Keywords
cognitive map, measures assessment, uncertainty, nitrogen diffuse pollution, agriculture
Start Date
1-7-2010 12:00 AM
Abstract
Expert elicitation processes are increasingly applied in environmental management, to deal with complex phenomena characterised by lack or scarcity of data. Information obtained from carefully defined elicitation protocols can overcome uncertainty limits and support decision-making processes. The work illustrates the potentials of an innovative methodological framework, built upon the integrated use of specific tools, such as cognitive maps and Bayesian networks (BNs). The methodology was applied to the Venice Lagoon Watershed (VLW) case study, to assess agri-environmental measures, adopted to reduce nitrogen diffuse pollution of water. The framework consisted of two main phases, both built upon processes eliciting judgments from experts. The first phase aimed at gaining a clear understanding of the problem, categorising the main factors of the system into a shared cognitive map, built during an interactive group elicitation exercise. The second phase focused on the construction of a BN, representing the probabilities of relationships between the variables of the system and assessing the effects of different measures on the environment. The Conditional Probability Tables (CPTs) underlying each node of the BN was populated with probabilities elicited through specific questionnaires. To minimize the elements of bias in the estimation of the experts, both processes of engagement of the experts were formalised, choosing the most indicated tools and procedures, into carefully designed protocols. The results of the research highlighted the potential of the methodology to incorporate the uncertainty in multidisciplinary and participatory contexts, for ex-ante and ex-post policy assessments. This methodology allows the integration of the available knowledge and of the uncertainty that characterise it, giving the possibility of knowledge updating whenever new information become available.
The Use of Experts’ Judgements to Assess Agri- Environmental Policies in the Venice Lagoon Watershed: a Bayesian Network Approach
Expert elicitation processes are increasingly applied in environmental management, to deal with complex phenomena characterised by lack or scarcity of data. Information obtained from carefully defined elicitation protocols can overcome uncertainty limits and support decision-making processes. The work illustrates the potentials of an innovative methodological framework, built upon the integrated use of specific tools, such as cognitive maps and Bayesian networks (BNs). The methodology was applied to the Venice Lagoon Watershed (VLW) case study, to assess agri-environmental measures, adopted to reduce nitrogen diffuse pollution of water. The framework consisted of two main phases, both built upon processes eliciting judgments from experts. The first phase aimed at gaining a clear understanding of the problem, categorising the main factors of the system into a shared cognitive map, built during an interactive group elicitation exercise. The second phase focused on the construction of a BN, representing the probabilities of relationships between the variables of the system and assessing the effects of different measures on the environment. The Conditional Probability Tables (CPTs) underlying each node of the BN was populated with probabilities elicited through specific questionnaires. To minimize the elements of bias in the estimation of the experts, both processes of engagement of the experts were formalised, choosing the most indicated tools and procedures, into carefully designed protocols. The results of the research highlighted the potential of the methodology to incorporate the uncertainty in multidisciplinary and participatory contexts, for ex-ante and ex-post policy assessments. This methodology allows the integration of the available knowledge and of the uncertainty that characterise it, giving the possibility of knowledge updating whenever new information become available.
