Comparison of first-order analysis and fuzzy set approach for the evaluation of imprecision in a pesticide groundwater pollution screening model

The results of a screening model are always approximate, lying within an imprecision range. This paper focuses first on various types of imprecision in such modelling results, introduced either by subjective or state-of-the-art estimates of coefficients or through measurement error, and second on the different types of methods able to take into account imprecision in the input parameters in order to evaluate imprecision in the simulation results. Emphasis is placed on the evaluation of imprecision in potential groundwater contamination by pesticides using the Attenuation Factor (AF) index, the Retardation Factor (RF) index and two different methods of uncertainty analysis: the classical first-order uncertainty analysis and a method based on the fuzzy set approach as an illustration of the basic ideas. The results of this comparison show that the fuzzy set approach is more suitable for evaluating imprecision in screening models than the classical technique. Furthermore, it furnishes a mean value imprecision range and adds a degree of confidence to this range.