Enhanced Data Models for Geographic Information systems

Abstract

Geographic Information Systems (GISs) are becoming essential tools for analyzing and newlinegeographically transferring the information about the world. GIS incorporates, both newlinespatial and non spatial data, and has the capability for searches, overlays, association newlineand it also displays output in 2D/3D. In addition to spatial data GISs contain georeferenced newlineinformation which enables near true representation of geography and newlinecapability to provide measurements, variations etc. These features make the technology newlinedifferent from the Computer Aided Design (CAD) systems and conventional database newlineapplications. The capabilities of GIS have brought a paradigm shift in the approach to newlinethe presentation of information. newlineIn order to enhance the outcome of the mentioned capabilities the GIS data representation newlineand storage should be appropriate. This demands proper database designing, leading to the newlinechoice of suitable data model, for the effective results. Various spatial data models at newlineconceptual and logical levels are available for the data representation and storage. Literature newlineshows research gaps at both the levels. At the conceptual level representation of network and newlinetopological geo-primitives for relational databases are not available and at logical level newlinegeometric and topological inconsistencies have not been taken care of appropriately. newlineThe present research has evolved an approach to represent spatial data at conceptual level and newlineto design vector data storage model at logical level. Spatial data representations have been newlineinvestigated further under two important aspects, namely geometric and topologic consistencies newlineusing the developed Terzatto tool. newlineAt conceptual level Topo-Net Spatial Entity Relationship Model has been proposed by newlineenhancing the representations in the existing spatial entity relationship model by introducing newlinerepresentations for topological and network geo-primitives being required for GIS applications.