The value of using Bayesian Networks in Environmental Decision Support Systems to support natural resource management

Abstract

Environmental Decision Support Systems (EDSS) are one of a suite of tools that natural resource management (NRM) practitioners may use to support the development of, or help report on the success, of NRM programs. This paper outlines two DSS that are being developed for state government or NRM agencies in Australia to assist them to develop, implement and manage programs aimed at improving resource condition. IBIS is an EDSS that models ecological outcomes of environmental flows. VegBN models the effectiveness of NRM interventions on native vegetation quality on private land in northern Victoria. The underlying model base of each EDSS is comprised of Bayesian Network (BN) models linked with other BN or component models. Bayesian networks have proved to be a flexible and highly valuable approach to modelling such highly complex and uncertain environmental systems. They are an approach that can add rigour and transparency to decision-making processes and have in Australia gained considerable interest from researchers as well as government and other organisations involved in the management of natural resources.