削度方程是描述树干干形的一种重要方法，过去几十年中得到广泛应用。但实际应用时，发现这种方法存在2个主要缺陷：第一、传统树干形状描述理论的建模思想是以不同的干曲线绕坐标轴旋转形成圆截面的原条数学模型。但由于截面形状不变，树干中心轴是直线，不论怎样改变干曲线，都不可能准确地表达实际树干形状。第二、削度方程只能反映整个林分或样地树木干形的平均变化，但实际情形中，每棵树木因自身生长状况，遗传因素、所处立地和生境条件不同，树干形状可能有较大差异。然而树木干形的确定是林业造材的一项关键技术，其精度高低决定了造材的成败。为此，本项目重点解决三个问题：（1）给出一套针对树干干形研究的三维激光扫描外业调查技术体系以及多株林木点云数据的参数提取；（2）利用近代统计模型方法构建三维树干曲面模型，并在ForStat中创建单株树干曲面模拟模块；（3）基于以上研究成果，以杨树为例，首次实现三维单木树干曲面模型的构建。

Taper equation is an important approach to reflect the tree trunk form and therefore was widely applied in forest in the past few decades. However, two major drawbacks of this approach were found in practical application, which are （1）The modeling idea of the descriptive theory for the traditional trunk form is established based on the pole-timber mathematical model of the circular section generated by different trunk form curve arounding the axis of rotation. It cannot describe the actual tree trunk form accurately no matter how to change the trunk curve due to invariable cross sectional shape and the straight line for the trunk center axis. (1) the taper equation just descripe the average variation of the tree trunk forms in the whole stand or sample plot, but in fact, the trunk form of different tree may be great difference due to the variances of their state of self-growth, genetic features, site and habit conditions. Determination of tree trunk form is a key technology for forestry conversion, and the precision of it would affect the conversion results.Therefore, this project would focus on the following three scientific problems: (1) Give a set of field investigation technology system for 3D laser scanning and an approach to extract the parameters from the point-cloud data for tree trunk form study. (2) Three-dimensional tree trunk surface models would be constructed based on the approaches of traditional least square regression and nonlinear mixed-effects models and the module for individual tree trunk surface simulation also would be developed in the ForStat software. (3) Based on the above research results, the three-dimension tree trunk surface models would be developed firstly using Populus as an example.