Research on Behavior of Land Origin Materials in Coastal Area by Ocean Color Remote Sensing

近年来海洋颜色遥感技术的发展使探测海面叶绿素a浓度成为可能。然而，在沿海水域，其他物质，如悬浮物（SS）和彩色溶解有机物（CDOM）干扰叶绿素的测定。本研究收集了大村湾、有明湾和东海海域的光学数据，探讨了利用光学性质检测浮游植物和陆源物质行为的技术。基于海面辐照度的经验水内算法检测大村湾水体叶绿素a浓度存在误差。而卫星叶绿素a的误差较大。这意味着大气校正可能是该海湾比水中算法更大的误差来源。在有明湾，除近岸混浊区域外，均可检测到赤潮的动态。在东海海域，盐度与CDOM呈负相关关系，通过海洋颜色可以探测到低盐度的长江水体。此外，除了淡水输入外，黑潮上涌和Trichodesmium水华可能是东海浮游植物变异的重要因素。海洋颜色遥感在东南亚沿海地区也很有用，但解释时应注意混浊的水。提出了基于叶绿素、SS和CDOM的光学模型的神经网络反演算法，并与实测数据进行了验证，表明基于大气校正方案的卫星反演算法具有广阔的发展前景。

Recent advancement of ocean color remote sensing makes it possible to detect sea surface chlorophyll a concentration. However, in coastal waters, other materials, such as suspended solid (SS) and colored dissolved organic matter (CDOM) interfere chlorophyll determination. In this study, optical data from Omura Bay, Ariake Sound, and East China Sea were collected, and technique to detect phytoplankton and behavior of land origin materials from the optical properties was investigated. Empirical in-water algorithm from sea surface irradiance had error to detect chlorophyll a concentration in the Omura Bay. However, error of satellite chlorophyll a was larger. This implies that atmospheric correction may be larger source of the error than in-water algorithm in this bay. In the Ariake Sound, the behavior of red tide could be detected except in turbid near coastal areas. In the East China Sea, there was a negative correlation between salinity and CDOM, and low salinity water from the Changjiang River may be able detect by ocean color. Furthermore, upwelling of Kuroshio water and bloom of Trichodesmium may be important for the phytoplankton variability in the East China Sea, other than freshwater input. Ocean color remote sensing was also useful in the coastal area of the South Eastern Asia although the interpretation should be careful for turbid water. Neural network inversion algorithm with an optical model with chlorophyll, SS and CDOM was developed, and the verification with field data indicates that future development of satellite algorithm with atmospheric correction scheme was promising.