Page 214 - Introduction to Mineral Exploration
P. 214

9: MINERAL EXPLORATION DATA  197



                  55,000






                  50,000






                  45,000
                                  5
                                         km
                        N
                       250,000     255,000      260,000     265,000     270,000      275,000     280,000


                                                         Probability                     0.40–0.55
                            Legend
                                                                no data                  0.55–0.65
                                     Prospect
                                     Coast                      < 0.25                   0.65–0.75
                                                                0.25–0.40                >0.75

                 FIG. 9.16  Probability map resulting from a fuzzy logic analysis. This analysis highlights any layer which has
                 possible association with gold mineralisation and is the result of a fuzzy “or.” This should include all
                 occurrences. (Source: Wang 1995.)

                 Brown et al. (2000) compared the method with  9.3  INTEGRATION WITH RESOURCE
                 fuzzy logic and weights of evidence on data      CALCULATION AND MINE PLANNING
                 from an area in New South Wales and found        SOFTWARE
                 that it gave superior results. They claim that
                 although neural networks require a training  Most resource calculations performed in explo-
                 set the number of occurrences required is not  ration, in feasibility studies, or in routine mine
                 as onerous as that of the weights of evidence  grade control and scheduling use a specialized
                 method.                                      computer package (or packages) that deals with
                   There have been a number of other studies  3D data. At the early stages of exploration, the
                 of using complex methods in integrating data  key features of the package will be to input drill
                 and the reader should examine some of these  information and relate this to surface features,
                 and decide which approach is suitable. A good  as discussed in Chapter 5 and section 10.4.
                 starting point is the volume of Gubins (1997).  When a resource is being calculated, the ability
                 The paper by Chinn and Ascough (1997)        of the package to model the shape of geolo-
                 details the effort of Noranda geologists in the  gical units and calculate volumes and tonnages
                 Bathurst area of Canada. They used a fuzzy   becomes more important. Subsequently in
                 logic approach to define areas for follow-up  feasibility studies (see section 11.4.4) the capa-
                 but concluded that the main advantage of     city to design underground or surface workings
                 computer-based methods was to get their      and schedule production is crucial.
                 deposit experts to sit down together and put   The package chosen will depend on the
                 their experience on paper. The methods are   finance available and the nature of the opera-
                 therefore not a substitute for excellent geology  tion, as there are packages specifically designed
                 but a complement.                            for both open cut and underground mining. The
   209   210   211   212   213   214   215   216   217   218   219