Point to Ponder “I think there is a world market for maybe five computers.” »Thomas Watson, chairman of IBM, 1943.

Point to Ponder

• “I think there is a world market for maybe five computers.”

» Thomas Watson, chairman of IBM, 1943

Spatial Data Representation

• Maps and GIS are models of reality

– They emphasize some aspects of reality in a cartographic (and database) representation while ignoring or greatly simplifying other aspects of reality.

Spatial Data Representation

• Abstraction

– the process of interpreting what can be sensed from the real world into representational symbols (e.g. the category forest)

Spatial Data Representation

• Data Modeling

– the process of abstraction from the real world for the purpose of representation in a GIS (or other information system)

Spatial Data Representation

Conceptual Model

Data Model

Data Structure

File Structure

Data

Modeling

Levels

Spatial Data Representation• What geographic entities can be identified in this

landscape?

Spatial Data Representation• The geographic entities (things) that are

represented in a GIS depend on:

– worldview/perspective– knowledge– purpose

of the GIS designer and the GIS user

– and the GIS representational structure

Spatial Data Representation

• Geographic entities can be conceptualized as:

– points – lines – areas – surfaces

[volumes]

Spatial Data Representation

• How geographic entities are conceptualized depends on:

– your worldview/perspective– your knowledge– your purpose– your scale of conceptualization

Spatial Data Representation

– your scale of conceptualization

State College in the context of its location in the state (point feature)

State College in the context of its size and shape (area feature)

Spatial Data Representation

• Geographic entities can be spatially (graphically) represented as spatial elements:

– points 0 dimensional– lines 1 dimensional– areas 2 dimensional– surfaces 2.5 dimensional– volumes 3 dimensional

Spatial Data Representation

• Most geographic entities exist in three dimensions through time

• The conceptualization of the geographic entity to be represented governs the choice of graphic representation as a (set of) spatial element(s)

Spatial Data Representation

• Some spatial representation choices are easier than others:– roads are represented as lines on road maps of

Pennsylvania

• Some spatial representation choices are more difficult:– especially the representation of events and

processes (e.g. soils)

Spatial Data Representation

Spatial Data Representation• Cartographic generalization - reducing information

so it can be more clearly cartographically represented

Feature elimination - eliminating portions of features

Spatial Data Representation

Feature smoothing - graphically simplifying features

Spatial Data Representation

• Can distinguish between two different conceptualizations of geographic phenomena:

– object view: empty space ‘littered’ with objects (e.g. lakes)

– field view: a geographic entity that ‘varies’ across a space (e.g. elevation)

Areas within 1 km of hazardous facility in the Philadelphia area

Elevation of the State College area

Object View Field View

Spatial Data Representation

Spatial Data RepresentationObject or field conceptualizations can be spatially represented in different ways, e.g. the elevation field represented as a surface of shading values (left) or as a set of discrete contour lines (right).

Spatial Data Representation

Conceptual Model

Data Model

Data Structure

File Structure

Data

Modeling

Levels

Elevation varies over the surface of the earth

Set of lines that represent elevation (contours)

Vector-Relational

Spatial Data Representation

• Geographic entities that are conceptualized as surfaces can be thought of as:

– discrete: spatial extent of an entity defined by sharp boundaries (e.g. population density per census block group)

– continuous: varies ‘smoothly’ over space (e.g. temperature)

Spatial Data RepresentationPopulation density by county in Pennsylvania changes in discrete steps

Elevation changes smoothly and continuously over space

Spatial Data Representation• Is land cover (forested, urban, agricultural, etc.)

conceptualized with an object view or a field view?

– Is land cover an assemblage of individual areas, each a certain type of cover (object view)? Or is land cover a spatially varying property that can be measured at any point in space (field view)?

• Is land cover spatially continuous or discrete?

Spatial Data Representation

• GIS is a model of the real world

• The representation of the real world in a GIS is determined by: – the conceptualization of the real world– the structure of the GIS database representation

• spatial elements GIS supports

• data structures underlying the conceptual level