高分辨率遥感数据云杉林林分冠幅估计

Estimation on Forest Stand Crown of the Picea schrenkiana Based on High Spatial Resolution Imagery

以西天山林区典型植被类型——云杉林为研究对象,选择高空间分辨率卫星遥感影像(QuickBird)作为数据源,采用空间统计学方法对林分冠幅进行定量估计。云杉林林分冠幅具有一定的空间结构性,在可以作为表征树冠大小及其空间分布特征的二维区域化随机变量的高空间分辨率卫星遥感影像上表现为均衡分布的简单斑块,并且间隔一定距离具有空间自相关特征。通过对影像特定方向上的半方差函数的变程与云杉林林分(样地)平均冠幅之间的相关分析,发现二者之间存在较强的线性相关关系,表明采用空间统计学半方差理论可以用作研究区范围云杉林林分平均冠幅定量估计的参考。

The object of Picea schrenkiana var tianshanica,the typical vegetation in forest zone of west Tianshan Mountains were chozen to study high spatial resolution satellite remote sensing image （Quick- Bird） for the experimental data,the forest stand crown information was estimated quantificationally with the method of spatial statistics in this paper.Forest stand crown of Picea schrenkiana var tianshanica has spatial structure, shows a simple patches with isostatic distribution in the image,as the two-dimensional re- gional random variables characterization of crown size and spatial distribution information, and shows spa- tial autocorrelation characteristics with a certain distance. Analysis indicates that there is a strong linear correlation between the rang of stand （or plot） image＇s semivariogram in particular directions and the av- erage crown size of stand （or plot）, the semivariogram method of spatial statistics may be as an effective means of estimating the forest stand crown of Picea schrenkiana var tianshanica.