Plate motion and a dipolar geomagnetic field at 3.25 Ga

Abstract

The paleomagnetic record is an archive of Earth's geophysical history, informing reconstructions of ancient plate motions and probing the core via the geodynamo. We report a robust 3.25-billion-year-old (Ga) paleomagnetic pole from the East Pilbara Craton, Western Australia. Together with previous results from the East Pilbara between 3.34 and 3.18 Ga, this pole enables the oldest reconstruction of time-resolved lithospheric motions, documenting 160 My of both latitudinal drift and rotation at rates of at least 0.55°/My. Motions of this style, rate, and duration are difficult to reconcile with true polar wander or stagnant-lid geodynamics, arguing strongly for mobile-lid geodynamics by 3.25 Ga. Additionally, this pole includes the oldest documented geomagnetic reversal, reflecting a stably dipolar, core-generated Archean dynamo.

Keywords: Archean geodynamics; geodynamo; hydrothermal alteration; paleomagnetism; plate tectonics.

Fig. 1.

Map of the Soanesville Syncline with sampling sites. (A) Reference map of the East Pilbara Craton. (B) The Kunagunarrina Formation (medium green) sampled in this study (filled points) is well preserved along the southeast limb of the Soanesville Syncline (SVS), along with the overlying Honeyeater Basalt (blue-green) from which the HEBh paleomagnetic pole was measured [hollow points (4)]. (C) Map details sites from locality KUA, as well as a dolerite sill (sampled at site KUA7) and a prominent marker bed of komatiitic volcaniclastic rocks (sampled at locality KUT). Note that our sampling at KUA is grouped into “Lower” and “Upper” groups of sites within the volcanic stratigraphy. Also noteworthy are two horizons associated with hydrothermalism, one in the central Kunagunarrina Formation (white) and another called the “Marker Chert” capping the overlying Kangaroo Caves Formation (medium blue).

Fig. 2.

Paleomagnetic results. (A–C) Orthographic plots of demagnetization (in situ coordinates) show multiple magnetization components (arrows). Magnetites hosting the H component in C have been directly identified by QDM mapping (SI Appendix, Figs. S9 and S10). (D and E) Stereonets of the highest-temperature “H” component. Site means converge upon tilt correction, indicating a prefolding H magnetization (SI Appendix, Fig. S2B). Site means from the uppermost sites (red) are antipodal to those below, reflecting a geomagnetic reversal and an H magnetization acquired shortly after deposition.

Fig. 3.

Constraints on the timing of magnetization. The H component was acquired during seafloor hydrothermal alteration between >3.223 ± 0.023 Ga (U-Pb titanite age of postmagnetization albitization in our samples, SI Appendix, Fig. S11 and Appendix S4.5) and <∼3.275–3.249 Ga [age range of Kunagunarrina Formation eruption from U-Pb zircon ages in the Kunagunarrina and Kangaroo Caves Formations (15, 21, 48)]. This matches the range of U-Pb zircon crystallization ages measured from the nearest granitic intrusions of the Cleland Supersuite from 3.257 to 3.235 Ga (15) and dates from their associated volcanic-hosted massive sulfide (VHMS) hydrothermal mineralization (16, 17, 21). This mineralization is bracketed to between 3.265 Ga, the upper bound on the oldest documented age of the VHMS deposits themselves [${3.257}_{-0.006}^{+0.008}$ Ga Pb-Pb galena model age (17)], and 3.235 Ga, the lower bound on the age of the epithermal Marker Chert that overlies our samples [the youngest permissible U-Pb zircon age of the 3.238 ± 0.003 Ga inner-phase Strelley Granite that drove the VHMS mineralization, as well as the mean U-Pb zircon age of a 3.235 ± 0.003 Ga rhyolite that immediately overlies the epithermal horizon (21)]. The H magnetization escaped full overprinting during all later events, including those preceding 2.93-Ga folding that is the basis of our fold test (constrained by U-Pb zircon dates from syn-kinematic granitoids (22, 49)). Thus, the H component dates to VHMS mineralization between 3.265 and 3.235 Ga.

Fig. 4.

Minimum-motion reconstruction of the East Pilbara from ∼3.34 to 3.18 Ga. (A) Apparent polar wander (APW) path constructed to minimize implied motions. Pole EBm samples the opposing reversal state relative to KUH-R and HEBh, and the SVS block is assumed to have rotated 20° clockwise, implying no rotational motion between 3.34 and 3.25 Ga (SI Appendix, Appendix S2). (B) The simplest motion reconstruction based on this path, starting with 95 My of ${0.55}_{-0.16}^{+0.19}$ °/My latitudinal motion followed by 65 My of ${0.55}_{-0.38}^{+0.46}$ °/My rotation. Other reconstructions are possible but require faster motions, most >1°/My; see SI Appendix, Fig. S6 for these less-plausible cases.

Fig. 5.

Comparison of measured East Pilbara motion rates with those of candidate drivers. Rates are expressed in both degrees/My and the equivalent value in cm/y, assuming measurement 90° away from the motion’s Euler pole. Measured rates (red, 2σ CIs) are lower bounds time averaged over the indicated intervals, documenting substantial motions between 3.34 and 3.18 Ga (see text and Fig. 4). These are comparable to time-averaged recent plate motions [green distributions (2)] but faster than time-averaged recent net rotations [blue distributions (50)] and expected net rotations of an Archean stagnant lid (blue shaded bar and arrow; SI Appendix, Appendix S5). The arrow indicates the expected highest permissible time-averaged rate of stagnant-lid net rotation in a perfect but unrealistic driving scenario, which we calculate as 40% of the expected highest instantaneous rate. This is based on the observation that the fastest 65-My-time-averaged net rotation rate on the modern Earth does not exceed 40% of the theoretical fastest permissible instantaneous rate (SI Appendix, Appendix S5.3). Net rotations are therefore most likely insufficient to explain the measured motions, suggesting a plate tectonic driver.

Fig. 6.

Constraints on geomagnetic field dipolarity at 3.25 Ga. The directional stability and polarity inclinations of the KUH-R pole constrain the axial quadrupole/dipole ($G2$, x axis) and octupole/dipole ($G3$, y axis) moments ratios to the 1σ and 2σ confidence regions in red. Insets show representative field geometries. The data are consistent with a pure geocentric axial dipole field (“GAD” inset, at 0,0 on this plot). While reproducing the observed polarity inclinations exactly without including uncertainty would only require a 15% relative contribution from an axial octupole (“$G3=-0.15$” inset), we cannot rule out the simpler explanation of a GAD field, since it remains statistically compatible with our polarity data based on this test and a traditional common-mean reversal test (14). The 3.25-Ga field was therefore strongly dipolar and consistent with previous constraints on Precambrian field geometry [dashed rectangle (–33)].