| 不同空间权重定义下中国人口分布空间自相关特征分析 |
| |
| 引用本文： | 吴珣,杨婕,张红.不同空间权重定义下中国人口分布空间自相关特征分析[J].地理信息世界,2017,24(2). |
| |
| 作者姓名： | 吴珣 杨婕 张红 |
| |
| 作者单位： | 西南交通大学 地球科学与环境工程学院,四川 成都,611756 |
| |
| 基金项目： | 国家自然科学项目面上项目，国家自然科学基金青年基金项目 |
| |
| 摘 要： | 人口分布的研究对实现人口、环境、资源可持续管理具有深刻意义。相对于洛伦斯曲线、基尼系数等传统度量指标,运用空间自相关分析可以较好地表达人口分布的集聚现象,揭示人口格局的空间结构和空间相互作用。空间自相关测度的关键因素之一为空间权重矩阵。当前人口空间自相关特征测度多采用单一邻近关系定义空间权重矩阵,忽视了不同空间邻近关系对自相关特征分析结果的影响。本文根据距离阈值和邻接关系定义了八种空间权重,包括新定义的高铁两小时可达性空间权重。基于2010年《中国人口年鉴》统计资料,结合GeoDa软件测度了中国人口空间分布自相关特征并分析空间权重定义对自相关特征分析结果的影响。发现:(1)在不同空间邻近关系下,中国省域人口密度分布在全局自相关上均呈现空间正自相关,但其显著性水平有差异;(2)局部自相关分析结果Moran's I表现出明显的区域差异。
|
| 关 键 词： | 中国人口分布 空间自相关 邻近关系 空间权重矩阵 |
Analyzing Spatial Autocorrelation of Population Distribution in Different Spatial Weights: A Case of China |
| |
| WU Xun,YANG Jie,ZHANG Hong.Analyzing Spatial Autocorrelation of Population Distribution in Different Spatial Weights: A Case of China[J].Geomatics World,2017,24(2). |
| |
| Authors: | WU Xun YANG Jie ZHANG Hong |
| |
| Abstract: | The study on population distribution contributes greatly to the sustainable management of population, environment and resources. Compared to conventional measures such as Lorenz curve and Gini coefficient, spatial autocorrelation analysis enables better expressing the aggregation of population distribution, as well as revealing the spatial structures and interactions of population distribution patterns. The spatial weight matrix is one of the key factors impacting on the measurement of spatial autocorrelation. Currently, however, the spatial autocorrelation is measured mostly through a spatial weight matrix featuring single proximity relation, which the influences of different spatial proximity relations on the autocorrelation characteristic analysis are ignored. This paper defines eight spatial weight matrices according to the distance thresholds and the proximity relations, including the spatial weight matrix based on 2 h accessibility of high-speed railway. The spatial autocorrelation characteristics of the population distribution in China are measured and the influences of spatial weights on the autocorrelation characteristic analysis are analyzed by using the GeoDa software with the Almanac of China's Population (2010). Results show: (1) In different spatial proximity relations, the provincial population density distributions all assume positive spatial autocorrelation in a global view, but the significance levels are varies; (2) In local autocorrelation analysis, Moran's I exhibits great difference. |
| |
| Keywords: | China population distribution spatial autocorrelation proximity relation spatial weight matrix |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
