欢迎访问林业科学,今天是

林业科学 ›› 2001, Vol. 37 ›› Issue (5): 24-30.doi: 10.11707/j.1001-7488.20010505

• 论文及研究报告 • 上一篇    下一篇

用岭估计研究以RS和GIS为基础的森林郁闭度估测

李崇贵1 石强4 赵宪文2 田永林3   

  1. 深圳职业技术学院,深圳518055;中国林业科学研究院资源信息研究所,北京100091
  • 收稿日期:2000-09-25 修回日期:1900-01-01 出版日期:2001-09-25 发布日期:2001-09-25

ON FOREST CANOPY DENSITY ESTIMATION RESEARCH BASED ON RS AND GIS WITH RIDGE ESTIMATION

Li Chonggui,Shi Qiang,Zhao Xianwen,Tian Yonglin   

  1. Shenzhen Polytechnic Shenzhen 518055;The Research Institute of Resource Information, CAF Beijing 100091
  • Received:2000-09-25 Revised:1900-01-01 Online:2001-09-25 Published:2001-09-25

摘要:

本文根据最小二乘估计(LS)及岭估计(R)原理,利用平均残差平方和准则,从样地对应的RS和GIS信息中,用计算机仿真方法筛选出影响郁闭度估测的主要信息。在分析LS估计存在缺陷的基础上,提出采用岭估计建立郁闭度估测方程。实例分析表明,当在影响郁闭度估测的变量间存在复共线性时,岭估计比LS估计具有较高的预报精度,可以满足生产的需要。

关键词: 郁闭度, 方差扩大因子, 平均残差平方和, 岭估计, 复共线性

Abstract:

In this paper, by means of the principles of LS estimation and ridge estimation, using the rule of residual mean squares(RMSq), the main information that effect the estimation of forest canopy density is screened out from the RS and GIS information of the ground sample plots through computer simulation. Then, based on the analysis of possible limitation of LS estimation, the ridge estimation is put forward to establish the equation of canopy density estimation. The practical example indicates that when there are some multi-collinearities between the RS and GIS information that effect canopy density estimation, ridge estimation is prior to LS estimation.

Key words: Canopy density, Variance inflation factor, RMSq, Ridge estimate, Multi-collinearity