Using belief networks in pollution abatement planning. Example from Morsa catrchment, South Eastern Norway

Abstract

A Bayesian belief network approach is used to conduct decision analysis under uncertainty of nutrient abatement measures in the Morsa catchment, South Eastern Norway, structuring available cost-effectiveness studies, eutrophication models and data in a DPSIR framework. The report demonstrates that Bayesian belief networks can be used to conduct cost-effectiveness and benefit-cost analyses under uncertainty, responding to the economic analysis. requirements of the EU Water Framework Directive (WFD). Forward and backward sensitivity analyses and information analysis are also demonstrated using the Hugin Expert software. Information analysis uncovers which parts of the DPSIR ‘chain’ contribute most to uncertainty in decision-making, and can be of used to optimise integrated environmental monitoring programmes in watershed management plans expected under the EU WFD