摘要
人口统计数据空间化是解决统计数据与自然要素数据融合分析的有效途径。随着RS和GIS技术的发展,人口空间化方法推陈出新,但不同的方法各有优缺点,方法选择和精度保证是两个重要的问题。本文以黄河三角洲高效生态经济区2010年县(市、区)单元人口统计信息为案例,分别基于土地利用数据和夜间灯光数据建立1km栅格的人口空间化模型,再分别以县(市、区)和乡(镇、街道办)为验证单元,比较两种方法模拟人口分布的优缺点。结果表明,基于土地利用数据的人口空间化模型整体优于基于夜间灯光数据的人口空间化模型,且前者更适合于模拟人口总数居中的县(市、区),后者在人口密度梯度较大的城镇边缘有更好的模拟效果,在人口较少的地区二者的模拟效果均不太理想。本研究能够为获取高精度的区域人口空间化模型提供参考。
In this paper, two models were tried to simulate population distribution in the Efficient Ecological Economic Zone of the Yellow River Delta, in specific, township population census data in 2010 was available, and two population spatialization models were developed at 1 km resolution based on land use data and DMSP/OLS data respectively. The result shows that the linear relationship between population data and land use data or DMSP/OLS data is relatively high（the correlative coefficient R2 is 0.921, 0.920 respectively）,however the model based on land use data, which usually should be revised by combing other information such as DEM and traffic network in spatial areas where the terrain varies dramatically or the transportation system is well-developed, is much better than the model based on DMSP/OLS data, especially in counties that have moderate population density. In spite of this, benefited from the positive effect of pixel overflow,the model based on DMSP/OLS data is more suitable for marginal areas of towns and cities where violent gradient of population density is found. In addition, both models perform weakly in less populated regions.
出处
《人文地理》
CSSCI
北大核心
2014年第5期94-100,共7页
Human Geography
基金
中国科学院烟台海岸带研究所"一三五"项目(Y254021031)
中国科学院战略性先导科技专项-应对气候变化的碳收支认证及相关问题(XDA05130703)
