This guide to creating a dynamic GIS data model helps database managers design a schema that has comprehensive and descriptive query definitions, a user-friendly cartographic display, and increased performance standards. The five steps for taking a data model through its conceptual, logical, and physical phases, including modeling the user's view, defining objects and relationships, selecting geographic representations, matching geodatabase elements, and organizing the geodatabase structure are studied in detail. A look at nine decision points that deal with concerns common to all data modeling exercises, such as validating feature geometries, modeling linear networks, managing raster data, and labeling map features help database managers fine-tune their GIS data models. Several design models for a variety of applications are considered including addresses and locations, census units and boundaries, stream and river networks, and topography and the base map.
David Arctur is a product specialist at ESRI and is part of the geodatabase development team. He has written for the Earth Observation & Mapping Magazine, IEEE Computer Magazine, and International Journal of Geographical Information Science. He lives in Austin, Texas. Michael Zeiler is a technical writer and data model specialist at ESRI with experience in diagramming GIS concepts, building data models, and programming. He is the author of Exploring ArcObjects, Inside ARC/INFO, and Modeling Our World. He lives in Santa Fe, New Mexico.