
Many aspects of solid Earth's evolution such as crustal growth, tectonic regime and the onset of plate tectonics remain debated. Here, the influence of water on Earth's dynamics is investigated using 2D global compressible convection models. These models show a tectonic regime fluctuating between mobile-lid, plutonic-squishy-lid and ridge-only during the Hadean and early Archean, accompanied by rapid crustal growth and intensive crustal recycling. As the water budget of the mantle and the surface remains poorly constrained, different ingassed values of water are tested, which show a strong correlation with continental crust production. Upon incorporating water-dependent density, these models showcase a significant drop in surface mobility and are less efficient at planetary cooling. This happens due to excess of water in the oceanic crust, which makes it buoyant and less prone to subduct. Once the surface rocks are partially saturated to better fit the petrological constraints and the viscosity of rocks is reduced in the presence of water, subduction becomes active and a fluctuating tectonic regime is observed again. Models initialised with the effective frictional strength of 0.1 and a partial saturation of surface rocks reproduce the Nb/U and Ce/Pb measurements from new geochemical data obtained from 3.2 Ga old melt inclusions in olivines from komatiites. Finally, the reduction in effective friction of the oceanic lithosphere during the mid-Archean, resembling the lubrication effect of continental sediments, helps to stabilise the mobile-lid regime, thereby indicating a transition to plate tectonics.
The main results from this study have been accepted as: Jain, C., Sobolev, S. V. (2026) Controls on the tectonic evolution of the early Earth: 2. Contrasted effects of water. Journal of Geophysical Research: Solid Earth
Additional results with geochemical proxies are published in:

In this study, the influence of three different rheological setups on the dynamics of early Earth is investigated using 2D global compressible convection models. In the first setup, viscous deformation includes only diffusion creep. In the second setup, deformation in the upper mantle is accommodated by composite rheology, which is a combination of diffusion creep and dislocation creep. In the third setup, based on classical modelling studies, a composite proxy rheology is prescribed using dislocation creep at a constant strain rate devised to stabilise and speed up computations. Irrespective of the rheology considered, all models exhibit a fluctuating mobile-lid regime with plume-induced lithospheric subduction for the initial 200-300 Myr. Afterwards, they transition to a ridge-only regime featuring short (from several Myr to a few tens of Myr) episodes of high mobility, sometimes with global resurfacing events. When dislocation creep is included, models have higher surface mobility, experience longer and more frequent resurfacing events, produce more continental crust, and are more efficient at planetary cooling when compared with diffusion creep only models. Models that combine diffusion creep with either a dislocation creep or a dislocation creep proxy, yield similar long-term thermochemical planetary evolution and continental crust production. At the same time, composite proxy rheology models are computationally less expensive and less prone to producing over-thickened oceanic crust in 2D geometry, and are therefore a reasonable approximation for composite rheology in large-scale global models.
The results have been accepted as: Jain, C., Sobolev, S. V. (2026), Controls on the tectonic evolution of the early Earth: 1. Composite rheology. Journal of Geophysical Research: Solid Earth

Geophysical, geochemical, and geological investigations have attributed the stable behaviour of Earth’s continents to the presence of their Archean cratonic roots. These roots are likely composed of melt-depleted, low density residual peridotite with high magnesium number (Mg#), while devolatilisation from the upper mantle during magmatic events might have made these roots more viscous and intrinsically stronger than the convecting mantle. Several conceptual dynamic and petrological models of craton formation have been proposed. Dynamic models invoke far-field shortening or mantle melting events, e.g., by mantle plumes, to create melt-depleted and thick cratons. Compositional buoyancy and rheological modifications have also been invoked to create long-lived stable cratonic lithosphere. However, these conceptual models have not been tested in a dynamically self-consistent model. In this study, we present global thermochemical models of craton formation with coupled core-mantle-crust evolution driven entirely by gravitational forces. Our results with melting and crustal production (both oceanic and continental) show that formation of cratonic roots can occur through naturally occurring lateral compression and thickening of the lithosphere in a self-consistent manner, without the need to invoke far-field tectonic forces. Plume impingements, and gravitational sliding creates thrusting of lithosphere to form thick, stable, and strong lithosphere that has a strong resemblance to the Archean cratons that we can still observe today at the Earth’s surface. These models also suggest the recycling of denser eclogitic crust by delamination and dripping processes. Within our computed parameter space, a variety of tectonic regimes are observed which also transition with time. Based on these results, we propose that a ridge-only regime or a sluggish-lid regime might have been active on Earth during the Archean Eon as they offer favourable dynamics and conditions for craton formation.
The results are published as: Jain, C., Rozel, A. B., van Hunen, J., Chin, E. J. and E. J. Chin and Manjón-Cabeza Córdoba, A. (2022) Building Archean cratonic roots. Frontiers in Earth Science, 10, (preprint on EarthArXiv)

The majority of continental crust formed during the hotter Archean was composed of Tonalite-Trondhjemite-Granodiorite (TTG) rocks. In contrast to the present-day loci of crust formation around subduction zones and intra-plate tectonic settings, TTGs are formed when hydrated basalt melts at garnet-amphibolite, granulite or eclogite facies conditions. Generating continental crust requires a two step differentiation process. Basaltic magma is extracted from the pyrolytic mantle, is hydrated, and then partially melts to form continental crust. In this project, we parametrised the melt production and melt extraction processes and showed self-consistent generation of primordial continental crust using evolutionary thermochemical mantle convection models. To study the growth of TTG and the geodynamic regime of early Earth, we systematically varied the ratio of intrusive (plutonic) and eruptive (volcanic) magmatism, initial core temperature, and internal friction coefficient. As the amount of TTG that can be extracted from the basalt (or basalt-to-TTG production efficiency) is not known, we also tested two different values in our simulations, thereby limiting TTG mass to 10% or 50% of basalt mass. For simulations with lower basalt-to-TTG production efficiency, the volume of TTG crust produced is in agreement with net crustal growth models but overall crustal (basaltic and TTG) composition stays more mafic than expected from geochemical data. With higher production efficiency, abundant TTG crust is produced, with a production rate far exceeding typical net crustal growth models but the felsic to mafic crustal ratio follows the expected trend. These modelling results indicate that (i) early Earth exhibited a ''plutonic squishy lid'' or vertical-tectonics geodynamic regime, (ii) present-day slab-driven subduction was not necessary for the production of early continental crust, and (iii) the Archean Earth was dominated by intrusive magmatism as opposed to ''heat-pipe'' eruptive magmatism.
The results are published as: Jain, C., Rozel, A. B., Tackley, P. J., Sanan, P. and Gerya, T. (2019) Growing primordial continental crust self-consistently in global mantle convection models. Gondwana Research, 73, 96-122 (preprint on EarthArXiv)

Continents influence the mantle's convective wavelength and the heat flow escaping from the planet's surface. Over the last few decades, many numerical and analytical studies have contributed to the debate about whether the continents can warm up the subcontinental mantle or not and if they do, then to what extent? However, a consensus regarding the exact nature and magnitude of this correlation between continents and elevated temperatures in the subcontinental mantle remains to be achieved. By conducting a systematic parameter study using 2D global mantle convection simulations with mobile continents, we provided qualitative and quantitative observations on the nature of this correlation. In our incompressible and compressible convection models, we observed the general processes of downwellings bringing cold material into the mantle along continental margins and a subsequent buildup of warm thermal anomalies underneath the continents. We computed the amplitude and degree of this correlation using spectral decomposition of the temperature and composition fields. The dominant degree of correlation evolves with time and changes with continental configuration. Using simple empirical fits, we observed that this correlation decreases with increasing core temperature, number of continents, internal heating, or decreasing reference viscosity. We also reported simple regressions of the time-dependence of this correlation. Additionally, we showed that decompression melting as a result of a mantle upwelling or small-scale sublithospheric convection leads to voluminous volcanism. The emplacement of this dense basalt-eclogite material breaks the continents apart and destroys the correlation.
The results are published as: Jain, C., Rozel, A., and Tackley P. J. (2019) Quantifying the correlation between mobile continents and elevated temperatures in the subcontinental mantle. Geochemistry, Geophysics, Geosystems, 20, 1358–1386

Geological and geochemical data suggest hotter Archean mantle temperature and more intense magmatism, however the global geodynamic regime during that time remains controversial. In this project, we tested the influence of melt-eruption efficiency towards generating primordial Tonalite–Trondhjemite–Granodiorite (TTG) continental crust using thermochemical convection models. We showed that volcanism-dominated heat-pipe tectonic regime results in cold geotherms and is unable to produce TTG crust. In contrast, plutonism-dominated squishy lid tectonic regime (eruption efficiency < 40%) results in hotter crustal geotherms and is able to produce the observed proportions of various TTG rocks.
The results are published as: Rozel, A. B., Golabek, G. J., Jain, C., Tackley, P. J., and Gerya, T. (2017) Continental crust formation on early Earth controlled by intrusive magmatism. Nature, 545, 332–335