从近年来兴起的网格(grid)技术和信息网格(inform ation grid)出发,研究地球空间信息(Geo-spatialinform ation)领域中如何在网格环境下实现从数据到信息再到知识的升华以及基于网格的空间信息服务。提出广义空间信息网格和狭义空间信息网格两个层次的概念。广义空间信息网格是指在网格技术支撑下空间数据获取、更新、传输、存储、处理、分析、信息提取、知识发现到应用的新一代空间信息系统。狭义空间信息网格则指在网格计算环境下新一代地理信息系统,是广义空间信息网格的一个组成部分。文中试图给出这两个层次的空间信息网格的定义、任务及其组成。并对需要解决的关键问题进行一些初步的探讨。

This paper discusses how to implement the procedure of information extraction from data and knowledge discovery from information in the grid computing environment, from the point of view of grid technology and information grid. Two hierarchical concepts, namely, generalized spatial information grid and special spatial information grid are proposed by the author. The former is defined as a new generation spatial information system embedded in the environment of grid technology where spatial data acquiring, updating, transferring, storing, processing, analyzing, information extracting and knowledge discovering are carried out. In order to do so, it is believed by the author that at least four functional components are needed: a smart sensor web, an intelligent data-information-knowledge processing system based on grid computing, a new generation GIS suitable to grid computing environment and Intelligent geo-service agents. The latter refers to the new generation of geographic information system in the environment of grid technology. Since conventional GISs have being developed for single person computers, four challenges caused by four types of inconsistencies are arisen when such systems are transferred to computing grid environment, they are: Challenges caused by temporal datum inconsistency, Challenges caused by geographic datum inconsistency, Challenges caused by data format inconsistency and Challenges caused by semantic inconsistency. Geospatial data representation should be re-considered if grid technology is concerned. An innovative concept, Spatial Information Multi-Grid (SIMG), is proposed by the author and regarded as the backbone characteristics of specialized spatial information grid. The core feature of SIMG is that demarcating the studied area into different level of grids. These grids are different in size and all attribute items describing them are various and a hierarchical structure is formed where coarse grids contains several finer grids. This paper tries to put forward the definition, tasks and components of these two concepts. Some tentative approaches for solving key problems arising from them are expounded. (