Geostatistical Problems in Marine Placer Exploration

Abstract

Current developments in in-situ geochemical analysis systems undoubtedly will result in significant cost-reductions in marine placer exploration programs. These systems cannot be used to full advantage, however, until more sophisticated mathematical techniques, to process and interpret the geochemical data they generate, are also developed and employed. In particular, methods are needed to facilitate the identification of specific geochemical anomalies associated with particular types of marine placers, to discern the relationship between geochemical signals used in exploration and the geochemistry of the source rocks from which the placer minerals are derived, and to improve the understanding of sediment dispersal patterns in nearshore areas. Simple correlation analysis of geochemical data has been used to gain insight into the formation of both gold and platinum placers. For platinum placers, it has been shown that the results of simple correlation analysis can be refined by employing a comparitive principal-components analysis in order to obtain a better understanding of the geochemical basis for any observed statistical correlations. Recent research results suggest that newly developed geostatistical techniques, such as end-member composition factor analysis and modelling with linear programming methods, may also be useful in solving the types of geostatistical problems encountered in marine placer exploration.