A Giant Impact Origin for the First Subduction on Earth

A Giant Impact Origin for the First Subduction on Earth
Qian Yuan1,2
, Michael Gurnis2
, Paul D. Asimow1
, and Yida Li2
1
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, USA, 2
Seismological
Laboratory, California Institute of Technology, Pasadena, CA, USA
Abstract Hadean zircons provide a potential record of Earth's earliest subduction 4.3 billion years ago. It
remains enigmatic how subduction could be initiated so soon after the presumably Moon‐forming giant impact
(MGI). Earlier studies found an increase in Earth's core‐mantle boundary (CMB) temperature due to the
accumulation of the impactor's core, and our recent work shows Earth's lower mantle remains largely solid, with
some of the impactor's mantle potentially surviving as the large low‐shear velocity provinces (LLSVPs). Here,
we show that a hot post‐impact CMB drives the initiation of strong mantle plumes that can induce subduction
initiation ∼200 Myr after the MGI. 2D and 3D thermomechanical computations show that a high CMB
temperature is the primary factor triggering early subduction, with enrichment of heat‐producing elements in
LLSVPs as another potential factor. The models link the earliest subduction to the MGI with implications for
understanding the diverse tectonic regimes of rocky planets.
Plain Language Summary Plate tectonics remains unique to Earth, but when and how it started is
debated. Earth's oldest minerals indicate a clement surface by 4.3 Ga, resembling the modern Earth with its
granitic crust and oceans. Granite is most easily explained as originating from subduction. However, the
mechanisms for subduction initiation, especially so soon after the Moon‐forming giant impact, remain elusive.
Earlier studies indicate that the core‐mantle boundary (CMB) temperature is increased due to accumulation of
the impactor's core during the impact. Our recent work further shows that the lower half of Earth's mantle
remains mostly solid after this impact and that parts of the impactor's mantle might have survived as the two
seismically‐observed large low‐shear velocity provinces (LLSVPs). In this study, we perform whole‐mantle
convection models to illustrate that strong mantle plumes can arise, weaken the lithosphere, and eventually
initiate subduction ∼200 Myr after the giant impact. Our systematic computations reveal that the hot CMB
temperature after the impact is the primary factor determining whether there is early subduction initiation, with
enrichment of heat‐producing elements in LLSVPs as another potential contributor. Our model ties the Moon's
formation to incipient subduction, providing insights for understanding the diverse tectonic regimes of rocky
planets.
1. Introduction
As planetary exploration continues, Earth remains unique as the only planet that is known to operate in the plate
tectonic regime. In plate tectonics, subduction is the major process affecting the geodynamic and geochemical
evolution of our planet, but when and how the first subduction occurred remains contentious. Although their
interpretation is debated (Kemp et al., 2010), analyses of Hadean detrital zircons show geochemical signals
consistent with subduction as early as 4.3 Ga (Valley et al., 2002; Watson & Harrison, 2005). Thus, the earliest
subduction may have followed less than ∼200 million years after the Moon‐forming giant impact (MGI) (Canup
& Asphaug, 2001). The collision of a protoplanet called Theia with the proto‐Earth at ∼4.51 Ga is expected to
largely reset the initial conditions of Earth's evolution, creating a need for systematic and quantitative exploration
of the planetary state established by MGI and its lasting influence on Earth's tectonic evolution. Here we propose
that the thermal and chemical structure established by a canonical MGI creates favorable conditions for the
initiation of subduction in Earth's earliest period, exemplifying the pivotal influence of initial conditions set by
giant impact processes for the tectonic evolution of terrestrial planets.
A number of mechanisms have been proposed to initiate early subduction (Stern & Gerya, 2018), but all except
three of these models require a pre‐existing weakness in the lithosphere. The exceptions are grainsize evolution
(Foley et al., 2014), impact‐driven (Borgeat & Tackley, 2022; O’Neill et al., 2017) and plume‐induced (Gerya
et al., 2015; Ueda et al., 2008) subduction initiation. However, large (>700 km) and energetic impacting bolides
capable of initiating subduction are rare and only occurred very early in Earth's history (Marchi et al., 2014;
RESEARCH LETTER
10.1029/2023GL106723
Key Points:
• The mantle thermochemical structure
left by the Moon‐forming impact trig-
gers strong plumes that may have
initiated the first subduction
• The strong mantle plumes may rise
from a heated core or from large low‐
shear velocity provinces enriched in
heat‐producing elements
• Early giant impact events may have a
profound influence on the diverse
tectonic regimes of rocky planets
Supporting Information:
Supporting Information may be found in
the online version of this article.
Correspondence to:
Q. Yuan,
qyuan@caltech.edu
Citation:
Yuan, Q., Gurnis, M., Asimow, P. D., &
Li, Y. (2024). A giant impact origin for the
first subduction on Earth. Geophysical
Research Letters, 51, e2023GL106723.
https://doi.org/10.1029/2023GL106723
Received 10 OCT 2023
Accepted 14 APR 2024
Author Contributions:
Conceptualization: Qian Yuan
Formal analysis: Qian Yuan
Funding acquisition: Qian Yuan,
Michael Gurnis
Investigation: Qian Yuan, Yida Li
Methodology: Qian Yuan
Software: Qian Yuan
Supervision: Michael Gurnis, Paul
D. Asimow
Validation: Qian Yuan, Michael Gurnis,
Paul D. Asimow
Visualization: Qian Yuan
Writing – original draft: Qian Yuan
Writing – review & editing: Qian Yuan,
Michael Gurnis, Paul D. Asimow, Yida Li
© 2024. The Authors.
This is an open access article under the
terms of the Creative Commons
Attribution License, which permits use,
distribution and reproduction in any
medium, provided the original work is
properly cited.
YUAN ET AL. 1 of 10

O’Neill et al., 2017), which may coincide with the period of magma ocean solidification. Recently, plume‐
induced subduction has been increasingly proposed as a viable mechanism for subduction initiation in the
early Earth (Baes et al., 2021). Nevertheless, it is unclear how to generate strong mantle plumes in a vigorous
convecting Hadean mantle (Korenaga, 2008).
The MGI is typically thought violent enough to melt a substantial portion or even the entire mantle, especially in
scenarios involving high energy and high angular momentum (Lock & Stewart, 2017; Nakajima & Steven-
son, 2015). Using two different giant impact computational methods with improved equations of state (Stewart
et al., 2020) and at high (Deng et al., 2019) and ultra‐high (Kegerreis et al., 2022) resolution, our recent work
(Yuan et al., 2023) demonstrates that the lower half of Earth's mantle would have remained mostly solid after a
canonical MGI. It also suggests that large intact domains of Theia's likely Fe‐rich mantle are candidates of the two
seismically‐observed large low‐shear velocity provinces (LLSVPs). Furthermore, the concomitant core formation
process during the MGI may substantially heat the core to raise the temperatures at the core‐mantle boundary
(CMB) (Canup, 2004; Deng et al., 2019). This unique thermochemical structure, featuring a solid lower mantle
and a hotter CMB resulting from the MGI, provides ideal conditions for the formation of strong mantle plumes, a
scenario less likely when the mantle and core are in equilibrium. Accretion scenarios built from only smaller‐scale
impacts dissipate more of their gravitational potential energy in the mantle and yield a more stable initial thermal
structure. Here, we test the hypothesis that such strong mantle plumes can be developed in the early Hadean,
weakening the lithosphere and eventually causing subduction initiation (Figure 1). We model this scenario using
2‐D and 3‐D whole‐mantle thermo‐chemical‐mechanical models with a visco‐elasto‐plastic rock rheology.
2. Geodynamic Models
We compute whole‐mantle convection models in 2D and 3D Cartesian geometries using Underworld2, an open‐
source, particle‐in‐cell, finite element code (Mansour et al., 2020; Moresi et al., 2007).
2.1. Governing Equations and Material Description
The computations solve the conservation equations of mass, momentum, and energy:
∇ · u = 0 (1)
∇ · σ′ ∇P + ρgẑ = 0 (2)
ρcp (
∂T
∂t
+ u · ∇T) ∇ · k∇T H = 0. (3)
Figure 1. A schematic illustration of the path from the Moon‐forming giant impact (MGI) to plume‐induced subduction in the
early Hadean. A canonical MGI (a) leads to a two‐layered mantle structure with a mostly solid lower mantle and an elevated
core‐mantle boundary temperature (b) (Deng et al., 2019; Yuan et al., 2023). After solidification of the upper magma ocean
within several million years (Hamano et al., 2013; Nikolaou et al., 2019), we assume a proto‐lithosphere gradually formed
over 50–100 million years (c), which was then destabilized and segmented by a mantle plume from an large low‐shear
velocity province made of the Theia mantle remnant (Yuan et al., 2023) (d).
Geophysical Research Letters 10.1029/2023GL106723
YUAN ET AL. 2 of 10
19448007,
2024,
9,
Downloaded
from
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023GL106723
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
(https://onlinelibrary.wiley.com/terms-and-conditions)
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

where u is velocity, σ′ is deviatoric stress, P is pressure, T is temperature, ρ is density, g is gravitational accel-
eration, ẑ is the vertical unit vector, cp is heat capacity, k is thermal conductivity and H is heat production.
An incompressible Maxwell‐rheological model is adopted to reflect viscoelastic behavior:
ε̇ij =
1
2Ke
σ̇ij +
1
2η
σij (4)
where ε̇ij is the strain rate tensor, σ̇ij is the stress rate tensor, Ke is the shear modulus for elasticity, and η is the
viscosity. Non‐Newtonian and temperature‐dependent rheology is specified by:
η(T,ε̇) = η0e
( E
nRT
E
nRT0
)
(
ε̇
ε̇0
)
1 n
n
(5)
where E is the activation energy, R is the gas constant, and n is the exponent for shear thinning in dislocation
creep. Material yielding is included by imposing an upper limit on the stress, thus the effective viscosity becomes
ηeff = min(η(T,ε̇),
τy
ε̇II
) (6)
where τy is the yield stress and ε̇II is the second invariant of the strain rate tensor. The effective viscosity is
restricted to the range between 1018
and 1024
Pa s. The yield stress follows the Drucker–Prager yielding criteria
with an upper cut‐off:
τy = min C + μyp,τy0) (7)
where τy is the yield stress, C and μy are cohesion and friction coefficient, p is pressure, and τy0 is the maximum
yield stress.
Similar to previous studies (Li & Gurnis, 2022), weakening processes are approximated by reducing the yield
stress with plastic strain, following a two‐stage process: prior to a strain saturation, C and μy linearly decrease as
follows:
C = (τyf C0) min(
εP
εP0
,1) + C0 (8)
μy = μy0 μy0 min(
εP
εP0
,1) (9)
where the variables are the accumulated plastic strain εP, the reference plastic strain εP0 that determines the rate of
weakening (thus, the inverse of εP0 is the weakening rate), minimum yield stress τyf, initial cohesion C0, and initial
friction coefficient μy0.
Three materials are present within the model domain: a 30‐km‐thick mafic crust on top that can metamorphose to
eclogite at depth (Hacker et al., 2003), LLSVP material, and background mantle.
2.2. Model Setup in 2D
Most 2D models have a domain that is 5,780 × 2,890 km with 256 × 128 linear, quadrilateral elements, with 30
Lagrangian particles in each element. We have verified the validity of 2D results with a higher‐resolution grid of
512 × 256 elements, and in longer aspect ratio of 14,450 × 2,890 km with 640 × 128 elements. A free‐slip
boundary condition is applied to all model domain boundaries. The temperature at the upper boundary is
maintained at 273 K, and the bottom boundary is fixed at a varying temperature (5273 K down to 3273 K). The
lithosphere's maximum yield strength in 2D is set between 80 and 180 MPa. As seismic observations require that
YUAN ET AL. 3 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

LLSVPs have a distinct bulk modulus (Tan & Gurnis, 2005), we use a depth‐dependent density profile within the
LLSVP. The values of physical parameters are listed in Table S1 of Supporting Information S1.
2.3. Model Setup in 3D
Three‐dimensional plume‐induced subduction computations are used to verify the 2D results and consist of a
Cartesian domain 2,890 km deep, 2,000 × 2,000 km horizontally. We use a resolution of 128 × 80 × 80 elements
with a uniform grid spacing in each coordinate direction. Each element initially has 30 material points randomly
located in each element. Free slip conditions are applied to all the boundaries. Maximum yield stresses of lith-
osphere between 60 and 100 MPa were tested in 3D.
3. Results
Following the MGI, a surface magma ocean extending to a certain depth is expected. Tracking magma ocean
solidification is beyond the scope of this study, but it is generally accepted that the solidification process would be
mostly complete within several million years (Hamano et al., 2013; Nikolaou et al., 2019). For simplicity, our
models begin 50 Myr after the magma ocean solidification; the initial state is derived from a 2D thermo-
mechanical model with a proto‐lithosphere defined by a half‐space cooling model with a prescribed age of 50 Ma.
Most model runs include an intrinsically dense layer of Theia mantle materials above the CMB. Current estimates
of the present‐day CMB temperature that account for a lower boundary layer range from ∼3473 to 4573 K
(Deschamps & Cobden, 2022), but because of Earth's secular cooling, the Hadean CMB temperature should have
been higher. Moreover, a hot CMB is anticipated after the MGI due to the deposition of Theia's iron core, as
shown in previous MGI simulations (Canup, 2004). Consequently, for most cases we used a higher CMB tem-
perature (5273 K) and a boundary‐layer contrast of 1273 K between the CMB and the initial LLSVP temperature.
Such high CMB temperature may lead to partial or complete melting in the boundary layer, but temperature
should drop quickly below the solidus of likely lower mantle compositions above the CMB. A thin partial melting
zone within the mantle does not substantially alter the dynamics, since the fully molten core sets the lower free
slip boundary condition in all cases. Furthermore, we note that similar or higher CMB temperatures were used in
previous convection simulations (Nakagawa & Tackley, 2010), and we confirm our results with lower CMB
temperatures (4273 and 3773 K). We mostly used a 80 MPa yield strength for the proto‐lithosphere (Rudi
et al., 2022).
Model runs with this initial configuration consistently show that a mantle plume nucleates atop the LLSVP‐like
thermochemical pile, ascends through the mantle, ruptures the lithosphere, and initiates subduction after ∼120
Myr (Figure 2). In detail, typical model evolution can be divided into several stages. First, the initial layer of
intrinsically dense material (Figure 2a) at the CMB concentrates into discrete piles due to a combination of strong
CMB heating and its depth‐dependent density (Tan & Gurnis, 2005). Meanwhile, the lithosphere grows thicker
due to inefficient cooling while in a stagnant‐lid convection, although the mode of magmatism may affect a
planet's tectonic regime (Lourenço et al., 2018). At 99 Myr, a strong mantle plume develops from the top of the
central pile (Figure 2b). This plume quickly rises to the upper mantle, becoming narrower due to the viscosity
decrease in the transition zone. As the plume approaches the surface, the lithosphere is locally weakened through
yielding above the plume and, at 109 Myr, the persistent localized thinning leads to plume penetration through the
lithosphere, forming a high‐temperature, low‐viscosity wedge structure (Figure 2c). Meanwhile, the proximal
lithosphere thickens around the plume head, resulting in two lithospheric‐scale shear bands on each side
(Figure 2c). The plume then starts to spread near the surface, overcoming the yield strength of the surrounding
lithosphere segments and forming shear zones along the subducting plate boundary that promote self‐sustaining
and retreating subduction at 123 Myr (Figure 2d).
Our reference case above shows that strong plumes can arise from the LLSVPs and initiate subduction when the
lithosphere's strength is less than 100 MPa, a threshold yield stress consistent with previous geodynamic studies
(Moresi & Solomatov, 1998; Van Heck & Tackley, 2008). Here, we aim to explore the maximum lithosphere
strength that the MGI‐caused mantle plumes can breach to initiate subduction and assess the influence of LLSVPs
on this capability (Figure 3). Initially, we increase the lithosphere yield strength from 80 to 100 MPa and find that
it becomes more challenging to induce subduction; subduction only occurs on one side of the plume head
(Figure 3a). When the lithosphere yield stress is further increased to 150 MPa, plumes from LLSVPs do not
penetrate through the lithosphere to induce subduction but instead spread at the base of the lithosphere
YUAN ET AL. 4 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

(Figure 3b). Next, for the same calculation after removing LLSVPs materials, we find that strong plumes rising
from the CMB can initiate subduction, as shown in Figure 3c for a case with a yield stress of 150 MPa, and in
another case with a yield stress of 180 MPa (Figure 3d).
As studied in previous research on plume‐induced subduction, temperature, size, and buoyancy of plumes play a
major role in subduction initiation. Therefore, we systematically explore the influence of CMB temperature,
which significantly affects all these factors in models where plumes are self‐consistently generated. We lower the
CMB temperature to 4773 K (Figure 3e) and 3773 K (Figure 3f), and find an increase in the time delay for plume‐
induced subduction initiation (Figures 3i and 3j) due to the lower buoyancy of the plumes. Nonetheless, sub-
duction still occurs as long as CMB temperature is ≥3773 K. When CMB temperature is 3273 K, no subduction is
initiated (Figures 3g and 3k). However, if we assume that LLSVP materials are enriched in heat‐producing el-
ements for the same calculation (Citron et al., 2020), the results indicate that plumes rising from hotter LLSVPs
can still generate subduction (Figures 3h and 3l).
Figure 2. Temporal evolution for the temperature (left column) and viscosity (right column) fields of the reference case showing a LLSVP‐sourced plume inducing
subduction initiation. (a) Initial set up of the model; (b) Plume rising from the LLSVP; (c) Plume penetrates the lithosphere; (d) Slab retreat and melting of crust. The
yellow contours in temperature field marks the boundary of different materials including crust, lithosphere, molten materials, and large low‐shear velocity province
materials.
YUAN ET AL. 5 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

Computations with different rheologies, compositions, lithosphere thicknesses, and model geometries showed
that these factors have a second‐order influence on outcomes compared to the maximum yield strength and CMB
temperature. Cases with different rheologies show that a smaller viscosity of the LLSVPs (Figures S1a in
Supporting Information S1), a larger reference viscosity (Figures S1b in Supporting Information S1), and a more
strongly temperature‐dependent viscosity (Figures S1c in Supporting Information S1) do not significantly alter
the plume‐induced subduction process, except for the onset time (77–320 Myr)—in general, higher mantle
viscosity leads to later plume‐induced subduction initiation. We considered cases with lower (Figures S1d in
Supporting Information S1) and higher (Figures S1e in Supporting Information S1) LLSVP density than the
reference case, as in our earlier work (Yuan and Li, 2022a, 2022b), as well as a more compressible (more depth‐
dependent density) LLSVP case (Figures S1f in Supporting Information S1). We find the overall dynamics of
these computations closely resemble those of the reference case. Different ages of the lithosphere from 25 to 100
Myr (Figures S2a–S2c in Supporting Information S1) and different weakening rates (Figures S2d and S2e in
Supporting Information S1) were also considered, and their results remain similar to the reference case. Models
with higher resolution and periodic boundary conditions (Figure S2f in Supporting Information S1), and a
different aspect ratio (5:1) (Figure S3 in Supporting Information S1) do not significantly differ in the results.
We further test our model using 3D whole‐mantle thermomechanical models, as plume‐induced subduction
necessitates a three‐dimensional perspective. As in the 2D geometry, we observe that strong mantle plumes can be
generated from the top of an LLSVP‐like pile (Figures 4a and 4b). As the plume rises and pierces the lithosphere
(Figure 4c), it spreads over broken lithosphere segments in a circular area, pushing them downward into the
mantle by overcoming their yield strength. As the plume continues to spread, the surface shear zones extend
deeper along the subducting plate boundary, which facilitates subduction and retreat of the circular subduction
zone (Figure 4d). The evolution resembles previous 3D models of plume‐initiated subduction in which the plume
Figure 3. Snapshots of viscosity and temperature field of eight representative cases with varying lithosphere yield stress (τy0) and core‐mantle boundary (CMB)
temperatures (T_cmb). Each panel from (a–h) represents a different calculation, with panels (i–l) corresponding to the viscosity field for cases in panels (e–h). In panels
(a–b), the lithosphere yield strength is 100, 150, 150, and 180 MPa, respectively. Note there are no large low‐shear velocity province (LLSVP) materials in panels (c–d).
For panels (e–g), the CMB temperature is 4773 K (e), 3773 K (f), 3273 K (g). Panel (h) has the same parameters as panel (g) except that the LLSVP materials
(Radi_LLSVP) are enriched in heat‐producing elements.
YUAN ET AL. 6 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

was imposed (Gerya et al., 2015). In another 3D case, we find that a slightly higher yield strength for the lith-
osphere (80 MPa) does not significantly influence the results (Figure S4 in Supporting Information S1), but
subduction is not initiated if the lithosphere strength is 100 MPa as shown in another case (Figure S5 in Sup-
porting Information S1). However, if we simply remove LLSVP materials, plumes rising from the CMB can still
cause subduction initiation (Figure S6 in Supporting Information S1), similar to what we identified in our 2D
models. In the next 3D case, we lower the CMB temperature to 3773 K, the plume does not penetrate the lith-
osphere or initiate subduction (Figure S7 in Supporting Information S1). Thus, the systematic trends we identified
in 2D computations—that a higher CMB temperature is crucial in plume‐induced subduction when the litho-
sphere strength is not very strong—remain unchanged in 3D, though the threshold values may differ due to effect
of geometry, plume size and morphology.
4. Discussion and Conclusion
Hadean detrital zircons and anomalies in isotope ratios derived from short‐lived radioactive decay systems
support the notion that subduction operated in the Hadean (Hyung & Jacobsen, 2020; Valley et al., 2002; Watson
& Harrison, 2005). Lithospheric dripping, suggested as an alternative mechanism in early Earth tectonics, may
arise from plume‐lid interaction (Fischer & Gerya, 2016) or high magmatic intrusion/extrusion ratios (Lourenço
et al., 2020). However, lithospheric dripping represents non‐plate‐tectonic subduction, and a recent experimental
study (Hastie et al., 2023) lends support to subduction over the dripping mechanism for the formation of Earth's
earliest continental crust. Their results indicate that the absence of garnet in the lower crust would complicate the
initiation of the gravitational instability typically necessary for crustal drips and delamination, though not
necessarily in grain‐damage scenarios. Other studies suggest an early surface ocean could thermally weaken the
Figure 4. Temporal evolution for the viscosity (upper row) and compositional (lower row) field of the 3D reference case showing LLSVP‐sourced plume‐induced
subduction initiation. (a–) Model snapshots at 0 Myr (a), 104 Myr (b), 108 Myr (c), and 111 Myr (d), respectively.
YUAN ET AL. 7 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

lithosphere, aiding subduction initiation (Korenaga, 2007). Yet, advanced models on crack formation and ser-
pentinization effects leading to subduction are still lacking. Plume‐induced subduction presents an appealing
mechanism for the earliest subduction initiation, as it does not necessitate pre‐existing weak zones or other
external forces (Baes et al., 2021). It also holds the potential to trigger plate tectonics (Gerya et al., 2015), but
developing such strong mantle plumes in an expected hot Hadean Earth is difficult. Our computations show that
there is a path to developing such plumes, and it is closely linked to the presumed MGI, which happened <150 Ma
before the oldest dated terrestrial zircons. Following the earliest subduction, the proto‐lithosphere can usually be
recycled globally, depending on lithosphere strength and plume‐slab interactions, with subsequent periods of
sluggish lid convection where the subducting slab periodically tears off, similar to the tectonic mode inferred for
early Earth (Foley, 2020).
The present model builds on results of high‐resolution MGI simulations that yield an end‐state with a largely solid
lower mantle (Yuan et al., 2023) and an expected increase in CMB temperature due to the gravitational deposition
of Theia's core during the agglomeration process (Asphaug, 2014). Although the exact CMB temperature increase
varies in different MGI models, it should be higher than that resulting from the core‐formation process during
previous smaller impacts (Nimmo, 2022). Pebble accretion may also increase CMB temperatures, while recent
studies suggest this mechanism is so efficient that it could melt the entire mantle (e.g., Johansen et al., 2023),
leaving no solid mantle to produce strong mantle plumes. Our results show that the unique context caused by the
MGI can give rise to strong mantle plumes capable of inducing subduction even when the maximum lithosphere
yield strength is >150 MPa, which is noticeably higher than inversions based on present‐day plate tectonics (Rudi
et al., 2022). In addition, our recent work suggests parts of Theia's iron‐rich mantle may persist over time and form
the seismically‐observed LLSVPs (Yuan et al., 2023). Here, we assess the influence of dense LLSVPs on sub-
duction initiation. Our results suggest that plumes rising from the CMB, rather than from LLSVPs, are more
capable of generating subduction and allow the mechanism to operate under a stronger lithosphere. In addition,
the timing of subduction due to the presence of LLSVPs is shown to be slightly later compared to the case where
plumes rise from the CMB, consistent with an earlier study (Kreielkamp et al., 2022). Nevertheless, our results
suggest that plumes rising from LLSVPs are still able to generate subduction to match the Hadean zircon records
as long as the CMB temperature is ≥3773 K. As the CMB temperature further decreases over time, plumes may
not be strong enough to induce new subduction (Figure 3e). However, if LLSVP materials are enriched in heat‐
producing elements, LLSVP‐sourced plumes may continue to trigger transient episodes of subduction (Figures 3h
and 3l). The enrichment of heat‐producing elements in LLSVPs can be created by Earth's internal differentiation
process (Citron et al., 2020). If LLSVPs are made of Theia's mantle (Yuan et al., 2023), they are likely also
enriched in heat‐producing elements, given that the canonical MGI suggests Theia resembles Mars, which has
been estimated to have higher heat‐producing elements than Earth (Table S2 in Supporting Information S1). Thus,
exploring the potential link between ancient plume‐induced subduction sites and the location of LLSVPs may
provide a powerful means to reconstruct global plate motions beyond 200 Ma.
Earth's unique convective regime, mobile‐lid plate tectonics, is distinctly absent on its twin planet, Venus. This is
true even though there are potential sites of subduction, at least on modern Venus, thought to be plume‐induced,
which do not apparently mature into full mobile‐lid tectonics. This may be due to the high surface temperature of
Venus, which allows rapid healing of damage along shear zones (Bercovici & Ricard, 2014) and arrests the
formation of continuous plate boundaries (Davaille et al., 2017). Alternatively, it may be harder to convert Venus
crust to eclogite, hindering development of planetary‐scale subduction (Chen et al., 2022). In addition to the
various ongoing differences between the two planets, another significant distinction may be the initial conditions.
The absence of a moon and a core‐generated dynamo suggest that Venus never experienced late Moon‐forming
giant impacts, based on planetary accretion models (Jacobson et al., 2017). Our proposed link between the MGI
and subduction initiation on Earth thus implies that the conditions for nucleation of early subduction on Earth may
have been absent on Venus. Given that a planet's tectonic development is history‐dependent (O’Neill et al., 2016),
the early giant impact events may play a crucial role in shaping their divergent evolutionary paths.
Data Availability Statement
The geodynamic code used in this work is available in Mansour et al. (2020). The model data on which this article
is based are available at figshare (Yuan, 2024). The code used to make the figures can be accessed at Tian
et al. (2023) and Ahrens et al. (2005).
YUAN ET AL. 8 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

References
Ahrens, J., Geveci, B., Law, C., Hansen, C., & Johnson, C. (2005). 36‐paraview: An end‐user tool for large‐data visualization 50038–1. (Version
5.11.0) [Software]. The Visualization Handbook, 717. https://www.paraview.org/download/
Asphaug, E. (2014). Impact origin of the Moon? Annual Review of Earth and Planetary Sciences, 42(1), 551–578. https://doi.org/10.1146/
annurev‐earth‐050212‐124057
Baes, M., Stern, R. J., Whattam, S., Gerya, T. V., & Sobolev, S. V. (2021). Plume‐induced subduction initiation: Revisiting models and ob-
servations. Frontiers in Earth Science, 9, 766604. https://doi.org/10.3389/feart.2021.766604
Bercovici, D., & Ricard, Y. (2014). Plate tectonics, damage and inheritance. Nature, 508(7497), 513–516. https://doi.org/10.1038/nature13072
Borgeat, X., & Tackley, P. J. (2022). Hadean/Eoarchean tectonics and mantle mixing induced by impacts: A three‐dimensional study. Progress in
Earth and Planetary Science, 9(1), 38. https://doi.org/10.1186/s40645‐022‐00497‐0
Canup, R. M. (2004). Simulations of a late lunar‐forming impact. Icarus, 168(2), 433–456. https://doi.org/10.1016/j.icarus.2003.09.028
Canup, R. M., & Asphaug, E. (2001). Origin of the Moon in a giant impact near the end of the Earth’s formation. Nature, 412(6848), 708–712.
https://doi.org/10.1038/35089010
Chen, J., Jiang, H., Tang, M., Hao, J., Tian, M., & Chu, X. (2022). Venus’ light slab hinders its development of planetary‐scale subduction. Nature
Communications, 13(1), 7647. https://doi.org/10.1038/s41467‐022‐35304‐3
Citron, R. I., Lourenço, D. L., Wilson, A. J., Grima, A. G., Wipperfurth, S. A., Rudolph, M. L., et al. (2020). Effects of heat‐producing elements on
the stability of deep mantle thermochemical piles. Geochemistry, Geophysics, Geosystems, 21(4), e2019GC008895. https://doi.org/10.1029/
2019GC008895
Crameri, F., Shephard, G. E., & Heron, P. J. (2020). The misuse of colour in science communication. Nature Communications, 11(1), 5444. https://
doi.org/10.1038/s41467‐020‐19160‐7
Davaille, A., Smrekar, S. E., & Tomlinson, S. (2017). Experimental and observational evidence for plume‐induced subduction on Venus. Nature
Geoscience, 10(5), 349–355. https://doi.org/10.1038/ngeo2928
Deng, H., Ballmer, M. D., Reinhardt, C., Meier, M. M. M., Mayer, L., Stadel, J., & Benitez, F. (2019). Primordial Earth mantle heterogeneity
caused by the Moon‐forming giant impact? The Astrophysical Journal, 887(2), 211. https://doi.org/10.3847/1538‐4357/ab50b9
Deschamps, F., & Cobden, L. (2022). Estimating core‐mantle boundary temperature from seismic shear velocity and attenuation. Frontiers in
Earth Science, 10, 1031507. https://doi.org/10.3389/feart.2022.1031507
Fischer, R., & Gerya, T. (2016). Early Earth plume‐lid tectonics: A high‐resolution 3D numerical modelling approach. Journal of Geodynamics,
100, 198–214. https://doi.org/10.1016/j.jog.2016.03.004
Foley, B. J. (2020). Timescale of short‐term subduction episodicity in convection models with grain damage: Applications to Archean tectonics.
Journal of Geophysical Research: Solid Earth, 125(12), e2020JB020478. https://doi.org/10.1029/2020JB020478
Foley, B. J., Bercovici, D., & Elkins‐Tanton, L. T. (2014). Initiation of plate tectonics from post‐magma ocean thermochemical convection.
Journal of Geophysical Research: Solid Earth, 119(11), 8538–8561. https://doi.org/10.1002/2014JB011121
Gerya, T. V., Stern, R. J., Baes, M., Sobolev, S. V., & Whattam, S. A. (2015). Plate tectonics on the Earth triggered by plume‐induced subduction
initiation. Nature, 527(7577), 221–225. https://doi.org/10.1038/nature15752
Hacker, B. R., Abers, G. A., & Peacock, S. M. (2003). Subduction factory 1. Theoretical mineralogy, densities, seismic wave speeds, and H2O
contents. Journal of Geophysical Research, 108(B1), 2029. https://doi.org/10.1029/2001JB001127
Hamano, K., Abe, Y., & Genda, H. (2013). Emergence of two types of terrestrial planet on solidification of magma ocean. Nature, 497(7451),
607–610. https://doi.org/10.1038/nature12163
Hastie, A. R., Law, S., Bromiley, G. D., Fitton, J. G., Harley, S. L., & Muir, D. D. (2023). Deep formation of Earth’s earliest continental crust
consistent with subduction. Nature Geoscience, 16(9), 816–821. https://doi.org/10.1038/s41561‐023‐01249‐5
Hyung, E., & Jacobsen, S. B. (2020). The 142
Nd/144
Nd variations in mantle‐derived rocks provide constraints on the stirring rate of the mantle
from the Hadean to the present. Proceedings of the National Academy of Sciences, 117(26), 14738–14744. https://doi.org/10.1073/pnas.
2006950117
Jacobson, S. A., Rubie, D. C., Hernlund, J., Morbidelli, A., & Nakajima, M. (2017). Formation, stratification, and mixing of the cores of Earth and
Venus. Earth and Planetary Science Letters, 474, 375–386. https://doi.org/10.1016/j.epsl.2017.06.023
Johansen, A., Ronnet, T., Schiller, M., Deng, Z., & Bizzarro, M. (2023). Anatomy of rocky planets formed by rapid pebble accretion: II. Dif-
ferentiation by accretion energy and thermal blanketing. Astronomy & Astrophysics, 671, A75. https://doi.org/10.1051/0004‐6361/202142142
Kegerreis, J. A., Eke, V. R., Massey, R. J., Sandnes, T. D., & Teodoro, L. F. A. (2022). Immediate origin of the Moon as a post‐impact satellite.
The Astrophysical Journal Letters, 937(2), L40. https://doi.org/10.3847/2041‐8213/ac8d96
Kemp, A. I. S., Wilde, S. A., Hawkesworth, C. J., Coath, C. D., Nemchin, A., Pidgeon, R. T., et al. (2010). Hadean crustal evolution revisited: New
constraints from Pb–Hf isotope systematics of the Jack Hills zircons. Earth and Planetary Science Letters, 296(1–2), 45–56. https://doi.org/10.
1016/j.epsl.2010.04.043
Korenaga, J. (2007). Thermal cracking and the deep hydration of oceanic lithosphere: A key to the generation of plate tectonics? Journal of
Geophysical Research, 112(B5), B05408. https://doi.org/10.1029/2006JB004502
Korenaga, J. (2008). Urey ratio and the structure and evolution of Earth’s mantle. Reviews of Geophysics, 46(2), 1–32. https://doi.org/10.1029/
2007RG000241
Kreielkamp, P., Stein, C., & Hansen, U. (2022). LLSVPs of primordial origin: Implications for the evolution of plate tectonics. Earth and
Planetary Science Letters, 579, 117357. https://doi.org/10.1016/j.epsl.2021.117357
Li, Y., & Gurnis, M. (2022). A simple force balance model of subduction initiation. Geophysical Journal International, 232(1), 128–146. https://
doi.org/10.1093/gji/ggac332
Lock, S. J., & Stewart, S. T. (2017). The structure of terrestrial bodies: Impact heating, corotation limits, and synestias. Journal of Geophysical
Research: Planets, 122(5), 950–982. https://doi.org/10.1002/2016JE005239
Lourenço, D. L., Rozel, A. B., Ballmer, M. D., & Tackley, P. J. (2020). Plutonic‐squishy lid: A new global tectonic regime generated by intrusive
magmatism on Earth‐like planets. Geochemistry, Geophysics, Geosystems, 21(4), e2019GC008756. https://doi.org/10.1029/2019GC008756
Lourenço, D. L., Rozel, A. B., Gerya, T., & Tackley, P. J. (2018). Efficient cooling of rocky planets by intrusive magmatism. Nature Geoscience,
11(5), 322–327. https://doi.org/10.1038/s41561‐018‐0094‐8
Mansour, J., Giordani, J., Moresi, L., Beucher, R., Kaluza, O., Velic, M., et al. (2020). Underworld2: Python geodynamics modelling for Desktop,
HPC and Cloud (Version v2.10.0b) [Software]. Zenodo. https://doi.org/10.5281/zenodo.3975252
Marchi, S., Bottke, W. F., Elkins‐Tanton, L. T., Bierhaus, M., Wuennemann, K., Morbidelli, A., & Kring, D. A. (2014). Widespread mixing and
burial of Earth’s Hadean crust by asteroid impacts. Nature, 511(7511), 578–582. https://doi.org/10.1038/nature13539
Acknowledgments
We appreciate the constructive comments
by Bradford Foley and an anonymous
reviewer, as well as the editor's efficient
handling of our study. Discussions with W.
Mao, M. Li. H. Deng, J. Kegerreis, Y.
Miyazaki, E. Asphaug are appreciated. Q.
Yuan acknowledges support from the O.K.
Earl Postdoctoral Fellowship at Caltech
and partial support by the NSF through
OCE‐2049086 and EAR‐2330810. We
acknowledge use of the Anvil
supercomputer at Purdue University
supported by the NSF ACCESS program
TG‐EAR160027. Some figures in this
study use the perceptually uniform color
map “batlow” (Crameri et al., 2020).
YUAN ET AL. 9 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

Moresi, L., Quenette, S., Lemiale, V., Mériaux, C., Appelbe, B., & Mühlhaus, H.‐B. (2007). Computational approaches to studying non‐linear
dynamics of the crust and mantle. Physics of the Earth and Planetary Interiors, 163(1–4), 69–82. https://doi.org/10.1016/j.pepi.2007.06.009
Moresi, L., & Solomatov, V. (1998). Mantle convection with a brittle lithosphere: Throughts on the global tectonic styles of the Earth and Venus.
Geophysical Journal International, 133(3), 669–682. https://doi.org/10.1046/j.1365‐246X.1998.00521.x
Nakagawa, T., & Tackley, P. J. (2010). Influence of initial CMB temperature and other parameters on the thermal evolution of Earth’s core
resulting from thermochemical spherical mantle convection. Geochemistry, Geophysics, Geosystems, 11(6), 2010GC003031. https://doi.org/
10.1029/2010GC003031
Nakajima, M., & Stevenson, D. J. (2015). Melting and mixing states of the Earth’s mantle after the Moon‐forming impact. Earth and Planetary
Science Letters, 427, 286–295. https://doi.org/10.1016/j.epsl.2015.06.023
Nikolaou, A., Katyal, N., Tosi, N., Godolt, M., Grenfell, J. L., & Rauer, H. (2019). What factors affect the duration and outgassing of the terrestrial
magma ocean? The Astrophysical Journal, 875(1), 11. https://doi.org/10.3847/1538‐4357/ab08ed
Nimmo, F. (2022). Formation, composition, and evolution of the Earth’s core. In Oxford Research Encyclopedia of Planetary Science.
O’Neill, C., Lenardic, A., Weller, M., Moresi, L., Quenette, S., & Zhang, S. (2016). A window for plate tectonics in terrestrial planet evolution?
Physics of the Earth and Planetary Interiors, 255, 80–92. https://doi.org/10.1016/j.pepi.2016.04.002
O’Neill, C., Marchi, S., Zhang, S., & Bottke, W. (2017). Impact‐driven subduction on the Hadean Earth. Nature Geoscience, 10(10), 793–797.
https://doi.org/10.1038/ngeo3029
Rudi, J., Gurnis, M., & Stadler, G. (2022). Simultaneous inference of plate boundary stresses and mantle rheology using adjoints: Large‐scale 2‐D
models. Geophysical Journal International, 231(1), 597–614. https://doi.org/10.1093/gji/ggac207
Stern, R. J., & Gerya, T. (2018). Subduction initiation in nature and models: A review. Tectonophysics, 746, 173–198. https://doi.org/10.1016/j.
tecto.2017.10.014
Stewart, S., Davies, E., Duncan, M., Lock, S., Root, S., Townsend, J., et al. (2020). The shock physics of giant impacts: Key requirements for the
equations of state. AIP Conference Proceedings (Vol. 2272). https://doi.org/10.1063/12.0000946
Tan, E., & Gurnis, M. (2005). Metastable superplumes and mantle compressibility. Geophysical Research Letters, 32(20), 1–4. https://doi.org/10.
1029/2005GL024190
Tian, D., Uieda, L., Leong, W. J., Schlitzer, W., Fröhlich, Y., Grund, M., et al. (2023). PyGMT: A Python interface for the generic mapping tools.
https://doi.org/10.5281/Zenodo.8303186
Ueda, K., Gerya, T., & Sobolev, S. V. (2008). Subduction initiation by thermal–chemical plumes: Numerical studies. Physics of the Earth and
Planetary Interiors, 171(1–4), 296–312. https://doi.org/10.1016/j.pepi.2008.06.032
Valley, J. W., Peck, W. H., King, E. M., & Wilde, S. A. (2002). A cool early Earth. Geology, 30(4), 351–354. https://doi.org/10.1130/0091‐7613
(2002)030<0351:ACEE>2.0.CO;2
Van Heck, H. J., & Tackley, P. J. (2008). Planforms of self‐consistently generated plates in 3D spherical geometry. Geophysical Research Letters,
35(19), 2008GL035190. https://doi.org/10.1029/2008GL035190
Watson, E. B., & Harrison, T. M. (2005). Zircon thermometer reveals minimum melting conditions on earliest Earth. Science, 308(5723),
841–844. https://doi.org/10.1126/science.1110873
Yuan, Q. (2024). A giant impact origin for the first subduction on Earth [Dataset]. figshare. Retrieved from https://figshare.com/articles/figure/
Figures/25262476
Yuan, Q., & Li, M. (2022a). Instability of the African large low‐shear‐wave‐velocity province due to its low intrinsic density. Nature Geoscience,
15(4), 334–339. https://doi.org/10.1038/s41561‐022‐00908‐3
Yuan, Q., & Li, M. (2022b). Vastly different heights of LLVPs caused by different strengths of historical slab push. Geophysical Research Letters,
49(17), e2022GL099564. https://doi.org/10.1029/2022gl099564
Yuan, Q., Li, M., Desch, S. J., Ko, B., Deng, H., Garnero, E. J., et al. (2023). Moon‐forming impactor as a source of Earth’s basal mantle
anomalies. Nature, 623(7985), 95–99. https://doi.org/10.1038/s41586‐023‐06589‐1
References From the Supporting Information
Hofmann, A. W. (1988). Chemical differentiation of the Earth: The relationship between mantle, continental crust, and oceanic crust. Earth and
Planetary Science Letters, 90(3), 297–314. https://doi.org/10.1016/0012‐821x(88)90132‐x
Lodders, K., & Fegley, B., Jr. (1997). An oxygen isotope model for the composition of Mars. Icarus, 126(2), 373–394. https://doi.org/10.1006/
icar.1996.5653
McDonough, W. F., & Sun, S. S. (1995). The composition of the Earth. Chemical geology. Chemical Geology, 120(3–4), 223–253. https://doi.org/
10.1016/0009‐2541(94)00140‐4
Morgan, J. W., & Anders, E. (1979). Chemical composition of Mars. Geochimica et Cosmochimica Acta, 43(10), 1601–1610. https://doi.org/10.
1016/0016‐7037(79)90180‐7
Taylor, G. J. (2013). The bulk composition of Mars. Geochemistry, 73(4), 401–420. https://doi.org/10.1016/j.chemer.2013.09.006
Taylor, S. R., & McLennan, S. M. (1985). The continental crust: Its composition and evolution. Blackwell Scientific Publications.
Van Schmus, W. (1995). Natural radioactivity of the crust and mantle. In Global earth physics: A handbook of physical constants (Vol. 1, pp.
283–291). American Geophysical Union. https://doi.org/10.1029/rf001p0283
Wänke, H., & Dreibus, G. (1994). Chemistry and accretion history of Mars. Philosophical Transactions of the Royal Society of London. Series A:
Physical and Engineering Sciences, 349(1690), 285–293.
Yoshizaki, T., & McDonough, W. F. (2020). The composition of Mars. Geochimica et Cosmochimica Acta, 273, 137–162. https://doi.org/10.
1016/j.gca.2020.01.011
YUAN ET AL. 10 of 10
19448007,
2024,
9,
Downloaded
from
by
CAPES,
Wiley
Online
Library
on
[24/05/2024].
See
the
Terms
and
Conditions
on
Wiley
Online
Library
for
rules
of
use;
OA
articles
are
governed
by
the
applicable
Creative
Commons
License

A Giant Impact Origin for the First Subduction on Earth

Recommended

Recommended

More Related Content

Similar to A Giant Impact Origin for the First Subduction on Earth

Similar to A Giant Impact Origin for the First Subduction on Earth (20)

More from Sérgio Sacani

More from Sérgio Sacani (20)

Recently uploaded

Recently uploaded (20)

A Giant Impact Origin for the First Subduction on Earth