Bayesian networks for a multi-objective evaluation of River Basin Management Plans.
Holzkaemper, A., Surridge, B., Paetzold, A., Lerner, D.N., Maltby, L., Wainwright, J., Niranjan, M., Harris, B.In: Conference proceedings of the 6th European Conference on Ecological Modelling, Trieste, Italy, Nov 2007, p. 223.
The European Water Framework Directive (WFD) sets out an integrated perspective to water management in river catchments and river basin districts and is a key driver in the movement towards Integrated River Basin Management. Integrated river basin management must deliver objectives related to the WFD in the wider context of various other stakeholder interests, for example related to flooding, water resources, employment and cost. In managing such complex systems, a specific objective can be achieved through different management actions. Likewise, a specific management action can have implications for multiple objectives. Synergies or conflicts between specific objectives and between specific actions are likely to occur, and need careful consideration in order to increase the efficiency of planned management actions. However, such integrated decision making is a very difficult and highly complex task, which cannot easily be accomplished by either single or groups of planners. Integrated modelling tools to facilitate and enhance communication within a group of decision-makers and inform a more objective and evidence-based multi-criteria decision-making process are required. The scope for the development of such an integrated tool is being tested by the Catchment Science Centre (CSC) at The University of Sheffield. The CSC and the Environment Agency are jointly developing a tool termed the Macro-Ecological Model (MEM). The MEM is developed as a consistent framework for the integration of knowledge and information about environmental, social and economic processes and process-interactions that are affected by management actions and have impacts on multiple management objectives. The MEM enables knowledge from various different resources to be integrated, including empirical data, model results and even expert knowledge using a Bayesian Belief Network (BBN) approach. BBNs have the advantage of representing system understanding in an intuitive, graphical format. Furthermore, the approach provides the ability to explicitly account for uncertainties in model predictions. Therefore, the model framework provides a good tool for visualising system understanding and communicating uncertainties. Applied in a participatory process, it can support robust decision making in river basin management. The conceptual model framework is illustrated with examples from the prototyping study. The prototype model captures the process interactions affecting the management objectives "Ecological Status" (composed of both Biological Quality and Physico-chemical Quality) and "Flood Risk". It is planned to be later extended to incorporate further environmental, and also social and economic objectives.