Paleomagnetism of Late Paleozoic series in Morocco and Argentina: implications for GAD Hypothesis and Pangea reconstruction

The configuration of Pangea during the upper Paleozoic is still strongly debated: the APWP for Gondwana and Laurussia should fit in the GAD hypothesis when Pangea is reconstructed using marine data. But they appear to be significantly distinct, by as much as 15° for certain periods. Motion between the two blocks has been proposed (B Pangea of Irving) but is geologically problematic. Erroneous age assignments, magnetic overprints, insufficient demagnetization, problems with the recording of geomagnetic field in sediments (f.i. flattening) and finally non-dipole contributions have all been invoked to reconcile apparently discrepant poles. In this talk, we investigate the geometry of the geomagnetic field during the Late Carboniferous-Early Triassic period (320-240 Ma). As a starting point, we use two recent surveys in Morocco and Argentina, which provide paleomagnetic tests and good age control that were missing in most previous studies and yield respectively paleolatitude data close to the equator and mid/high latitudes. Using a compilation of poles from the GPMDB, we analyze the positions of mean poles for mid-northern and southern, and equatorial latitudes, searching for the distinctive antisymmetrical pattern expected for a dipole with an octupole contribution. We also discuss the main causes of errors, such as the occurrence of lithospheric deformation, which induces important rotations at various scales, particularly in future rift or mountain zones (Colorado, South of France, South American cordilleras, east of Australia, etc.). Even when using only sampling sites close to the Paleo-equator (which minimizes any octupolar effect), the APWP of Gondwana remains offset from that of Laurussia. The plate configuration inferred is a classical A Pangea reconstruction at about 260 Ma, but data do not rule out the possibility of a B Pangea before 270 Ma, which would account for a large number of geological constraints. In contradiction with recently proposed persistent octupolar components of the geomagnetic field through the Paleozoic, we find no statistical evidence for a significant octupolar component of the field (or flattening) for most periods in the 200-320 interval, with exceptions around ~250 Ma, when more data are required for a robust analysis.