<p> The DC resistivity method has been an important tool for mineral exploration for the direct detection of conductive bodies with economic value. It has also been used for the structural mapping of lithology and alteration where boundaries are zones of economic interest and the detection of edges is the primary goal of surveying. Edge preserving inversion has been explored extensively within the context of potential field methods but has seen relatively little attention for the DC resistivity method. The focus of this thesis is to develop and implement methods which employ specific advantages of the DC method to aid the recovery of edges in the earth’s resistivity distribution.</p><p> I begin by utilizing sparse a priori geologic knowledge to create a geologic concept of pervasive blocky resistivity. <i>l</i><sub>1</sub> and <i> l</i><sub>0</sub> approximating measures of model values and model gradients are used as a vehicle to inject the a priori knowledge into a regularized inversion. An iterative method is used to solve for the model that minimizes a total objective function using these general measures.</p><p> A series of synthetic modelling and inversion scenarios demonstrate the effectiveness of <i>l</i><sub>1</sub> and minimum gradient support regularization to recover boundaries when compared to traditional sum-of-squares regularization. These blocky inversion schemes also exhibit an improved recovery of the resistivity value of distinct bodies. Additionally, I recognize that the various regularization types have different strengths and weaknesses. I exploit this property to create a new composite regularization that combines smooth model and blocky model regularization. This composite regularization exhibits the strengths of both regularization styles and less of the weaknesses.</p><p> A case study on field data from the Sabajo gold deposit was performed utilizing this methodology. Sharp lithologic contacts from drillholes informed the creation of a sharp resistivity concept. A blocky inversion was performed to recover a blocky model that was consistent with this concept and compared to the results of a smooth model inversion. Important differences were noted with their economic implications. I observed that the blocky regularized inversion may have recovered better estimates of the conductivity of features and this can greatly aid prioritization of targets for drilling. Finally, the differences between the inversions utilizing diverse regularization styles provided a proxy for model uncertainty.</p><p>
| Identifer | oai:union.ndltd.org:PROQUEST/oai:pqdtoai.proquest.com:10744014 |
| Date | 18 April 2018 |
| Creators | Par, Andrew R. |
| Publisher | Colorado School of Mines |
| Source Sets | ProQuest.com |
| Language | English |
| Detected Language | English |
| Type | thesis |
Page generated in 0.0108 seconds