摘要
为了弥补传统水质监测的不足,加强赤潮的风险管控,基于Landsat 8 OLI遥感影像和实地监测数据,分析了遥感影像光谱特征对叶绿素a浓度变化的响应关系,统计各种水体指数(FAI、VB-FAH、NDCI等)与叶绿素a实测浓度之间的相关关系,对最佳的相关性指数FAI进行改进;再分别基于相关系数较高的水体指数(改进的FAI、FAI、VB-FAH)建立叶绿素a浓度反演模型,并评价反演精度。结果表明:改进的FAI模型和叶绿素a浓度的相关性最高,相关系数R达到0.74;构建的一阶线性反演模型的精度最高(R 2为0.79,RMSE为1.66,MAPE为26%)。利用改进FAI建立的一阶线性反演模型对整个研究区域的叶绿素a浓度进行反演,结果表明:叶绿素a浓度与到岸距离呈负相关关系,河流入海口叶绿素a浓度较高,南流江入海口的叶绿素a浓度值相较于其他入海口更高,河流流域水质整体较差。改进的FAI建立的一阶线性反演模型计算简单、精度较高,可以快速有效地反演北部湾近岸海域的叶绿素a浓度。
To improve water quality monitoring and reduce the risk of red tide,the correlation between changes of chlorophyll-a concentration and spectral characteristics of remote sensing images was examined based on the Landsat 8 OLI remote sensing images and field monitoring data.The correlation between several water indexes and measured concentration of chlorophyll-a was analyzed.The most relative index(FAI)was improved.The chlorophyll-a inversion model is established based on the water body index(improved FAI,FAI,VB-FAH)with high correlation coefficient,and the inversion accuracy was evaluated.The improved FAI model and chlorophyll-a concentration have the highest correlation.The correlation coefficient(R)is 0.74,the first-order linear inversion model is accurate.The decision coefficient(R 2)is 0.79,RMSE is 1.66,MAPE is 26%.The firstorder linear inversion model established by the improved FAI inverted chlorophyll-a concentration of the whole study area.The results showed that chlorophyll-a concentration was negatively correlated with shore distancethe.The concentration value of chlorophyll-a in the mouth of Nanliu River is higher than other estuaries,the river basin water quality is generally poor.The first-order linear inversion model established by improved FAI is simple in calculation and high in precision.It can effectively reverse the chlorophyll-a concentration in the nearshore water of the Beibu Gulf.
作者
姚焕玫
廖鹏任
韦毅明
钟炜萍
纳泽林
黄以
文可
YAO Huan-mei;LIAO Peng-ren;WEI Yi-ming;ZHONG Wei-ping;NA Ze-ling;HUANG Yi;WEN Ke(School of Resources,Environment and Materials,Guangxi University,Nanning 530004,China)
出处
《桂林理工大学学报》
CAS
北大核心
2022年第2期469-476,共8页
Journal of Guilin University of Technology
基金
广西自然科学基金面上项目(2018GXNSFAA281100)。
