JOINT USE OF MULTIPLE ENVIRONMENTAL ASSESSMENT MODELS BY A BAYESIAN META-MODEL - THE BALTIC SALMON CASE

An approach to construct a meta-modeling framework using Bayesian calc ulus is presented. It allows the inclusion of several - both causal an d empirical - models in a Bayesian belief network that is capable of r eal-time updating of uncertainties in the system. The approach can be used for diagnosis and forecasting purposes. It is illustrated with th e assessment problem of the threatened, wild salmon stocks in the Balt ic Sea. As usual, the institutions responsible for the management are prone to accept well-known standard assessment models. In this case, t he Virtual Population Analysis approach and a set of regression models have been used. A debate on possible superiority of one or more of th ese models over the others is met at each annual international negotia tion among stock assessment experts. The need has arisen to develop a methodology that would allow analytic fusion of these models to learn from the properties of the models and from the problem setting in gene ral, as well as to facilitate probabilistic predictions that would uti lize as much of the available information as possible. Moreover, the p robabilistic approach is very compatible with and supportive to the us e of the precautionary principle (risk-averse attitude). The belief ne twork approach has shown appropriateness for handling of this type of problem in environmental and resource management. (C) 1997 Elsevier Sc ience B.V.