基于林木多期直径测定数据的异龄林年龄估计方法  

Age estimation method of uneven-aged forest based on data of multistage diameter measurement

作　　者：龙时胜 曾思齐[1] 甘世书[2] 肖化顺[1] 刘洵 向博文 LONG Shisheng;ZENG Siqi;GAN Shishu;XIAO Huashun;LIU Xun;XIANG Bowen(Central South University of Forestry and Technology,Changsha 410004,Hunan,China;Central South Forest Inventory and Planning Institute of State Forestry Administration,Changsha 410014,Hnnan,China)

机构地区：[1]中南林业科技大学,湖南长沙410004 [2]国家林业局中南林业调查规划设计院,湖南长沙410014

出　　处：《中南林业科技大学学报》2018年第9期1-8,共8页Journal of Central South University of Forestry & Technology

基　　金：国家林业公益性行业科研专项(201504301);中南林业科技大学研究生创新基金项目(20181001)

摘　　要：基于固定样地多期直径测定数据,拟合直径与年龄相关的生长模型,为异龄林年龄结构的确定提供估计方法。通过为第一期林木设置不同年龄初值的方法采集样地内各树种的直径与年龄坐标信息,采用不同直径生长方程对数据进行模拟,依据树高生长至1.3 m时的年龄t0,将拟合曲线在坐标内准确定位并用生长方程估计单木、径阶以及林分的平均年龄,对结果进行适用性检验和评价。结果表明:运用Richards方程分别拟合青冈栎、杉木、马尾松的直径与年龄的生长关系时,模型的相关指数R2分别为0.93、0.91、0.97,剩余标准差分别为1.18、0.78、0.37,平均绝对误差分别为0.80、0.61、0.31;该估计方法预测林分平均年龄的绝对误差最大值为2 a,相对误差最大值为8.2%;预测径阶年龄的绝对误差最大值为3 a,相对误差最大值为12.5%,平均绝对百分误差为5.1%;预测单木年龄的绝对误差最大值为7.0 a,相对误差最大值为29.3%,平均绝对百分误差最大值为14.8%。Richards方程在参数生物学意义以及模型拟合精度方面能较好地反映出各树种的直径与年龄生长的关系。研究提出的依据林木多期直径测定数据估计异龄林年龄的方法,预测林分平均年龄与径阶年龄的平均绝对百分误差小于10%,预测效果较好,其估计方法可以应用于科研和生产实践。Based on the data of fixed sample multiphase diameter measurement, the growth model of fitting diameter and age was proposed to provide an estimation method for the determination of the age structure of uneven-aged forest. A method of set initial values of different ages for the first stage forest was used to collect the information about the diameter and age of each tree species in the sample plots, the data was simulated by using different diameter growth equations, based on the age of tree height to 1.3 m, the fitting curve was accurately positioned in the coordinates, the growth equation was used to estimate the age of the single tree, the diameter and the stand,and the results were tested and evaluated. The Richards equation was used to fit the relationship between diameter and age growth, the correlation index R2 of the model were 0.93, 0.91, 0.97, the residual standard deviations were 1.18, 0.78, 0.37, and the mean absolute errors were 0.80, 0.61, 0.31. The maximum of absolute error was 2 year, and the maximum of relative error was 8.2%, when using the estimation method to predict the average age of the stand. The maximum absolute error was three year, the maximum of the relative error was 12.5%, the average absolute percentage error was 5.1%, when using the estimation method to predict the age of diameter class. The maximum absolute error was seven year, the maximum of the relative error was 29.3%, the maximum of the mean absolute percentage error was 14.8%, when using the estimation method to predict the age of individual tree. The Richards equation can reflectd the growth relationship between the diameter and age in terms of the biological significance of parameters and the precision of model fitting. The average absolute percent error of predicting the age of the stand and diameter class was less than 10%, the prediction results of the method were better, and the estimation method can be applied to the scientific research and production practice.

关 键 词：异龄林 林木多期直径 年龄 方法 RICHARDS方程 

分 类 号：S758[农业科学—森林经理学]