以太湖为研究区域,测试水体的反射光谱和表层水质并分析叶绿素的光谱特征,对比MODIS波段并找出与叶绿素a质量浓度相关性较好或对叶绿素a质量浓度变化较为敏感的波段,根据主成分变换后的MODIS图像色调对太湖进行分区并分别用主成分、单波段及波段组合因子建立各区内相应的叶绿素a质量浓度估测模型.结果表明:对太湖合理分区可以提高叶绿素a质量浓度的遥感估测精度;经过主成分变化后的第2主成分,MODIS波段1,3-4,10~14的反射率及波段反射率组合可以较好地估测分区后的叶绿素a质量浓度。
With Taihu Lake as the area of study, the reflection spectrum of water body and water quality of surface layer of Taihu Lake were detected, and the spectrum characteristics of chlorophyll were analyzed. By comparison of MODIS bands, the bands which are of high correlation with or sensitive to the variation of chlomphyll-a(Chla) concentration were selected. Then, Taihu Lake was divided into two subareas by principal component analysis (PCA), and Chla concentration estimation models for each subarea were developed based on principal components, MODIS single band, and band combination factors, respectively. The result shows that the rational regionalization of Taihu Lake can improve the estimation precision of Chla concentration by RS, and that the Chla concentration of each subarea can be better evaluated by means of the second principal component, the reflectivity of MODIS bands 1, 3-4 and 10-14, and their combinations.