Modelling of Water Pollution in the Thermaikos Gulf with Fuzzy Parameters

Scientists used fuzzy set theory to help characterize the imprecise parameters used in a water pollution model. Techniques used to express the physico-chemical coefficients and the loads of pollution sources as triangular fuzzy numbers are discussed. The advection-dispersion equations for ten different water quality variables were resolved using a 2-D finite element algorithm combined with fuzzy logic analysis. Temporal discretization was achieved using the characteristic Galerkin technique. Interval operations were enacted for the solution of the algebraic system of equations with fuzzy coefficients. Findings are presented from a sample application of the model to the assessment of pollution in the Gulf of Thermaikos, Greece.