摘要
为提高杉木林蓄积量的估测精度,筛选出对蓄积量有显著解释意义的变量集,本研究采用岭迹分析和逐步回归方法对福建省闽侯白沙国有林场杉木林蓄积量预估模型的主要评价指标因子进行多重相关性研究.结果表明:2种方法对因子的优选结果非常相近,能有效地剔除显著相关的变量,筛选出影响蓄积估测的主要因子信息,为准确估测杉木林蓄积量以及建立预测模型提供方法上的支持.
In order to improve the estimatton accuracy of Chinese fir volume and select the variables set to significantly explain the meaning of the volume, this paper used the ridge trace analysis and stepwise regression methods to research the multiple correlations between major evaluation factors of Chinese fir volume estimated model of the state-owned forest farms in Baisha, Minhou County of Fujian province. The results indicated that the choice of factors using two methods is very similar to each oth-er and both can effectively eliminate the significant related variables and sieve the main factor informa-tion out that affect the estimate of volume. In addition, the choice can provide a methodological sup-port for establishing the prediction model and accurately estimating Chinese fir volume.
出处
《亚热带资源与环境学报》
2012年第2期76-83,共8页
Journal of Subtropical Resources and Environment
基金
福建省林业厅科技研究项目(DH-397)
福建省自然科学基金项目(2009J01210)
福建省教育厅科技项目(JA09059)
关键词
蓄积量
多重相关性
因子优选
岭迹分析
逐步回归
volume
multiple correlations
factor optimization
ridge trace analysis
stepwise re-gression
