A multicriteria stochastic optimization framework for sustainable forest decision making under uncertainty

Abstract

A core process in forestry planning corresponds to the design of optimal harvesting policies along with road network layouts. In the most common setting, decision makers seek for solutions that maximize the profit of the forest while respecting operative and market constraints. Due to the long-term nature of the industry, the inherent uncertainty in both forest growth and market conditions should be taken into account. Nowadays, forest planning must target towards a sustainable management; the maximization of carbon sequestration and the minimization of land erosion are two common environmental goals.The planning challenge addressed in this paper integrates uncertainty of future forest growth and timber prices with the need for considering three criteria; net-present value, carbon sequestration, and land erosion caused by the road construction within the forest. By using mathematical programming tools and stochastic optimization techniques, we develop a stochastic multicriteria model that enables decision makers to have not only one, but a pool of long term planning policies.Moreover, a risk-averse variant of the framework is also considered. To the best of our knowledge, this is the first time that this type of forestry planning setting, which responds to the new challenges of the industry, is addressed.The proposed tool is used on an eucalyptus forest located in Portugal; the obtained results show the benefit of the proposed framework for producing a pool of sustainable forest plans with efficient trade-offs among the three considered criteria.