Lee生物光学模型在不同水体组分特性下的适用性  

Accuracy assessment of bio-optical model in turbid case II waters

作　　者：王胜强[1] 陈晋[1] 杨伟[1] 梁涵玮[1] 朱晶晶[1] 

机构地区：[1]北京师范大学地表过程与资源生态国家重点实验室,北京100875

出　　处：《湖泊科学》2011年第2期217-222,共6页Journal of Lake Sciences

基　　金：国家自然科学基金项目(40871162)资助

摘　　要：辐射传输模型和生物光学模型均可用于模拟水体遥感反射率.前者模拟精度高,但计算复杂,不利于水质参数的反演;后者简便易反演,但在浑浊水体中的模拟精度还有待进一步检验.本文通过设计大量不同组分浓度组成的水体,以辐射传输模型(即Hydrolight模型)模拟结果为真值,对生物光学模型(即Lee模型)模拟二类水体遥感反射率的精度进行了检验.结果表明:(1)除藻类悬浮物主导型水体外,Lee模型在所有波段的总体模拟误差在10%以内;(2)叶绿素a浓度反演算法常用波段光谱模拟误差较大;(3)刚好在水面以上与刚好在水面以下的遥感反射率显示出相似的模拟精度.Both radiative transfer model and bio-optical model can be used to simulate the remote sensing reflectance for waters.Generally,the radiative transfer model is of high accuracy,but it is too sophisticated to retrieve water quality parameters（WQPs）.The bio-optical model is simple and easy to be implemented in retrieving WQPs,but its accuracy in simulating turbid case II waters is still unknown.In this study,the radiative transfer model（i.e.Hydrolight model） and bio-optical model（i.e.Lee model） were both used to generated a large number of spectra data with different combinations of WQPs.Using the spectra yielded by Hydrolight model as true value,we assessed the simulation accuracy of the Lee model in turbid case II waters.Results show that（1） except for the non-phytoplankton suspended substance dominated waters,results of Lee model can be acceptable with overall relative error less than 10%;（2） larger errors in Lee model occurred in the bands for Chl.a estimation than other bands;（3） simulation accuracy of remote sensing reflectance just below water surface is consistent with that just above surface.

关 键 词：遥感反射率 Hydrolight模型 Lee生物光学模型 

分 类 号：P237[天文地球—摄影测量与遥感]