Page 37 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

Spatial Data Models, Management and Operations 33

TABLE 2-III

Relational database terminology.

Analogue version Digital version Relational database
Table File Relation
Row Record/case Tuple
Column Field Attribute
Number of rows Number of cases Cardinality
Number of columns Number of fields Degree
Unique ID Primary key Index
Possible values Domain
(Adapted from Date, 1990)

common attributes designated as keys, through which retrieval, search or query (or so-
called join operations) of data from one table to another is made possible and efficient.
Attributes designated as keys must have entries that are unique; however, there is no
standard to predefine keys. In GIS, location (i.e., coordinate pairs) is usually but tacitly
designated as primary key. On the one hand, for data or tables pertaining to point geo-
objects such as sample locations, unique identifiers of each location (e.g., sample
number) can be designated as key. On the other hand, for data or tables pertaining to
linear or polygonal geo-objects, unique identifiers to be used as keys must first be
created and assigned to each linear or polygonal geo-object. The assignment of unique
identifiers to geo-objects usually takes place during spatial data capture and is
considered during the database modeling stage of spatial database creation. This allows
for an efficient link between maps of different types of geo-objects and their
corresponding attribute tables and therefore allows for efficient analysis of relationships
between geo-objects of interest through suitable spatial data operations. Extensive
explanations about operations on relational databases (or tables) of geoscience
exploration data and their relationships with maps can be found in Bonham-Carter
(1994).

OPERATIONS ON SPATIAL DATA

The focus of the following introductory discussions is on generic concepts of certain
operations rather than on algorithms behind such operations, which could vary in vector-
or raster-based GIS software packages.

Spatial query and selection
In the early stages of spatial data analysis, it is common to (a) visualise and query the
spatial database and then (b) select locations of spatial data or geo-objects of interest.
Fig. 2-6 shows an example how spatial data or information can be visualised and
queried. Clicking the cursor on any portion of a map view would indicate data attributes
at specific locations, meanwhile an interactive table view could indicate map coordinates