The geometry and location of objects in world space must be modeled and stored. There are three components to this:

- spatial concepts,
- a geometric model, and
- a spatial data structure.

Spatial concepts involve how the user thinks about space and what kinds of spatial objects are used. For example, one may think of the space as a 2D plane, with data available at every point on the plane, or at least at certain regular intervals. One can assume a regular grid and fit the data to that grid. An alternative would be irregular objects, each associated with some property. Other users might think in terms of political boundaries. Some applications require a 3D space.

A geometric model is a set of formalized concepts that are used to describe a particular application. Spatial data structures, and their associated operations, are implementations of geometric models. There are three major groups of geometric models:

- cartographic models, that are oriented towards rendering in map form
- regular spatial models which use a regular grid and are called "raster models"
- irregular spatial models that model real world objects and describe them using coordinates and are called "vector models".

A research area is the cognitive aspects of space. GIS usually use a 2D Euclidean space, which is not how most people perceive space, e.g., in a valley. So it is important to not only model the physical space but to understand and model the cognitive space of the users. This might be different for different users.

GIS: Table of Contents

HyperVis Table
of Contents