Predicting Lake Acidification and Regionalization of Predictions: Two Conference Papers

(1) Only in recent years with the fast development of modern computers has the analysis of ecological systems become a practical possibility. Systems analysis implies a description of the system and its processes, i.e. a development of a model representing the real system, the behavior of which closely resembles that of the real system. Two broad categories of models can be distinguished capable of predicting acidification: "Research models" and "Management models". In acidification research, models of the first type have proven useful in their hypothesis-generating role as well as in improving our understanding of the important mechanisms of acidification. Management models, in turn, can significantly assist in the evaluation of emission control strategies. There are two basic ways of using a prediction model in a decision-making context; scenario analysis and optimization analysis. The different ways to use prediction models are discussed. (2) The principal objective of the RAINS model, being developed at IIASA, is to assist in the evaluation of policies for controlling the acidification of Europe's environment. As part of this task, a dynamic model has been developed for describing the key processes assumed to be important in determining the long-term dynamics of surface water acidification. The input data available on a large regional scale are few. The model is regionalized by selecting input combinations from feasible ranges or frequency distributions. Monte Carlo techniques are used to determine those combinations of parameters that produce the observed present-day lake acidity distribution for each individual region, when the model is driven by a specified historical deposition. The ensembles obtained in this filtering procedure for each lake region are then used for the scenario analysis. The model runs, assuming different future energy-emission scenarios, suggest that all reductions in emissions are likely to reduce also the number of lakes being threatened by acidic deposition. In conclusion, the Monte Carlo method seems to provide a working tool for the application of catchment models on a large regional scale