Assessing model uncertainty for environmental decision making: A case study of the coherence of independently derived environmental quality objectives for air and water

Many decisions in environmental management are based on model predictions t hat are plagued by uncertainty. It is sensible to consider this uncertainty because it can-influence the outcome of the decision-making process. This paper outlines a method to assess uncertainty in model predictions in situa tions in which validation research is difficult or impassible. It is illust rated in a case study of the coherence of independently derived environment al quality objectives (EQOs) for air and water involving the application of a multimedia fate model (SimpleBox). Distinction was made between operatio nal and fundamental uncertainty, which were assessed separately. Operationa l uncertainty was assessed by quantifying the uncertainties in the input pa rameters of the model and performing Monte Carlo simulations. Fundamental u ncertainty was assessed tentatively by means of a postal questionnaire amon g scientists and scientifically trained policy makers and representatives o f interest groups in the Netherlands. The results indicate that the perceiv ed fundamental uncertainty tends to exceed the calculated operational uncer tainty. These results emphasize the importance of considering fundamental u ncertainty in environmental decision making. Efforts to improve the quality of coherence testing of independently derived EQOs should aim at reducing the perceived fundamental uncertainty, for example; by defining the coheren ce criteria of EQOs in measurable attributes, thus facilitating validation research and enabling a more detailed assessment of the fundamental uncerta inty.