This paper process spectral data of the water based on normalization and Envelope elimination methods, by correlation analysis between land surface spectral data of single band or combining band and concentration of chlorophyll a and Suspended Solid to select band with max correlation coefficients, construct the best linear regressive models between spectral data and chlorophyll a and suspended solid. Firstly, the reflectance of the land surface of Hyperion images were obtained through geometric rectification and atmospheric modification adopted the 6S model, then the maps of spatial distributions of the concentration of chlorophyll a and suspended solid in the Nansi Lake were obtained using the best linear regressive model of chlorophyll a and the best linear regressive model of suspended solid. The results show that, data process of envelope elimination can remove the effect of the background spectral effectively and stand out the features of spectral absorption and reflection, and the combination of spectral R700/R560 and R736/R759 can be used to retrieved the concentration of chlorophyll a and suspended solid in the Nansi Lake.
Read full abstract