Chlorophyll concentration estimation model based on leaf spectra of salt marsh vegetation

The nutrient status of Spartina Anglica should be monitored dynamically in order to manage scientifically it, enhance conserve level, and improve its productivity. The author chose leaf spectra of Spartina Anglica located in the same sample plot and tested leaf reflectance and extracted chlorophyll a, b, and chlorophyll a and b concentration for leaf, stem, and mixture of leaf and stem. Seven single variable characteristics parameters which were red edge position (REP), red well position (RWP), red edge area (REA), yellow edge (YE), blue edge (BE), Leaf Chlorophyll Index (LCI) and Water Index (WI) respectively were calculated based on leaf reflectance. Correlation analysis (Table I) between these seven parameters and corresponding chlorophyll concentration was conducted. Based on correlation analysis results, we made chlorophyll concentration estimation model by applying suitable curve simulation methods. Above research results showed that high negative correlation between chlorophyll concentration and red well position, red edge area were found. The largest correlation coefficients were -0.614 and -0.682 respectively. While positive correlation between chlorophyll concentration and LCI and WI were grasped. The largest correlation coefficients were 0.514 and 0.778 respectively. Chlorophyll concentration estimation models (formula 3, 5, 7, and 9) based on leaf level reflectance by applying red well position, red edge area, LCI and WI were established. The accuracy of models was very high and R² were 0.478, 0.563, 0.413, and 0.764 respectively. The researches provide important reference data with further discuss hyper-spectral map of salt marsh vegetation. Thus, it has important theoretical value.