Optical classification of contrasted coastal waters

• Published in: Remote sensing of environment Vol. 123; pp. 306 - 323
• Main Authors: Vantrepotte, V., Loisel, H., Dessailly, D., Mériaux, X.
• Format: Journal Article
• Language: English
• Published: New York, NY Elsevier Inc 01.08.2012; Elsevier
• Subjects: algorithms; Animal, plant and microbial ecology; Applied geophysics; Biological and medical sciences; Classification; coastal water; Coastal waters; coasts; color; data collection; Earth Sciences; Earth, ocean, space; English Channel; Exact sciences and technology; French Guiana; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; hydrodynamics; Internal geophysics; monitoring; North Sea; Oceanography; Optical; phytoplankton; Reflectance; remote sensing; Sciences of the Universe; statistical analysis; Teledetection and vegetation maps; South America; French Guiana; Atlantic Ocean; English Channel; Northeast Atlantic; North Atlantic; North Sea; Coastal waters; Optical; Classification; Reflectance; turbidity; standardization; coastal zone; Daily; Reflection spectrum; phytoplankton; plankton; Coastal water; magnitude; dissolved materials; particles; Littoral zone; Load; Character; classification; biogeochemistry; in situ; Type; materials; Environment; Parameter; Typology; statistical analysis; seasonal variations
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2012.03.004

Optical water types were identified from an in situ data set of concomitant biogeochemical and optical parameters collected in contrasted turbid coastal areas of the eastern English Channel, southern North Sea and French Guiana at different seasons (211 stations). Four optical classes have been defined using a clustering approach performed on the spectrally normalized reflectance spectra. Normalization of the reflectance spectra was carried out during the statistical analysis to emphasize the shape of the reflectances rather than their magnitude. Each optical water type is associated with a specific bio-optical environment, in agreement with previous works. Two classes present a very marked optical character, one being mostly determined by strong phytoplankton and dissolved material loads, and the other one by a high proportion of mineral particles. The two remaining classes are related to optically mixed situations although there are some differences in the relative proportion of particulate mineral material. Applying this optical typology to the SeaWiFS daily reflectance data, we emphasized the high representativeness of these 4 optical water types which allow to describe about two thirds of the reflectance spectra found within the development sites whatever the season. The adequacy of optical water type definition for monitoring the spatio-temporal variability of coastal water masses optical quality, which reflects the impact of biological and hydrodynamic processes occurring at different time scales (i.e. from high frequency to seasonal processes), has been demonstrated. The four optical classes' typology has been shown to remain highly representative at global scale. This underlines the effective optical vicinity of some parts of the coastal ocean during some periods of the year, in spite of the recognized high optical diversity of coastal waters. This further demonstrates the high pertinence of class-based approach for large scale coastal applications. Finally, the potential for class-based inversion algorithms for improving ocean color products retrieval, as well as the applicability of such class-specific algorithms with respect to satellite information have been illustrated from the estimation of the suspended matter concentration. This work provides very encouraging evidence of the potential and adequacy of class-based inversion methods for deriving bio-optical products in optically complex waters such as the coastal ocean. ► We demonstrate the applicability of class-based approach in coastal waters. ► Using optical classification is of interest for monitoring coastal water dynamics. ► Using class-based algorithms can ameliorate coastal water ocean color products.