Lessons worth highlighting

The main lesson is examining the interesting conductor at the south-east edge of the model as seen in Fig. 461. The feature is present in the original inversion in Figures 4a and 5a of Rutley et al [ROS01]. Given the known geologic structure and placement of the body at the edge of the data, the conductor may be an artifact of the inversion. The most unfortunate aspect of the conductor is in fact its large conductivity values that detract from the recovered Breakaway shale. We show the testing of the inversion artefact hypothesis below.


Fig. 461 The initial 3D DCR inversion recovered model. The south-east conductor reaches conductivity values similar to that of the Breakaway shale.

In order to assess the validity of the conductor, we change the initial reference model from the inversion from a 0.04 S/m normal space to a modified version of the recovered model. The goal is to see if the data force the solution to deviate from the reference model. The zone east of the resistive feature (i.e., East Creek volcanics) is set back to 0.4 S/m (Fig. 462). The data were re-inverted with the new reference model also set to the initial model. One interesting observation was that simply removing the conductor had an initial misfit of twice the desired misfit. The recovered model still required a conductor, but one at much less conductivity (Fig. 453). The conductor now spans two lines and is removed from the third and fourth lines.


Fig. 462 The conductivity reference model after the conductor to the southeast was removed and replaced with 0.4 S/m.


Fig. 463 Sections through the recovered conductivity model and a volume rendered image of conductivities above 1 S/m. The topographic surface and electrode locations (white dots) are shown for reference.

In this case study, multiple physical properties are important. Therefore, we carry out the 3D IP inversion for thoroughness. The new recovered conductivity model is used for the inversion. The recovered chargeability models with and without the conductor are shown in Fig. 464. There are some subtle differences between using the different conductivity models, but nothing that would affect the final IP interpretation. This result was to be expected because the initial inversion did not put any chargeable material in the conductor.


Fig. 464 : Plan-view sections through the recovered chargeability model (left) with the initial conductivity model and (right) the final conductivity model. Due to the lack of recovered chargeability where the large southeast conductor was located, the final interpretation of this physical property did not change.