A Methodology for Uncertainty Analysis in Integrated Assessment Models

This project focuses on uncertainties in Integrated Assessment (IA) models with an emphasis on transboundary air pollution. IA models can be considered tools for the evaluation of strategies for environmental pollution control. Ideally, such models cover the whole range of the problem from emissions of pollutants to their environmental effects. IA models typically answer questions of the "what if…" type. By defining different scenarios for human activities, the models explore a variety of possible future developments, thus illustrating possible consequences of alternative strategies. Some IA models also include options for optimisation. There are different ways to deal with uncertainties in IA models. In general, a distinction can be made between deterministic and stochastic modeling approaches. Stochastic models explicitly account for uncertainties in (some) model components, usually in terms of probability frequencies, while deterministic models do not. The uncertainties in deterministic models may be assessed in different ways. Uncertainty assessments may focus on the uncertainties in the model structure and/or in parameter values. These uncertainties can be assessed by, for instance, identification of potential biases, analysis of the model structure, sensitivity analyses or Monte Carlo Simulation techniques. RAINS, the Regional Air Pollution Information and Simulation model for Europe and Asia, is one of the deterministic IA models for acidification which has been widely used for policy analysis by the European Commission and the UN Economic Commission for Europe. The overall purpose of this project is to develop a methodology for the analysis of uncertainties in Integrated Assessment models for transboundary air pollution problems that are used by policy makers. This methodology includes an analysis of model uncertainties by combining both quantitative and qualitative methods, and taking into account both the expert judgement of the model builders and the possible differences in perceptions by the model users. The associated research objectives are: · To provide an overview and evaluation of different tools and techniques available for qualitative and quantitative analysis of uncertainties in IA models · To develop a methodology for the analysis of structural and parameter-related uncertainties in IA models for transboundary air pollution problems, that takes into account both the expert judgement of the model builders and the possible differences in perceptions by the model users · To apply this methodology to the acidification model RAINS as an illustrative case study. The innovative aspects of this study are associated with the integrative approach to assess both structural and parameter uncertainties in IA models.