Proterozoic GAD Hypothesis: Reliability Test Using Dyke Swarms

Pre-Mesozoic continental reconstructions and paleoclimatic inferences from paleomagnetism rely critically upon the assumption of a time-averaged geocentric axial dipole (GAD) magnetic field. A shallow inclination bias has been discovered in the paleomagnetic database for the Paleozoic and Precambrian [Kent and Smethurst 1998]. Possible explanations include sampling bias of low latitude continents, lumped results from far-traveled terranes, quadrupole or octupole field components (at surprisingly high values of 10% and 25%, respectively, of the dipole at the Earth's surface), or some other unrecognized bias. A 25% octupole is found to be inconsistent with evaporite paleolatitudes for most of the last two billion years [Evans 2006]. This result suggests that continents have preferentially sampled the equator or there might be some other issue with the database-wide inclination test. We have been testing the GAD assumption and localized non-dipole components in a different manner, by observing directional variations within the Matachewan and Mackenzie dyke swarms. Large dyke swarms, commonly emplaced within a few million years, provide the necessary broad areal coverage for testing time-averaged geomagnetic field topology. We vary the quadrupole and octupole values of the generalized paleolatitude equation to determine a minimal angular dispersion and maximum precision of paleopoles. Current results of the Matachewan and Mackenzie swarms have best fits that are non-dipolar but with less than a 25% octupole contribution. Further analysis will include the Franklin dyke swarm, and Central Atlantic Magmatic Province (CAMP) volcanics as a test case for the method.