利用GOCI卫星数据开展秦皇岛海域叶绿素a浓度反演

利用2016年4月23日、4月24日、6月1日和7月6日采集的水体实测数据及同期的GOCI（Geostationary Ocean Color Imager）卫星遥感影像，开展秦皇岛海域的叶绿素a（chlorophy a，Chl a）浓度研究。通过归一化暗像元指数（normalized dark pixel index，NDPI）对影像8个波段组合的辐射率与Chl a浓度进行相关性分析，选择相关系数≥70%的波段组合作为自变量，构建Chl a浓度反演模型（R2=0.956）。经独立时相实测数据检验，模型平均相对误差为25.3%。利用该模型反演秦皇岛海域2017年12月—2018年11月的Chl a浓度，结果显示，秦皇岛海域近岸的Chl a浓度总体较高，秋、冬季节总体浓度较低，夏季出现浓度峰值。

Inversion of Chlorophyll a concentration in Qinhuangdao sea area based on GOCI satellite data

Based on the measured data of water bodies collected on april 23rd, 24th, June 1st and July 6th, 2016 and the GOCI (Geostationary Ocean Color Imager) satellite remote sensing images of the same period, the Chl a concentration in Qinhuangdao sea area is studied. The NDPI (normalized dark pixel index) method is used to analyze the correlation between the radiance and Chl a concentration of eight band combinations of the images. The band combinations with correlation coefficient≥70% are selected as independent variables to construct the Chl a concentration inversion model (R2=0.956). The average relative error of the model is 25.3% verified by independent phase measured data. The Chl a concentration in Qinhuangdao sea area from December 2017 to November 2018 was retrieved by using this model. The results show that the concentration of Chl a in the coastal area of Qinhuangdao is generally higher, the overall concentration in autumn and winter is lower, and the peak concentration appears in summer.