Examples for each type of geographic information.
ware. Storage of positions is the most common, as locations usually best cor-
respond to the cartographic representation of a map, but certain fields are
more likely to use networks (transportation scheduling and routing) or fields
(environmental modeling). The storage of GI usually approximates GI repre-
sentation types, but when not, it is often possible to transform GI between
various representation types.
Its ability to transform GI is the underlying reason why GIS has become
such a worldwide success for so many human endeavors. It is always possible
to transform one representation type to another using a GIS. For example,
cities shown as points in a position representation (Figure 9.1) can be trans-
formed into areas in a field representation. The transformation into a field
can also take into account that cities have fuzzy boundaries, not the sharp
edges of a point.
The type of GI representation has consequences for quality and accu-
racy. Transforming the GI representation of cities as points into areas in a
field may make it possible to show how cities diffuse into the surrounding
area, but cities, even in a fuzzy form, are never perfectly round like the circle
that represents them at small scales. Generally, we can distinguish between
intra- and interunit qualities in relationship to the geographic information
type. Intraquality describes how well the differences between properties of
units are represented—for example, population of cities in the categories less
than and greater than 100,000 or with the exact count of the population.
Interunit quality refers to the reliability that things and events are accurately
represented—for example, the extent of a forest or a marsh. The boundaries
of a wetland or city created at large scale will be much more accurate than a
small-scale state map showing the location of wetlands and cities.
GI Representation Types
The three types of GI representation refer to concepts used by most GIS to
represent things and events. Each representation type uses specific storage