Inversion of oceanic chlorophyll concentrations by neural networks

Neural networks (NNs) for the inversion of chlorophyll concentrations from remote sensing reflectance measurements were designed and trained on a subset of the SeaBAM data set. The remaining SeaBAM data set was then applied to evaluating the performance of NNs and compared with those of the SeaBAM empirical algorithms. NNs achieved better inversion accuracy than the empirical algorithms in most of chlorophyll concentration range, especially in the intermediate and high chlorophyll regions and Case II waters. Systematic overestimation existed in the very low chlorophyll (<0.031 mg/m³) region, and little improvement was obtained by changing the size of the training data set.