Page 197 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

Knowledge-Driven Modeling of Mineral Prospectivity 199

the predicted non-prospective areas. The significance of this performance of the Boolean
epithermal Au prospectivity map of the case study area can be appreciated by comparing
it with the performances of other mineral prospectivity maps derived via the other
modeling techniques.

Binary index overlay modeling

In binary index overlay modeling, attributes or classes of attributes of spatial data
that satisfy a prospectivity recognition criterion are assigned a class score of 1;
otherwise, they are assigned a class score of 0. Therefore, a binary index map is similar
to a Boolean map, except that the values in the former are both symbolic and numeric.
So, a binary index map is amenable to arithmetic operations. Consequently, each binary
evidential map B i (i=1,2,…,n) can be given (i.e., multiplied with) a numerical weight W i
based on ‘expert’ judgment of the relative importance of a set of indicative geological
features represented by an evidential map with respect to the proposition under
examination.
The weighted binary evidential maps are combined using the following equation,
which calculates an average score, S, for each location (cf. Bonham-Carter, 1994):

n
W
¦ i B i
S = i (7.1)
n
W
¦ i
i

where W i is weight of each B i (i=1,2,…,n) binary evidential map. In the output map S,
each location or pixel takes on values ranging from 0 (i.e., completely non-prospective)
to 1 (i.e., completely prospective). So, although the input maps only have two classes,
the output map can have intermediate prospectivity values, which is more intuitive than
the output in Boolean logic modeling. Examples of mapping mineral prospectivity via
binary index overlay modeling can be found in Bonham-Carter (1994), Carranza et al.
(1999), Thiart and De Wit (2000) and Carranza (2002).
Assignment of meaningful weights to individual evidential maps is a highly
subjective exercise and it may involve a trial-and-error procedure, even in the case when
‘real expert’ knowledge is available particularly from different experts. The difficulty
lies in deciding objectively and simultaneously how much more important or how much
less important is one evidential map compared to every other evidential map. This
difficulty may be overcome by making pairwise comparisons among the evidential maps
in the context of a decision making process known as the analytical hierarchy process
(AHP). The concept of the AHP was developed by Saaty (1977, 1980, 1994) for
pairwise analysis of priorities in multi-criteria decision making. It aims to derive a
hierarchy of criteria based on their pairwise relative importance with respect to the
objective of a decision making process (e.g., evaluation of the mineral prospectivity
proposition). Most GIS-based applications of the AHP concern land-use allocations (e.g.,