摘要:
随着高分辨率遥感数据、LiDAR 技术在森林资源调查中应用的日益普遍,以及自动快速提取树木冠幅、树高 方法的日益成熟,需要建立新的树木模型来估测其他林分因子,以适应林业调查手段更新和发展的需求。在参考 国内外大量文献的基础上,选择常用的胸径鄄冠幅模型和树高鄄胸径曲线模型,根据北京市178 块森林样地调查数 据,建立了北京市13 个常见树种的胸径鄄冠幅模型,胸径鄄树高模型,胸径鄄冠幅、树高综合模型。结果表明:并不是 所有树种的冠幅、树高都与胸径有高相关性;13 个树种中,只有臭椿、雪松、栾树和加杨的胸径鄄冠幅模型的R2 大于 0郾7,胸径鄄树高模型R2 大于0郾5,胸径鄄冠幅、树高综合模型达到0郾8 以上;油松、杨树、槲栎、圆柏4 个树种的胸径鄄 冠幅模型的R2 小于0郾3;核桃、油松、火炬树、柳树、国槐5 个树种的胸径鄄树高模型R2 低于0郾3。并提出在下一步工 作中,把林龄,林分密度,立地条件如坡度、坡向、海拔、地位级等林分因子与树木冠幅、树高联合建立估测模型来提 高胸径估测的精度。该方法可用于现代遥感技术快速获取树木冠幅、树高之后,根据已有的数据库资料,用树种的 冠幅、树高估测胸径,再推算其他林分因子,实现森林资源的快速调查与更新。
关键词:
-
胸径估测模型
/
-
冠幅
/
-
树高
/
-
回归模型
Abstract:
As high spatial resolution remote sensing images and LiDAR(light detection and ranging)technology applied in forestry to extract tree crown and height automatically, new models are needed to predict forest stand information. Based on the investigation of 178 sample plots distributed in Beijing, the most common used crown diameter(C)-diameter at breast height(D) models and tree height(H)-D models were chosen to predict D from C and H for thirteen common tree species in Beijing. The results showed that not all tree species'C and H had close relationship with D. Tree species of Ailanthus altissima, Cedrus deodara, Koelreuteria paniculata and Populus canadensis had high correlation of D-C and D-H equation, with determination coefficient (R2) value higher than 0.7 and 0.5, and above 0.8 by combination C and H to estimate D. But the C of aspen, China savin, Oriental white oak, Chinese pine and the H of walnut, willow, Chinese scholar tree, torch tree, Chinese pine were poor correlated with D, both with a R2 value lower than 0郾3. Regional site conditions and stand variables (site index, stand age and density) were suggested to join with C and H to improve the D predicting accuracy in the next step work. This method can be used in estimating D and other stand information with C and H automatically extracted by modern remote sensing technology, which can realize quick and economical forest resource investigation and update.
本站查看
下载:
