Page 33 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

Spatial Data Models, Management and Operations 29

nominal or categorical scales of measurements. Qualitative variables, unlike quantitative
variables, usually have to be represented numerically before they can be integrated in
mathematical operations. For example, lithology, which is a qualitative variable, can be
integrated quantitatively in the estimation of local background by representing it
numerically as an areal proportion of a drainage sample catchment basin (see Chapter 5).
A simple numerical representation of qualitative variables is the assignment of discrete
values in either binary or ternary scales of measurement according to a particular
proposition. For example, for a proposition that “this site contains a mineral deposit”,
lithologic units that are unfavourable and favourable host rocks according to genetic
models of the deposit-type sought can be assigned a value of [0] and [1], respectively.
Other examples of types of numerical representations of qualitative variables (but also of
quantitative variables) are fuzzy membership and probability, which range in the interval
[0,1] reflecting degrees of non-ambiguity and certainty, respectively, with respect to a
proposition (see Chapter 7).

MANAGEMENT OF SPATIAL DATA
In a GIS, management of spatial data is concerned with (1) storing data in the
computer (i.e., spatial data capture) and (2) organising data in the computer (i.e., spatial
database creation). Management of spatial data takes a major proportion of resources
(personnel, time and money) in any GIS-based project.

Spatial Data Capture
The first fundamental step in spatial data capture is to choose a coordinate system,
into which all geo-objects or data are geographically-registered or georeferenced. A
coordinate system consists of a spheroid (or an ellipsoid) representing the Earth’s
surface and a map projection to convert spherical or geographical coordinates (latitudes,
longitudes) to planar or map (metric) coordinates. The choice of an appropriate
coordinate system can benefit from the authoritative discussions on spheroids and map
projections given by Maling (1992) and Snyder (1993). On the one hand, the choice of
an ellipsoid depends on global surface curvatures, such that for every region or country
there is a commonly used ‘best fit’ ellipsoid (Table 2-I). On the other hand, the selection
of a map projection depends on (a) geographic position of region or country, (b) size and
shape of region or country where a study area is situated and (c) requirements or
objectives of the study. These three factors must be considered together if the primary
aim is to obtain minimum geometric distortions in terms of either shape or area.
There are different types of map projections and each map projection creates
geometric distortions but guarantees a known relationship between locations on a map
and their true locations on the Earth. It is essential to use a map projection because
geographical coordinates are not planar coordinates and most spatial data are visualised
as 2-D features using planar coordinates. Although spatial data can be stored and
manipulated using geographical coordinates, storing spatial data using map projections