A framework for rock property data acquisition, visualization and analysis: an example from the Bathurst Mining Camp, Northern New Brunswick

Petrophysical studies provide the link between geophysical grids and interpretive geological maps. In this study we assess the reliability of some physical property measurements: density, magnetic susceptibility, IP/Resistivity, and conductivity. Comparing susceptibility measurements obtained with different instruments shows there is a need for careful selection of sensor coil frequency and use of a suite of reference standards. For density measurements we show more reliable estimates require the sensitivity of the weighing scale be adjusted relative to sample weight and that a vacuum saturation tank is needed when the sample may have significant porosity. Galvanic resistivity measurements on hand samples are limited by the ability to properly estimate the effective cross-sectional area. Inductive conductivity instruments generally have a limited sensitivity range. We recommend the study of physical properties based on the concept of populations associated with stratigraphic units, rather than based on lithology. We present a suite of data visualizations which highlight different aspects of the sampled lithologies. Cross plots, ordered data plots, and violin plots permit the recognition of populations, and outliers. Henkel plots using templates with known mineralogical input parameters allow the identification of geological factors such as enhanced feldspar content, serpentinization, alteration, etc. Analysis of the sample data from the Bathurst Mining Camp, New Brunswick reveals which units have sufficient petrophysical contrast to permit pseudo-geological mapping. More specifically, not all basalts have high susceptibility and low density; it is impossible to discriminate between the various packages of rhyolites, and there is little difference between the granites sampled.