Automatic extraction of fuzzy rules from MERIS data to identify sea water optically active constituent concentration

Determining the concentrations of dissolved organic matter and suspended non-chlorophyllous particles in the sea water is basic to the sludy of the impact of anthropic activity in coastal areas. As these concentrations aflect the spectral distribution of the solar light back-scattered by the water body. their estimation can be computed by using a set of measures of average subsuface rejlectances over spectral channels centered around prefied wavelength of a MEdium Resolution Imaging Spectrometer (MERIS) on board a satellite. In this paper, the relation between the concentrations of interest and the average subsuface rejlectances is modeled by a set of firtzy rules extracted automatically fiom MERIS data through a two-step procedure. First, a compact initial rule base is generated by projecting onto the input variables the clusters produced by a Iuny clustering algorithm. ?%en, a genetic algorithm is applied to optimize the rules. Appropriate constraints maintain the semantic properties of the initial model during the genetic evolution Results of the application of t h e w model are shown and discussed