论文部分内容阅读
普查数据是地理学空间分析的重要数据源。由于受到数据与计算机处理能力的限制,以往的研究对普查数据空间分析的不确定性未给予足够重视,也未形成成熟的研究方法。在建筑物单元的人口普查数据支持下,本文基于多边形统计数据的可塑面积单元问题(Modifiable areal unit problem,MAUP)特征,设计了一种该类数据空间分析不确定性的研究方法,采用不同的尺度(Scale)及分区(Zoning)系统对多边形的统计数据空间分析的准确性进行了分析。实验引入尺度与形态指数,利用可视化分析和数据拟合的研究方法,对尺度及分区对空间分析结果的影响模式进行了模拟。研究结果表明:(1)以统计小区的空间分析,其结果受统计小区空间形态的影响较大,不确定性强,不能充分反映统计数据本身的空间特征;(2)规则格网能较好地保持原始统计数据的空间分布特征,但仍然受尺度及分区影响;(3)规则格网的空间分析结果及其准确性与尺度有较好的拟合关系,不同尺度下的分析结果不确定性是原始数据不同尺度特征的体现;(4)分区效应受空间分析方法的计算尺度影响,两者共同对空间分析结果产生影响。对于固定尺度的规则格网,其邻接多边形数目是分析结果不确定的主要原因。本文研究结果表明,在多边形统计数据空间分析时,应该对其使用规则格网重新聚合,并根据实际应用的需求选择多尺度分析方法,以达到实际应用目的。
Census data is an important data source for geospatial analysis. Due to the limitations of data and computer processing capacity, the previous studies did not pay enough attention to the uncertainty of spatial analysis of census data and did not form a mature research method. With the help of Census data of building units, this paper presents a method of research on the uncertainty of spatial data analysis based on the features of the polyformal areal unit problem (MAUP) Scale and Zoning systems analyze the accuracy of spatial data analysis of polygons. The scale and morphological index were introduced into the experiment, and the impact of scales and zoning on the spatial analysis results were simulated by means of visual analysis and data fitting methods. The results show that: (1) With the spatial analysis of the statistical community, the results are greatly influenced by the spatial morphology of the statistical community, and the uncertainty is strong, which can not fully reflect the spatial characteristics of the statistical data; (2) The regular grid can be better (3) The results of spatial analysis of the regular grid and its accuracy have good fitting relationship with the scale, the results of the analysis under different scales are not sure (4) The partition effect is affected by the calculation scale of the spatial analysis method, and the two together affect the spatial analysis results. For fixed-scale regular grids, the number of adjacent polygons is the main reason for the uncertainty of the analysis results. The results of this paper show that in the polygon statistical data space analysis, the grid of rules should be re-aggregated, and the multi-scale analysis method should be chosen according to the practical application to achieve the practical application.