Summary: | An improved method has been developed for carrying out 2-plate reconstructions, in which fracture zone locations are fitted to synthetic flowlines and magnetic anomaly picks are rotated and fitted to great circles representing other, not necessarily conjugate, anomaly isochrons. This enables the determination of finite rotation poles for regions with sparse data coverage, or where much of one or both plates has been subducted. Misfits and partial derivatives are calculated for each type of data, and combined in a single iterative inversion, allowing the direct calculation of confidence intervals. This method is then extended to a 3-plate reconstruction, taking closure into consideration. The South American - African - Antarctic plate system is then studied. Fracture zone locations are identified from a gravity map constructed from GEOSAT altimeter data, and magnetic anomalies are identified from ship profiles. Two-plate reconstructions are carried out for each plate pair, giving good fits to the observed data, and then all three datasets are combined in a 3-plate reconstruction. Comparison of the results reveals a discontinuity in spreading in the Weddell Sea, believed to be related to pseudo-asymmetric spreading caused by ridge re-organisation in the Paleocene and early Eocene. A revised 3-plate inversion, taking this discontinuity into account, produces an internally consistent set of poles, indicating a closed 3-plate system since anomaly 34 (83Ma), with no evidence for a Malvinas Plate extending into the Weddell Sea in the Late Cretaceous. Disruption to the system from anomaly 32 (71Ma) until anomaly 24 (52Ma), appears to be related to the collision of Africa with Eurasia. A study of the past motion, configuration and stability of the Bouvet Triple junction suggests that for the majority of the past 50Ma it has been in a RFF configuration, in theory considerably less stable than RRR, the other possible configuration.
|