A geographical information “system”
is composed of a combination of software, data, operators and users, processing
and treatments which underlie a certain organization.
We will discuss here the technical fundamentals of
the system: the management (capture, storage, extraction), the analysis
(interrogation and processing), the communication (map production, plans and
reports), and finally the simulation (modeling and production of scenarios ).
As technical fundamentals we include the definition
of a series of words that take on a particular meaning when it comes to GIS.
The geographical entities
A system manages, by definition, customized
entities: a banking system will have the bank accounts as entities, the Health
Insurance manages the entities “social insured”, etc. In the case of a geographical information system the entities will be the
objects of the same type (buildings, roads, hydrography, monuments, …).
These entities will be grouped into
sets called information layers.
Compared to other types of management
of geographical objects (CAD-CAD, …) the geographical objects consist of
three different conceptual levels:
- the geometric level: the object is represented as a point, a line, a
surface, a volume. Each of these representations is created
by one or more points that constitute the object, each of these points,
necessarily, having a location in the space (geographical coordinates). - the attribute level: each object has alphanumeric information describing
its characteristics (name of the commune, address, type of building, …) - the relational level: the geographical objects maintain spatial and
semantic relations between them. The spatial relations are of the type building in plot, plot in commune,
plot on the right side of the road, river crosses the commune, etc. This is called topological relations. Semantic relationships may include land belonging
to an owner, areas with special protection, and so on. At finer levels, there are other types of
GIS-specific relationships such as aggregation (of communes forming a department)
or inheritance (each section of a watercourse inherits the name of the
watercourse). .
Real world representations
The set of geographical objects can be
represented as two main forms: the image form (raster) or the vector form
(point, line, polygon).
- the raster mode: the real world is represented by a series of
regular pixels each containing one or more values that characterize the
object located in this pixel. The most
common raster data are aerial photographs, scanned maps or plans, digital terrain
models or satellite images. Thus, the
real world is represented by a regular matrix of cells where each object
corresponds to a cell or a group of adjacent cells. - the vector mode: each object is represented in an identified way and
independent of other objects. It is
represented by a point, a line, a polygon or a volume having a unique
identifier. This unique identifier is also
present in the attributes table and makes it possible to link the geometry of
the object with its attribute properties.
The GIS data
There are two major types of data in a GIS:
- the reference data
- the proper or business data
The reference data will help you, when creating new
data, to position them accurately in the real world. This “accuracy”
will be relative to the chosen reference. The production of the most common
geographical references is the responsibility of public institutions such as
IGN (National Geographic Institute), INSEE (National Institute of Statistics
and Economic Studies), DGFIP (General Directorate of Public Finance) , MEDDE
(Ministry of Ecology, Sustainable Development and Energy), the BRGM (Bureau of
Geological and Mining Research). At municipal level, the two most frequently
used digital map funds are the cadastral plan and the orthophotoplan
(orthorectified aerial photography).
Relying on a given reference source makes it possible to ensure the consistency
of the data produced and the possibility of exchanging data with other
partners. But you should know that the different reference sources are not
always completely consistent among them: the positioning accuracy of each one can
be different and the objects present depend on the rate of update of each source.
In any case, it must be acknowledged that efforts have been made in recent
years to bring together all the available references.
Business data may have an external origin (service providers, thematic
data providers, etc.) or internal, in the latter case, special attention should
be given to metadata information. These famous metadata will be the subject of
the next article in this series.
