Page 307 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

310 Chapter 8

predictive modeling of volcanic lahar-inundation zones based on the same sets of
predictor variables.
In closing, two recommendations can be made for further studies. Firstly, further
research is required in the proper representation and evaluation of uncertainties
associated with predictive models of geochemical anomalies and/or mineral
prospectivity. Although the application of EBFs is useful in representation of evidential
uncertainty, further studies are needed in monitoring the propagation of uncertainty from
the input maps to the output map(s). This is a challenging task because, for example,
most, if not all, geological maps invariably do no contain information about their
accuracy. Secondly, although cross-validation strategies allow us to estimate empirically
the likelihood of discovery of a new deposit-type location within the predicted
prospective zones in a study area and although mineral prospectivity models can be used
to estimate undiscovered mineral endowment (e.g., McCammon and Kork, 1992),
further research is required to answer the following logical question regarding the
efficacy of a mineral prospectivity map: “Where, in the predicted (most) prospective
zones, should targets for further exploration of undiscovered deposit-type locations be
focused?”. This is also a challenging task because, after all, the ultimate goal and thus
real validation of mineral prospectivity mapping is finding undiscovered mineral
deposits. Thus, because methods for mapping mineral prospectivity are now mostly
well-established, scientific progress in developing new techniques for improving the
accuracy and utility of mineral prospectivity maps is more desirable than scientific
progress in developing new techniques for creating and integrating predictor maps for
mineral prospectivity mapping.