摘要
首先将山东省各县市按级别分类,利用回归方法得到不同级别县市的城镇居住密度与农村居住密度,然后根据乡镇驻地位置,构造Voronoi图来替代乡镇边界。在此基础上利用各乡镇人口数据与居民点数据计算得到各 Voronoi图形的居住密度,并将该密度赋予每个图形的发生点(即乡镇驻地位置),而后利用反距离插值算法,并结合窗口移动平均处理,得到全省的居住密度分布图。使用居住密度分布图和居民点分布数据,计算人口分布数据。并根据每个县市的图上汇总人口和统计人口,计算得到各县市的调整系数,使用该系数对每个栅格的人口数进行调整,保证了各县市总人口不会出现误差。最终得到山东省1km×1km栅格的人口分布数据。最后选取某些样本县, 在样本县内对比每个乡镇图上的人口与实际统计人口,结果表明该人口分布数据有较高的精度。
Firstly, the 108 counties in Shandong province are classified which are categorized into three types, then the different urban residential and rural residential densities of each type are calculated by regression with land use data (1:10 000) interpreted from TM remote sensing image. Secondly, the Voronoi map of the location of township seat is constructed, and is taken as the township boundary, the residential density of every Voronoi part can be calculated by dividing the statistical population and the number of the settlements. Let every location of the township seats have the corresponding Voronoi's density, then the residential density surface of the whole province is calculated by IDW interpolation and neighborhood statistics method in ArcGIS. With this density surface and the land use data, the lkmxlkm grid population data is calculated supported by ArcGIS. Considering the difference between the statistical pop adjusting coefficient for every county is calculated. ulation and the calculated population of all counties, an After the adjustment, the statistical population and the calculated population are equal. Finally, the comparison of the statistical population density and the actual population density of the selected sample counties proves the higher precision of the results.
出处
《地球信息科学》
CSCD
2005年第4期54-58,共5页
Geo-information Science
基金
国家"十五"攻关子课题(2004BA608B-1-11)中国科学院地理科学与资源研究所特别支持项目"人口空间数据库"(Y4900)。
关键词
人口
VORONOI图
居住密度
空间化
栅格
population
voronoi map
density of settlement
spatial distribution
grid