基于四波段半分析算法和Hyperion遥感影像反演太湖叶绿素a浓度

水体叶绿素a浓度不仅是水质状况的重要指标,也是制定水环境保护和水资源开发利用方案的重要依据。以2004年8月19日太湖水质浓度实验数据和同步的Hyperion影像为数据基础,研究适用于Hyperion影像的四波段半分析算法。由模型参数标定数据集(37组)对四波段半分析算法参数的拟合分析和模型检验数据集(5组)对算法精度的评估可知,基于指数拟合方法获取的四波段半分析算法具有较高的叶绿素a浓度估算精度(相关系数为0.8913,平均绝对误差为1.1109μg/L,对应的平均相对误差为5.69%,其对应的4个波段波长分别为671.02nm、701.55nm、711.72nm和742.25nm)。用以上四波段半分析算法从Hyperion影像中提取的叶绿素a浓度呈湖心低、沿湖区域高的格局。与22.23 μg/L的年均叶绿素a浓度相比较,2004年8月19日的叶绿素a浓度处于年际较高水平。

Chlorophyll-a Concentration Estimation in Lake Taihu based on Four Band Semi-analytical Algorithm and Hyperion Imageries

The chlorophyll\|a concentration of water bodies was not only an important indicator for water quality,but also a significant basis of scheme established for water environment protection and water resources development and utilization.based on analyzing of the water qualities experimental data and Hyperion image collected on 19 August,2004,the study discussed the optimal bands of the four\|bands semi\|analytical algorithms for estimating the chlorophyll\|a concentration from Hyperion image.According to the regression analysis based on model calibration dataset (37 stations) and the accuracy estimation based on the model validation dataset (5 stations),it was found that four\|bands semi\|analytical algorithm based on the exponential fitting methods had the highest accuracy for estimation chlorophyll\|a concentration (the correlation coefficient was 0.8913,the mean absolute error was 1.1109 μg/L,the average relative error was 5.69%,and the corresponding wavelength of four bands were 671.02 nm,701.55 nm,711.72 nm and 742.25 nm).Additionally,the study used the four\|bands semi\|analytical algorithm discussed in this study to inverse the chlorophyll\|a concentration from Hyperion imageries.The inversion result showed that the chlorophyll\|a concentration was lower in the central lake and higher around the banks of lake.Comparing with the annual average chlorophyll\|a concentration,22.23 μg/L of Taihu Lake,it was known that the chlorophyll\|a concentration on 19 August,2004,was at the relative high level during the inter\|annual.