A scenario approach to stochastic anticipatory optimization in stand management

A flexible model for stochastic optimization is developed that can be used with forest stand simulation models. Stochasticity is represented by a large set of scenarios, each of which is an outcome of stochastic processes. A stochastic environment is described by random yearly growth rate levels and catastrophes, such as wild fire or windthrow. The optimization model is defined in control variable space and includes the timing, intensity, and type of thinning, and rotation length for an even-aged stand. Single-tree growth and mortality models are used. Numerical results in a risk-neutral case show that the optimum rotation is shortened with an increasing probability of a catastrophe. Further, an increasing growth rate variation has mixed and weak effects that depend, in particular, on the tree mortality model. If a stand cannot be thinned, increasing risk-taking shortens the optimum rotation, given the model set used.