Introduction to Global Topcut Analysis

Global topcut analysis, also known as topcutting or capping, can be used to identify and manage outliers in mineral resource estimation. Positively skewed distributions are characterised by a relatively small number of high-grade data points that can cause overestimation in an area around the outlying samples in a process known as ‘smearing’. This is particularly important in gold exploration where a small number of "nuggety" high-grade assay results can artificially inflate the estimated gold content within the deposit.

Global topcut analysis involves applying a high-grade cut-off, where all values above this cut-off are reduced to a specified limit. In Supervisor, topcutting involves selecting a topcut value by analysing the histogram, probability plot, mean and variance plot, and cumulative metal plot using the ‘disintegration method’. This allows you to jointly analyse the impact of various topcuts using the statistical distribution, as well as the cumulative amount of metal. Setting a topcut value results in all values above it being reduced to that value.

Notes:

The disintegration method is the point at which the distribution ‘disintegrates’. This is a visual cue that varies from plot to plot.

Data points above the topcut value are not removed from the dataset. After a global topcut analysis has been applied to a data node in Supervisor, all further analyses beneath that node will use the topcut value in place of any data points that initially exceeded it, meaning that the number of data points in further analyses is maintained. The topcut can also be removed at any time and all data points will return to their original value.

The objective of global topcut analysis is to limit the adverse effects of extreme values during variography analysis and estimation.