Orientation uncertainty goes bananas: An algorithm to visualise the uncertainty sample space on stereonets for oriented objects measured in boreholes

Abstract

Measurements of structure orientations are afflicted with uncertainties which arise from many sources. Commonly, such uncertainties involve instrument imprecision, external disturbances and human factors. The aggregated uncertainty depends on the uncertainty of each of the sources. The orientation of an object measured in a borehole (e.g. a fracture) is calculated using four parameters: the bearing and inclination of the borehole and two relative angles of the measured object to the borehole. Each parameter may be a result of one or several measurements. The aim of this paper is to develop a method to both calculate and visualize the aggregated uncertainty resulting from the uncertainty in each of the four geometrical constituents. Numerical methods were used to develop a VBA-application in Microsoft Excel to calculate the aggregated uncertainty. The code calculates two different representations of the aggregated uncertainty: a 1-parameter uncertainty, the ‘minimum dihedral angle’, denoted by Ω; and, a non-parametric visual representation of the uncertainty, denoted by χ. The simple 1-parameter uncertainty algorithm calculates the minimum dihedral angle accurately, but overestimates the probability space that plots as an ellipsoid on a lower hemisphere stereonet. The non-parametric representation plots the uncertainty probability space accurately, usually as a sector of an annulus for steeply inclined boreholes, but is difficult to express numerically. The 1-parameter uncertainty can be used for evaluating statistics of large datasets whilst the non-parametric representation is useful when scrutinizing single or a few objects.