具有地理参考特征的信息是社会信息的主要组成部分，统计数据的空间化对统计数据空间分布特征的表现、多源数据融合、基于统计数据的空间分析与决策有比较重要的意义，是当前空间信息科学的重要研究任务之一。统计数据空间化的目标在于对统计对象空间特征的还原，将不同空间分布特征的统计数据根据其空间分布模式直观地反映到地理基础框架下。因此，建立一套标准的可以被推广使用的统计数据空间化方法体系，为各种统计数据，如人文数据、社会经济数据以及土地利用／土地覆被等自然要素数据的集成提供一个基础信息平台，使不同来源、不同用途、不同专题的统计数据均可以在这一公共的基础平台上实现信息融合，具有十分重要的现实意义。本文在国内外相关工作的基础上，以北京市朝阳区乡镇统计数据为研究对象，对统计数据空间化模型和方法进行了研究。首先，介绍了统计数据指标体系及空间分布模式的基本理论，将研究区的统计指标划分为点状、线状、离散面状、连续面状四种分布模式；其次，介绍了当前主要空间化模型，包括面域加权、点与多边形叠加、基于遥感影像的二值化方法、层次分析法等，并依据不同模型的特点构建了一套统计数据空间化方法体系；最后，应用本文提出的方法体系，对北京市朝阳区22个乡镇的2002年统计数据进行试验。本论文在归纳出北京市朝阳区乡镇统计数据的四种空间分布模式的基础上，将各分布模式同与之适应的空间化模型对应起来，总结出了一套相对完善的统计数据空间化方法体系，并成功地将该方法体系运用到试验中，实现了统计数据从行政单元到规则网格中的转换。实验结果表明本文提出的方法体系能够比较有效的实现统计数据的空间化，与传统的统计数据空间表达方式相比，空间化后的数据更好的反映了统计数据在空间上的连续性与渐变性，同时具有更详细的空间尺度。

The information with geographical features is the main part of the information for our society. The statistical data spatialization plays an important role in the presentation of statistical data’s spatial attributes, the multi-sources data integration, and the spatial analysis and decision-making, its goal is to show the spatial characteristics of the statistical data, and display them in the geographical basis of the framework in accordance with their spatial distribution pattern. Therefore, to provide a basis information platform for a variety of statistical data, such as the humanities data, socio-economic data , land use / land cover and other natural elements , the establishment of a system of statistical data spatialization methods is of great practical significance.This paper explores methods of statistical data spatialization based on Domestic and international research, aiming at Chao Yang District in Beijing. First, the theory of spatial pattern are presented, including point, line, continuous surface, discrete surface. Second, the spatial patterns of statistical data of Chao Yang District in Beijing are constructed on the above theory. Third, several spatialization models are introducted, such as areal weighting; point-in-polygon; line-in-polygon; AHP, and establish a system of statistical data spatialization methods . At last, the method system is tested with the statistical data of 2002 from 22 towns in Chao Yang District in Beijing.This paper meets the four spatial distribution patterns of statistical data with different spatialization model, constructing a relatively comprehensive system of statistical data spatialization methods. The experiment shows that the method system is effective for the statistical data of the study area. According to the original data, the spatialized data has a more evident continuity and gradualness and a more detailed spatial scale.