Uncertainty in Model Predictions: Does it Preclude Effective Decision Support?

The uncertainty in the predictions of models for the development of environmental systems is usually very large. In many cases the width of predictive probability distributions for variables of interest is significantly larger than the difference between the expected values of the results for different policy alternatives. This seems to lead to a serious problem for model-based decision support, as policy actions appear to have an “insignificant” effect on decision variables relative to the predictive uncertainty. However, in some cases it is evident that some of the alternatives at least lead to changes with the desired trend. A formal analysis of this argument is made based on the dependence of the probability distribution of the variables of interest for different policy alternatives. This analysis leads to the conclusion that the uncertainty in the difference of model results based on different policy alternatives may be significantly smaller than the uncertainty in the results themselves. The knowledge about the width of this distribution is useful to assess the effectiveness of the alternative especially in the presence of high uncertainty of the predictions. This general argument is illustrated for phosphorus reduction scenarios in a simple, didactical model of phosphorus loading to a lake.