利用第八次全国森林资源清查青海省2013年清查的4万多株保留木前后期实测胸径数据,分析了林木胸径生长量、比率和生长率模型外在表现和内在本质的异同,并建立了4个主要树种组的立木胸径生长率模型。研究结果表明,林木胸径生长量、比率和生长率模型只是反映林木生长信息的不同数学表现形式,尽管各类模型的统计指标存在很大差异,但用于胸径生长量及期末胸径预估时,不同模型之间的差异很小。当用于建模的样木数量很大且分布极不均匀时,可将相同胸径的样木进行合并处理、采用平均数建模,但模型统计指标的计算,应当基于合并前的全部样木数据,而不是合并后的建模数据,否则会对模型的拟合优度给出过高的评价。利用青海省4个树种组的胸径生长率模型预估林木的期末胸径时,确定系数R²能达到0.93以上,平均预估误差MPE均在0.3%以下。所建模型可为第九次全国森林资源清查青海省2018年清查外业调查时判定林木胸径生长量是否过大提供定量依据。

Based on the mensuration data of diameter at breast height (dbh)from about 40 thousand survival trees in 2013 Qinghai forest inventory of the 8^th National Forest Inventory in China,the differences in external expression and similarities in inner essence between individual tree dbh growth,ratio,and rate models were analyzed,and growth rate models for 4 tree species were developed.It is showed that:1)individual tree dbh growth,ratio,and rate models are only different mathematical expressions reflecting tree growth information,even though the statistical indices of developed models exhibit great differences,when they are used to estimate dbh growth or dbh at the end of interval period,the differences are very few.2)When the sample trees are too many and distributed unevenly by diameter class,it is suggested to merge the data of trees with same diameter,and use the average values for modeling.However,the calculation of statistical indices of fitted models should be based on the data of all sample trees,not the merged data for modeling,otherwise the goodness-of-fit of models will be over-rated.3)The determination coefficients (R²)are more than 0.93,and mean prediction errors (MPE)are less than 0.3%,when dbh growth rate models for 4 tree species in Qinghai are used to estimate diameters at the end of interval period.The developed dbh growth rate models will provide quantitative basis for determining whether the dbh growth of trees is too large in field survey of Qinhai forest inventory of the 9^th National Forest Inventory.