A Caltech Library Service

Decoding Earth’s plate tectonic history using sparse geochemical data

Tetley, Michael G. and Li, Zheng-Xiang and Matthews, Kara J. and Williams, Simon E. and Müller, R. Dietmar (2020) Decoding Earth’s plate tectonic history using sparse geochemical data. Geoscience Frontiers, 11 (1). pp. 265-276. ISSN 1674-9871. doi:10.1016/j.gsf.2019.05.002.

[img] PDF - Published Version
Creative Commons Attribution Non-commercial No Derivatives.

[img] PDF (Supplementary data) - Supplemental Material
Creative Commons Attribution Non-commercial No Derivatives.


Use this Persistent URL to link to this item:


Accurately mapping plate boundary types and locations through time is essential for understanding the evolution of the plate-mantle system and the exchange of material between the solid Earth and surface environments. However, the complexity of the Earth system and the cryptic nature of the geological record make it difficult to discriminate tectonic environments through deep time. Here we present a new method for identifying tectonic paleo-environments on Earth through a data mining approach using global geochemical data. We first fingerprint a variety of present-day tectonic environments utilising up to 136 geochemical data attributes in any available combination. A total of 38301 geochemical analyses from basalts aged from 5–0 Ma together with a well-established plate reconstruction model are used to construct a suite of discriminatory models for the first order tectonic environments of subduction and mid-ocean ridge as distinct from intraplate hotspot oceanic environments, identifying 41, 35, and 39 key discriminatory geochemical attributes, respectively. After training and validation, our model is applied to a global geochemical database of 1547 basalt samples of unknown tectonic origin aged between 1000–410 Ma, a relatively ill-constrained period of Earth's evolution following the breakup of the Rodinia supercontinent, producing 56 unique global tectonic environment predictions throughout the Neoproterozoic and Early Paleozoic. Predictions are used to discriminate between three alternative published Rodinia configuration models, identifying the model demonstrating the closest spatio-temporal consistency with the basalt record, and emphasizing the importance of integrating geochemical data into plate reconstructions. Our approach offers an extensible framework for constructing full-plate, deep-time reconstructions capable of assimilating a broad range of geochemical and geological observations, enabling next generation Earth system models.

Item Type:Article
Related URLs:
URLURL TypeDescription
Tetley, Michael G.0000-0002-2320-4239
Li, Zheng-Xiang0000-0003-4350-5976
Matthews, Kara J.0000-0003-3249-3869
Williams, Simon E.0000-0003-4670-8883
Müller, R. Dietmar0000-0002-3334-5764
Additional Information:© 2019 China University of Geosciences (Beijing) and Peking University. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( Received 10 October 2018, Revised 18 March 2019, Accepted 9 May 2019, Available online 17 May 2019. This research was supported by the Science Industry Endowment Fund (RP 04-174) Big Data Knowledge Discovery Project. MGT received additional support from a CSIRO-Data61 Postgraduate Scholarship. ZXL acknowledges the support of the Australian Research Council through a Laureate Fellowship grant (FL150100133). We thank Derrick Hasterok and an anonymous reviewer for their considered and constructive reviews, Craig O'Neill, Peter Cawood, Thomas Bodin and Andrew Merdith for helpful comments that improved the paper, and Jason Ash and Daniel Steinberg for their combined technical support in developing these analyses. This is a contribution to IGCP project 648. Analyses were conducted using the following open source tools: GPlates and pyGPlates (, Python (, and Project Jupyter (
Group:Seismological Laboratory
Funding AgencyGrant Number
Science and Industry Endowment FundRP 04-174
Commonwealth Scientific and Industrial Research Organisation (CSIRO)UNSPECIFIED
Australian Research CouncilFL150100133
Subject Keywords:plate tectonics; geochemistry; geodynamics; supercontinents; rodinia; big data
Issue or Number:1
Record Number:CaltechAUTHORS:20190517-100807069
Persistent URL:
Official Citation:Michael G. Tetley, Zheng-Xiang Li, Kara J. Matthews, Simon E. Williams, R. Dietmar Müller, Decoding earth's plate tectonic history using sparse geochemical data, Geoscience Frontiers, Volume 11, Issue 1, 2020, Pages 265-276, ISSN 1674-9871, (
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:95562
Deposited By: Tony Diaz
Deposited On:17 May 2019 17:18
Last Modified:16 Nov 2021 17:13

Repository Staff Only: item control page