| Is there more information in the color of the sea than just a simple measure of the chlorophyll pigment concentration? To answer this question, two basic tools are employed; statistical analysis and mathematical modeling. First, an extensive data set of in situ particulate absorption spectra is analyzed to assess the potential of using ocean color imagers to examine variability in the structure of the near-surface ocean planktonic ecosystem. The important result from this study is that ocean color will reflect only three statistically significant components: the total amount of particulate material, the relative amounts of chlorophyll-containing biomass and detrital materials. Thus, it is unlikely that robust global algorithms for determining particular phytoplankton groups can be developed from remotely sensed ocean color data. These results led the development of a nonlinear ocean color inversion model to maximize the information found in ocean color observations for the analysis of biogeochemical variability. The implementation of this model for the inversion of ocean color spectra results in estimates of concentrations of relevant dissolved and suspended materials found in the ocean including; (1) phytoplankton and phytoplankton pigments, (2) colored dissolved and detrital materials, and (3) inorganic suspended and particulate materials. Accurate estimates of these quantities can then be used in the analysis of biogeochemical variability in the worlds oceans. This inversion approach has been successfully applied to in situ ocean color measurements from the Sargasso Sea. Last, the IOP ocean color inversion model is applied to a global data set of in situ ocean color observations. The results indicate that model performance is strongly dependent upon the IOP shape functions assumed. The simple methods for parameterizing these spectral shapes result in a globally applicable inversion model that can be applied to satellite imagery. |
PDF Full Text Request