文献类型：期刊文献

中文题名：杉木人工林林分断面积生长模型的贝叶斯法估计

英文题名：Application of Bayesian Method in Stand Basal Area Prediction of Chinese Fir Plantation

作者：张雄清[1,2] 张建国[1,2] 段爱国[1,2]

第一作者：张雄清

机构：[1]中国林业科学研究院林业研究所国家林业局林木培育重点实验室;[2]南京林业大学南方现代林业协同创新中心

年份：2015

卷号：28

期号：4

起止页码：538-542

中文期刊名：林业科学研究

外文期刊名：Forest Research

收录：CSTPCD;;Scopus;北大核心:【北大核心2014】；CSCD:【CSCD2015_2016】；

基金：国家自然科学基金(31300537);中国林业科学研究院林业研究所科研院所基本科研业务资金项目(RIF2013-09);中国林业科学研究院科研院所基本科研业务资金项目(CAFYBB2014QB002)

语种：中文

中文关键词：贝叶斯法;传统法;林分断面积;杉木

外文关键词：Bayesian method, stand basal area, Chinese fir, Cunninghamia lanceolata

分类号：S711

摘要：以江西杉木人工林为例,以Korf型、Richards型和Hossfeld型3种模型为基础,通过广义代数差分法(GADA)分别建立杉木林分断面积生长模型。结果表明:以Richards型为基础的杉木林分断面积预测精度最高,以Richards型模型为最优模型,分别基于贝叶斯法和传统法(非线性最小二乘法)估计杉木林分断面积生长模型。研究发现,利用贝叶斯法估计杉木林分断面积生长模型,预测精度相当且预测值的可靠性比传统法好。
Chinese fir （ Cunninghamia lanceolata （ Lamb. ） Hook. ）, an endemic tree species in China＇ s subtropi- cal area, is one of the most important fast-growing tree species for timber production in southern China. Based on the periodic data of the Chinese fir in Jiangxi province, three stand basal area models （ Korf-based model, Richards - based model, and Hossfeld-based model） were developed using generalized algebraic difference approach （ GA- DA）. The results showed that Richards-based model was the best for modeling the stand basal area of Chinese fir in the study. Additionally, Bayesian method and Classical method （nonlinear least squares method） were used to esti- mate the Richards-based model. Although the precision of Bayesian method was nearly equal to that of the classical method, the model reliability using Bayesian method was better than using classical method.