摘要
湖泊水色遥感为大范围、长时序监测湖泊水质提供了可能性,遥感反射率对构建水环境参数反演模型具有重要意义。该研究以太湖和乌梁素海为研究区,结合Landsat-8 OLI数据及已有的算法对透明度、悬浮物、藻华等参数进行验证,比较了不同级别反射率产品(大气顶部反射率(TOA)、经瑞利校正后的反射率(R_(rc))、地表反射率(SR))在参数反演中的不确定性。结果表明:(1)浑浊水体反射率高于清澈水体,TOA在可见光波段反射率明显提升,R_(rc)和SR更符合水体光谱曲线特征,且R_(rc)稳定性优于SR;(2)透明度模型SDD_(Song)在乌梁素海的效果优于太湖,SDD_(Zhang)模型则相反,SPM_(Cao)模型在2个湖泊的精度均高于SPM_(Wen)模型,FAI_(Rrc)和FAI_(SR)结果具有一致性;(3)不同级别的反射率产品误差表现为TOA<R_(rc)<SR,经误差传递后基于TOA和R_(rc)构建水环境参数反演模型的不确定性均低于SR,效果较好,且太湖各模型的不确定性均低于乌梁素海,在浑浊水体的应用效果优于清澈水体。该研究结合反射率产品及模型对比不同类型的水体参数计算结果,评价不同级别的遥感反射率产品在光学特性复杂的湖泊水环境参数反演中的不确定性,为湖泊水色遥感经验算法提供理论参考,以促进湖泊水质监测的可持续发展。
Remote sensing of lake water color provides the possibility for monitoring lake water quality in a large range and long time series,and remote sensing reflectance is of great significance for the construction of water environment parameter inversion model.In this study,taking Lake Taihu and Lake Wuliangsuhai as study region,combined with Landsat-8 OLI(operational land imager)data and existing algorithms,the results of transparency,suspended solids and algal blooms are verified for the uncertainty of three reflectance products(top of atmosphere reflectance(TOA),Rayleigh corrected reflectance(R_(rc)),surface reflectance(SR))in parameter inversion.The results show that the reflectance of turbidity water is higher than that of clear water.R_(rc) and SR is more consistent with the characteristics of real water spectral curve,but TOA is significantly increased in visible bands,and the stability of R_(rc) is better than that of SR.The results of SDD_(Song) in Lake Wuliangsuhai are better than that of Lake Taihu,while the results of SDD_(Zhang) are opposite.The accuracy of SPM_(Cao) in the two lakes is higher than that of SPM_(Wen),and besides,the results of FAI_(Rrc) and FAI_(SR) show great consistency.The reflectance errors of different levels are shown as TOA<R_(rc)<SR.Through error propagation analysis,the uncertainty of water environment parameter model constructed based on TOA and R_(rc) is lower than that of SR,and the uncertainty of models in Lake Taihu is lower than that in Lake Wuliangsuhai,with the precision in turbid water better than that in clear water.This study combined different reflectance products and models to compare the results of three types of water parameters,and evaluated the uncertainty of different levels of remote sensing reflectance products in the inversion of water environmental parameters of lakes with complex optical properties,which would provide theoretical reference for the empirical algorithm of lake water color remote sensing,so as to promote the sustainable development of lake water quality monitoring.
作者
郭羽羽
李含含
隗晓琪
黄泽晖
薛坤
马荣华
胡召玲
GUO Yuyu;LI Hanhan;WEI Xiaoqi;HUANG Zehui;XUE Kun;MA Ronghua;HU Zhaoling(School of Geography,Geomatics&Planning,Jiangsu Normal University,Xuzhou 221116,China;State Key Laboratory of Lake Science and Environment,Nanjing Institute of Geography and Limnology,Chinese Academy of Sciences,Nanjing 210008,China;School of Remote Sensing and Geomatics Engineering,Nanjing University of Information Science and Technology,Nanjing 210044,China;University of Chinese Academy of Sciences,Beijing 100049,China)
出处
《环境科学与技术》
CAS
CSCD
北大核心
2023年第8期15-26,共12页
Environmental Science & Technology
基金
江苏高校优势学科建设工程资助项目(PAPD0705)
中国科学院STS项目(KFJ-STS-QYZD-2021-01-002)
内蒙古自治区科技成果转化专项资金项目(2021CG0013)
国家自然科学基金青年科学基金项目(41901281)
博士后创新人才支持计划(BG519003)
江苏师范大学研究生科研与实践创新计划项目(2022XKT0074)
