陕西省主要树种单木及林分生长率模型构建

王含予; 程 炜; 胡云云; 郝家田; 何学高; 魏江涛; 侯晓巍

文章摘要

王含予,程炜,胡云云,郝家田,何学高,魏江涛,侯晓巍.陕西省主要树种单木及林分生长率模型构建[J].林业调查规划,2026,51(2):1-7

陕西省主要树种单木及林分生长率模型构建

Development of Individual Tree and Stand Growth Rate Models forMajor Tree Species in Shaanxi Province

DOI：

中文关键词: 单木生长率模型林分生长率模型最优模型二元林分蓄积量生长率模型非线性回归陕西省

英文关键词: individual tree growth rate model stand growth rate model optimal model two-variable stand volume growth rate model nonlinear regression Shaanxi Province

基金项目:陕西省重点研发计划(2022ZDLSF06-01).

作者	单位
王含予	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048
程炜	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048
胡云云	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048
郝家田	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048
何学高	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048
魏江涛	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048
侯晓巍	国家林业和草原局西北调查规划院/ 旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安 710048

摘要点击次数: 513

全文下载次数: 74

中文摘要:

基于陕西省第六次至第九次全国森林资源清查固定样地数据,采用非线性回归分析,构建15个树种(组)的单木胸径和材积生长率模型,以及9个树种(组)的林分蓄积量生长率模型,利用赤池信息准则(AIC)和贝叶斯信息准则(BIC)对一元(平均胸径或平均年龄)和二元(平均胸径、平均年龄) 林分生长率模型进行最优模型筛选。结果表明,各树种(组)单木生长率模型决定系数(R²)均大于 0.89,总体相对误差(TRE)均在±5%以内(冷杉除外),平均预估误差(MPE)均小于5%,在平均百分标准误差(MPSE)中,胸径生长率模型均小于10%,材积生长率模型基本小于20%。大部分树种(组)林分生长率模型的R²均高于0.9,TRE 均在±5%以内,MPE不超过 5%,而MPSE小于20%,与一元林分蓄积量生长率模型相比,二元模型精度更高,推荐使用二元林分蓄积量生长率模型(马尾松除外)。

英文摘要:

Based on the fixed sample plot data from the 6th to 9th National Forest Inventory in Shaanxi Province, nonlinear regression analysis was mployed to develop individual tree diameter at breast height (DBH) and volume growth rate models for 15 tree species (groups), and stand volume growth rate models for 9 tree species (groups). The Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to select the optimal models between the single-variable (mean DBH or mean age) and two-variable (mean DBH and mean age) stand growth rate models. The results showed that the coefficients of determination (R²) for all individual tree growth rate models were greater than 0.89. The total relative errors (TRE) were all within ±5% (except for Abies), and the mean prediction errors (MPE) were all less than 5%. Regarding the mean percent standard errors (MPSE), the values for DBH growth rate models were all less than 10%, and those for volume growth rate models were mostly below 20%. For most tree species (groups), the R² values of the stand growth rate models were higher than 0.9, the TRE values were within ±5%, the MPE values did not exceed 5%, and the MPSE values were less than 20%. Compared to the single-variable stand volume growth rate models, the two-variable models demonstrated higher accuracy. Therefore, the twovariable stand volume growth rate models are recommended for application (except for Pinus massoniana).

查看全文查看/发表评论下载PDF阅读器

关闭