Page 194 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

196 Chapter 7

or incomplete and/or when the accuracy or resolution of available spatial data is poor.
We now turn to the individual techniques for knowledge-based binary representation and
integration of spatial evidence that can be used in order to derive a mineral prospectivity
map.

Boolean logic modeling

Ample explanations of Boolean logic applications to geological studies can be found
in Varnes (1974) and/or Robinove (1989), whilst examples of Boolean logic applications
to mineral prospectivity mapping can be found in Bonham-Carter (1994), Thiart and De
Wit (2000) and Harris et al. (2001b).
In the application of Boolean logic to mineral prospectivity mapping, attributes or
classes of attributes of spatial data that meet the condition of a prospectivity recognition
criterion are labelled TRUE (or given a class score of 1); otherwise, they are labelled
FALSE (or given a class score of 0). Thus, a Boolean evidential map contains only class
scores of 0 and 1. Every Boolean evidential map has equal weight in providing support
to the proposition under examination; that is because in the concept of Boolean logic
there is no such thing as, say, “2×truth”. Thus, the class scores of 0 and 1 in a Boolean
evidential map are only symbolic and non-numeric.
Boolean evidential maps are combined logically according to a network of steps,
which reflect inferences about the inter-relationships of processes that control the
occurrence of a geo-object (e.g., mineral deposits) and spatial features that indicate the
presence of that geo-object. The logical steps of combining Boolean evidential maps are
illustrated in a so-called inference network. Every step, whereby at least two evidential
maps are combined, represents a hypothesis of inter-relationship between two sets of
processes that control the occurrence of a geo-object (e.g., mineral deposits) and/or
spatial features that indicate the presence of the geo-object. A Boolean inference
network makes use of set operators such as AND and OR. The AND (or intersection)
operator is used if it is considered that at least two sets of spatial evidence must be
present together in order to provide support to the proposition under examination. The
OR (or union) operator is used if it is considered that either one of at least two sets of
spatial evidence is sufficient to provide support to the proposition under examination.
Boolean logic modeling is not exclusive to using only the AND and OR operators,
although the other Boolean operators (e.g., NOT, XOR, etc.) are not commonly applied
in GIS-based knowledge-driven mineral prospectivity mapping. The output of
combining evidential maps via Boolean logic modeling is a map with two classes, one
class represent locations where all or most of the prospectivity recognition criteria are
satisfied, whilst the other class represents locations where none of the prospectivity
recognition criteria is satisfied.
For the case study area, Boolean evidential maps are prepared from the individual
spatial data sets according to the prospectivity recognition criteria given above. Fig. 7-4
shows the inference network for combining the evidential maps based on the given
spatial data sets. The Boolean evidential maps of proximity to NNW-trending
faults/fractures and proximity to NW-trending faults/fractures are first combined by