PhD Thesis (2022-2025)

Title:  Data constrained stochastic modelling of alteration & mineralization envelope structure : methodology & application to Athabasca uranium deposits

Keywords:  Stochastic modelling, alteration, mineralization, structure, Athabasca uranium deposits

Supervisors: Guillaume Caumon (RING Team), Pauline Collon (RING Team)

This study proposes to approach the modelling of faults/fractures based on a hybrid approach between pixel-based and object-based methods.

The principle of object-based methods is to generate geological objects defined by parametric equations (e.g. a fault is a line of length l1 with an ellipsoidal alteration halo of width l2 ...). This principle ensures geological realism by building an explicit model of the geometry of the objects.

Pixel-based methods are based on a matrix filling of properties and offer a better conditionning to the data at the cost of geological consistency.

The proposed hybrid approach aims to provide a methodology for stochastic modeling of the structure of alteration envelopes and mineralization, constrained by drilling data, as well as an estimate of the uncertainty on the position of structures.