SYNTHETIC DATA GENERATION METHOD FOR OCEAN COLOR REMOTE SENSING CONSIDERING COMLEX OPTICAL PROPERTIES OF COASTAL AREAS

Abstract

 In this study, we proposed a method for generating synthetic data for ocean color remote sensing that takes into account the complex optical properties of coastal areas. In order to calculate the synthetic data, radiative transfer simulations were carried out based on inherent optical properties (IOPs), and measured water quality data collected in Tokyo Bay, and the effectiveness of the radiative transfer model was confirmed. In addition, the IOPs and water quality items were categorized by hierarchical clustering, and the model was incorporated into a bio-optical model to develop the synthetic data set that can represent phytoplankton species changes, colored dissolved organic matter (CDOM) from terrestrial and internal production sources, and differences in optical properties between inorganic suspended solids and detritus. The synthetic data of Rrs and IOPs were compared with the measured data and applied to the existing IOPs estimation algorithm, and it was confirmed that the synthetic data were generally in accordance with reality.