In addition to optimising search parameters Snowden’s Supervisor software now also provides a process that ensures areas supported by several higher grade samples are not unfairly overly downgraded by a global topcut.
The traditional approach to resource estimation is to apply the same estimation parameters to each block and the same topcut to each sample within a domain even though block estimates are independent of one another. Local Kriging Neighbourhood Optimisation (LKNO) was introduced to Snowden’s Supervisor software in 2019 and optimises search parameters ensuring that each block is estimated with the best possible combination of parameters presented. Whilst LKNO optimises sample selection relative to the variogram model it does not account for the variability of informing sample grades.
The local topcut transition model provides a method by which samples are topcut on a block by block basis. The degree of topcutting applied is determined by the distance between the sample and the block centroid. The closer the sample is to the block centroid the lower the degree of topcut applied. The degree of topcut increases until a specified distance is reached, beyond which the global topcut is applied. The rate of transition from the centroid to sample distance is controlled by a user defined transition rate. The impact of the transition rate on the topcut is illustrated below:
Local topcut transition modelling allows you to evaluate the impact of extreme grades on each block’s estimate and when used in conjunction with the global topcut analysis and LKNO to optimise the search parameters the resultant resource model will reflect not only the global conditions, but also account for local grade variability.