Gravity model - Cross-section 3

A possible gravity model for cross-section 3 is illustrated below. The starting model (shown as red dashed outline) was that derived by geological reasoning (Smith et al., 1967), and appropriate densities and magnetic susceptibilities measured on drill-core from the deep holes were assigned to the various units. The primary objective was match the observed gravity field, and this has been achieved in the model above. A number of changes from the initial model were made. The main funnel-shaped portion of the intrusion has been widened, and a western slab-like extension has been added, imparting a marked asymmetry to the cross-section. Units near the eastern margin have been thinned, resulting in westward migration of this margin, and the keel area has also been displaced (~2 km) to the west. Gabbroic layers have been thickened and a gentle synformal geometry has been imparted to the eastern funnel. Problematically, the match of magnetic profiles is extremely poor. At this stage of modelling it was becoming obvious that the observed magnetic field could not be reproduced from magnetic units having relatively flat-lying attitudes.

Gravity model - Cross-section 5

A residual gravity high derived for cross-section 5 has an amplitude of only about 2 mGal - not surprising given the small volume of gabbro present. In a model constrained by the geological cross-section and measured densities, the amplitude of the model anomaly was ~3 times larger than that of the residual anomaly. A more reasonable match between the anomalies, shown below, was achieved by modifying this starting model: (1) densities were modified, (2) units in the western down-dropped block and eastern marginal zones were thinned, and (3) the eastern contact, marginal zones and feeder zone were moved variously 1-2 km to the west. A noticeable mismatch remaining along the western margin, may be rectified by introducing density variations within country rocks. Note that the elimination of the eastern portion of the Muskox intrusion may be artificial, since this portion may have a density similar to that of the country rocks and would be invisible to the gravity method. There is little similarity between the model and observed magnetic profiles.

Matching the magnetic profile of Cross-section 5

The mismatch between observed and magnetic profiles for both cross-section 3 and 5 is puzzling, since reasonable gravity models have been derived. However, it is clear that the narrow magnetic highsof the profile cannot be reproduced from the relatively flat-lying units of the models, and that geological units with much steeper attitudes are required to reproduce the observed magnetic signatures. With this in mind a simplistic approach to modelling the magnetic anomalies was undertaken. In the model below a synformal geometry was adopted and the dunites of the Muskox intrusion were designated as serpentinized (an upper layer) and unserpentinized (a lower layer). This is not unreasonable, since densities in the lower part of Muskox South hole are > 3.0 g/cm3, suggesting that the rocks are not highly serpentinized, and may therefore be less magnetic. Measurements of magnetic susceptibility will be carried out on drill-core to test this. If this model is indeed reasonable it will be necessary to remodel the gravity profile adopting a similar synformal approach.

Total magnetic field - Cross-section 8 & 10

Magnetic profiles for cross-sections 3 and 5 are characterized by more than one magnetic high, but the picture is much simpler in the southern part of the Muskox intrusion, where a single strong magnetic high is present. This magnetic high lies along the central part of the intrusion, and is, apparently, a singular feature related to a single source. Modelling here should be simpler and should provide insights for modelling profiles crossing the northern part of the intrusion. Cross-sections 8 and 10, displayed on the image below, are selected for modelling.

Magnetic model - Cross-section 10

The geological model constructed by Smith et al. (1967) has been used as a starting point with magnetic susceptibilities based on measurements made on similar lithologies from Muskox North drill-hole being assigned to appropriate units. Most of the contribution to the magnetic high, depicted in the model below, is made by the steep picritic keel of the intrusion, and in spite of the suggestion of gentle synformal layers in the uppermost part of the section, no difficulty was encountered in matching the observed anomaly. Magnetic units were modelled in the adjacent country rocks to explain anomalies outside the intrusion.

Magnetic model - Cross-section 8

Again a geological model (Smith et al., 1967) was used as the basis for modelling the central magnetic high, and no geometrical changes were made to this model. Picrite and serpentinized dunite units make major contributions to the magnetic high. It was necessary to assign a fairly strong magnetic susceptibility (0.018 SI) to semipelite and quartzite to the east of the Muskox intrusion to achieve the level of the field in that area. Once again the magnetic anomaly is reproduced by magnetic units that have steep structural attitudes, though it must be noted that at the top of the intrusion in this section, Smith et al. (1967) show a thin layer of olivine clinopyroxenite which is flat-lying (too thin to show in the figure below).

Synformal model

From a magnetic perspective, a model in which layering is synformal and steep in at least part of the intrusion is a requirement. Detailed gravity does not exist in the southern part of the intrusion to test this from a gravity perspective. In the northern part of the intrusion introduction of synformal structure must be reconciled with the structural layering as mapped from surface geology and constrained by drill-holes. Are the horizontally-layered cross-sections portrayed by Smith et al. (1967) indeed compatible with drill-hole information? Or is there some latitude in the constraints used by Smith et al. (1967) that would allow a more synformal structure to be applied? The great dyke of Zimbabwe (see illustration below) is a layered intrusion having similar layered lithologies, and a width and depth similar to those of the Muskox intrusion, indicating that a synformal aspect is not out of the question.

Conclusions: Utility of gravity and magnetic data.

1. Gravity and magnetic images permit the Muskox Intrusion to be traced and modelled under basaltic cover to the north.
2. Gravity modelling is effective only in areas unaffected by serpentinization, i.e. in gabbro-rich northern section.
3. Gravity modelling suggests (1) that Muskox Intrusion widens from ~10 to ~19 km, and thins significantly, under cover sediments near its northern extremity, (2) position of keel is located up to 2 km west of that proposed on geological grounds, (3) layering of intrusion in northern segment is gently synformal, and (4) that basal region of intrusion in north is not serpentinized.
4. Magnetic modelling indicates (1) that magnetic field over southern third of intrusion can be reconciled with geometry of intrusion predicted from geological considerations, i.e. funnel-shape with steep layering, and (2) that nature of layering in northern and central parts may be synformal and relatively steep.
5. Serpentinized portions of Muskox intrusion having densities similar to those of country rocks may be invisible to gravity modelling, hence it is cautioned that westward displacement of eastern margin and axial zone of intrusion implied by gravity models for cross-sections 3 and 5 may be unrealistic.

Future Directions

1. Obtain additional rock property data; examine effect of remanent magnetization.
2. Model additional sections and link laterally.

Acknowlegments

This data was first presented as a poster display during the 2003 Yellowknife Geoscience forum.

I am grateful to the following individuals for various forms of assistance, in the form of data, advice or information, and in particular I would like to thank Gerry Diakow for continued support.

Gerry Diakow, Trilogy Metals inc., Vancouver
Richard Ernst, Geological Survey of Canada, Ottawa
Mariette Henderson, M & H Geolink Ltd., Ottawa
Bob Hornal, Vancouver
David Mark, Geotronics Surveys Ltd., Vancouver
Trent Pezzot, SJ Geophysics Ltd., Delta