An evaluation of indicators of prospectivity and potential mapping of porphyry deposits in middle and southern British Columbia by a GIS study of regional geochemical and other geoscientific data.

This thesis presents mineral potential maps for prospecting porphyry deposits at the regional scale. The study area covers certain geological units in southern and middle British Columbia (from 48°N to 56°N) that are known for hosting major porphyry deposits. Four main types of data were used in this thesis, namely geochemical, geophysical, tectonic and distribution of known porphyry deposits. Several "evidence" maps were generated based on these data sets. These maps can indicate areas with porphyry mineralizations inside the study area. The "evidence" maps were combined to create potential maps for porphyry deposits. Such potential maps show the gradation of favourability within different areas for prospecting porphyry deposits. The fuzzy logic method was utillised for representing the relative importance of maps and map combinations. The stream geochemical dataset showed a "boundary shift" for some elements between a number of the National Geochemical Reconnaissance (NGR) surveys map sheets. In such cases, the data for stream sediment and water samples were levelled prior to statistical or GIS analysis. For levelling the geochemical data, a method was presented to estimate the required width for the bands that must be selected at the contact of the map sheets with a "boundary shift". The geochemical data were disaggregated based on the geological environment and univariate and multivariate analyses were performed on these subsets of data. It was noticed that the median of Cu in stream sediments was often a good predictor of porphyry deposits. In many cases parameters measuring the spatial variability of Cu and Mo, such as variance or slope (maximum first spatial derivitive) of these elements in a moving cell, were good indicators of potentially mineralized zones. In addition, the results of multivariate techniques such as factor and discriminant analysis could often be used as good indicators of mineralized areas. Residual total field magnetics and vertical derivative (gradient) of magnetics were also good indicators of porphyry Cu-Au and Cu-Mo mineralizations at the regional scale. At this scale areas with high density of faults show also strong association with the location of known porphyry deposits. Areas close to known porphyry deposits and mining camps were subtracted from the final potential maps to introduce only those areas predicted by this model, with no known porphyry deposits. In application of GIS at the regional scale, diversity of geology along the study area must be considered. It was noticed that evidences introduced by the univariate and multivariate analysis of geochemical data can be very useful at the regional scale if the geochemical data is first separated based on the similarity of geological environment. Such a separation enhances any statistical analysis at the regional scale