浮游植物类群遥感算法PHYSAT在台湾海峡的适用性研究  

Study of the applicability of PHYSAT method to detect phytoplankton groups from space in the Taiwan Strait

作　　者：吴璟瑜 商少凌[1] 柳欣[3] 商少平 WU Jingyu;SHANG Shaoling;LIU Xin;SHANG Shaoping(Key Laboratory of Underwater Acoustic Communication and Marine Information Technology Ministry of Education,Xiamen University,Xiamen 361005,China;College of Ocean and Earth Sciences,Xiamen University,Xiamen 361102,China;College of the Environmen & Ecology,XiamenUniversity,Xiamen361102,China)

机构地区：[1]厦门大学水声通信与海洋信息技术教育部重点实验室,福建厦门361005 [2]厦门大学海洋与地球学院,福建厦门361102 [3]厦门大学环境与生态学院,福建厦门361102

出　　处：《厦门大学学报（自然科学版）》2019年第1期70-78,共9页Journal of Xiamen University：Natural Science

基　　金：国家重点研发技术海洋环境安全保障专项(2017YFC1404804);国家自然科学基金(41776146);厦门大学校长基金(20720180106)

摘　　要：浮游植物类群遥感是海色遥感的热点问题,关乎全球变化生态响应研究及有害藻华的辨识.针对目前广泛应用的浮游植物类群遥感全球算法PHYSAT,应用台湾海峡夏季表层浮游植物光合色素与SeaWiFS同步卫星遥感数据,探讨其区域适用性.结果显示两种主要类群(硅藻(Diatom)和聚球藻(Synechococcus))的遥感光谱异常(Ra)分布交错,且同一类群的Ra在不同航次、不同站位之间也存在差异,用PHYSAT算法阈值标准均不能得到有效识别.在建立归一化离水辐射率nLwref(λ,Chla)台湾海峡区域查找表的基础上,重新生成硅藻和聚球藻的Ra,不同类群的Ra依旧混杂.这可能与建立PHYSAT算法的标准海域和台湾海峡水体光学组分差异及台湾海峡的水体光学组分时空差异,尤其是颗粒后向散射系数bbp的变动有关.采用K-means和FCM(Fuzzy c-means)方法对443nm归一化的Ra进行聚类,准确率超过70%.该结果说明在类似台湾海峡的区域水体,浮游植物类群的遥感分辨可能需要更多考虑光谱谱形上的差异,而非如PHYSAT算法进行量值范围区分.PHYSAT,one of the widely-accepted methods to detect multiple phytoplankton groups from space,was tested for its applicability in the Taiwan Strait(TWS)with 27 match-ups of in situ photosynthetic pigments collected during three cruises in summer and SeaWiFS daily data.Reflectance anomalies(Ra)spectra of the two dominant phytoplankton groups(Diatom and Synechococcus)in the TWS were completely mixed.Significant temporal and spatial variabilities in Ra distribution were also observed among samples of a specific phytoplankton group.The phytoplankton group in the TWS was not identified successfully using Ra with PHYSAT threshold and additional spectral criteria,even Ra based on the regional lookup table of nLwref(λ,Chla)for the TWS,which may be due to different bio-optical characteristics in waters,especially particulate backscattering coefficients(bbp).Clustering methods,K-means and FCM(Fuzzy c-means),were applied with Ra normalized at 443 nm.More than 70% of the samples were successfully identified.It seems that the remote sensing of phytoplankton groups may pay more attention to the difference in Raspectra shape rather than in the magnitude by PHYSAT method in regional areas such as the TWS.

关 键 词：PHYSAT 浮游植物类群 遥感 台湾海峡 阈值 光谱谱形 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]