Optimisation modelling of sustainable forest management at the regional level: an Australian example

Modelling environmental sustainability is a challenging task because of the spontaneous complex behaviour that emerges from the interaction of a large number of components of that environment. To reduce the complexity it is common to model such ecological systems at different scales with differieng degrees of detail. Incompatibilities between models at different planning levels may create implementation problems. This research shows that modern optimisation algorithms and fast computing can solve integrative forest resource models at a resolution suitable for operational planning for forest regions where normally at least two levels of planning would be applied. Two packages are evaluated, SPECTRUM, a linear programming (LP) package, and HABPLAN, which uses the Metropolis algorithm to converge on the optimum solution. Preliminary indications are that both packages could handle the problem size, but the HABPLAN package had some additional features which makes it attractive for solving multi-objective problems with spatial components as in modelling sustainable forest management.