Application of ant colony optimization for the risk management of wind damage in forest planning

Ant colony optimization (ACO) is still quite a new technique and seldom used in the field of forest planning compared to other heuristics such as simulated annealing and genetic algorithms. This work was aimed at evaluating the suitability of AGO for optimizing the clear-cut patterns of a forest landscape when aiming at simultaneously minimizing the risk of wind damage and maintaining sustainable and even flow of periodical harvests. For this purpose, the AGO was first revised and the algorithm was coded using the Visual Basic Application of the ArcGIS software. Thereafter, the performance of the modified AGO was demonstrated in a forest located in central Finland using a 30-year planning period. Its performance was compared to simulated annealing and a genetic algorithm. The revised AGO performed logically since the objective function value was improving and the algorithm was converging during the optimization process. The solutions maintained a quite even periodical harvesting timber while minimizing the risk of wind damage. Implementing the solution would result in smooth landscape in terms of stand height after the 30-year planning period. The algorithm is quite sensitive to the parameters controlling pheromone updating and schedule selecting. It is comparable in solution quality to simulated annealing and genetic algorithms.