基于综合信息元的多源数据集成

Multiple Data Integration based on the Integrated Information Units

基于地理信息系统的空间分析模型一般采用传统空间叠加方法进行数据集成 ,这一方法在具体操作过程中叠加产生的单元在大小和数目上具有很大的不确定性 ,同时还会产生大量无意义的细小图斑 ,影响空间分析模型的有效实施 .本文引入综合信息元作为地理信息系统空间数据采集、存储、管理以及空间分析模型的基本单元 ,它在一定研究目的和空间尺度下具有相对稳定的大小、均一的内部属性和清晰的边界 .针对不同的相关因子以及数据类型 ,综合信息元可以采用数据叠加法、综合法和网格划分法等进行构建 .一旦综合信息元得以确定 ,便可以此为基础 ,进行多源数据的集成 ,为空间分析模型的实施提供数据准备 .本文进一步探讨了ARC/INFO环境下基于综合信息元的多源数据集成方法及其程序实现 .结果表明 ,这一集成方法在精度满足分析要求的同时 ,集成效率大大提高 ,同时与传统数据叠加方法相比 ,不但单元数目明显减少 ,而且单元边界相对固定 。

The conventional overlay methods used to perform spatial analysis supported by GIS were found to have shortcomings in variability and uncertainty in unit scales associated with a large number of spurious polygons. In this paper, the concept of integrated information units (IIU) was introduced as primary units resulted from certain rules and marked by steady scales with characteristics of homogeneous attributes, clear boundaries and adaptation to variations of scales and relevant geographic features. The integrated information units could be identified by a variety of methods according to different data types and issues including overlay operation for Vector GIS, cell partition for Raster GIS, and for some cases, synthetical consideration based on special treatment or expert knowledge. After the identification of IIU, the data from different sources or scales should be integrated in terms of proper way so as to perform the spatial analysis efficiently. 14 data integration methods were recommended and a comprehensive program was presented under the environment of ARC/INFO. It was shown that this method not only fulfills the requirements in data precision, but also improves the data integration processes significantly. Meanwhile, it has advantages of providing invariable unit number and steady boundary in the modeling analysis and averting the problems associated with spurious polygons.