基于时空融合算法的水体叶绿素a反演研究  

Inversion of Chlorophyll a in Water Based on Spatio-temporal Fusion Algorithm

作　　者：陈玲[1,2] 董晓华[1,2] 马耀明[3,4,5,6,7,8] 章程焱 薄会娟[1,2] CHEN Ling;DONG Xiaohua;MA Yaoming;ZHANG Chengyan;BO Huijuan(College of Hydraulic and Environmental Engineering,China Three Gorges University,Yichang 443002,China;Engineering Research Center of Eco-environment in Three Gorges Reservoir Region,Ministry of Education,Yichang 443002,China;Land-Atmosphere Interaction and Its Climatic Effects Group,State Key Laboratory of Tibetan Plateau Earth System,Environment and Resources(TPESER),Institute of Tibetan Plateau Research,Chinese Academy of Sciences,Beijing 100101,China;College of Earth and Planetary Science,University of Chinese Academy of Sciences,Beijing 100049,China;College of Atmospheric Science,Lanzhou University,Lanzhou 730000,China;National Observation and Research Station for Qomolongma Special Atmospheric Processes and Environmental Changes,Dingri 858200,China;Kathmandu Center of Research and Education,Chinese Academy of Sciences,Beijing 100101,China;China-Pakistan Joint Research Center on Earth Sciences,Chinese Academy of Sciences,Islamabad 45320,Pakistan)

机构地区：[1]三峡大学水利与环境学院,湖北宜昌443002 [2]三峡库区生态环境教育部工程研究中心,湖北宜昌443002 [3]中国科学院青藏高原研究所青藏高原地球系统与资源环境国家重点实验室地气作用与气候效应团队,北京100101 [4]中国科学院大学地球与行星科学学院,北京100049 [5]兰州大学大气科学学院,甘肃兰州730000 [6]西藏珠穆朗玛特殊大气过程与环境变化国家野外科学观测研究站,西藏定日858200 [7]中国科学院加德满都科教中心,北京100101 [8]中国科学院中国-巴基斯坦地球科学研究中心,伊斯兰堡巴基斯坦45320

出　　处：《水文》2024年第2期26-33,共8页Journal of China Hydrology

基　　金：第二次青藏高原综合科学考察研究项目(2019QZKK0103);湖北省教育厅科学技术研究项目(Q20221209);欧洲空间局、中国国家遥感中心项目(58516)。

摘　　要：为了准确反演水体中叶绿素a浓度,以黄柏河东支流域为例,采用STNLFFM时空融合算法,对2017年GF-4和Sentinel-2反射率数据进行融合,以重构Sentinel-2影像的时间序列数据,并对应用算法前后获取的水质参数-光谱特征响应关系建立多元线性回归模型,比较模型对叶绿素a的预测效果以验证时空融合算法的可行性,利用重构后影像光谱特征与水质参数的响应关系建立人工神经网络模型,反演2017年黄柏河东支流域各水库水体叶绿素a浓度。结果表明:利用时空融合算法生成的影像接近真实影像,提高了多元线性回归模型预测叶绿素a的效果,R2从融合前0.659提高至融合后0.844,且基于时空融合算法获取的水质参数-光谱关系建立的人工神经网络模型模拟精度较好,R2和MRE达到0.925和9.461%,反演的叶绿素a浓度空间差异性明显。证明了时空融合算法在水质参数反演过程中具有较好的应用前景。In order to accurately invert the concentration of chlorophyll a in water,taking the eastern branch of Huangbai River as a case,the STNLFFM space-time fusion algorithm was used to fuse the reflectance data of GF-4 and Sentinel-2 in 2017 to reconstruct the time series data of Sentinel-2 image.A multiple linear regression model was established for the response relationship between water quality parameters and spectral characteristics obtained before and after the application of the algorithm,and the prediction effect of the model on chlorophyll a was compared to verify the feasibility of the space-time fusion algorithm.The artificial neural network model was established by using the response relationship between the reconstructed image spectral characteristics and water quality parameters to invert the chlorophyll a concentration of each reservoir in the eastern branch of Huangbai River in 2017.The results show that the image generated by the spatio-temporal fusion algorithm is close to the real image,which improves the effect of multiple linear regression model to predict chlorophyll a.The R2 is increased from 0.659 before fusion to 0.844 after fusion,and the artificial neural network model based on the water quality parameters-spectral relationship obtained by the spatio-temporal fusion algorithm has better simulation accuracy.The R2 and MRE reach 0.925 and 9.461%,and the spatial difference of retrieved chlorophyll a concentration is obvious.It is proved that the spatio-temporal fusion algorithm has a good application prospect in the process of water quality parameter inversion.

关 键 词：STNLFFM时空融合算法 黄柏河 人工神经网络 水质反演 叶绿素A 

分 类 号：X824[环境科学与工程—环境工程] TP79[自动化与计算机技术—检测技术与自动化装置] TV1[自动化与计算机技术—控制科学与工程] P33[水利工程]