摘要
叶绿素是我国近海水质监测的主要参数之一,其浓度的遥感反演是监测水体光学特性、评价水体污染的重要指标。本研究以Landsat-8/OLI、FY-3A/MERSI和HJ-1B/CCD遥感影像为数据源,结合2016年实测的Chl a浓度和水体光谱特征,建立胶州湾Chl a浓度的半经验/半分析反演模型。研究表明:基于Landsat-8建立的反演模型,整体Pearson相关系数最高,最优模型的预测值与实测值的决定系数R^2>0.86,反演效果最好,能较好的适应于胶州湾Chl a浓度的反演研究。Landsat 8最佳波段组合为:2月份Band4/Band2,R^2=0.83;5月份[(Band3)^(-1)-(Band4)^(-1)]*Band5,R^2=0.80;8月份[(Band2)^(-1)-(Band3)^(-1)]*Band4,R^2=0.78;11月份[(Band4)^(-1)-(Band2)^(-1)]*Band1,R^2=0.86。
Chlorophyll is one of the main parameters of water quality monitoring in coastal waters of China.The remote sensing inversion of its concentration is an important index to monitor the water optical characteristics and evaluate water pollution.In this study,quasi-synchronous Landsat-8/OLI,FY-3A/MERSI and HJ-1B/CCD remote sensing images are used,and combined with the in-situ chlorophyll-a concentration and water spectral characteristics in 2016,a semiempirical and semi-analytical inversion model of chlorophyll-a concentration in Jiaozhou bay is established.The research shows:the inversion model with the highest correlation coefficient is derived from Landsat-8,the determinant coefficient,R2,between the in-situ data and predicted values of the optimal model is>0.86.Furthermore,the inversion effect is satisfied,and the results can be well adapted to the inversion of chlorophyll-a concentration in Jiaozhou bay.The optimum band combination of landsat 8 is:February Band4/Band2,R2=0.83;May[(Band3)-1-(Band4)-1]*Band5,R2=0.80;August[(Band2)-1-(Band3)-1]*Band4,R2=0.78;November[(Band4)-1-(Band2)-1]*Band1,R2=0.86.
作者
王爽
马安青
胡娟
侯琳琳
徐建强
WANG Shuang;MA An-qing;HU Juan;HOU Lin-lin;XU Jian-qiang(College of Environmental Science and Engineering,Ocean University of China,Qingdao 266100,China)
出处
《海洋环境科学》
CAS
CSCD
北大核心
2019年第1期75-83,共9页
Marine Environmental Science
基金
海河南系子牙河流域下游湿地生态系统恢复关键技术(931407022)
关键词
CHL
a
实测光谱
反演模型
胶州湾
chlorophyll-a
field spectral
inversion model
Jiaozhou bay