The last 20 years have seen a dramatic rise in the use of 3D visualisation and modelling for geological data, but there has been relatively little effort to do the same for the variogram models that underpin the standard estimation algorithms used by the mining industry.
For kriging, one of the most used geostatistical estimation methods, specifying a variogram model is a critical part of the modelling process. Experimental variography is the primary method for approximating the theoretical variogram function from the available sample data, and then determining an appropriate variogram model from the approximations.
The traditional tools used for experimental variography are limited to looking at subsets of the data and are often presented without reference to geographic space and geology. It can be difficult to know what input data is being used for a particular experimental variogram, and whether there is enough support in the experimental variograms to propose an appropriate model.
To overcome these limitations, an innovative approach called situational variography has been proposed that visualises variograms in 3D alongside contextual geological and geometric data. Presenting variograms in this manner aids understanding of how the spatial variance relates to the underlying data.
A 3D extension of the 2D radial variogram maps will be investigated which allows for dynamic exploration of the full experimental variogram cloud. These 3D experimental variograms are much easier to visualise situationally.
A workflow for situational variography will be outlined and the advantages of the situational approach will be illustrated using a porphyry copper deposit.
Gleeson, P, McLennan, T, Suresh, P and Coombes, J, 2017. Situational 3D variography – determining the variogram in the context of geology, in Proceedings Tenth International Mining Geology Conference 2017, pp 191–198 (The Australasian Institute of Mining and Metallurgy: Melbourne).