Determination of chlorophyll content and trophic state of lakes using field spectrometer and IRS-1C satellite data in the Mecklenburg lake district, Germany

Multitemporal data sets of the LISS-III sensor mounted on the Indian Remote Sensing Satellite (IRS-1C) and field relectance spectra have been evaluated for estimating chlorophyll-a content in lakes. The results were compared to laboratory analyses of in situ water samples. Quantification from field reflectance spectra was carried out using the 678 nm absorption maximum and the 705 nm reflectance peak. For the evaluation of the LISS-III satelite data three approaches were compared: spectral height of the green peak, supervised maximum likelihood classification, and linear spetral unmixing. The latter gave the best results with the highest certainty measure of R-2=0.85 and was applied to all five LISS-III data sets. The results are maps of the chlorophyll-a content in 10 mu g/l classes for each for each of the five dates. For comparison of the accuracy of the different methods for water quality analysis, the trophic state index was calculated based on chlorophyll-a determination from laboratory, field spectra, and satellite data. Regarding the five lakes for which all the data were available, each method shows similiar results for the estimation of trophic state. (C) Elsevier Science Inc., 2000.