基于耦合干扰效应的珠江口叶绿素a自适应遥感反演模型研究

Remote sensing is an important method to gain in-water chlorophyll-a information rapidly and effectively. However, the overall accuracy, stability and versatility of chlorophyll-a remote sensing retrieval model is not high due to effects of non-pigment suspended solids of turbid water in estuaries, lakes. There are many studies about how to eliminate the non-pigment suspended solids effect on chlorophyll-a remote sensing retrieval model, but there is still lack of theoretical study about some questions, for example how to choose optimal model and whether or how to partition modeling in view of the different effects of non-pigment suspended solids. Therefore, this project takes Pearl River turbid and productive estuary as study area. And it is systematically analyzed by use of the methods of field observation ,indoor water simulation experiment, radiation transfer simulation, that non-pigment suspended solids coupling interference effects on optical properties of chlorophyll-a and dynamic changes of chlorophyll-a sensitive characteristic bands. Then an adaptive remote sensing inversion model of Chlorophyll-a is put forward based on non-pigment suspended solids coupling interference effect in order to provide a new idea for improving the accuracy of chlorophyll-a retrieval. This study includes three main contents as follows. Firstly, inherent optical properties of water components, spatial and temporal variations of water quality parameters are researched. Secondly, non-pigment suspended solids coupling interference effect on chlorophyll-a remote sensing retrieval model is researched. Lastly, it is proposed, namely an adaptive remote sensing retrieval adaptive model for chlorophyll-a based on non-pigment suspended solids coupling interference effect.

遥感是快速、有效获取水体叶绿素a信息的重要手段。然而，河口、湖泊浑浊水体中非色素悬浮物影响常常导致叶绿素a遥感反演模型整体精度不高、稳定性与通用性不强。对于如何消除非色素悬浮物影响虽已有不少研究，但针对不同非色素悬浮物影响如何选择最优模型？怎样自动分区/分段建模最优，均缺乏系统深入地机理研究。为此，本项目以珠江口浑浊高产水域为研究对象，从物理响应机理角度，利用野外观测、室内水样模拟试验、辐射传输模拟等多种方法，系统深入地剖析非色素悬浮物对叶绿素a光学性质的耦合干扰效应及叶绿素a敏感特征波段动态变迁规律，进而提出一种基于耦合干扰效应的珠江口浑浊高产水域自适应叶绿素a遥感反演方法，为叶绿素a反演模型的改进提供一种新的思路。具体内容包括：珠江口水体固有光学特性及水质参数时空规律研究；非色素颗粒物对叶绿素a遥感反演的耦合干扰效应研究；基于耦合干扰效应的珠江口自适应叶绿素a遥感反演研究。

叶绿素 a 浓度是衡量水体初级生产力和富营养化状况的一项重要参考指标。对于珠江河口浑浊高产水体，非色素悬浮物对叶绿素 a 光学特性的耦合干扰效应直接制约了叶绿素 a 反演模型精度。项目通过对珠江口水体典型区域实地光学特性观测和室内水样实验测定，分析了CDOM、浮游藻类和非色素悬浮颗粒吸收与散射光谱时空变化规律；并从水质参数-水体组分固有光学特性-水体表观反射率三者之间内在响应机理角度，提出了非色素悬浮物耦合干扰区间的阈值判定；进而建立了不同耦合干扰区间下的叶绿素a浓度反演模型，得出以下结论： .首先，结合珠江口典型水域水样的室内化验结果和区域水动力条件，分析了CDOM、非色素颗粒物、浮游植物藻类的吸收光谱特征变化规律，构建了不同潮位条件下的水色组分固有光学特性模型。同时，基于非色素颗粒物对叶绿素a吸收光谱的耦合干扰特征，建立了受耦合干扰最弱的676nm色素吸收与叶绿素a浓度的经验关系。.其次，通过水池配比实验和Hydrolight模拟实验，分析了不同悬浮颗粒物和叶绿素a浓度条件下，两者的反射光谱特征变化；建立了各自特征光谱迁移变化规律的数学表达。.最后，根据高光谱实测数据和多光谱遥感数据的应用需求，提出了耦合干扰指数和悬浮颗粒物浓度两种非色素颗粒物对叶绿素a耦合干扰的阈值指标；并以此探究了各自的阈值范围，建立了不同耦合区间的叶绿素a反演模型，以Landsat8-OLI遥感数据为例给出了珠江口叶绿素a浓度空间分布，取得了与实测数据较一致的空间分布结果。.项目实施期间：① 发表了1篇SCI学术论文、2篇EI论文、1篇中文核心期刊论文，并有1篇EI论文和1篇中文核心已接受待刊；② 联合南京师范大学培养了1名水质遥感方向的硕士研究生；③ 构建了一套基于非色素悬浮物耦合干扰效应的珠江口叶绿素 a 遥感反演理论及方法。 .项目以水色组分之间的耦合干扰机理为出发点，研究成果丰富了珠江口水体光学特性的理论内容，提高了该区域叶绿素a遥感反演模型的精度与稳定性，可为区域化叶绿素a或相关水色组分浓度遥感反演模型的构建提供参考。