基于TM数据的太湖叶绿素A浓度定量反演

利用TM(ETM)数据与准实时地面采样数据,建立太湖叶绿素浓度反演模型。结果表明,TM3/(TM1+TM4)与叶绿素A浓度的相关性最好,并以此建立了太湖叶绿素A浓度线性反演模型,但反演精度并不高,因此,建立了一个两层BP神经网络模型反演太湖的叶绿素A浓度,结果表明,神经网络模型的反演精度远高于线性反演模型,16个测试样本表明,神经网络模型反演的相对误差小于30%的有15个点,占总测试样本93.75%,而线性反演模型反演相对误差在30%以下的仅有3个点,这表明对于太湖这样一个光谱特征复杂的二类水体,可以利用神经网络模型反演水质参数。

Quantitative Retrieval of Chlorophyll-a by Remote Sensing in Taihu Lake based on TM data

Based on TM(ETM) data and in-situ measurements of chlorophyll-a concentration(Chla) in Taihu Lake,analysis was conducted to decide the correlation between Chla and the ratios of different reflectance corrected by the 6S model.The results show that Chla is closely related to TM3/(TM1 TM4) and the inverse model to infer Chla in Taihu Lake can be written as ln(ChlA)=-9.247*(TM1 TM4)/(TM2 TM3)-27.903*TM3/(TM1 TM4) 24.518.However,the accuracy of this model can not be enssured due to the complexity of spectral reflectance strongly depending on water quality in Taihu Lake.Thus a further 2-layer BP neural net model based on 4 input nodes,7 hide nodes and 1 output node was made to decide Chla in the lake.The derived results reveal that the BP model has much higher accuracy than the linear model.A test was made based on the chosen 16 samples and the results suggest that the maximum relative error(RE) of BP model was only 35.43%.Of all the samples,15 ones had a RE of less than 30%,accounting for 93.7% of the total samples.However,there were only 3 with RE less than 30% from the results derived from the linear model.The comparison shows that the BP model has high availability for inferring Chla of surface water having complex spectral reflectance.