Constellation, le dépôt institutionnel de l'Université du Québec à Chicoutimi

Résumé

Many geology, mining, and geotechnical applications require or depend upon some form of modeling of bedrock topography. Optimizing the manner with which bedrock topography is modeled poses a significant challenge because of the unpredictable or erratic presentation of the surface shape of bedrock. Unlike surface topography, bedrock topography is more difficult to determine because direct observation points are often not readily or directly accessible, unless the bedrock outcrops at the surface and is exposed, a relatively rare occurrence. When bedrock is covered by granular deposits, the only methods that allow practitioners to objectively establish the location of the top of the bedrock are to drill boreholes or conduct geophysical surveys. This makes the determination of bedrock topography not only difficult but also expensive. This study proposes a new approach for optimizing the modeling of complex bedrock topography, whose originality is based on the addition of “virtual” data points derived from cross-sections located between known boreholes. The proposed methodology is thus composed of four steps: gathering the maximum amount of relevant surface and subsurface data (from observation points), selecting the most appropriate technique for interpolating the observed bedrock elevations that will be entered into the dataset to be modeled, enriching the quantity of modeling data by adding “virtual” data elements based on geological interpretations of cross-sections (inserted into the model alongside the original objective data), and finally the modeling itself. The proposed approach is illustrated using data from a study area located in the Canadian Shield. Thousands of borehole records and surficial geological data as well as geological cross-section records were integrated to construct a three-dimensional bedrock topography model. The new proposed methodology can be applied to other regions worldwide.