Tropical forest decision support system: A prototype model
Abstract
Between 1980 and 1986, approximately 5% (7200 km$\sp2)$ of West Africa's rainforests were transformed into fallow territory and an additional 9% opened up by timber exploitation. The biological diversity and ecological equilibrium of the entire region is endangered and the timber industry has played a key role in the process. This study aims to strengthen forest sector institutions for effective implementation of sustainable resource management programs. A prototype decision support system, comprising of three application modules, is developed to assist with the rational choices for the development of long-term forest management policies and strategy. Optimization techniques are used to impose economic, biological and political constraints on the current status of the resource and subsequently simulate future impact of current inputs. It also simulates optimum resource yield in product terms given a predetermined sustainable development policy. In the forest module a technique, using hierarchical clustering and canonical Discriminant procedures, was developed to pool 112 species with similar growth increment characteristics into 7 groups suitable for the construction of timber growth and yield models. The technique minimized the variations within each group and hence the bias in the estimation of model. Compatible growth and yield models were subsequently developed for each species group by the solution of a system of 77 differential equations expressing the rate of change of ingrowth, mortality and survival growth components within a forest stand. The solution provided the means to project the status of the timber stand at any future time given well specified initial stand conditions, and also future yields in terms of product assortments which are necessary to link the forest and the industrial modules. The industrial module incorporated traditional supply-demand, least squares, autoregressive integrated moving average and simulation techniques to develop an econometric model of the forest product commodity market. The model was used to simulate future production and consumption under various political, economic and timber resource sustainability assumptions. The modules were linked together through wood supply and demand interactions. A planning horizon was defined to correspond to a felling cycle subdivided into 5 year periods. The states of the forestry and industry systems with regards to the wood inventories in the forest, production capacities, fixed assets and liabilities of the industrial units were simulated at the end of each period and an optimization model formulated to account for the selected objective function and associated constraints. The results were consistent with observable characteristics of growing timber resource and product commodities; and the simulations provided extremely useful information concerning the future of the timber resource and wood commodities market.
Degree
Ph.D.
Advisors
Moser, Purdue University.
Subject Area
Forestry|Computer science|Agricultural economics|Public administration
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.