Journal
FOREST POLICY AND ECONOMICS
Volume 83, Issue -, Pages 107-120Publisher
ELSEVIER
DOI: 10.1016/j.forpol.2017.07.006
Keywords
Wildland fire; Spatial; Ecological disturbance; Risk; Approximate dynamic programming; Reinforcement learning; Forestry
Categories
Funding
- National Science Foundation under CyberSEES Computing and Visualizing Optimal Policies for Ecosystem Management [1331932]
- Direct For Computer & Info Scie & Enginr
- Division of Computing and Communication Foundations [1331932, 1521687] Funding Source: National Science Foundation
Ask authors/readers for more resources
Forest management in the face of fire risk is a challenging problem because fire spreads across a landscape and because its occurrence is unpredictable. Accounting for the existence of stochastic events that generate spatial interactions in the context of a dynamic decision process is crucial for determining optimal management. This paper demonstrates a method for incorporating spatial information and interactions into management decisions made over time. A machine learning technique called approximate dynamic programming is applied to determine the optimal timing and location of fuel treatments and timber harvests for a fire-threatened landscape. Larger net present values can be achieved using policies that explicitly consider evolving spatial interactions created by fire spread, compared to policies that ignore the spatial dimension of the inter-temporal optimization problem.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available