基于特征分区的山东省人口数据空间化研究

Modeling the population density of Shandong province based on population regionalization

相对于人口统计数据,人口空间分布数据更能反映人口的实际分布情况.本文以我国人口密集地区—山东省为研究区,在人口分布特征分区的基础上进行了人口空间数据化处理,并探讨了人口分布特征分区的效果和基于研究结果的山东省人口集中地区(Densely Inhabited District,DID)分布特征.结果表明:经过人口分布特征分区,各分区模型的调整R^2均在0.9以上,人口空间数据总体平均误差值较分区前降低了14.58%,表明该方法在提高人口数据空间化模型精度的基础上能够展现人口空间分布规律.DID分析结果表明青岛市、烟台市、潍坊市、济南市、临沂市等城市有较强的人才、产业、资源的凝聚能力.

Compared with the statistical population data at administrative regions,spatial distribution of population data can reflect the actual distribution of the population better.Taking Shandong Province as the sample,based on population regionalization,spatialization of statistical population was conducted,the precision of the spatial distribution of statistical population data was discussed,and densely inhabited district(DID)of Shandong Province was analyzed.The results showed that after population regionalization,R^2 of partition models were higher than 0.9 and the overall average relative error reduced 14.58%compared with before population regionalization.It indicated that the population regionalization method can improve the accuracy of spatialization of statistical population to reveal the actual distribution regularity of population.The results of DID analysis showed that Qingdao,Yantai,Weifang,Jinan,Linyi had stronger capacity on unity for talent,industry and resources.