Remote sensing of chlorophyll in Lake Kinneret using highspectral-resolution radiometer and Landsat TM: spectral features of reflectance and algorithm development

High-resolution reflectance spectra in the range of 400–850nm were obtained from Lake Kinneret during a period when densepopulations of the dinoflagellate Peridinium gatunense dominatedthe phytoplankton. Chlorophyll (Chl) concentrations ranged from5.1 to 185 mg m^–3 and from 2.4 to 187.5 mg m^–3 inthe samples of two independent experiments. The most prominentfeatures of the reflectance spectra were: (i) a wide minimumfrom 400 to 500 nm; (ii) a maximum at 550–570 nm, whichdid not surpass 3% in samples with high Chl concentration (>20mgm^–3), indicating a strong absorption by pigments in thegreen range of the spectrum; (iii) a minimum at 676 nm; thiswas {small tilde}1% and was almost insensitive to variationin Chl concentration >10 mg m^–3; (iv) a maximum reflectanceshowed near 700 nm; its magnitude and position were highly dependenton chlorophyll concentration. High-spectral-resolution datawere used as a guideline for selection of the most suitablespectral bands for chlorophyll remote sensing. Models were devised,based on the calculation of the integrated area above the baselinefrom 670 to 850 nm and the reflectance maximal height withinthis range. Some algorithms already used m previous studieswere tested and showed a plausible degree of accuracy when appliedto the current data base. However, novel models devised in thisstudy improved substantially the accuracy of Chl estimationby remotely sensed data, by reducing the estimation error from>11 to 6.5 mg m^–3 Those models were validated by anindependent data set where Chl concentration ranged over twoorders of magnitude. The use of three relatively narrow spectralbands was sufficient for Chl mapping in Lake Kinneret. Therefore,a relatively simple sensor, measuring only a few bands willbe employed in future applications for Chl monitoring in inlandwaters. Radiometric data were also used to simulate radiancesin the channels of TM Landsat and to find the algorithm forChl assessment. The ratio of channel 4 to channel 3 was usedand enabled Chl estimation with an error of <15mg m^–3This algorithm was employed to map Chl in the entire area ofLake Kinneret with 10 gradations.