Modeling of apparent hydro-optical properties and retrievals of water quality in the Great Lakes for SeaWiFS: A comparison with in situ measurements

Assessment of water quality parameters from SeaWiFS data proved to be nowadays operational for case I waters, i.e., oceanic and marine open waters. Relatively simple algorithms based on ratios of water-leaving radiance at two or more SeaWiFS wavelengths are successfully used for retrieval of algae pigments in such waters. However, water quality retrieval presents considerable difficulties when processing satellite (e.g., SeaWiFS) data pertinent to case II waters which are generally inland and marine coastal waters. Apart from the atmospheric correction problem for such waters (this issue is discussed in the authors' companion paper), case II water optical complexity necessitates application of sophisticated water quality retrieval algorithms which require a priori knowledge of a set of optical characteristics (i.e., usually referred to as a hydro-optical model) appropriate for major optically active constituents indigenous to a targeted water body/water mass. Since hydro-optical models are being only available for a limited number of case II water bodies, it appears important to investigate translatability of those hydro-optical models developed for specific water bodies to a larger variety of case II waters. Once the extend of this translatability is assessed, the solution to the inverse problem (i.e., retrieval of water constituents abundance) could be attained via application of a suitable bio-optical algorithm/retrieval procedure.