摘要
利用GIS软件ARC／INFO将北京地区的1：20万植被图数字化，并提取各斑块的面积信息。该图包含72个基本的斑块类型，它们又分属于森林、灌丛、草地、果园、农田和水体6大类型。这些大类又分别包含20、28、4、7、11和2个基本类型。选用斑块个数、总面积、平均斑块面积、标准差、变异系数、中值、最大斑块面积、最小斑块面积、极差和偏态系数等几个描述统计量，以及厂一分布、对数正态分布、Weibull分布、指数分布和正态分布等5个概率分布来刻画斑块大小的分布特征。结果表明：除具有很少斑块的少数基本类型以外，其它基本类型以及所有6个大类的斑块大小的分布都不是对称的，而是右偏的。因此，普通的正态分布不能对它们加以刻画；其他4种概率分布也只能刻画部分类型，并且服从对数正态分布的类型最多，服从负指数分布的斑块类型最少。
The vegetation map on the scale of one to two bunded thousandth in Beijing was digitized and analyzed with GIS software ARC/INFO. In this map, there are 72 lower level patch types in total, and they belongto 6 higher level Patch types, viz. forest, shrubland, grassland, orchard, cropland and waters, which include20, 28, 4, 7, 11 and 2 lower level patch types respeCtively. Sevenal basic descriptive statistics and five typesof probability distributions, which include Gamma, negative exponential, lognormal, weibull, and normaldistributions, were chosen to characterize the patch size distributions for all the patch types in Beijing area.The results show that the patch size distributions for most of the 72 lower level types and all the 6 higher leveltypes are right-skewed. Therefore, normal distribution cannot be used to characterize the patch size distributions in this area; and each of the other four probability distributions can only depict part of the types. Thepatch types that can be characterized with lognormal distribution are relatively more, and those with negativeexponential distribution are fewer.
基金
国家自然科学基金!100139770131
中国科学院特别支持项目!STZ-1-17
关键词
植被景观
斑块大小
概率分布
地理信息系统
Vegetation landscape, Patch size, Patch type, Probability distribution, GIS
