马尾松人工林相对树高曲线模型构建

Study on relative tree height curve model of Pinus Massoniana plantation

在广西马尾松Ⅰ类、Ⅱ类和Ⅲ类产区设置279块马尾松人工林标准样地,采集每木检尺数据,数据经预处理后选取理查德函数变形的可变参数式与固定参数式模型分别构建马尾松相对树高曲线模型,并以参数变动系数和综合修正决定系数(R^(2))、总相对误差(TRE)、平均系统误差(ASE)、平均百分标准误差(MPSE)及预估精度(P)等指标进行综合评价;同时,对该模型进行分区和分胸径、树高段及独立样本检验。结果显示:构建的最佳广西马尾松人工林理查德函数变形的固定参数相对树高曲线方程为RH_(i)=(1-e^(-c_(1)))^(-c_(2))×(1-e^(-c_(1)RD_(1)))^(c_(2))。其中c_(1)取值为2.2546、变动系数为3.902%,c_(2)取值为1.1982、变动系数为4.953%。该相对树高曲线模型的拟合效果佳,在通用性和适用性方面均表现良好,达到了编制林业数表模型的控制精度。

In first-class,second-class,and third-class Pinus massoniana plantation areas of Guangxi Zhuang Autonornous Region,279 standard plots were established in P.massoniana plantation to collect tree measurement data.After data preprocessing,the variable parameter formula and fixed parameter formula of the Richard function were selected to construct the relative height curve model of P.massoniana,and comprehensive evaluation was carried out using indicators such as the coefficient of parameter variation,the coefficient of determination(R^(2))with comprehensive correction,the total relative error(TRE),the average systematic error(ASE),the mean percentage standard error(MPSE),and prediction accuracy(P).At the same time,the model was partitioned and tested using independent samples based on chest diameter,tree height section,and other factors.The results showed that the best fixed-parameter relative height curve equation for P.massoniana plantations in Guangxi,constructed using the transformed Richard function,was:RH_(i)=(1-e^(-c_(1))^(-c_(2)))×(1-e_(-c_(1)RD_(1)))^(D_(1)).The value of c_(1) was 2.2546 with a coefficient of variation of 3.902%,while the value of c_(2) was 1.1982 with a coefficient of variation of 4.953%.The relative tree height curve model had good fitting performance and performs well in both generality and applicability,achieving the control accuracy of compiling forest inventory models.