中文摘要：

利用ArcGIS 9.3软件从2009年全国第二次土地调查成果数据库中提取丰县农村居民点数据作为数据源,采用中心属性值原则得到20种空间粒度下的农村居民点景观格局,借助Fragstats 3.3软件对选取的8个景观指数进行计算,分析农村居民点景观格局的粒度效应并据此确定适宜粒度范围。研究表明：除景观面积指数及景观分离度指数外,其他景观指数的粒度效应较明显,即景观格局对空间粒度的依赖性较强,若避开粒度谈论景观格局意义不大;空间粒度由10m增大到200m,斑块个数、斑块密度、景观形状指数、景观结合度指数、景观聚集度指数均下降,最大斑块指数上升,说明最大斑块的优势度上升,斑块破碎化程度降低,斑块形状趋于规则,景观斑块间连接程度降低且分布趋向分散;综合各个景观指数的第一尺度域,研究确定丰县1∶10 000比例尺精度下农村居民点景观格局分析的适宜粒度范围为50-60m,最佳粒度为60m。

英文摘要：

Taking the landscape pattern of the rural settlement in Fengxian of Jiangsu Province as the research subject, this paper extracted the rural settlement data of Fengxian in Jiangsu Province from the national results of the second land survey data as the main data source by using ArcGIS 9.3 spatial analysis software to find out how the different grinds of administrative center influenced the landscape pattern of land use by using Fragstats 3.3 as the landscape pattern analysis software, then got twenty kinds of rural settlement landscape pattern based on Rule of Centrie Cell. With the help of ArcGIS 9.3 spatial analysis software, we changed the vector data of rural settlement to raster data. By using Fragstats 3.3 as the landscape pattern analysis software, the eight selected landscape indices were calculated. The selected landscape indices could reflect the changes in the characteristics of amounts and scales, the shape, the spatial configuration and the dominance index. According to the results of the calculation, we analysed the effects of changing grain size on landscape indices of rural settlement at county scale, based on which we chose the appropriate grain of landscape pattern analysis. The results showed as follows. （1） Most of the rural settlements in Fengxian were small-scale and the per capita area was large （204.04 m^2 per capita）, at the same time the shape of landscape was fragmentized and anomalous. The distribution features of rural settlement in Fengxian were ＂low aggregation, high dispersion＂ and ＂small scale, large density＂. （2） Apart from the two landscape indices of Total Area and Landscape Division Index, the remaining six landscape indices all had strong scale effect. In other words, the landscape pattern was exposed to the space scale strongly. There was little significance talking about the landscape pattern without the space scale. （3） With the space scale increasing from 10m to 200m, all the five landscape indices of Number of Patches, Patch Density, Landscape Sha