| The stand growth and yield model in Heilongjiang province, biomass and individual-tree Model in Jilin province and Daxing'anling region for the main tree species (including Korean Pine, Picea asperata Mast., Abies, Larix gmelinii Rupr.,Pinus sylvestris var mongolicaLitv.,) of the Northeast Forest were established in this research using Five continuous forest inventory data in Heilongjiang province during the year of1986,1990,1995,2000and2005, as well as516main tree species regional biomass data and the continuous forestry resource inventory data from Jilin and Daxing'anling. These models for users'applications were implemented as computer procedure through the software of Visual C++. Afterwards, the stand dynamic growth model of northeast forestry was made. The research including:1. The forest survey data in five periods in Heilongjiang Province were divided into six regions, taking into account of different geographic location. Specific model was developed in each area that would lead to the incompatible between the regional models. To solve this problem, this study introduced a variable containing regional information called dummy variables in the process of model fitting. Then the compatible basal area growth model, stand volume estimation model, the dynamic stand density estimation model, gloomy degree model and the died ration of volume are independently established in the6region and the whole province, taking the independent samples as test data for assessing the accuracy of these models. The result shows that the fitted models have high accuracy and can be widely usedit is preferable to using dummy variable to solve the compatible problem. And the fitted models are very useful for some regions to be renewed and made dynamic estimation as well as provide the technology support and dynamic variation information for the forest sustainable management in Heilongjiang province.2.516analytic tree data and the biomass of every tree species from the5regions in Heilongjiang province were summarized. During the process of establishing biomass equation, the biomass models in four components, including roots, stem, branch and leaf are modeled respectively to get a higher accuracy. To sum up, the equations have nice fitting results. When using these estimated results to make a comparison to the total biomass, a significant gap can be found. And then the compatibility between the components and total is put forward. Within the further study, the method of nonlinear measurement error simultaneous equation variable model is adopted. Four fitting equations for responding biomass were built, coupled with the equation of the total biomass for limitation. The method for solving this equation is not simple least square,but two-stage or three-stage least square method. ForStat2.0is the tool we chose in the whole procedure. The fitting results and the accuracy of this method turn out well. 3. Current individual tree model was tested and studied based on Literature reference and the achievements from my supervisor. Using the third forestry resource inventory data in Daxing'anling and the second inventory data in Jilin province, stepwise regression was made to model individual tree biomass. The variables were introduced gradually to the equation until the fittings reaches the requirement. Because the factor of affecting growth is either large or small and the variables of various tree species are always the same, so the equation formation is not the same, However, most of the fitting results have high accuracies. Lacking enough data of some tree species, the equation is not significantly accurate.4. The fitting models are effective and can be widely used after getting positive results from assessing the accuracies of those mentioned models above. However, a Mathematical formula can not explain the growth dynamic variation directly. Considering this problem, the dynamic stand growth simulation model is constructed in this work integrating the whole programming models.
|
PDF Full Text Request