ANALYSIS OF UNCERTAINTY OF FUTURE TIMBER YIELDS IN FOREST MANAGEMENT

We consider the effect that the uncertainty of future yields has on fo rest management. One way of modeling this problem is through chance co nstrained linear programming, where constraints are represented as pro babilistic statements. Under normality conditions, an equivalent deter ministic nonlinear program can be solved in an efficient way using a c utting plane algorithm, which takes advantage of the characteristics o f the problem. To deal with uncertainty, we propose a forest planning approach, based on chance constrained linear programming. We also anal yze the importance of including the consideration of uncertainty in th e planning process. For this purpose, we simulate two scenarios, a det erministic one and another where uncertainty is included in the models . Results of a test case show that not considering uncertainty in the models when production demand constraints have small slack can lead to management situations with infeasible solutions.