article ">Figure 5
<p>Variation in coefficient of determination (<span class="html-italic">R</span><sup>2</sup>; blue lines) and mean relative error (MRE; red lines) with the number of hidden nodes in the nitrate and phosphate networks. Yellow dots indicate the point when the model achieved stability.</p> Full article ">Figure 6
<p>Comparisons of in situ samples and model estimates for nitrate and phosphate. The training, validation, and test data are represented by different symbols.</p> Full article ">Figure 7
<p>Evaluation of Level-2 data of the Geostationary Ocean Color Imager (GOCI). The colors of points represent different GOCI bands.</p> Full article ">Figure 8
<p>Evaluation of Soil Moisture Active Passive (SMAP) data. (<b>a</b>) Comparison of measured sea surface salinity (SSS) with matching SMAP products. (<b>b</b>) Monthly average SSS of SMAP in August 2016.</p> Full article ">Figure 9
<p>Monthly average results for concentrations of nitrogen (<b>a</b>) and phosphate (<b>b</b>) in the Changjiang River plume in the ECS from 2015 to 2017.</p> Full article ">Figure 10
<p>Comparisons between different nutrient models using (<b>a</b>,<b>b</b>) sea surface salinity (SSS) as the input parameter (SSS-based model), (<b>c</b>,<b>d</b>) remote-sensing reflectance (<span class="html-italic">R<sub>rs</sub></span>) as the input parameter (spectrum-based model), and (<b>e</b>,<b>f</b>) combined SSS and <span class="html-italic">R<sub>rs</sub></span> as input parameters (mixed model).</p> Full article ">Figure 11
<p>(<b>a</b>,<b>b</b>) Comparisons of monthly average nutrient concentration from model estimates and the World Ocean Atlas data (WOA13, V2) in Zone A (<a href="#remotesensing-10-01896-f001" class="html-fig">Figure 1</a>). (<b>c</b>,<b>d</b>) Monthly average changes in chlorophyll concentration, photosynthetically active radiation (PAR), sea surface temperature (SST), and sea surface salinity (SSS) in Zone A.</p> Full article ">
20 pages, 3999 KiB  
Article
Improved MODIS-Aqua Chlorophyll-a Retrievals in the Turbid Semi-Enclosed Ariake Bay, Japan
by Meng Meng Yang, Joji Ishizaka, Joaquim I. Goes, Helga Do R. Gomes, Elígio De Raús Maúre, Masataka Hayashi, Toshiya Katano, Naoki Fujii, Katsuya Saitoh, Takayuki Mine, Hirokazu Yamashita, Naoki Fujii and Akiko Mizuno
Remote Sens. 2018, 10(9), 1335; https://doi.org/10.3390/rs10091335 - 21 Aug 2018
Cited by 23 | Viewed by 5899
Abstract
The accurate retrieval of chlorophyll-a concentration (Chl-a) from ocean color satellite data is extremely challenging in turbid, optically complex coastal waters. Ariake Bay in Japan is a turbid semi-enclosed bay of great socio-economic significance, but it suffers from serious water quality problems, particularly [...] Read more.
The accurate retrieval of chlorophyll-a concentration (Chl-a) from ocean color satellite data is extremely challenging in turbid, optically complex coastal waters. Ariake Bay in Japan is a turbid semi-enclosed bay of great socio-economic significance, but it suffers from serious water quality problems, particularly due to red tide events. Chl-a derived from the MODerate resolution Imaging Spectroradiometer (MODIS) sensor on satellite Aqua in Ariake Bay was investigated, and it was determined that the causes of the errors were from inaccurate atmospheric correction and inappropriate in-water algorithms. To improve the accuracy of MODIS remote sensing reflectance (Rrs) in the blue and green bands, a simple method was adopted using in situ Rrs data. This method assumes that the error in MODIS Rrs(547) is small, and MODIS Rrs(412) can be estimated from MODIS Rrs(547) using a linear relation between in situ Rrs(412) and Rrs(547). We also showed that the standard MODIS Chl-a algorithm, OC3M, underestimated Chl-a, which was mostly due to water column turbidity. A new empirical switching algorithm was generated based on the relationship between in situ Chl-a and the blue-to-green band ratio, max(Rrs(443), Rrs(448)/Rrs(547), which was the same as the OC3M algorithm. The criterion of Rrs(667) of 0.005 sr−1 was used to evaluate the extent of turbidity for the switching algorithm. The results showed that the switching algorithm performed better than OC3M, and the root mean square error (RMSE) of estimated Chl-a decreased from 0.414 to 0.326. The RMSE for MODIS Chl-a using the recalculated Rrs and the switching algorithm was 0.287, which was a significant improvement from the RMSE of 0.610, which was obtained using standard MODIS Chl-a. Finally, the accuracy of our method was tested with an independent dataset collected by the local Fisheries Research Institute, and the results revealed that the switching algorithm with the recalculated Rrs reduced the RMSE of MODIS Chl-a from 0.412 of the standard to 0.335. Full article
(This article belongs to the Special Issue Satellite Monitoring of Water Quality and Water Environment)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Location of Ariake Bay, Japan and sampling stations for this study. Water depth of the bay is shown in color. Station locations of data collected by Nagoya and Nagasaki universities, and Saga, Kumamoto, and Fukuoka Fishery Research institutes (<a href="#remotesensing-10-01335-t001" class="html-table">Table 1</a>) are shown by color symbols.</p> Full article ">Figure 2
<p>Scatter plots of in situ Chl-a and standard MODerate resolution Imaging Spectroradiometer (MODIS) Chl-a for (<b>a</b>) Nagoya and Nagasaki universities datasets, and (<b>b</b>) Fisheries Research institutes. The dotted lines are Y = X, Y = 2X, and Y = X/2.</p> Full article ">Figure 3
<p>Scatter diagrams of in situ Rrs versus MODIS Rrs for (<b>a</b>) 412 nm, (<b>b</b>) 443 nm, (<b>c</b>) 488 nm, (<b>d</b>) 547 nm, (<b>e</b>) 667 nm, and (<b>f</b>) an OC3M band ratio. Triangles and circles in (<b>a</b>–<b>c</b>,<b>f</b>) represent cases where the standard MODIS Rrs(412) was smaller or larger than the Rrs(412) estimated from Rrs(547), respectively. Unfilled and filled symbols in (<b>a</b>–<b>c</b>,<b>f</b>) represent standard and recalculated data, respectively. Dashed black line is Y = X.</p> Full article ">Figure 4
<p>Relationship between in situ Rrs(412) and Rrs(547) used to improve the retrieval of MODIS Rrs(412).</p> Full article ">Figure 5
<p>Spatial distributions of the difference between recalculated and standard MODIS Rrs(488), and comparison of Rrs and normalized Rrs for in situ, standard, and recalculated data indicated by blue, red, and green color, respectively; (<b>a</b>) 6 August 2003, (<b>b</b>) 10 August 2004, (<b>c</b>) 14 May 2010, and (<b>d</b>) 10 February 2016. Right panels showed the examples of the spectra of match-up locations; (<b>e</b>,<b>f</b>) 6 August 2003, (<b>g</b>,<b>h</b>) 10 August 2004, (<b>i</b>,<b>j</b>) 14 May 2010, and (<b>k</b>,<b>l</b>) 10 February 2016. The black symbol in (<b>a</b>–<b>d</b>) represents the locations from where the Rrs spectra was derived.</p> Full article ">Figure 6
<p>(<b>a</b>) Comparison of in situ and OC3M estimated Chl-a. (<b>b</b>) Relation between in situ Chl-a and max(Rrs443, Rrs488)/Rrs547 (R). Data are from Nagoya and Nagasaki universities datasets. The dash and dotted lines in (<b>a</b>) are Y = X, Y = 2X/Y, and Y = X/2, respectively. The dashed line in (<b>b</b>) is the OC3M algorithm.</p> Full article ">Figure 7
<p>(<b>a</b>) Ternary plot of a<sub>ph</sub>(443), a<sub>y</sub>(443), and a<sub>npp</sub>(443) for data from Ariake Bay, the East China Sea, and Ise Bay. The value on each side represents the ratio of the corresponding water constituents of absorption to the total absorption. (<b>b</b>) Relation between Chl-a and OC3M band ratio. (<b>c</b>) Relation between in situ Rrs(667) and a<sub>npp</sub>(443). Red, green, dark blue, yellow, light blue, purple, and black symbols represent the waters of TSM-dominated, phytoplankton-dominated, CDOM-dominated, a mixture of TSM-dominated and phytoplankton-dominated water, a mixture of phytoplankton-dominated and CDOM-dominated water, a mixture of CDOM-dominated and TSM-dominated water, and a mixture of TSM-dominated, phytoplankton-dominated, and CDOM-dominated water, respectively. Circles and triangles represent the Ariake Bay dataset and combined data from Ise Bay and the East China Sea, respectively.</p> Full article ">Figure 8
<p>Relation between in situ Chl-a and max(Rrs443, Rrs488)/Rrs547 (R). Red and blue symbols represent the subsets of non-turbid and turbid waters, respectively, from Ariake Bay. The dashed lines with lower and higher slope represents the regression for non-turbid waters and turbid waters, respectively. The equations of the second order polynomial and linear regressions represent the switching algorithm for non-turbid and turbid waters, respectively.</p> Full article ">Figure 9
<p>Scatter plots of corrected MODIS Chl-a versus in situ Chl-a collected by Nagoya and Nagasaki universities (<b>a</b>,<b>c</b>), and Fisheries Research institutes (<b>b</b>,<b>d</b>). The solid and dotted lines are Y = X, Y = 2X, and Y = X/2, respectively. Chl-a calculated using OC3M algorithm (<b>a</b>,<b>b</b>) and using new switching algorithm (<b>c</b>,<b>d</b>).</p> Full article ">Figure 10
<p>Comparison between standard and improved MODIS Chl-a. (<b>a</b>) 6 August 2003; (<b>b</b>) 10 August 2004; (<b>c</b>) 14 May 2010; (<b>d</b>) 10 February 2016.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-12
Back to TopTop