Management of the risk of wind damage in forestry: a graph-based Markov decision process approach

This study deals with the problem of including the risk of wind damage in long-term forestry management. A model based on Graph-Based Markov Decision Processes (GMDP) is suggested for development of silvicultural management policies. The model can both take stochastic wind events into account and be applied to forest estates containing a large number of stands. The model is demonstrated for a forest estate in southern Sweden. Treatment of the stands according to the management policy specified by the GMDP model increased the expected net present value (NPV) of the whole forest only slightly, less than 2%, under different wind-risk assumptions. Most of the stands were managed in the same manner as when the risk of wind damage was not considered. For the stands that were treated differently, however, the expected NPV increased by 3% to 8%.

[1]  Timo Pukkala,et al.  The use of heuristic optimization in risk management of wind damage in forest planning , 2007 .

[2]  B. Thorsen,et al.  Optimal spatial harvest planning under risk of windthrow , 2001 .

[3]  Alan T. Murray,et al.  HANDBOOK OF OPERATIONS RESEARCH IN NATURAL RESOURCES , 2007 .

[4]  Charles ReVelle,et al.  The grid packing problem : selecting a harvesting pattern in an area with forbidden regions , 1996 .

[5]  Régis Sabbadin,et al.  Approximate Linear-Programming Algorithms for Graph-Based Markov Decision Processes , 2006, ECAI.

[6]  Christian P. Robert,et al.  Monte Carlo Statistical Methods , 2005, Springer Texts in Statistics.

[7]  J. Borges,et al.  Combining random and systematic search heuristic procedures for solving spatially constrained forest management scheduling models , 2002 .

[8]  H. L. Scheurman,et al.  Techniques for Prescribing Optimal Timber Harvest and Investment Under Different Objectives—Discussion and Synthesis , 1977 .

[9]  B. Gardiner,et al.  Field and wind tunnel assessments of the implications of respacing and thinning for tree stability , 1997 .

[10]  H. Peltola,et al.  A mechanistic model for assessing the risk of wind and snow damage to single trees and stands of Scots pine, Norway spruce, and birch , 1999 .

[11]  Ola Sallnäs,et al.  WINDA—a system of models for assessing the probability of wind damage to forest stands within a landscape , 2004 .

[12]  Jonas Fridman,et al.  Modelling probability of snow and wind damage in Scots pine stands using tree characteristics , 1997 .

[13]  Miguel Constantino,et al.  A column generation approach for solving a non-temporal forest harvest model with spatial structure constraints , 2005, Eur. J. Oper. Res..

[14]  F. Barahona,et al.  A column generation algorithm for solving general forest planning problems with adjacency constraints , 1994 .

[15]  Richard Hartman,et al.  The Harvesting Decision When a Standing Forest Has Value , 1976 .

[16]  P. Wikström,et al.  A solution method for uneven-aged management applied to Norway spruce , 2000 .

[17]  Martin L. Puterman,et al.  Markov Decision Processes: Discrete Stochastic Dynamic Programming , 1994 .

[18]  K. Blennow,et al.  Decision support for identifying spruce forest stand edges with high probability of wind damage , 2005 .

[19]  Hans Daduna,et al.  Control of Spatially Structured Random Processes and Random Fields with Applications , 2006 .

[20]  C. Quine.  Forests and wind : management to minimise damage , 1995 .

[21]  Alan T. Murray,et al.  Analysing cliques for imposing adjacency restrictions in forest models , 1996 .

[22]  Nikos A. Vlassis,et al.  Collaborative Multiagent Reinforcement Learning by Payoff Propagation , 2006, J. Mach. Learn. Res..

[23]  J. Buongiorno,et al.  Economic Harvesting of Uneven-Aged Northern Hardwood Stands Under Risk: A Markovian Decision Model , 1987 .

[24]  Erika Olofsson,et al.  Supporting Management of the Risk of Wind Damage in South Swedish Forestry , 2006 .

[25]  D. Boychuk,et al.  A multistage stochastic programming model for sustainable forest-level timber supply under risk of fire , 1996 .

[26]  Gert-Jan Nabuurs,et al.  Natural disturbances in the European forests in the 19th and 20th centuries , 2003 .

[27]  E. Gunn.  Models for Strategic Forest Management , 2007 .

[28]  F. Helles,et al.  Windthrow probability as a function of stand characteristics and shelter , 1986 .

[29]  E. Valinger,et al.  Wind and snow damage in a thinning and fertilization experiment in Picea abies in southern Sweden , 1996 .

[30]  Nathalie Peyrard,et al.  Mean Field Approximation of the Policy Iteration Algorithm for Graph-Based Markov Decision Processes , 2006, ECAI.

[31]  Leif Martin Schroeder,et al.  Attacks of bark- and wood-boring Coleoptera on snow-broken conifers over a two-year period. , 1993 .

[32]  Michail G. Lagoudakis,et al.  Coordinated Reinforcement Learning , 2002, ICML.

[33]  Richard Hartman,et al.  THE HARVESTING DECISION WHENA STANDING FOREST HAS VALUEA , 1976 .

[34]  C. Revelle,et al.  Dynamic Selection of Harvests with Adjacency Restrictions: The SHARe Model , 1997 .

[35]  Richard S. Sutton,et al.  Planning by Incremental Dynamic Programming , 1991, ML.