巨尾桉人工林林分二元形高模型的研建  

Research and Construction of Binary High Model of Eucalyptus grandis × E. urophylla Plantation

作　　者：高德祥[1] 韦文长 GAO Dexiang;WEI Wenchang(Yunnan Forestry Survey and Planning Institute,Kunming 650051,Yunnan,China;Forestry and Grassland Bureau of Shuangjiang County,Shuangjiang 677399,Yunnan,China)

机构地区：[1]云南省林业调查规划院,云南昆明650051 [2]双江县林业和草原局,云南双江677399

出　　处：《桉树科技》2022年第1期24-29,共6页Eucalypt Science & Technology

基　　金：云南省临沧市科技创新人才培养项目(202004AC100001-B14)。

摘　　要：为研究巨尾桉人工林林分蓄积量与林分平均胸径和平均树高的相关关系,建立巨尾桉林分二元形高模型,为巨尾桉人工林伐区设计和森林资源资产评估提供科学实用的林分形高表,以云南省临沧市双江县巨尾桉伐区小班的标准地每木检尺及造材数据作为研究对象,分92个建模样本和31个检验样本。以标准地胸径实测值计算林分平均胸径和每公顷断面积,以双江县巨尾桉胸径-树高数学模型导算单株木树高值并计算林分平均树高,以双江县巨尾桉二元材积模型计算单株木材积并计算每公顷蓄积,进而计算林分形高值。应用92个标准地林分平均胸径、平均树高和形高值作为建模样本,选择有代表性的8个二元形高数学模型作为备选模型,使用SPSS21.0软件非线性回归分析方法对各模型进行拟合评价,以拟合优度(R^(2))、均方差(RMSE)、平均绝对偏差(MAD)、总相对误差(RS)和预估精度(P)作为评价指标,8个备选模型的评价结果表明:建模样本拟合效果最好的模型为FH=sH^(b)D^(c)和FH=sH^(b)/D^(c),经综合比较,选定模型FH=sH^(b)/D^(c)作为双江县巨尾桉人工林林分二元形高模型,代入参数值后的模型表达式为FH=4.555080H^(-0.479807)/D^(-0.749612)。使用31个检验样本进行适用性检验的结果表明:模型总相对误差(RS)为-0.6%,在±5%的误差范围内;F检验的F统计量为2.12,小于F_(0.05)(1,29)的查表值4.18,通过F检验。各项评价指标及适用性检验结果表明,本模型适合作为双江县巨尾桉人工林林分二元形高模型,根据本模型导算的林分二元形高表可以在森林蓄积量调查生产实践中应用。This study was undertaken to examine correlations between Eucalyptus forest stock and forest average breast height diameter and average tree height and establish a Eucalyptus forest binary high model for the design of forest and forest resource asset evaluation,in Binjiang county,Lincang city,Yunnan province.To do this,sample trees were divided into two sets comprising 92 and 31 trees.The average breast height diameter and basal area per hectare of forest stands were calculated based on the measured values of standard breast height diameter.The height value of single trees and the average tree height of forest stands was calculated by the mathematical model based on Eucalyptus breast height diameter-tree height,and the wood accumulation per hectare was calculated by the binary timber volume model of E.grandis×E.urophylla in Shuangjiang County,and the fractal height value was also calculated.The average breast height diameter,average tree height and shape height value of the 92 standard sample trees were used as building books,Eight representative binary high mathematical models were selected as alternative models.The fitted models were evaluated using the non-linear regression analysis method of SPSS 21.0 software,and the parameters of best of fit(R2),mean variance(RMSE),mean absolute deviation(MAD),relative error(RS),and estimated accuracy(P)were used as evaluation indicators,The evaluation results for the 8 alternative models showed that the best fitting model was:FH=sH^(b)D^(c)andFH=sH^(b)/D^(c).After a comprehensive comparison,the modelFH=sH^(b)/D^(c)was selected as the binary high model for Eucalyptus plantation forests in Shuangjiang County.The model expression after substituting appropriate parameter values was:FH=4.555080H^(-0.479807)/D^(-0.749612).The results of the applicability test using 31 test sample trees showed that the model’s total relative error(RS)was-0.6%,within the±5%error range;the F statistic of the F test is 2.12,which was less than the table value of 4.18 for F_(0.05)(1,29),and hence

关 键 词：巨尾桉 人工林 林分 二元形高模型 

分 类 号：S758.5[农业科学—森林经理学]