Forest Planning under Climate Uncertainty

Abstract

Forest management planning aims at selecting the set of stands that should receive treatments such as harvest (harvest scheduling) in different time periods to provide economic, environmental, social, etc., services. However, since forest planning spans several decades, it is subject to many uncertainties including forest growth uncertainty due to climate change. Climate change is of paramount interest because it is arguably one of the most important challenges that contemporary forest managers need to address. It is reported to heavily affect forest growth.Failure to incorporate growth uncertainty in forest management planning can be detrimental to natural resources sustainability since forest managers might implement policies that are not optimal to different climate scenarios. In this dissertation, I argue that climate change uncertainty needs to be explicitly addressed while developing strategic harvest plans. I propose to model forest harvest scheduling in the face of forest growth uncertainty due to climate change as a stochastic optimization problem. This modeling framework allows to make robust decisions without knowing the actual forest growth and yield. However, the resulting optimization models belong to the family of mathematical models deemed hard to solve especially when a high number of scenarios (future growth possibilities) is considered. Moreover, one may argue that the added value of stochastic harvest scheduling models may not be warranted given how their development and implementation is more involved compared to a deterministic model which ignores uncertainties. This research proposes to study how the uncertainty in forest growth due to climate change can be incorporated in forest harvest scheduling. Specifically, in this research 1) I showed that overlooking climate change in harvest scheduling models affects the net present value and the conservation of natural resources; 2) I demonstrated that out of the four climate paths that may affect forest growth in the Pacific northwest, only two climate paths significantly affect the strategic forest harvest scheduling; and 3) I developed a heuristic that allows to handle large size stochastic optimization problems such as the ones in the case of harvest scheduling with climate uncertainty.