Paleosecular variation over the last 10 Ma from a new global dataset (PSV10): Evidence for long-term hemispheric asymmetry

Accurate paleosecular variation (PSV) models of the geomagnetic field require good temporal sampling of globally distributed paleomagnetic directional and intensity data from discrete lava flows. Recent interpretations of PSV based on directional data have offered conflicting conclusions as to whether virtual geomagnetic pole (VGP) dispersion increases with latitude. These PSV studies have drawn on compilations of paleodirectional data that lack equatorial and high latitude sites and use latitudinal VGP cutoffs designed to remove transitional field directions. VGP cutoffs have been shown to artificially suppress scatter, especially at high latitudes. We present a new selected global dataset (PSV10) of paleodirectional data spanning the last 10 Ma. PSV10 has a more robust global distribution than previous compilations, comprising 1869 sites from 66 studies with site latitudes ranging from 78°S to 78°N. We include all directional results calculated with modern laboratory methods (complete stepwise demagnetization, principle component analysis) that meet our site level criteria (minimum of 5 samples, Fisher precision parameter estimate, κ, of at least 50) regardless of VGP latitude. We exclude sites from studies that specifically target transitional field states or exhibit significant tectonic effects. All sites are rotated to zero age latitude and longitude in order to correct for plate motion: the maximum latitudinal correction is 1.34°. Site VGPs are recalculated for plate-corrected latitudes and VGP dispersion is then determined in 10° latitude bins. The VGP scatter increases with latitude in both hemispheres but is significantly higher in the Antarctic than in the Arctic data. Current PSV models (e.g., Model G [1] and TK03 [2]) do not account for this asymmetry and preferentially fit northern hemisphere data while underestimating Antarctic VGP scatter. Time-averaged field contributions from terms other than a geocentric axial dipole are required to predict the observed hemispheric asymmetry. The addition of a 6% axial-quadrupole component to the simplified statistical PSV model, TK03, accounts for secular variation in both hemispheres. An axial-quadrupole of this magnitude is consistent with model CP88 [3] although CP88 does not predict latitudinal variations in VGP dispersion. These observations, coupled with asymmetric paleointensity estimates at high latitudes, indicate that the long-term geomagnetic field may contain non-zero contributions from terms in addition to the axial-dipole when averaged over million year timescales. [1] McElhinny, M.W., and P.L. McFadden, Palaeosecular variation over the past 5 Myr based on a new generalized database, Geophys. J. Int., 131, 240-252, 1997. [2] Tauxe, L., and D.V. Kent, A Simplified Statistical Model for the Geomagnetic Field and the Detection of Shallow Bias in Paleomagnetic Inclinations: Was the Ancient Magnetic Field Dipolar?, in Timescales of the Paleomagnetic Field, edited by e.a. Channell, J.E.T., vol. 145, pp. 101-116, American Geophysical Union, Washington D.C., 2004. [3] Constable, C., and R.L. Parker, Statistics of the Geomagnetic Secular Variation for the Past 5 m.y., J. Geophys. Res., 93, 11,569-11,581, 1988.