Optical characterisation and reflectance modelling in Case II waters : quantitative tools for investigations of coastal environments

A methodology is proposed here to satisfy a community requirement for a regional solution to developing algorithms in coastal waters. A three component model, which allows maximum variability of the non-chlorophyllous components, was developed. This model was then systematically applied to a range of input values of the forward model defined using lognormal distributions, yielding a dataset of reflectances for the studied area. A principal component analysis was then performed on the simulated dataset. A multiple regression was carried out on the resulting eigenvectors and the input values of the optically active components to define the algorithms for the chosen region for the retrieval of chlorophyll, total suspended matter and yellow substance.A case study was made applying the method to the South Baltic and a validation dataset from the same region showed the capacity of this method to retrieve the concentrations of the individual optically active components to within the accuracy defined by NASA for Case I water chlorophyll retrieval. The sensitivity of the method, to inelastic processes and bottom reflectance, was considered. The study was made through the influence on the absolute signal as well as the influence on the shape of the characteristic vectors that are retrieved from the simulated reflectance in an attempt to assess the impact on the algorithm development procedure. In both cases the influence is shown to be minimum for wavelengths below 650nm.An alternative methodology is also presented (Direct Inversion Algorithm) which is based on the inversion of absolute reflectances for the retrieval of total absorption. This method was validated for an in-situ dataset and shown to retrieve chlorophyll "a" to within 30% accuracy. The applications of such a technique were considered in the context of primary production in coastal waters and chemotaxonomy.