|Title||Method development for green-fields to deposit-scale 3D geologic modelling illustrated with case studies from the Abitibi, Flin Flon and Bathurst TGI-3 project areas|
|Author||Schetselaar, E; de Kemp, E; Hillier, M|
|Source||TGI-3 Workshop: Public geoscience in support of base metal exploration programme and abstracts; by Geological Association of Canada, Cordilleran Section; 2010 p. 1-2|
|Links||Online - En ligne|
|Links||Full abstracts volume (27 MB)|
|Alt Series||Earth Sciences Sector, Contribution Series 20090394|
|Meeting||TGI3 Workshop: Public Geoscience in Support of Base Metal Exploration; Geological Association of Canada, Cordilleran Section; Vancouver; CA; March 22, 2010|
|Subjects||stratigraphy; economic geology; mineral deposits; sulphide deposits; volcanogenic deposits; modelling; lithostratigraphy; stratigraphic correlations; Precambrian|
|Program||Deep Search TGI-3, Targeted Geoscience Initiative (TGI-3), 2005-2010|
|Abstract||One of the objectives of the Geological Survey of Canada's third Targeted Geoscience Initiative Program (TGI3) was to update the geoscience knowledge base in VMS exploration camps across Canada through
3D geologic modelling. This work has resulted in a number of green-fields to deposit scale query able 3D models in VMS camps. In addition, R&D initiatives were undertaken, to address fundamental problems and limitations identified through these 3D
modelling efforts. This presentation gives an overview of the methods under development and illustrates their applicability for green-fields to deposit scales VMS settings in the Bathurst, Flin Flon and Abitibi TGI3 project
Traditionally, 3D structural modelling in green-fields (areas with sparse or no drill holes) relies on labour-intensive compilation of multiple cross-sections. The resulting 3D models are highly subjective, irreproducible and difficult
to update when new subsurface information from drilling or geophysical surveys become available. We are developing reproducible vector field interpolation methods that constrain the regional 3D subsurface structure of green-fields domains using the 3
directional cosines of outcrop structural elements (bedding/foliation). Multiple 3D form surface solutions of lithostratigraphic surfaces and structures can be instantly generated by user-defined parameters that weigh local versus global structural
geometry and styles. This work can be expanded to include additional geological parameters describing thickness variations of strata as well as constraints from topological relationships with other lithostratigraphic surfaces and structures.
The hole-to-hole correlation of lithologic or lithostratigraphic markers in tectonically replicated VMS-hosting volcanic and volcaniclastic successions with rare diagnostic lithostratigraphic horizons or horizons of limited lateral extent is
a highly underconstrained problem. To tackle this problem, we are developing algorithms that match user-defined lithostratigraphic patterns with drill hole lithology logs to support hole-to-hole correlation and identify diagnostic transitions in
lithofacies in the volcanic and volcaniclastic successions. Lateral lithofacies changes can be taken into account by sequencing multiple patterns. If numerical multivariate attributes, such as geochemistry or geophysical log data are available, the
1D spatial along hole pattern recognition can be combined with statistical pattern recognition in a contextual lithologic log classifier.
Traditional lithofacies models of ore-hosting horizons have been based on correlating interpreted
contacts from adjacent mine sections or plans. Geostatistic lithofacies modeling offers an alternative that casts the problem in a probabilistic framework that yields reproducible multiple scenarios and assesses uncertainty. As opposed to 3D surface
models, the resulting 3D grid models provide direct input to 3D forward modelling and inversion routines facilitating the reconciliation of geophysical and geological data.
Legacy data of giant VMS deposits are often the most spatially
extensive and voluminous subsurface datasets available in minerals exploration camps and record relative large continuous tracts of observed underground geology. These analogue datasets are often poorly archived, registered to poorly-specified mine
coordinate systems and captured in a format that renders them unsuitable for constraining 3D models. Since many of these legacy datasets have been acquired by mining enterprises that do not exist any longer, there is a significant risk for loosing
these knowledge assets for good. We have developed input routines that capture the geological interpretation from drill hole logs and mine workings across multiple mine levels. The polygons of the 2D mine interpretations in shape files are vertically
projected to their appropriate depth levels using batch routines. Lithologic contacts and structures are systematically encoded to facilitate the 3D reconstruction process. The resulting 3D models provided insights in the geological setting of giant
ore deposits that were never possible before.
The essence of a head frame to green-fields 3D mapping strategy is to conduct simultaneous data reconciliation at all scales and from multiple sources. Although in its infancy we believe we
increase the potential for mineral discovery when we focus on developing a workflow that supports full 3D multi-scale integration as opposed to higher risk single scale and single data source targeting. Future work will be undertaken to support
upscaling methods, cross disciplinary data reconciliation and 3D interpretative environments with this theme in mind.