More than 60 individual paleomagnetic poles have been obtained by various workers in the last 20 years from late Precambrian Keweenawan rocks of the Lake Superior region. Nearly all major formations and intrusive units have been subject to at least one paleomagnetic study. Keweenawan rocks thus represent paleomagnetically the world’s most intensely studied rock sequence, one that may span a time interval from about 1.2 to 1.0 b.y. ago. The large amount of paleomagnetic data coupled with locally excellent stratigraphic and structural control allows an examination of the extent to which factors other than continental displacement determine the distribution of Precambrian paleopoles.
Keweenawan paleomagnetic poles of both normal and reversed polarity plot along a northeast-southwest trending band in the North Central pacific. Stratigraphic and radiometric evidence suggests that within this polar distribution there is a hairpin-shaped path open to the southwest (the so-called Logan Loop) along which there appears to be an anticlockwise polar movement with time. After filtering of the pole population using certain reliability criteria, the width of the better documented western arm of the loop decreases from 20 to 10 degrees of arc along an arc length of about 70 degrees. A smooth narrow polar path is thus produced by selecting those poles for which errors due to sampling density, structural correction, and unremoved secondary components are considered to be a minimum.
Although much of the dispersion in pole position may be caused by uncertainties in the paleomagnetic data and associated geological constraints, the gross form of the loop appears to result from two superimposed effects: an apparent movement of the pole relative to the North American continent and a fictitious one arising from a violation in the assumption of a geocentric axial dipole to calculate pole positions. The latter effect is revealed by successive asymmetric reversals that can be explained neither by the presence of an unremoved secondary component nor by continental motion.
The Keweenawan apparent polar wander path and that for a contemporaneous sequence from the Grand Canyon, Arizona, agree closely if only normal poles are used. In this case both paths have a similar form to the Logan Loop but are more subdued. While the Keweenawan reversed data also appear to follow an arcuate path, the arc is displaced to the northeast of the normal one as a possible consequence of non-geocentric dipole field behavior. However, both paleointensity and paleosecular variation results from Keweenawan igneous rocks are compatible with the usual assumption of a geocentric dipole and with a change to higher paleolatitudes during times of reversed polarity, but it is possible that some non-geocentric dipole model could also explain these data.
Although a regional secondary component can be discounted as the cause of Keweenawan reversal asymmetry, other generally minor components are present with different directions and origins. They may be due to late Keweenawan igneous activity, burial of the Keweenawan sequence, Grenville tectonism, emplacement of copper-bearing ores, and in one instance, possible meteorite impact. Some of these magnetic overprints appear to have formed within a time period of about 1.0 to 0.8 b.y. ago and are thus important as they lie in an age interval poorly represented in North American paleomagnetic data.