适于FVS的杉木单木模型构建  

Establishment of Individual Growth Model of Cunninhamia lanceolata in FVS System

作　　者：衣旭彤 孙玉军[1] 

机构地区：[1]省部共建森林培育与保护教育部重点实验室(北京林业大学),北京100083

出　　处：《东北林业大学学报》2017年第7期12-17,共6页Journal of Northeast Forestry University

基　　金：国家林业局"948"项目(2015-4-31);林业科技成果国家级推广项目([2014]26)

摘　　要：以福建省将乐国有林场杉木(Cunninghamia lanceolata)人工林标准地和解析木数据为依托,使用R语言,构建杉木单木胸径-树高模型、冠幅模型、胸径生长量模型(大树、小树)、树高生长量方程(大树)以及材积模型,并计算树皮因子。通过赤池信息准则(AIC)、贝叶斯信息准则(BIC)和对数似然值,结合R2值选择最优模型,并对所建模型进行精度检验。结果表明:各模型的预估精度最高的为树皮因子模型,精度达到99.01%,预估精度最低的模型为直径生长量模型,精度为83.54%,其余模型精度均达到95%以上;经t检验,所有模型估计值与实际值差异不显著,模型均可用于森林植被模拟系统。Data from Cunninghamia lanceolata plantation permanent stands and analytical trees taken from Jiangle state forest farm in Fujian Province were used to establish the height-diameter model, crown model, DBH increment model （ including big tree and small tree） , height increment （for 10 a） mode] （big tree） and volume model, and the bark factor was calcu- lated. Optima] equations were selected according to Akaike information criterion, Bayesian information criterion ancl Log Likelihood value for determining coefficient, accuracy test was conducted afterwards. Among all the models, the bark factor model was the most accurate one with P of 99.01%, and the least accurate model was DBH increment model with P of 83.54%. Accuracy of other models were higher than 95%. By the t-test, estimated value had no significant difference than the actual value, and all the models could be used for Forest Vegetation Simulator system （FVS）.

关 键 词：杉木 单木模型 生长量 森林植被模拟系统 

分 类 号：S757.1[农业科学—森林经理学]