Last year, Geovariances launched Isatis.neo Mining Edition, the geostatistics software that it says boosts performances and enables confident resource estimation and comprehensive geostatistical studies. The company has just released version 2020.02 of Isatis.neo with tools derived from the former Isatis that were most popular with users.
Among them are the following:
- Model sub-celling for a sub-block model adjusted to domain boundaries
- Grade capping for more robust estimation
- Resource estimation through Multiple Indicator Kriging (MIK)
- Grade variation analysis from either side of domain boundaries
- The smart selection of a representative subset of simulations for efficient risk analysis
It is now possible to create sub-block models from scratch. Sub-celling is done from either side of domain boundaries. Users can run estimation on sub-block models or migrate estimates or simulation results from regular block to sub-block models for further processing.
It allows capping of the higher or lower values in heavy-tailed distributions to ignore outliers and make the estimation more robust. Isatis.neo lets users test a range of bottom or top-cut values to help identify the most appropriate one for the deposit or dataset. Statistics before and after capping are computed for each test value to support decision making.
Users can now assess resources through Multiple Indicator Kriging (MIK). This non-linear resource estimation technique is commonly used when the sample grade distribution is skewed, or drilling is wide-spaced. It consists of kriging several indicators, from which grade-tonnage values (quantity of metal Q, tonnage T, and mean grade M above cutoffs) and the e-type estimate (the expectation of the distribution defined by the kriged indicators and the histogram interpolation parameters) are derived.
You can select a representative subset of simulations to characterise the risk attached to a project with the Simulation Reduction tool. This tool, first developed in Isatis, allows reducing the original number of realisations into a few more manageable items while preserving the statistics of the original simulation set.