Multiparametric classification of trophic conditions. The OECD methodology extended: combined probabilities and uncertainties — application to the North Adriatic Sea

A new approach to trophic classification of water bodies, based on combining the exact probabilities from the OECD (Organization for Economic Cooperation and Development) lognormal frequency distributions of chlorophyll, nitrogen, and phosphorus concentrations and Secchi depths, is presented, using a sample data set from monitoring activities carried out in the North Adriatic Sea in the summer of 1992. This probabilistic approach is shown to be effective in solving uncertain or even contradictory classifications which may often arise from membership probabilities, based on single sample variables, as in the present OECD system. Such an extension to multiparametric sample space is open to possible modifications of trophic variables, and allows comparisons among samples with inhomogeneous or even unmatching types of measurements. Classification uncertainties, caused by the uncoupling of trophic variables, are shown to be related to environmental disturbances given by discharges, inflows and turbulent diffusion. Their role in trophic classification procedures, and environmental importance and meaning is stressed in relation to the applicability of the extended OECD methodology to non-steady state systems.