Current plate velocities relative to hotspots: implications for hotspot motion, mantle viscosity and global reference frame

https://doi.org/10.1016/S0012-821X(01)00351-XGet rights and content

Abstract

The HS2-NUVEL1 model being used for current plate velocities relative to hotspots is shown to be generally inconsistent with the observed hotspot data sampled from non-Pacific regions. Instead, we determine the T22A model, which provides a good and consistent fit to the trends of globally distributed hotspot traces, but predicts plate velocities at hotspots systematically lower than the observed rates of hotspot volcanic migrations. As a result, it implies that the return flow in the lower mantle has a velocity about 1/4 of the plate velocity, and that the lower mantle is about 20 times more viscous than the upper mantle. Although hotspots are not relatively fixed, they do define a global reference frame for plate motion and mantle convection.

Introduction

Plate motion relative to the deep mantle is usually referred to as absolute motion to distinguish it from relative motion between plates. Based on the hypothesis that hotspots are fixed in the deep mantle and stationary with respect to each other [1], [2], [3], [4], [5], hotspots are commonly used as a reference frame for defining the absolute plate motion, and a number of absolute velocity models [6], [7], [8], [9] incorporating current plate motions have been determined consequently by inversion of observed hotspot data. The results of these models, conversely, provide a direct test to assumed hotspot fixity, since the current relative motion between plates is well known. In fact, these models have been widely accepted as evidences to support the fixed hotspot hypothesis, because they provide a good fit to the adopted hotspot data. However, the test of these models on the hypothesis is not sufficient. Models AM1 [6] and P073 [7] are constrained by observed trends of hotspot traces only. Therefore, it is impossible to distinguish whether the hotspots are stationary or they are moving in a direction opposite to the plate motion owing to a return flow in the lower mantle [10]. In models AM1-2 [8] and HS2-NUVEL1 [9], although five observed rates of hotspot volcanic migrations are used as constraints together with nine hotspot trends, all the data are sampled from hotspots distributed in the Pacific region. We think additional data from other geographical regions need to be incorporated.

Section snippets

Models and results

If hotspots are indeed stationary with respect to each other, all the observed hotspot trends and rates should be satisfied by a single model. Here we shall compare the HS2-NUVEL1 model [9], which is presently being used, with the hotspot data used by the AM1 model [6]. Among the 20 trend data from the AM1 data set, with the exception of six that have already been used by the HS2-NUVEL1 model and a questionable one [6] from the Iceland hotspot under the Eurasian plate, the remaining 13 trends

Discussion

Our preferred T22A model is determined completely based on the NUVEL-1A model [13] and the 22 hotspot trend data. The NUVEL-1A model can be considered as a reliable model for current relative motions between plates, not only because it provides a good fit to globally sampled spreading rates, transform fault azimuths and earthquake slip vectors, but also because its predictions on plate motions are well consistent with the space-based geodetic observations [14]. As the relative motions between

Acknowledgements

First of all, I thank my coauthor and former supervisor, Prof. Ren Wang, who passed away during the revision of this contribution, for his support, encouragement and enthusiasm. I will miss him.

We thank Yinting Li and Xunying Sun for helpful discussion, and two anonymous reviewers for constructive comments. This research was supported by the NSF of China.[SK]

References (36)

  • J.B. Minster et al.

    Numerical modelling of instantaneous plate tectonics

    Geophys. J. R. Astron. Soc.

    (1974)
  • J.B. Minster et al.

    Present-day plate motions

    J. Geophys. Res.

    (1978)
  • A.E. Gripp et al.

    Current plate velocities relative to the hotspots incorporating the NUVEL-1 global plate motion model

    Geophys. Res. Lett.

    (1990)
  • P. Olson

    Drifting mantle hotspots

    Nature

    (1987)
  • C. DeMets et al.

    Current plate motions

    Geophys. J. Int.

    (1990)
  • D.F. Argus et al.

    No-net-rotation model of current plate velocities incorporating plate motion model NUVEL-1

    Geophys. Res. Lett.

    (1991)
  • C. DeMets et al.

    Effect of recent revisions to the geomagnetic reveral time scale on estimates of current plate motions

    Geophys. Res. Lett.

    (1994)
  • R.G. Gordon, Present plate motions and plate boundaries, in: T.J. Ahrens (Ed.), Global Earth Physics: A Handbook of...
  • Cited by (37)

    • The trials and tribulations of the Hawaii hot spot model

      2021, Earth-Science Reviews
      Citation Excerpt :

      Matthews et al. (2016) used the plate chain through Antarctica but with relative plate motions in Zealandia as additional constraints, and in this way, they predict the Hawaii-Emperor bend much better than for a plate chain through Antarctica without this additional constraint (Torsvik et al., 2019). Wang and Liu (2006) suggested that the motion of hot spots may be systematically opposite to the direction of plate motions, and moving hot spots still form a useful global reference frame for defining absolute plate motion and mantle convection (Wang and Wang, 2001). Hence, Doubrovine et al. (2012) defined a new moving hot spot reference frame that reconciles plate tectonic and plume motions in the Pacific, Atlantic and Indian oceans back to Early Cretaceous time.

    • Absolute plate motions relative to deep mantle plumes

      2018, Earth and Planetary Science Letters
      Citation Excerpt :

      Such a mechanism deserves further research. An alternative explanation for the systematic discrepancy between hotspot rates and plate velocities is inter-hotspot motions, as suggested by Steinberger and O'Connell (1998), Steinberger (2000), Wang and Wang (2001), and Wang and Liu (2006). If the hotspot rates are reliable, the difference velocity obtained by subtracting a hotspot migration velocity from the T25M plate velocity will represent the hotspot motion relative to the T25M reference frame which is determined by hotspot trends.

    • Tectonically asymmetric Earth: From net rotation to polarized westward drift of the lithosphere

      2015, Geoscience Frontiers
      Citation Excerpt :

      Nonetheless, the hotspot reference frame continues to be a convenient – and easy to visualize – framework in which to study plate kinematics and dynamics. Many absolute present-day plate kinematic models have been proposed incorporating plate motions from magnetic anomalies (O'Connell et al., 1991; Wang and Wang, 2001; Gripp and Gordon, 2002; Cuffaro and Doglioni, 2007) and geodetic data (Crespi et al., 2007) into the hotspot frame. In other models, plate motions and plate reconstructions were attempted for the past (Gordon and Jurdy, 1986; Müller et al., 1993; Steinberger et al., 2004; O'Neill et al., 2005; Torsvik et al., 2010).

    • Hotspots, Large Igneous Provinces, and Melting Anomalies

      2015, Treatise on Geophysics: Second Edition
    View all citing articles on Scopus
    View full text