Studying Oceanic Plate Motions with Magnetic Data

The geocentric axial dipole hypothesis states that the geomagnetic field, when averaged over tens to hundreds of thousands of years, corresponds to that of a dipole located at Earth's center and aligned with Earth's rotation axis. Accordingly, paleomagnetic poles, which are estimates of the position of the ancient dipole axis, give past positions of the Earth's rotation axis relative to a sample locality from which the paleolatitude of the sample can be derived. Using this hypothesis, paleomagnetists have established detailed paleogeographies for the continents that span several hundreds of million years. Such information is essential for understanding plate and terrane motions, mantle convection, paleoclimates, and geomagnetism, as well as many other subjects.

Paleomagnetic data are typically obtained from the magnetization directions estimated from oriented samples collected from rock exposures, such as the layers of rocks exposed along stream or road cuts or the layers of volcanic flows exposed along the flanks of volcanos. For oceanic plates, such as the Pacific plate, rock exposures are limited to a few young islands. Even for continental plates, most of which have large oceanic components, rock exposures may be very limited in the range of ages they span. In these cases, paleomagnetic poles must be estimated by other techniques.

One technique with great promise uses the shapes of marine magnetic anomalies. Marine magnetic anomalies are the highs and lows in the intensity of the magnetic field measured by magnetometers that are towed behind ships or low-flying airplanes, typically along profiles roughly perpendicular to seafloor spreading centers or roughly parallel to the direction of seafloor spreading. The highs and lows are produced by the upper portion of the oceanic lithosphere, which becomes magnetized parallel to the ambient field direction upon its formation at a seafloor spreading center. The information we extract from the shapes of these anomalies is referred to as skewness data because the asymmetry or skewness in a marine magnetic anomaly is related to paleolatitude.

This method is especially practical because magnetic anomaly profiles already exist from cruises covering large parts of the oceans and can be inexpensively obtained from the U.S. National Geophysical Data Center; oceanic plates generally have anomalies that are distributed over large regions and span long intervals of time; and the relative ages of magnetic anomalies are known exactly and the apparent or numerical ages are known well.

The skewness method consists of finding a phase shift - an angle that describes the asymmetry of a signal in spectral analysis - which gives the anomaly a shape that would be expected if the anomaly had been produced by oceanic lithosphere that formed and remained at the North pole. A single skewness datum restricts the paleomagnetic pole to lie along a great semi-circle or within a skewness lune - the lunar-shaped region bounding the great semi-circle - when uncertainties in the phase shift are incorporated. A single paleomagnetic pole is obtained from two or more skewness data giving skewness lunes that should intersect. From a set of skewness data, we estimate a best-fitting paleomagnetic pole, a 95% confidence ellipse, and data importances from a maximum-likelihood inversion.

The inversion procedure [Petronotis et al., 1992] also allows us to estimate the size of anomalous skewness and remove biases caused by it. The additional asymmetry in an anomaly caused by anomalous skewness is thought to arise from non-vertical reversal boundaries in the oceanic lithosphere or from lateral variations in the magnetization intensity within a polarity interval. For example, if the intensity of the geomagnetic field decayed over a time interval when the field had a constant normal or reversed polarity, then oceanic lithosphere formed during this interval would give rise to magnetic anomalies with anomalous skewness. Thus, the estimates of anomalous skewness help to better understand the source of marine magnetic anomalies and properties of the geomagnetic field.

Our new paleomagnetic poles suggest that Earth's largest plate, the Pacific plate, moved about 1500 km northward over a 26-m.y. period from the Late Cretaceous to the late Eocene and another 500 km northward over a 39-m.y. period since then (Figure 1). In detail, the poles indicate that the Hawaiian hotspot moved about 900 km southward between 45 and 33 m.y.a., which is consistent with a large, fairly rapid episode of true polar wander [Petronotis et al., 1994; Petronotis, 1993].

Fig. 1. Three new paleomagnetic poles define the northward motion of the Pacific plate relative to its present-day position. Black dots (with standard error bars) show the change in latitude predicted for a point in the Hawaiian-Emperor island and seamount chain. For example, it is estimated that the point moved 17.8 degrees (1980 km) northward in the past 65 m.y. The Pacific plate seafloor age maps at each time interval (43 Ma, 57 Ma, and 65 Ma) illustrate the growth of the Pacific plate while the position and orientation of the equator through time shows that the dominant motion has been northward with a small clockwise rotation between 57 and 43 Ma and a small counterclockwise rotation between 43 Ma and the present.

The new poles are also being used in paleomagnetic tests of global plate reconstructions [Acton and Gordon, 1994], which indicate that there is a significant error in reconstructions through the South Pacific and Antarctica. The implications are far reaching for circum-Pacific plate and terrane motion studies and hotspot motion studies that use these reconstructions. In particular, prior estimates that suggested that Pacific hotspots move relative to other hotspots at rates of 10-20 mm/yr are probably too large by a factor of 2 or more.

Our skewness analyses have also shown that magnetic anomaly profiles from near paleoequatorial sites are very sensitive to the position of the paleomagnetic pole and to geocentric axial quadrupole components. These quadrupole components are likely the dominant non-dipole part of the time-averaged geomagnetic field and thus probably contribute the most to the deviations from a pure geocentric axial dipole field observed in some paleomagnetic data. The large gradients in skewness that occur near the paleoequator may also explain why Late Cretaceous to early Tertiary magnetic anomalies recorded between the Clarion and Clipperton fracture zones in the Pacific had not been recognized previously [Acton and Gordon, 1991; ]. Petronotis et al., 1994

Where comparisons can be made, the independent results being obtained from skewness data agree well with other types of paleomagnetic data. Where other types of data are sparse or nonexistent, skewness data have the potential to give high quality paleomagnetic observations with fine age resolution.

Acknowledgments

References

Petronotis, K. E., R. G. Gordon, amd G. D. Acton, A 57-Ma Pacific plate paleomagnetic pole determined from a skewness analysis of crossings of marine magnetic anomaly 25r, Geophys. J. Int., in press, 1994.