Comparison of Semi-analytical Algorithms for Estimating Inherent Optical Properties in the Korean Seas

Abstract

Ocean color remote sensing is to analyze the color change and properties of the ocean by detecting light reflected from the ocean, it is playing a very important role in ocean research. The properties of the ocean include Inherent optical properties(IOPs) representing the unique properties of seawater and apparent optical properties(AOPs) that can change due to external light environment, IOPs are important factor to characterize marine optical environments. IOPs include attenuation c(λ), absorption a (λ), scattering b(λ), and back-scattering bb(λ) coefficients, many algorithms have been developed to calculate IOPs. However, the Quasi-Analytical Algorithm(QAA), which is one of the semi-analytic algorithms(SAAs), was mainly used in the study for Korean Seas. This is algebraic and step-wise algorithm that first derive bulk IOPs and decompose a(λ) into absorption subcomponents. Since this is an algorithm that utilizes empirical formulas, input and model uncertainties are propagated in the retrieval. In addition, in the case of the Korean waters, which have various characteristics, a large error occurs in the southwest sea area with high turbidity. In this study, QAA and Generalized Inherent Optical Property(GIOP) algorithms applied to derive IOPs in the waters surrounding the Korean Peninsula using 2nd Geostationary Ocean Color Imager(GOCI-II). The GIOP algorithm is another SAAs that falls into the category of non-linear optimization. This is a bottom-up algorithm that derives a(λ) and b(λ) by simultaneously deriving bulk and subcomponent IOPs. It is intended to contribute to the development of algorithms suitable for the characteristics of the Korean Peninsula sea area through comparison of algorithm accuracy using in situ data.