A case history of regional geoscience data integration and analysis for the purpose of mineral resource appraisal, kasmere lake area, Manitoba

Abstract

The methods used to generate a unified cell structured data base for geological, mineral occurrence, geochemical, and three different types of geophysical data are reviewed. The extremes of both excess and paucity of data were encountered and procedures for data compression and interpolation between known values were required. In the latter case a weighted moving average was used to take advantage of knowledge of the regional geological structure in an informal kriging. Classical approaches based on training data sets were investigated and found to be wanting in the area due to the inadequate mineral occurrence data base. The few data cells containing mineral occurrences were known with certainty; however, many of the cells classified as non-mineral occurrence containing, due to the immature exploration history, probably contained mineral occurrences of equal or greater significance. This inadequacy in the mineral occurrence data base confounded formal statistical analysis. An empirical scoring approach was developed which draws on general exploration criteria. This methodology, based on the geochemical and geophysical data, was used to prepare an Athabasca type uranium deposit favorability index. The favorable areas are examined in terms of the known geology. Results show that the majority fell in areas that would also be considered favorable by geologic models for Athabasca type uranium deposits. This empirical approach, based on geochemical and geophysical interpretive procedures, may have particular potential in poorly explored areas lacking good training sets.