ASSESSING FULL RESOLUTION MERIS PRODUCTS IN MONITORING CHLOROPHYLL_A, AND SUSPENDED MATTER IN THE WESTERN BASIN OF LAKE ERIE

In this work we investigate the efficiency of Medium Resolution Imaging Spectrometer (MERIS) full resolution (FR) data in monitoring optical properties in highly eutrophic and turbid waters, such as the Western Basin of Lake Erie (WBLE). Proper orthogonal decomposition of FR-MERIS water leaving radiance (R_rs) values indicate that attenuation coefficients (absorption and backscattering) in the WBLE are determined by color producing agents (CPAs), i.e. phytoplankton pigments , suspended matter (SM) and dissolved organic matter. The high spectral resolution (15 narrow bands), high spatial resolution (300m) and high temporal resolution (revisit time 2-3 days) of FR-MERIS has enabled the sensor to identify the CPAs and map the spatial and temporal dynamics of the biophysical properties of the optically complex WBLE. Concentrations of Chlorophyll_a and SM are derived based on MERIS Case-2 Processer (C2R) which uses a neural network approach that emulates inverse as well as forward models. Inherent optical properties (IOPs), such as absorption and backscattering coefficient are computed by inverse procedures from derived products. Atmospheric correction for Rayleigh and aerosol scattering was performed based on the principle of bright pixel inversion model. Derived absorption coefficient for phytoplankton at 443nm show a positive correlation with the Chlorophyll_a concentrations measured from samples collected in the field at 18 stations during the same period of MERIS data acquisition, R=0.50. Results from a Malvern2000 laser particle size analyzer show strong correlation, R=0.78, between the concentration of suspended sediments and derived backscattering coefficient. Inclusion of samples with sand sized particles > 10% reduced the correlation coefficient considerably, R=0.56. Comparison of FR data to ground-truthing data showed that, in the WBLE, the MERIS C2R underestimated and overestimated the concentrations of Chlorophyll_a and SM, respectively. Spatial and temporal variability of Chlorophyll_a and SM have been mapped over WBLE based on locally established empirical algorithms.