Page 36 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

32 Chapter 2

In order to achieve the useful functions of a spatial database, data are handled through a
database management system (DBMS), which is a collection of software for creation and
maintenance of a database. Creation of a spatial database takes a major proportion of
resources in a GIS-based mineral exploration project. There are four major stages in
database creation (Oxborrow, 1989).
ƒ Data investigation. The task in this stage is to consider, according to needs or
objectives of a project, the type, quantity and qualities of spatial data to be
included in a database.
ƒ Database modeling. The task in this stage, which can be performed independently
of the software used, is to develop a conceptual or logical database model by
analysing characteristics of entities and relationships between entities and
attributes.
ƒ Database design. The task in this stage, which depends on the software and its
data model, is to create a practical design for a database. That means, the logical
design of a database produced in the data modeling stage is translated into a
practical design using a DBMS.
ƒ Database implementation. The task in this stage is to populate the database with
attribute data. This is followed always by monitoring and upkeep, including fine
tuning, modification and updating of attribute data.
The database models that have been used in different GIS applications are the
hierarchical, network, relational and object-oriented models. The hierarchical model
was useful in geochemical applications around the mid-1980s (Mattiske, 1983). The
hierarchical and network models have some importance in GIS applications to
geological studies (e.g., Lopez-Blanco and Villers-Ruiz, 1995; Wielemaker et al., 2003).
The object-oriented model became an emerging trend and topic of many researches in
GIS applications to geological studies (e.g., Alms et al., 1998; Breunig et al., 1999). The
relational model is now predominantly used in GIS and is most widely used in GIS
applications for geological exploration (e.g., Schmatz et al., 1995; Laxton and Becken,
1996; Colman-Sadd et al., 1997; Pouch, 1997; Diepenbroek et al., 2002). The relational
database model is given some supplementary treatment here; the other database models
are not discussed further.
The relational database model was developed by Codd (1970) based on the relational
algebra. In such a database model, data and relations between them are organised in a 2-
D structure (i.e., table). The terminology of relational databases as compared to the
terminology of traditional tables or simple spreadsheets is given in Table 2-III. Each
table represents data for one entity or theme (e.g., stream sediment geochemical data).
Each row or tuple is a record of different properties of a geo-object (e.g., a sample). Each
column represents an attribute or property. The order of tuples or attributes is trivial as
long as the contents of a table are not changed or the significance of relations is not
altered. Each cell (i.e., intersection between a row and a column) can have only one
value or is left blank if values are unknown or unmeasured. Where relations exist
between different data sets, tables representing such data sets are linked by certain