Simulation of forest management strategies under uncertainty: Application of dynamic programming method

Extended Abstract
 
1-The issue of forest depletion in the north of Iran has always been a subject of debate, especially in recent years, and the need for sound planning and proper management has been increasingly emphasized. In order to determine the optimal forest management strategy in northern Iran under uncertainty, this study has attempted to estimate forest growth model, ARIMA model for predicting wood price, Harvesting Cost Model and a Dynamic Planning Model, Management Strategies in Three Scenarios, Maximizing Net Present Value, Maximizing timber harvesting and maximizing growing stock. It is worth noting that the present study aims at selecting different management strategies in order to determine optimal harvesting strategy and evaluation of sustainability of simulated strategies is a new step in choosing appropriate management policy in northern forests.
2-Materials and Methode
In order to simulate forest management strategies, firstly it needs to estimate the price equation, the growth equation, the cost equation, and the net present value equation. ARMA models are a combination of Autoregressive (AR) and moving average (MA) variables. This model was applied to predict timber price and estimate price equation. The quadratic cost equation with intercept is assumed to be the function of the harvest cost and the squared of the harvest variable. In order to estimate the forest increment equation, we used a logistic growth model. This model, despite the low observations, shows well the relationship between growing stock and growth rate, and theoretically and mathematically describes the model well. Given the equation of price, cost equation, growth equation and net present value equation, we evaluated the defined strategies in the dynamic programming model using GAMS software. After simulating strategies to maximize timber harvest, maximize net present value, and maximize forest timber inventory, a hierarchical analysis model was used to evaluate the sustainability of forest resources in each strategy.
3-In the strategy of maximizing harvest, there was an almost decreasing trend for the growth variable, with a slight decrease, and some variations were observed for the harvest variable. Changes in harvest rate per hectare and growth rate per hectare follow each other in this strategy, which means that the harvest rate should be determined by the rate of growth per hectare. In the strategy of maximizing the net present value, the growth rate is at an acceptable level and therefore the forest is not damaged. By applying this strategy, the harvest rate will increase in the early years and then continue to be somewhat steady until the end of the period. This means that in this case, the situation of the stakeholders will improve economically, which is one of the goals of sustainability. In the second strategy Harvest rate follows the growth rate in this strategy, and there is a correlation between harvest rate per hectare and stock availability per hectare. In the third strategy of maximizing forest stock, the amount of growth was acceptable. Harvesting variable during the period of study has an increasing trend. The stock per hectare increased at the beginning of the study period and has been somewhat steady since the first few years. It is remarkable that in this strategy, which aimed to maximize the amount of growing stock per hectare, the growth rate was almost acceptable and the harvest rate per hectare followed the pattern of stock availability per hectare.
4-Suggestion
In general, it is suggested that dynamic planning methods be used in forest management, especially those of commercial and economic importance in the northern forests, to achieve appropriate management strategies. As the results showed, economic and environmental objectives will be achieved simultaneously with dynamic planning and proper management of forest resources.
JEL Classification: C61, C63, Q23