Page 255 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

258 Chapter 8

spatial information content in a map D. A N(•) about the transition from fine resolution
N(•) to coarse resolution N(•) represents a threshold N(•) that can be considered the
most suitable N(•). A N(•) that is either much finer or much coarser than the most
suitable N(•) is likely an impractical representation of D. Thus, according to the results
shown in Fig 8-3, the most suitable N(•) is either the coarsest fine resolution N(•) or the
finest coarse resolution N(•). For data-driven modeling of prospectivity for epithermal
Au deposits in the Aroroy district (Philippines), the results suggest that the most suitable
N(•) is 100 m (Fig. 8-3A). For data-driven modeling of prospectivity for epithermal Au
deposits in the Cabo de Gata area (Spain), the results suggest that the most suitable N(•)
is 90 m (Fig. 8-3B). For data-driven modeling of geothermal prospectivity in West Java
(Indonesia), the result suggest that the most suitable N(•) is 400 m (Fig. 8-3C). For data-
driven modeling of prospectivity for alkalic porphyry Cu-Au deposits in British
Columbia, the result suggest that the most suitable N(•) is 800 m (Fig. 8-3D). Based on
the analyses of the graphs in Fig. 8-3, it seems that the most suitable N(•) is
approximately an inflection point in each of the curves of [N(D)] : [N(T)–N(D)] versus
N(•) shown in Fig. 8-2. Provided that it is so, visual inspection of an inflection point in a
curve of [N(D)] : [N(T)–N(D)] versus N(•) is, however, difficult because such a curve is
very smooth (Fig. 8-2). The technique of deriving the curves shown in Fig. 8-3 aids,
therefore, in identification of an inflection point in a curve of [N(D)] : [N(T)–N(D)]
versus N(•) and in selection of a most suitable N(•).
In contrast to and notwithstanding of the results of the analyses illustrated in Figs. 8-
2 and 8-3, the following previous works of GIS-based data-driven modeling of mineral
prospectivity each used a N(•) based on subjective judgment in view of the distance-
probability relation. In data-driven modeling of prospectivity for epithermal Au deposits
in the Cabo de Gata area (Spain), Carranza et al. (2008a) chose and used a N(•) of 100
m, which is within the range of distances in which there is zero probability of one
neighbour epithermal Au deposit location situated next to another epithermal Au deposit
location (Fig. 8-1B) and which is slightly coarser than the most suitable N(•) of 90 m
suggested by the results presented in Fig. 8-3B. In data-driven modeling of geothermal
prospectivity in West Java (Indonesia), Carranza et al. (2008c) selected and used a N(•)
of 500, which is within the range of distances in which there is zero probability of one
neighbour geothermal location situated next to another geothermal location (Fig. 8-1C)
and which is slightly coarser than the most suitable N(•) of 400 m suggested by the
results displayed in Fig. 8-3C. In data-driven modeling of prospectivity for alkalic
porphyry Cu-Au deposits in British Columbia, Carranza et al. (2008b) used a N(•) of 1
km, which is outside the range of distances in which there is zero probability of one
neighbour alkalic porphyry Cu-Au deposit location situated next to another alkalic
porphyry Cu-Au deposit location (Fig. 8-1A) and which is slightly coarser than the most
suitable N(•) of 800 m suggested by the results shown in Fig. 8-3D. In these three case
studies, the common reason for selecting and using a N(•) that is slightly coarser than the
most suitable N(•) suggested by the results shown in Figs. 8-3B to 8-3D is simplicity of