【摘要】 针对传统的人口空间化建模辅助因子单一、方法选择不够准确、结果精度低等问题,该文以长沙市为例,提出了利用NPP/VIIRS DNB夜间灯光数据、土地利用数据、DEM数据和POI数据作为建模辅助因子,通过MLR模型、GWR模型和SLM模型,建立了该区域1 km×1 km格网人口空间化模型,并比较了模型的优缺点。结果表明:1)三种模型的决定系数R2分别为0.89、0.90、0.92,均通过了显著性检验(P<0.01); 2)三个模型的模拟结果人口分布相似,即在经济发达的城区人口趋向聚集状态,而在地形地势较复杂的农村地区人口较分散; 3)通过乡镇级人口统计数据对三种模型空间化结果进行比较以及误差检验得到基于GWR模型的模拟结果精度最高,其次是SLM模型,MLR模型精度最低。更多还原

【Abstract】 Aiming at the problems of single auxiliary factor of traditional population spatialization modeling,insufficient selection of methods,and low accuracy of results,this article takes Changsha as an example and proposes the use of NPP/VIIRS DNB night light data,land use data,DEM data and POI. The data is used as a modeling auxiliary factor. Through MLR model,GWR model and SLM model,a spatialized population model of 1 km×1 km grid in this area is established,and the advantages and disadvantages of the models are compared. The results show that: 1) The coefficients of determination R2 of the three models are 0.89,0.90,and 0.92 respectively,which have passed the significance test( P<0.01); 2) The simulation results of the three models have similar population distributions,that is,in economically developed urban areas the population tends to be agglomerated,and the population is more dispersed in rural areas with more complex topography; 3) The spatialization results of the three models are compared through the township-level demographic data and the error test shows that the simulation results based on the GWR model have the highest accuracy,followed by SLM model,MLR model has the lowest accuracy.更多还原