Study On Biomass Of Quercus Variabilis Planting In The Beijing Xishan Forest Farm

Forest productivity and biomass provide important information for Sustainable Management of Commercial Plantation and forest ecosystem assessment. The main content of this thesis is as follows:selection of fitting models, evaluation models and checking models. Based on survey data which is obtained from 50 sample trees in 7 sample plots of Quercus variabilis planted in the Beijing Xishan National Forest Park, this paper compares biomass models in way of linear models, allometric models and polynomial models with independent variables like D2H and D, and establishes 70 above-ground biomass models, stem biomass models, crow biomass models, branch biomass models and leave biomass models. Models produced in the thesis are evaluated by seven indicators which are the sum of squared residuals named Q, residual deviation named S, coefficient of determination named R2, relative error named Rs, mean relative error named E1, the mean relative absolute error named E2 and Coefficient of variation parameter named Ci%. The conclusion is that the Best Models of biomass about aboveground, stem, crow, leave, branch are respectively: W=0.057(D2H)0.96, W=0.032(D2H)0.985; W=0.079D2.122, W=0.004(D2H)0.894, W=0.068D2.123.How to obtain the compatibility is a problem for biomass models. The method of adjustment in proportion is proposed in the thesis. By the research on the biomass models, this article builds the tree biomass tables of Quercus variabilis planted in the Beijing Xishan National Forest Park and makes an analysis of these five models with Chi-square(x2) test. The results show that the biomass models are effective and credible.