A GIS consists of a data base coupled with a graphics output.
There are three major types of operations in a GIS:
The user will want to access data elements with certain properties. There are three questions involved here. The first is whether the requested data is available in the system. The second question is how long it will take to find the data. Spatial data structures pose unusual indexing problems and are a topic unto themselves Usually some variation of a quadtree (for 2D) or octree (for 3D) is used. The third question is how the user accesses the data. The conventional query language SQL is sometimes used.
Analytical operations allow the user to combine data to produce new data. The most important operation is the overlay of different spatial data sets to produce a new data set. For example, a biologist might want to determine what variables affect the population of dolphins. For a given part of the ocean, there may be measurements for water temperature, different fish populations, including predators (sharks), different types of pollution levels, etc. The biologist could then try to find correlations. Note that this is a multivariate visualization problem. The analytical operations can also include computations to produce a new data set.
Rendering operations show the data as some sort of map with the additional information displayed. This is where Visualization techniques can be applied to GIS.
Last modified on March 02, 1999, G.
Scott Owen, firstname.lastname@example.org
GIS: Table of Contents
HyperVis Table of Contents