Current absolute plate motion from seismic anisotropy and hotspot tracks; bounds on the latitudinal shift of the Hawaiian hotspot during formation of the Hawaiian island and seamount chain
Gordon, Richard G.
Doctor of Philosophy
Hotspots, the volcanic regions thought to be fed by underlying mantle that is anomalously hot compared with the mantle elsewhere, can serve as a reference frame to track the motion of the plates relative to the mantle beneath the asthenosphere. The application of the hotspot reference frame, however, is limited by the relative hotspot motion ranging from a few mm a-1 [Morgan, 1971; Duncan, 1981; Muller et al., 1993; Koivisto et al. 2013] to 80 mm a-1 [Raymond et al., 2000]. This dissertation aims to evaluate the hotspot reference frame. In the first part, I estimate the plate motion relative to the sub-asthenospheric mantle from seismic anisotropy data [Kreemer, 2009]. Prior studies based on seismic anisotropy assume that errors in the azimuths inferred from shear-wave splitting are uncorrelated. In this dissertation, I show that the residuals of azimuths inferred from shear-wave splitting beneath any one tectonic plate are strongly correlated with other residuals from the same plate. I account for these correlations in an inversion for absolute plate angular velocity by adopting a two-tier analysis of plate absolute velocities. First I find the pole of rotation and confidence limits for each plate individually. Then I perform a global inversion in which each plate is represented not by multiple individual estimates of the orientation of seismic anisotropy but by a single best-fitting pole and confidence limits. In the second part, I estimate plate motion from two hotspot data-sets: T57, consisting of 57 trends of hotspot tracks distributed on ten major plates [Morgan and Phipps Morgan, 2007]; and HS3, consisting of 11 trends and 2 propagation rates of hotspot tracks distributed on four major plates [Gripp and Gordon, 2002]. I test the assumption of uncorrelated errors in hotspot trend, following the same approach as previously used for seismic anisotropy data. I estimate the best-fitting angular velocities after removing the correlated errors within plates in T57 hotspot data-set. Volcanic age dates used in HS3 tend to be younger than true ages due to the bias inherent in K-Ar measurements. Therefore, the volcanic propagation rates tend to be too high. I use the difference between an astrogeochronologic-based, and a K-Ar-based, geomagnetic reversal time scale as a proxy to recalibrate the K-Ar age dates, which reduces the volcanic propagation rates of the Hawaii hotspot track and the Society hotspot track by 8% and 4% respectively. The global set of best-fitting angular velocities estimated from seismic anisotropy data in the first part, are compared with those estimated from hotspot data-sets in the second part. In the third part, I test the hotspot mobility from the magnetic anomaly data due to seafloor spreading. I estimate the location of the Pacific paleomagnetic pole of anomaly 20r (44 Ma B.P.) by analyzing the skewness of marine magnetic anomalies. Then I reconstruct this paleomagnetic pole in the Pacific hotspot reference frame [Andrew et al., 2006; Koivisto et al., 2013], and compare it with prior paleomagnetic pole at the similar age in the Indo-Atlantic hotspot reference frame [Besse and Courtillot, 2002].
Plate tectonics; Seismic anisotropy; Hotspot reference frame; Palaeomagnetism; Geophysics