摘要
本文利用SeaDAS软件的三种标准业务化叶绿素a浓度反演算法OC2、OC3和OC4进行了黄海和东海的叶绿素a浓度反演,并利用2003年的黄海和东海实测叶绿素a浓度数据进行验证,发现OC3和OC4算法反演结果明显偏大,OC2算法反演结果略微偏大。针对上述三种算法的不足,本文提出了一种基于岭回归的叶绿素a浓度反演模型,并利用2003年的MODIS遥感数据进行了黄海和东海的叶绿素a浓度反演,经与实测数据对比及F检验(F=25.893,p=0.000<0.05),模型预测值与实测表层叶绿素a浓度值之间的平均绝对百分比误差(MAPE)为21.8%,均方根误差(RMSE)为0.325mg/m^(3)和决定系数(R^(2))为0.847。该模型可有效地克服OC2、OC3和OC4算法存在的共线性问题,准确性和拟合度优势明显。因此,基于岭回归的叶绿素a浓度反演模型能够有效地反演近海的叶绿素a浓度。
This paper used the three standard operational chlorophyll a concentration inversion algorithms OC2,OC3,and OC4 of SeaDAS software to carry out chlorophyll a concentration inversion in the Yellow Sea and East Sea of China and validate the results by using the measured chlorophyll a concentration data of the Yellow Sea and East Sea of China in 2003.The inversion results of OC3 and OC4 algorithms were obviously large,and the inversion results of OC2 algorithm were slightly large.In view of the shortcomings of these algorithms,this paper proposed a chlorophyll concentration inversion model based on ridge regression,and utilized the MODIS remote sensing data in 2003 to carry out chlorophyll a concentration inversion in the Yellow Sea and East Sea.After comparison with the measured data and F-test(F=25.893,p=0.000<0.05),the mean absolute percentage error(MAPE)between model predictions and measured surface chlorophyll a concentrations was 21.8%,the root mean square error(RMSE)was 0.325mg/m^(3)and the coefficient of determination(R2)was 0.847,while the mean absolute percentage error(MAPE)between model predictions and measured surface chlorophyll a concentrations was 0.325 and the coefficient of determination(R^(2))was 0.847.The results showed that the inversion model can effectively overcome the deficiencies of OC2,OC3,and OC4 algorithms,with obvious advantages in accuracy and fitting degree.Therefore,the inversion model of chlorophyll a concentration based on ridge regression can effectively invert the chlorophyll a concentration in the offshore.
作者
李磊
张红梅
熊志玲
包云
LI Lei;ZHANG Hongmei;XIONG Zhiling;BAO Yun(School of Hydraulic Engineering,Nanchang Institute of Technology,Nanchang Jiangxi,330099,China)
出处
《江西水利科技》
2024年第4期235-241,共7页
Jiangxi Hydraulic Science & Technology
基金
南昌工程学院研究生创新计划项目资助(YC2023-S985).
关键词
SeaDAS大气校正
叶绿素a反演算法
岭回归
黄海和东海
SeaDAS atmospheric correction
Chlorophyll a inversion algorithm
Ridge regression
Yellow Sea and East Sea
