Practical unfolding for geostatistical modeling of vein-type and complex tabular mineral deposits

摘要

Geostatistical modeling works best in Cartesian coordinates. The basic approach is to take large-scale curvilinear structure, flatten it, model the variability, and put the values back in original coordinates. A practical one-to-one reversible transformation is proposed with a straightforward set of functions to permit simple implementation. The approach allows for (1) improved variogram modeling: the directions for variogram calculation follow geological continuity, (2) improved characterization of the limits to geologic mineralization: the hanging wall and footwall can be modeled more realistically, (3) improved uncertainty characterization: it is easier to simulate uncertainty in the thickness, rock types, and grades in the unfolded coordinate space than it is to assess uncertainty in a sold/wireframe, and (4) improved trend modeling: it is possible to compute simplified trend models perpendicular to structure, along strike, and down dip.