空间数据库已经成为实现海量空间数据管理的高效手段，在空间数据库设计和实现中，定义GSQL是其关键。目前，针对于矢量数据访问已经有许多学者进行了研究，但对于地理栅格数据，其访问和操作接口却尚无相关规范，为此参考OpenGIS，对该方面进行了研究，即首先定义了与栅格相关的抽象数据类型，其分别为Pixel(栅格点)、RasterRegion(栅格区域)、RasterCoverage(栅格覆盖)，它们之间存在聚合关系；然后对每种数据类型进行了定义，包括其数据对象以及相关操作；进而，基于上述抽象数据类型，给出了支持栅格数据访问的GSQL—R描述以及相关实例。通过实例表明，GSQL—R能较好支持矢量／栅格一体化的数据访问和操作。

Spatial database has become one of the dominating techniques tomanage spatialdata in geographical information systems, and geographic structured query language (GSQL) is the key issue ofspatialdatabase design and implementation. Atpresent, GSQL for vector data has been widely studied. However, when considering the raster data, little researchwas made for their specifications on visiting and operationa.l Based on OpenGIS○R, a GSQL extension named GSQL2R supporting raster data is discussed in this paper. At first, some abstractdata types (ADTs), includingPixe,l RasterRegion and RasterCoverage, are defined. The relationships between Pixel/RasterRegion and RasterRegion/RasterCoverage are part/whole. In these definitions, data objects and operations for each ADT are described in a formalized way. Then, GSQL2R syntax and some instances are presented. The GSQL2R description includes three parts, .i e. data definition language (DDL), data table schema and data manipulation language (DML). The example indicates that GSQL2R provides an efficient and integrated way for raster data access and operation.