Forest harvest scheduling is subject to a variety of risk and uncertainty in natural, financial, and social environments. Consequently, the embodied complexity and high dimension render its optimization extremely difficult, often unattainable. Markov Decision Process (MDP) models are capable of handling multitudes of risk and uncertainty and provide straightforward decision tables linking harvests to forest conditions. New extensions of MDPs are being developed that deal with risk attitude of landowners, nonstationary processes induced by climate change, and spatial management of neighboring stands affected by diffusing disturbances.
Zhou, M*. and J. Buongiorno, J. 2019. Optimal forest management under financial risk aversion with discounted Markov decision process models. Canadian Journal of Forest Research.
Buongiorno, J.*, Zhou, M.,Johnston, C. 2017. “Risk Aversion and Risk Seeking in Multi-criteria Forest Management: A MDP approach”. Canadian Journal of Forest Research 47: 800 – 807.
Buongiorno, J. and Zhou, M*. 2017. “Multi-criteria Forest Decision Making under Risk with Goal-programming Markov Decision Process models”. Forest Science. DOI: https://doi.org/10.5849/FS-2016-078R2
Zhou, M*.2015.Adapting Forest Management to Climate Policy Uncertainty: a Conceptual Framework”. Forest Policy and Economics 59, 66-74.