Long-term average non-dipole fields; how large or how small?

Paul Louis Mercanton suggested already in the late 1920's that paleomagnetism might provide a test of continental drift. However, the absence of an adequate understanding of the ancient (!) geomagnetic field structure hampered such a test until some 25 years later. But then, the results of the paleomagnetic study of Neogene Icelandic lavas by Hospers in the early 1950's, provided a breakthrough. Two very important findings were: (1) that the field in the Neogene was predominantly dipolar, implying that higher-order fields (quadrupoles, octupoles) averaged to near-zero, and (2) that the dipole axis remained on average aligned with the rotation axis, during normal- as well as reversed-polarity fields intervals. The last conclusion prompted Creer, Irving, and Runcorn to remark that "The coincidence of the magnetic and rotation axes [ . . .] covering many reversals is explained by the dominance of the Coriolis force". The geocentric axial dipole (GAD) hypothesis remained ever after the main guiding principle of paleomagnetic analysis, allowing declination anomalies to be interpreted as rotations and inclinations as representative of paleolatitudes. It is generally agreed upon that the long-term averaged field structure is largely, but not perfectly, dipolar. The critical question about non-dipole fields is "how large" (or, perhaps, "how small"). Analysis of the magnitude of non-dipole fields is restricted to zonal fields of degree (n) two or three, i.e., axial quadrupole and octupole fields, characterized by Gaussian coefficient ratios (Gn) where G is the ratio of the appropriate higher-order field coefficient and the axial dipole field coefficient. For the last 5 million years G2 and G3 are small, but not zero (Johnson et al., 2008, G-cubed), and for earlier geological times (Permian, Triassic) some speculations by some of us have considered values up to 0.2, on the basis of inclination patterns. The underlying assumption that inclination anomalies were attributable entirely to the geomagnetic field structure, though, was shown to be false; inclination shallowing due to early sedimentary consolidation processes plays a large role, which was not fully appreciated some twelve years ago. Of great influence in this development was the introduction of TK03, a geomagnetic field model of Tauxe and Kent, which provides (for rather ideal cases only) a method to separate the look-alike effects of inclination-shallowing and octupole fields. The enigmatic and troublesome Permo-Triassic paleomagnetic results from the Atlantic-bordering continents, which do not match the expected continental reconstruction for those times (Pangea A), have also been explained (and remedied) by corrections for sedimentary inclination shallowing models (Domeier et al., 2012, Tectonophysics). The developments of the ideas in this paragraph will be highlighted in this presentation.