基于遥感和GIS的人口数据空间化研究进展及案例分析

Advance and Case Analysis in Population Spatial Distribution Based on Remote Sensing and GIS

人口数据对于全球、洲际、区域尺度的人与环境交互作用研究的重要性已经被广泛地认知。但基于行政单位的人口统计数据难以与基于自然单位的环境数据相匹配,必须通过建模对人口数据进行空间化分布。人口数据空间化建模的基本思路是将人口数据和地球表面的地理因子关联起来,遥感和地理信息系统提供了有效的工具。介绍了国内外基于遥感和GIS的人口空间化研究的主要项目和方法思路。以黑河流域为研究区,在流域尺度上把典型的人口估计结果GPW、UNEP/GRID、LandScan和中国1km格网的人口资料与政府人口统计数据进行比较分析,可以看出国内外的研究机构和研究人员已做了大量的相关工作。国外研究主要包括从遥感解译信息反演人口数据、从DMSP-OLS夜间灯光数据反演人口数据和从遥感获取的光谱特征直接反演人口数据。国内研究尽管起步较晚但发展很快,主要是根据土地利用数据和其它地理因子(如高程、道路、居民区等)建立回归模型。黑河流域人口空间分布的比较结果表明中国1km格网人口数据是几种数据中与实践情况最相符合的。

It is widely recognized that population data are very important to integrate research among the human beings and environmental factors in global, continental and regional scales. But the population census data collected based on the administrative units do not match the environmental data collected based on the natural units. So studies on the population spatial distribution are necessary by modeling and estimating the population data. The basic method of the population estimation is linking population data with geographical variables. Remote Sensing and Geographic System Information (GIS) are the most effective research tools. The advance in population spatial distribution during the last two decades based on the Remote Sensing and GIS was introduced by presenting international and internal population spatial distribution projects and methods. Then the typical population estimation results of GPW, UNEP/GRID, LandScan and Chinese 1 kilometer population data were compared and analyzed in a case of Heihe River Basin with the registration statistics data provided by the government. The results show there are a lot of study works developed by the international and internal organizations and researchers. International researches mainly include population estimation from the land cover data, DMSP-OLS night light data or spectrum information directly afforded by the Remote Sensing images. Chinese researchers also make many distinguished works in despite of the later starting. Construction the regression correlation among the population data with land use data and other geographical factors (e.g. elevation, road, and residential area) is the most used method for these works. The comparing results show Chinese 1 kilometer population data has the best estimation accuracy for the Heihe River Basin population spatial distribution.