采用林分平均高和密度估计人工林蓄积量  

Estimation of Stand Volume Based on Stand Average Height and Density of Plantation

作　　者：周梅[1] 李春干[2] 代华兵 

机构地区：[1]广西大学计算机与电子信息学院,南宁530004 [2]广西大学林学院,南宁530004 [3]广西林业勘测设计院,南宁530011

出　　处：《广西林业科学》2017年第3期319-324,共6页Guangxi Forestry Science

基　　金：广西林业勘测设计院基本科研业务费专项(GXLKYKJ201601)

摘　　要：采用2015年广西森林资源连续清查第9次复查中人工林样地调查数据,按树种组(杉木、松树、桉树)建立林分每公顷蓄积量与林分平均高、林分密度(郁闭度、每公顷林木株数)二元非线性模型(不变参数和可变参数),用确定系数(R^2)和平均预估误差(MPE)等6个指标对模型进行评价和检验。结果显示:全部12个模型的总相对误差(TRE)、平均系统误差(MSE)均小于15%,MPE均小于10%,表明采用林分平均高和密度估计林分单位面积蓄积量可取得较好的效果,但可变参数模型的参数的变动系数太大,不宜采用;3个树种组中,不论是不变参数模型还是可变参数模型,以平均高和每公顷林木株数构建的模型的R^2均大于由平均高和郁闭度构建的相应模型的R^2,而剩余标准差(SEE)、MPE则相反,说明每公顷林木株数对林分每公顷蓄积量变动的解释能力优于郁闭度。综合考虑6个统计指标和参数的稳定性,3个树种组的每公顷蓄积量的最优估计模型均为由每公顷林木株数、平均高构建的不变参数模型。Using plantation data from the ninth review of continue forest inventory(CFI) of Guangxi in 2015,two-way nonlinear model(invariant parameter and variable parameters) based on stand volume per hectare,and stand average height and stand density(canopy density and number of trees per hectare) according to three species groups include Chinese fir,pine(most of them were Pinus massoniana) and Eucalyptus was established. Six indices including coefficient of determination(R^2) and mean prediction error(MPE) and so on were used to evaluate. The results indicated that for all twelve models,the total relative error(TRE) and the mean systemic error(MSE) were less than 15% ,and the MPE was less than 10% . It was better to estimate stand volume of unit area with stand average height and stand density,but all variable parameter models were not recommended for variation coefficients,because their parameters were too big. In three species groups,the R^2 values of the models based on the average height and number of tree per hectare were greater than those of the models based on the average height and canopy density in either the invariant parameter model or the variable parameter model. In contrast,the residual standard deviation(SEE),MPE of former model were smaller than those of the latter model. It showed that the number of tree per hectare had higher efficient to explain the variety of the stand volume per hectare than that of canopy density. Considering six statistic indexes and stability of parameters,the optimal estimated models of stand volume per hectare in three tree groups were the invariant parameter models based on average height and number of tree per hectare.

关 键 词：蓄积量 平均高 林分密度 杉木 松树 桉树 

分 类 号：S771.8[农业科学—森林工程]