HR: 11:45h
AN: V21F-06 [PDF]
TI: Revised Estimate of the Cenozoic Motion Between the Tristan da Cunha Hotspot and Pacific
Hotspots
AU: * Andrews, D L
EM: dandrews@rice.edu
AF: Department of Earth Science, Rice University, 6100 Main St. MS-126, Houston, TX 77005 United States
AU: Gordon, R G
EM: rgg@rice.edu
AF: Department of Earth Science, Rice University, 6100 Main St. MS-126, Houston, TX 77005 United States
AB:
We present a new method for objectively estimating the uncertainties of finite rotations of plates relative to hotspots. The
input for our analysis are coeval dated locations along two or more ancient hotspot tracks, as well as the present locations
of the hotspots. The output is a covariance matrix that can be used for testing hypotheses and for constructing confidence
regions. Unlike prior methods, our method incorporates the uncertainties in all 3 degrees of freedom of a finite rotation.
We apply our method to the Hawaiian and Louisville tracks on the Pacific plate and used the results to predict the track (and
uncertainties) of the Tristan da Cunha hotspot on the Nubian plate. The Tristan da Cunha track provides a critical test for
the fixed hotspot hypothesis because it has a fairly continuous track and has remained beneath the Nubian plate for the past
70 Myr and thus has not (in that time interval) greatly interacted with a mid-ocean ridge, which can severely complicate the
track.
In contrast to the results of prior studies, the predicted and observed Tristan da Cunha tracks have no significant
differences for the past 48 Myr, which includes reconstructions for 11 Ma (chron 5), 20 Ma (chron 6), 33 Ma (chron 13), and
39 Ma (chron 18). For example, the predicted location of the Tristan hotspot at 48 Ma (chron 21) is about 100 km ($\pm$ 250
km) from the nearest point on the Walvis Ridge, corresponding to motion between hotspots of 2 mm/yr $\pm$ 5 mm/yr. Numerical
experiments indicate that the improvement to the fit to the fixed hotspot assumption for reconstructions over the past 48
Myr come from many sources including improvements to the geomagnetic reversal time scale, improvements in relative plate
reconstructions, improvements in Pacific-hotspot rotation parameters, and improvements to age dates along the
Hawaiian-Emperor chain.
Unlike the reconstructions for the past 48 Myr, reconstructions for 56 Ma (chron 25) and 68 Ma (chron 31) have huge misfits
between predicted and observed tracks. For example, the misfit between the predicted and observed point at 68 Ma (chron 31)
is 1000 $\pm$ 250 km. A possible, but unlikely, explanation is that from 68 to 48 Ma, the Tristan da Cunha hotspot moved
relative to the Pacific hotspots at a rate of 50 $\pm$ 12 mm/yr. An alternative explanation is that the misfit is largely
accommodated by a "missing" plate boundary in the early Tertiary and Late Cretaceous global plate motion circuit, as is
independently indicated by global paleomagnetic data [Acton \& Gordon, 1994].
DE: 3040 Plate tectonics (8150, 8155, 8157, 8158)
DE: 8121 Dynamics, convection currents and mantle plumes
DE: 8155 Plate motions--general
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting