Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Jul 23;121(30):e2405160121.
doi: 10.1073/pnas.2405160121. Epub 2024 Jul 8.

Sediment subduction in Hadean revealed by machine learning

Affiliations

Sediment subduction in Hadean revealed by machine learning

Jilian Jiang et al. Proc Natl Acad Sci U S A. .

Abstract

Due to the scarcity of rock samples, the Hadean Era predating 4 billion years ago (Ga) poses challenges in understanding geological processes like subaerial weathering and plate tectonics that are critical for the evolution of life. The Jack Hills zircon from Western Australia, the primary Hadean samples available, offer valuable insights into magma sources and tectonic genesis through trace element signatures. However, a consensus on these signatures has not been reached. To address this, we developed a machine learning classifier capable of deciphering the geochemical fingerprints of zircon. This allowed us to identify the oldest detrital zircon originating from sedimentary-derived "S-type" granites. Our results indicate the presence of S-type granites as early as 4.24 Ga, persisting throughout the Hadean into the Archean. Examining global detrital zircon across Earth's history reveals consistent supercontinent-like cycles from the present back to the Hadean. These findings suggest that a significant amount of Hadean continental crust was exposed, weathered into sediments, and incorporated into the magma sources of Jack Hills zircon. Only the early operation of both subaerial weathering and plate subduction can account for the prevalence of S-type granites we observe. Additionally, the periodic evolution of S-type granite proportions implies that subduction-driven tectonic cycles were active during the Hadean, at least around 4.2 Ga. The evidence thus points toward an early Earth resembling the modern Earth in terms of active tectonics and habitable surface conditions. This suggests the potential for life to originate in environments like warm ponds rather than extreme hydrothermal settings.

Keywords: Hadean; S-type granite; machine learning; subduction; zircon.

PubMed Disclaimer

Conflict of interest statement

Competing interests statement:The authors declare no competing interest.

Figures

Fig. 1.
Fig. 1.
Traditional discriminant diagram of non-S-type (including TTG) and S-type zircon. (A) Diagram showing the boundary of I-(TTG) and S-type zircon and the trace element distribution of detrital, I-(TTG) and S-type zircon. Green shadow, upper line and lower line show the “traditional S-type region” and related I/S boundary (20), that zircon with P > 15 (μmol/g) and 0.77*P < (REE+Y) < 1.23*P is S-type zircon, and otherwise is non-S-type zircon. The inset shows the cumulative frequency of all Hadean–Archean detrital zircon. (B) P contents of detrital zircon and I- and S-type zircon used in this study (Dataset S1). The Inset shows P-content time series for igneous and sedimentary rocks from (33, 34) as bootstrapped averages in 10-million-year bins.
Fig. 2.
Fig. 2.
Trace element fingerprints of zircon. These bars show the composition and availability of the top 30 most common trace elements in the zircon dataset (SI Appendix, Table S1) and a detrital zircon dataset from ref. (Dataset S2). Box-and-whiskers plot shows the maximum, 3rd quartile, median, 1st quartile, and minimum values. The highest and lowest 20% of values of each element are excluded to reduce scatter. Trace element availability is calculated by n/N, where n is the available trace element data and N is the size of the dataset. Those elements used as input features for machine learning are indicated with stars.
Fig. 3.
Fig. 3.
Comparison of the machine-learning method (TSVM model) and the traditional P criterion in recognizing S-type zircon. (A) The accuracy of both methods for non-S- and S-type zircon with different P contents in the dataset. The distribution of P contents for non-S- and S-type zircon are shown in the red and blue histograms, respectively. The accuracy of each model in our dataset (both training and test sets) is calculated using a bin size of 5 μmol/g (Dataset S2). (B) Clear-cut discrimination of non-S- and S-type zircon by TSVM value = 0, where TSVM values mean the prediction output value of TSVM (defined in Materials and Methods). Blue open circles are S-type zircon, red open circles are non-S-type zircon, and orange filled circles are detrital zircon with Hadean to Archean ages. TSVM value is calculated based on the decision function of the machine-learning model (Materials and Methods), TSVM value = (5.66) * P + (0.54) * Y + (–1.66) * Ce+ (0.49) * Sm+ (–1.47) * Eu+ (–0.69) * Dy+ (–3.90) * Lu+ (0.42) * Th+ (0.58) * U, where >0 means S-type zircon and <0 indicates non-S-type zircon.
Fig. 4.
Fig. 4.
S-type detrital zircon on early Earth and the evolution of magma sources. (A) Hf isotopes of the Jack Hills zircon (9, 62, 63). The isotope trajectories of the putative depleted mantle (64), mafic crust (65), and upper continental crust (UCC) (66) are shown for reference, assuming silicate Earth differentiation at 4.5 Ga. Arrows mark the trend of prolonged internal crustal recycling of Hadean to Eoarchean crust and the transition at 3.8 to 3.7 Ga toward increasing juvenile contributions to magma sources. (B) Detrended Hf isotope of Hadean–Archean detrital zircon (67). (C) Bootstrapped average of S-type zircon proportion through time with 1 SE (SI Appendix, Table S4) and crustal recycling rate (68). (D) Stacked histogram of the distribution of I- and S-type zircon through time as classified by machine learning, including detrital zircon from the Jack Hills, Australia, and 12 other locations. The Inset shows δ18O step-change of Jack Hills zircon (61) and a histogram of zircon older than 4.0 Ga. The dashed line shows the probability density plot of detrital zircon along the traverse through the Jack Hills belt (69). See SI Appendix, Table S3, for the individual distributions of S-type zircon in each of the 12 locations studied. All the zircon presented have U-Pb ages that are <10% discordant.
Fig. 5.
Fig. 5.
Secular variation of the proportion of S-type detrital zircon throughout Earth history. The proportion of S-type zircon shown as grey curve is calculated by bootstrap method, involving resampling 100 times to calculate the S-type proportion within a bin size of 100 My, as error bars denote ±2 SEM. Jack Hill zircon is shown in orange and other detrital zircon is shown in dark blue. The green trend line shows the peaks decline from Hadean to Phanerozoic. The blue square shadow indicates the “Boring Billion” [a.k.a. “Balanced Billion” (84) ranging from 1.8 to 0.9 Ga. Upper bars indicate supercontinent (gray) and megacontinent (blue) time periods (83). See SI Appendix, Table S4, for the calculated proportion of S-type zircon from Hadean to the present day. All the zircon presented have concordant ages (<10% discordance).

Similar articles

Cited by

References

    1. Palin R. M., Secular change and the onset of plate tectonics on Earth. Earth-Sci. Rev. 207, 103172 (2020). 10.1016/j.earscirev.2020.103172. - DOI
    1. Maruyama S., Santosh M., Azuma S., Initiation of plate tectonics in the Hadean: Eclogitization triggered by the ABEL Bombardment. Geosci. Front. 9, 1033–1048 (2018).
    1. Tang M., Chen K., Rudnick R. L., Archean upper crust transition from mafic to felsic marks the onset of plate tectonics. Science 351, 372–375 (2016). - PubMed
    1. Shirey S. B., Richardson S. H., Start of the Wilson cycle at 3 GA shown by diamonds from subcontinental mantle. Science 333, 434–436 (2011). - PubMed
    1. Stern R. J., Leybourne M. I., Tsujimori T., Kimberlites and the start of plate tectonics. Geology 44, 799–802 (2016).