| Forest ecosystem is the most important part in terrestrial ecosystem. It also plays a decisive role on global change research.Forest biomass is the most essential quantitative properties in forest ecosystems, it shows the level of forest management and the value of exploitation, but also reflects the complex relationships between forest and environment in the material circulation and energy flow.In this paper, the nonlinear model for forest biomass has been presented, due to the mass of factors which influence the forest biomass and the certain correlations among factors, how to confirm the rational factors and build more comprehensive and rational model, become the key on the study of forest biomass modeling. In this paper, the dataset comes from Jin Gou Ling forest of Wang Qing county, Jilin Province. In factors selection, we mainly use the stepwise regression method to determine which factors impact relatively on biomass, reject the weak factors, and ultimately select the tree age, tree height and tree diameter as factors of modeling. The accuracy of model has been improved.We compare the robust estimation and least-squares estimation, mainly on the resistance to the gross error ability. Finally, considering gross errors, robust estimation performs obviously better than least-squares estimation on accuracy, and using robust estimation can also get ideal regression model. While, it is a reasonable and feasible method to appropriately introduce robust estimation in least-squares regression.In forest management, it is also very important to effectively control deforestation rate. We begin with deforestation rate, using optimal control theory to analyze equation of forest harvesting intensity. Finally we get the equation. Different areas and different species can be substituted into the data calculation, thus get the value of optimal harvesting intensity. Present a computable method of optimal harvesting intensity for the forest sustainable management and utilizing.
|
PDF Full Text Request