An operational, spatially constrained harvest scheduling model

Abstract

A Monte Carlo integer programming algorithm was developed to generate short-term (25-year), spatially feasible timber harvest plans for a New Brunswick Crown license. Solutions for the short-term plan are considered feasible if they meet spatial and temporal harvest-flow and adjacency constraints. The solution search procedure integrates a randomly generated harvesting sequence and checks of harvest-flow and adjacency constraints. The model was used to determine the annual allowable cut under three constraint formulations. The three formulations represented increasing levels of adjacency constraints, from no constraints to levels similar to current provincial requirements. The annual allowable cut under the most strict constraint formulation was reduced by 9% from the unconstrained formulation, for a given mapping strategy of a long-term harvest schedule. These applications of the model indicate that it is suitable for spatially constrained harvest scheduling on Crown licenses in New Brunswick.