用分析模型方法反演水体叶绿素的浓度

Inversing Chlorophyll Concentration of Taihu Lake by Analytic Model

水体叶绿素浓度是水质评价的一个重要指标。建立水体光学传输的分析模型,进而用分析模型反演水体叶绿素浓度,对于提高遥感反演水质参数的精度和实用性具有重要意义。作者于2004年6月初对太湖18个点位进行了同步水体波谱实测和水体取样分析。由这18个点位的实测数据,利用Gordon模型建立了R(0-)的模拟模型,并进而用优化函数的方法反演水体叶绿素浓度。反演值与实测值的相关系数达到0.99,当叶绿素浓度高于30mg/m3时,反演的相对误差小于20%。进而用该模型反演2005年7月太湖7个点位的叶绿素浓度,反演值与实测值的相关系数为0.94,其中,有6个点位的反演值相对误差小于60%。

Chlorophyll concentration is an important index for evaluating water quality. To make the inversion of remote sensing more accurate and more reality, it is significant to build analytic model for simulating water R（0^-） , and by the analytic model to inverse chlorophyll concentration. 18 points in Taihu lake were chosen, and at those points, the spectra of the water and water chlorophyll concentration were measured at the same time in June, 2004. The simulation model of R （0^-） was built using Gordon model, and the chlorophyll concentration was inversed using optimization method. The correlation of measured and inversed R （0^-） is 0. 99. When the chlorophyll concentration is higher than 30mg/m^3 , the relative error of inversed value is less than 20%. Then, the chlorophyll concentration is inversed by this model for seven measured samples in July, 2004. The correlation of measured and inversed R（0^-） is 0. 94. And the relative error of the majority points is less than 60%.