首页>>学术检索

不同生长势杉木成熟林的生物量生长模型研究

标题: 不同生长势杉木成熟林的生物量生长模型研究
英文标题: Biomass Model of Cunninghamia lanceolata with Different Growth Potentials
作者: 姜鹏,庞立欣,邵思祺,杨冲,郭金堂,张绍轩
英文作者: JIANG Peng,PANG Li-xin,SHAO Si-qi,YANG Chong,Zhang Shao-xuan,Agricultural University of Hebei
出版时间: 2016-05-15
机构: 河北农业大学
关键词: 杉木,生长势,生物量模型
英文关键词: Cunninghamia lanceolata,growth potential,biomass model
刊名: 西北林学院学报
英文刊名: Journal of Northwest Forestry University
ISSN: 1001-7461
期号: 03
卷号: v.31;No.139
国内刊号: 61-1202/S
基金: 河北省社会科学基金(HB15YJ051)
主题: 不同生长势杉木成熟林的生物量生长模型研究
页码: 42-46+55
分类号: S791.27
出版单位: 西北林学院学报
是否核心: 1
摘要: 以福建省将乐县国有林场杉木成熟林为研究对象,通过建立16块杉木标准地,获取各标准地基础数据。根据克拉夫特林木分级法以及胸径、树高的统计分析,将标准地内林木按生长势不同分为优势木、中庸木和被压木3类。结果表明:1)通过林木实测数据拟合不同生长势的杉木各器官生物量(干、根、叶、枝)估算模型、全株生物量模型,以D、DH为自变量,分别拟合了各个器官(树干、树根、树叶、树枝)和全株的Logistic方程和幂函数,其模型的相关判定系数R2均在0.881~0.932之间。2)根据残存平方和SSE、总相对误差RS、平均相对误差E1、平均相对误差绝对值E2、AIC、BIC6个检测评价指标进行方程的4选1筛选,筛选出最优模型中,幂函数模型9个(模型自变量D1个,自变量DH8个),Logistic函数模型6个(自变量D的2个,自变量DH4个)。
英文摘要: In this paper,the mature Cunninghamia lanceolata plantation of state-owned forest farms in Jiangle County of Fujian Province was taken as the research object,in which 16 sample plots were established to extract the basic data.According to the Kraft forest grading and statistical analysis of DBH and tree height,trees in sample plots were divided into dominant,intermediate,and suppressed trees by different growth potentials.1)Models that were used to predict biomass of whole plant and different organs(trunk,root,leaf,and branch)were established by fitting the actual measured data and different growth potentials.By using Dand DH as independent variables,the Logistic equation and power function of different organs and whole-plant were simulated.The determination coefficient R2 of the model was between 0.881to0.932.2)Based on the 6evaluation indicators,such as residual sum of squares,total average relative error,average relative error,average absolute relative error,Akaike information criterion,and Bayesian information criterion,the best model was selected from four models.The best model included 9power function models with independent variables of 1 Dand 8 DH,6Logistic models with independent variables of 2 D and 4 DH.