Cretaceous DSDP and ODP Basalt Cores from the Pacific Plate and Implications for Apparent Polar Wander

Uncertainties of the Pacific plate apparent polar wander path continue to have a negative impact on studies of true polar wander, relative motions of hotspots, plate tectonics and the shape of the geomagnetic field. Much debate about the APWP arises because non-traditional methods have been used to get around the fact that oriented paleomagnetic samples are difficult to obtain from beneath the ocean. For ∼32 years DSDP and ODP have recovered basalt cores from the Pacific and these should provide reliable measurements of paleomagnetic inclination (declination being indeterminate in azimuthally-unoriented cores). At the end of DSDP, there were too few cores, and fewer still that were well-dated, to allow precision in determining age and pole location. At the end of ODP can we do better? I have compiled Pacific basalt core data up to Leg 192 and analyzed the inclinations following the methods of Cox and Gordon [Rev. Geophys. Space Phys., 22, p. 47, 1984]. The data come from 35 sites and yield 173 independent units spanning late Jurassic to latest Cretaceous. Data distribution is poor for most of the Jurassic and later Cretaceous. Of 22 time windows, 5-Myr in width, 15 are empty or have only data from one site. Only the period 110--130 Ma has a large number of units and sites (125 and 20, respectively). These data were divided into two groups (110--119 and 122--129 Ma) and combined to determine paleomagnetic poles. Colatitude circles from the 110--119 Ma group are mostly in good agreement and cluster in the North Atlantic at 50--60^o. As expected, control on pole longitude is poor because of the lack of azimuthal orientation. Using paleodeclinations from three dated seamount magnetic anomaly inversions, the longitude errors are cut in half. The result is a pole at 53.9^oN, 334.2^oE (95% confidence ellipse semi-axes: 10.1^o, 5.1^o; major-semi axis azimuth 79^o). At first glance, the scatter in 122--129 Ma paleocolatitude circles seems large (>30^o) and data from Ontong Java Plateau appear inconsistent. However, the colatitude circles fall into two groups. Those that imply more northerly pole locations are from Ontong Java or seamounts, whereas those indicating southerly pole locations are from ocean crust (mostly in the Japanese magnetic lineations). This discrepancy suggests that ocean crust basalts may be biased by rotation of fault blocks away from the spreading axis. Excluding crustal basalts, remaining colatitude circles agree with two dated seamount poles and combine to give a pole at 61.8^oN, 337.2^oE (95% confidence ellipse semi-axes: 6.2^o, 3.6^o; major semi-axis azimuth: 86^o). Error ellipses for the two poles barely overlap and the pole positions imply southward motion of the pole moving forward in time. Late Jurassic data are inconsistent and cannot be interpreted to yield a pole or reliable paleolatitude. Late Cretaceous data are few, but hint at significant polar wander. Colatitude circles of Chron 33r age (79--83 Ma) cluster near previously published latest Cretaceous poles, ∼410^o north of an 82 Ma pole determined from seamounts. The basalt core data agree with skewness and sediment core data of the same age and imply the seamount poles are biased by overprint. The few colatitudes of 88--94 Ma age are consistent with the 110--120 Ma pole, displaced ∼10--15^o southwestward from the Chron 33r data. This gap is consistent with prior findings of rapid polar wander at approximately 84 Ma.