摘要
水色遥感作为目前获取大范围水体固有光学特性的唯一方法,其精度取决于所用的反演算法,而基于辐射传输理论的半分析算法凭借坚实的理论基础和良好的可解释性成为近年的研究重点,但其在不同类别水体中的反演性能未见系统分析。通过选取曼德海峡和马六甲海峡分别作为一类和二类水体的代表水域,并基于OLCI遥感影像数据,使用K_(d) 490作为间接评价参量对GSM、LMI和QAA三类主流半分析算法的反演性能进行了分析研究。结果表明三者一类水体中的反演性能LMI≈QAA>GSM,而在二类水体中GSM>QAA>LMI;在水体适应性方面,以QAA为最佳,GSM和LMI受光谱斜率参数的影响很大。因此,根据水体类型选取合适的半分析算法对于提高水体固有光学特性反演精度是十分必要的。
Ocean color remote sensing is currently the only method to obtain the inherent optical properties of large⁃scale waters,but its accuracy directly depends on the retrieval algorithms.The semi⁃analytical algorithm based on the radiative transfer theory has become the focus of research in recent years with its solid theoretical foundation and interpretability.However,there is less systematic analysis about its inversion performance to different cases of waters.In this paper,the Mandalay Strait and Malacca Strait were selected as typical representatives of Case Ⅰ and Case Ⅱ waters respectively,and remote sensing images and K_(d) 490 product of Sentinel⁃3 OLCI were used,to evaluate the inversion performance of GSM,LMI and QAA.The results show that the inversion performance of three algorithms in Case Ⅰ waters is LMI≈QAA>GSM,while in the Case Ⅱ waters,it is GSM>QAA>LMI;in terms of adaptability,QAA is the best,and GSM and LMI are greatly affected by the regional spectral slope parameters.It is necessary to select suitable semi⁃analytical algorithms for different types waters to improve the inversion accuracy of inherent optical properties.
作者
邢帅
刘宸博
王丹菂
李鹏程
焦麟
张鑫磊
XING Shuai;LIU Chenbo;WANG Dandi;LI Pengcheng;JIAO Lin;ZHANG Xinlei(School of Geospatial Information,Strategic Support Force Information Engineering University,Zhengzhou 450001,China;91937 Troops,Zhoushan 316000,China)
出处
《海洋测绘》
CSCD
北大核心
2022年第3期66-71,82,共7页
Hydrographic Surveying and Charting
基金
国家自然科学基金(41876105,41371436)。
关键词
海洋光学遥感
半分析算法
吸收系数
后向散射系数
漫衰减系数
海洋和陆地彩色成像仪
ocean optical remote sensing
semi⁃analytical algorithm
absorption coefficient
backscattering coefficient
diffuse attenuation coefficient
ocean and land color instrument
