The Forest Management Optimization Model Based On The Variation Of Trees’ Age Structure

Abstract

In order to maximize the comprehensive value of forests for human-beings, our team developed a method to acquire an optimized forest management plan, which not only consider the carbon sequestration of the forest, but also balances various needs of society. In this paper, we first consider the influence of our management plan on forest carbon sequestration. We divide the carbon sequestration into two parts: the trees in the forests and the existing wood product made from the forest. To predict the carbon sequestration in trees over time, we need the current carbon sequestration of forest, the age structure and scale of the forest over time and the carbon sequestration in trees as it grows. First, we use Lasso Regression to determine the significant forest characteristics, then we use Multiple Linear Regression Model to analyze how forests' characteristics affect current carbon sequestration. Then, we use improved Leslie Model to predict the variation of age structure and scale of the forest with the application of management plan. Then we use the improved exponential model to describe the variation of carbon sequestration as a tree grows. To predict the carbon sequestration in wood product over time, we calculate the amount of wood each time we harvest and establish a model describing how wood products are made and decay. Then we set the mean value of the total carbon sequestration over time as the optimization objective, the parameters in forest management plans as decision variables to establish a single-objective optimization problem. We give reasonable intervals to the variables and determine the optimized forest management plan through Grid Search Method and Monte Carlo Method. Second, we consider the influence of forest management plan on other aspects of the society. By using AHP, we set the weight of forests’ value in various aspects and by using TOPSIS Model, we established a model to evaluate the comprehensive value of forests. Then we set forest’s mean scores from TOPSIS over time as the optimization objective, the parameters in forest management plans as decision variables to establish an optimization problem. Computer simulation can be used to get an optimized management plan. Furthermore, we discussed the traits of the forests in which managements are suitable, and the transition strategy in management plans. Also, by System Hierarchy Clustering we classified all the management plans into 3 types which apply to different situations.