基于树高和冠幅的广西桉树二元立木材积模型研建

Study on Binary Volume Model for Eucalyptus Based on Tree Height and Crown Width in Guangxi

关键词:  树高  冠幅  桉树  材积模型  广西

Key words:  tree height  crown width  Eucalyptus  volume model  Guangxi

摘要:

激光雷达在获取树高和冠幅等森林资源信息方面具有无可比拟的优越性，构建基于树高和冠幅的二元立木材积模型，可为激光雷达技术在森林蓄积量估测应用中提供计量依据。通过测定广西桉树（Eucalyptus）典型分布区448株样木的树高、冠幅、胸径等因子，采用非线性回归估计方法建立树高冠幅二元材积模型、树高一元材积模型、胸径树高二元材积模型、胸径一元材积模型和冠幅一元材积模型，并对模型进行检验评价。建立的5个模型确定系数分别为0.969，0.875，0.994，0.945和0.588，总体误差分别为0.29%、-1.94%、-0.26%、1.88%和-2.82%，模型预估精度分别为97.75%、95.38%、99.14%、96.95%和91.72%；树高冠幅二元材积模型的模型总体检验、分树种检验、分段检验、分区检验和五折交叉检验，均符合林业数表编制的相关要求。树高冠幅二元材积模型各项指标显著优于树高一元材积模型和冠幅一元材积模型，与胸径树高二元材积模型接近，完全符合林业数表编制要求，可应用于基于机载激光雷达的森林资源调查和监测。

Abstract:

Laser radar has incomparable superiority in obtaining forest resources information such as tree height and crown width.The construction of binary volume model based on tree height and crown width can provide measurement basis for the application of laser radar technology in forest volume estimation.By measuring the factors such as tree height, crown width and diameter at breast height (DBH) of 448 sample trees in the typical distribution area of Eucalyptus in Guangxi, the binary volume model of tree height and crown width, the unitary volume model of tree height, the binary volume model of tree height and diameter at breast height, the unitary volume model of tree diameter at breast height and the unitary volume model of crown width were established by nonlinear regression estimation method, and the models were tested and evaluated.The determination coefficients of the five models were 0.969, 0.875, 0.994, 0.945 and 0.588, respectively.The overall errors were 0.29%, -1.94%, -0.26%, 1.88% and-2.82%, respectively.The model prediction accuracy was 97.75%, 95.38%, 99.14%, 96.95% and 91.72%, respectively.The model overall test, tree species test, segmentation test, zoning test and five-fold cross test of the binary volume model of the high crown width of the tree all meet the relevant requirements for the compilation of forestry tables.The indexes of the binary volume model of tree height and crown width are significantly better than those of the unitary volume model of tree height and crown width, which are close to the binary volume model of DBH and tree height.It fully meets the requirements of forestry data table compilation and can be applied to forest resource survey and monitoring based on airborne laser radar.