Minestis Domain Modeling Application has been released

July 22, 2016

Early July, Minestis Domaining pack has been made available to all following 4 years of research and sponsor exclusivity.

Minestis Domaining pack lets you perform:


Grouping boreholes samples into domains has never been so easy and fast than with Minestis Domaining application.

The software automatically identifies zones with homogeneous grade profiles and assigns each drillhole sample to a given domain.

The process is based on an innovative machine learning algorithm derived from the geostatistical hierarchical clustering algorithm (GHC). First samples, and then cluster of samples, are compared two by two and grouped according to their spatial dependency and their degree of similarity. Similarity is based upon a difference value calculated from input information including grade, lithology and distance between samples; the importance of each variable being moderated via a weighting scheme.

Minestis dendograms

Minestis delivers dendrograms showing the hierarchical relationship between samples. Similar samples are linked together

The procedure is iterative and merges gradually the least dissimilar samples into domains and ends when the number of domains you have specified is reached. Post-processing allows possible refining and merging operations if required.

Minestis sample clustering

Minestis automatically assigns a domain to each borehole sample combining grade, lithology and structural information


Minestis Domain Modeling application uses previous domaining information to interpolate domain boundaries.

Minestis bases its process on the Potential Field method which uses drillhole intercepts and orientation data to model domain boundaries through co-kriging. The model benefits from additional geological or structural information for more realistic envelopes.

Domain modeling with Minestis

Domain shape is captured through the variogram model

A key point of interest is that, besides the cokriging variance, the technique allows mapping the probability that a specific location lies within the domain.

Minestis probability to be inside a domain

Minestis maps the probability to be inside a domain


Whether built with Minestis or imported from external mining packages, possible overlaps or voids between domain envelopes are verified and can be fixed for improved consistency prior to resource estimation.

Besides, Minestis allows genuine uncertainty assessment on the domain volumes based on conditional simulations.

Minestis domain envelope volume curve

Minestis computes volume curves from simulated envelopes