Improving geological logging of drill holes using geochemical data and data analytics for mineral exploration in the Gawler Ranges, South Australia

Geochemical data are frequently collected from mineral exploration drill-hole samples to more accurately define and characterise the geological units intersected by drill hole. However, large multi-element sets slow challenging interpret without using some form of automated analysis, such as mathematical, statistical or machine learning techniques. Automated analysis techniques also have advantage in that they repeatable can provide consistent results, even for very sets. In this paper, an litho-geochemical interpretation workflow is demonstrated, which includes preparation appropriate compositional data-analysis Multiscale a modified wavelet tessellation has been applied coherent domains. Unsupervised (clustering) used first-pass classification. The results compared with detailed geologist’s logs. comparison shows how integration geochemical be enhance traditional logging demonstrates identification new were not apparent visual but geochemically distinct.KEY POINTSTo reduce computational complexity facilitate interpretation, subset elements selected, then centred log-ratio transform applied.The method domain holes into rock at range scales.Several clustering methods tested identify distinct multiscale domains classification.Results logs assess inform improve geology