湖泊水体叶绿素a含量估算的波段宽度变化影响分析--以太湖为例

The Influence of Bandwidth's Variety on Estimating Chlorophyll-a Concentration of Lake Waters: Taking Taihu Lake as an Example

为研究不同波段宽度遥感数据对监测水体叶绿素a含量的影响，以太湖水体实测高光谱遥感反射率数据为基础，分析计算不同波段宽度下遥感反射率的归一化值与叶绿素a浓度之间的相关系数。随着波段宽度在75．93nm范围内不断递增，最大相关系数逐渐减小，最大正相关波段向长波方向移动，最大负相关波段向短波方向移动。而波段宽度在31．6nm范围内变化时，最大正相关波段和最大负相关波段都会保持相对稳定。通过对不同波段处相关系数平均值和标准差的对比分析认为，718．77～734．58nm为叶绿素a遥感监测的最佳波段范围。这将对遥感传感器的波段设置，以及实际水体叶绿素a遥感监测时的波段选择，具有重要的参考价值。

In order to study the influence of different bandwidth＇s remote sensing data on estimating chlorophyll-a concentration in waters, this paper analyzes and calculates the correlation coefficient between normalized reflectance of different bandwidth and chlorophyll-a concentration based on real-measurement hyperspectral reflectance. With constant increase of bandwidth in a scope of 75.93nm, it has found that the maximum correlation coefficient decreases by degrees, the maximum positive correlation band moves to long wave, the maximum negative correla- tion band moves to short wave while estimating chlorophyll-a concentration in waters. As the bandwidth changes in the scope of 31.60nm, both the maximum positive correlation band and the maximum negative correlation band will keep a relative stable state. Finally, through contrasting the mean and standard deviation of different bandwidth＇s correlation coefficient, we think 723.51 nm -733.00nm is the best band range for estimating chloro- phyll-a concentration. This will provide important reference value for band setting of remote sensing sensor and band choosing of real chlorophyll-a estimation by remote sensing in water.