Abstract
For resource estimation, domains are interpreted in a deposit to define homogenous areas for grade estimation. The conventional approach is to interpret the domains and then to separately model the distribution of the grade (i.e., partial grades) within each domain. The problem is that independent modeling of partial grades ignores the cross-dependency that exists between the partial grades within the defined domains. In this study, an alternative is proposed to first model the grade domains using a stochastic approach and then to model the partial grades within each domain using a co-simulation algorithm that incorporates their cross-correlation structures. The proposed enhanced co-simulation methodology has been applied at an iron deposit and a single grade threshold. This case study shows that the enhanced co-simulation methodology is capable of reproducing the global statistics and spatial continuity characteristics of the partial grades and results in improved domaining, grade estimation and definition of ore.