The sub-plate mantle flow traction has been considered as a major driving force for plate motion; however, the force acting on the overlying plate is difficult to be well constrained. One reason lies in the variable rheological flow laws of mantle rocks, e.g. linear versus power-law rheology, applied in previous studies. Here, systematic numerical models are conducted to evaluate the mantle flow traction under variable rheological, geometrical and kinematic conditions. The results indicate that mantle flow traction with power-law rheology is much lower than that with linear rheology under the same mantle/plate velocity contrast. In addition, the existence of lithospheric root in the overlying plate enhances the mantle flow traction. In a regime with reasonable parameters, the mantle flow traction with power-law rheology is comparable to the ridge push on the order of 1012 N/m, whereas that with linear rheology is comparable to the slab pull of 1013 N/m.

Oceanic-plates vertical tearing is seismically-identified in the present-day Earth. This type of plate tearing is frequently reported in horizontally-oblique subduction zones where transform-faulted oceanic plates are subducting (or subducted). However, the mechanisms behind vertical slab tearing are still poorly understood, thus we utilize 3D time-dependent Stokes’ flow thermo-mechanical models to further study this problem. We find that (i) the age offset of transform fault and (ii) the horizontal obliqueness of subduction fundamentally control the tearing behavior of two generic, materially-homogeneous oceanic slabs separated by a low-viscosity zone. The two slabs sequentially bend, which combined with the age-thickness difference between slabs, causes the differential sinking of them. Based on the modeling results, well-developed slabs vertical tearing would happen when the oblique angle of subduction is ≥30° or the age ratio of the secondly-bent to firstly-bent slab being ~<0.6. Quantifying the horizontal distance-vector between sinking slabs, we find that subduction at medium-low horizontal-obliqueness angles (≤40°) of young lithosphere (slabs-average ~15 Myr) tends to produce fault-perpendicular tearing. Contrastingly, old-age slabs (average ≥ 30 Myr) with medium-large obliqueness angles (~>20°) tend to produce fault-parallel tearing, related to differential slab-hinge retreat or rollback. Correlations between slabs’ (i) computed tearing horizontal-width and (ii) scaling-theory forms of their subduction-velocity differences, are reasonable (0.76-0.97). Our numerically-predicted scenarios are reasonably consistent with plate-tear imaging results from at least 4 natural subduction zones. Our modeling also suggests that continual along-trench variation in subduction dip angle may be related to a special case of oblique subduction.

Oceanic slab vertical tearing is prevalently identified in the present Earth. More general background for vertical slab tearing is the transform-fault subduction during horizontally-oblique tectonic convergence. However, its geodynamic mechanisms are poorly understood to date. This work introduces a full numerical 3-D time-dependent Stokes’ thermo-mechanical flow model to investigate the characteristics and mechanism of vertical tearing of active transform-faulted oceanic slab during oblique subduction. We find that (i) transform-fault ages-offset and (ii) subduction horizontal obliqueness have the first-order control, even without the lateral physical-property differences. The overriding plate enforces (surface contact interaction) bending of one slab first, which superimposes the differential sinking driven by slabs-age-thickness differences. For obliqueness angles ≥30° and/or age-ratios of the secondly-bent to the firstly-bent slab being <0.6, well-developed slab vertical tearing is unavoidable inside the mantle. Quantifying the horizontal distance vector between sinking slabs, we find that young overall lithosphere (average <30 Myr, for any age ratio) at high subduction obliqueness angles (>~25°) tends to produce trench-parallel slab tearing. In contrast, combinations of small-intermediate obliqueness angles (0-30°) and age ratios with the slab that bends at the trench first being relatively older-thicker, tend to produce trench-perpendicular tearing, which is related to differential slabs-hinge retreat or rollback. These numerically-predicted scenarios and phenomena are consistent with plate-tear imaging results from subduction zones. Our modeling results also suggest that the continual along-trench variation in subduction dip angle may be related to oblique subduction’s early stages of evolution.

Removal and thinning of cratonic lithosphere is believed to have occurred under different tectonic settings, for example, near subduction zones and above mantle plumes. Subduction-induced cratonic modification has been widely discussed; however, the mechanisms and dynamic processes of plume-induced lithospheric removal remain elusive and require further systematic investigation. In this study, we conduct a series of 2-D thermo-mechanical models to explore the dynamics of the removal and thinning of cratonic lithosphere due to the interaction between a mantle plume and a weak mid-lithosphere discontinuity (MLD) layer. Our modeling results suggest that the interaction between a mantle plume and weak MLD layer can lead to a large-scale removal of the cratonic lithosphere as long as the connection between the hot upwelling and weak MLD layer is satisfied. The presence of a vertical lithospheric weak zone and its closeness to the plume center play critical roles in creating a connection between the weak MLD and hot plume/asthenosphere. Furthermore, delamination of cratonic lithosphere is favored by a larger plume radius/volume, a higher plume temperature anomaly, and a lower viscosity of the MLD layer. A systematic comparison between subduction-induced and plume-induced lithospheric thinning patterns is further conducted. We summarize their significant differences on the origin and migration of melt generation, the water content in melts, and topographic evolution. The combination of numerical models and geological/geophysical observations indicates that mantle plume-MLD interaction may have played a crucial role in lithospheric removal beneath South Indian, South American and North Siberian Cratons.

The subducting slab morphology in the mantle transition zone (MTZ) is strongly affected by the mantle viscosity and density variations at the 660-km discontinuity (D660). Besides the negative Clapeyron slope of phase transition and the viscosity increase, a possible thin weak layer at around D660 is proposed to play a key role in the slab stagnation, which is however not well constrained. In this study, a series of numerical models are systematically conducted, which reveal that a weak layer beneath D660 does not change the slab mode selection (penetration versus stagnation). However, it will contribute to longer slab flattening at the bottom of the MTZ, when slab sinking is strongly resisted by either the viscosity increase or a large Clapeyron slope at D660. The role of a weak layer on slab flattening is dependent on the lubrication effect that promotes sub-horizontal slab movement at the bottom of the MTZ.

Subduction-induced deep hydration and water cycling may play significant roles in the modification and destruction of the overriding cratonic lithosphere. Two contrasting modes are generally proposed: (1) flat subduction (FS) regime with slab subducting sub-horizontally beneath the overriding lithosphere, and (2) big mantle wedge (BMW) regime with slab flattening in the mantle transition zone. Here, systematic petrological-thermomechanical models are conducted to investigate the fluid/melt activities in the contrasting subduction regimes as well as their effects on the modification of overriding lithosphere. The model results indicate that the dehydration process in the FS regime can significantly modify the overriding lithosphere for a region of about 600 km from the trench. During the progressive flat subduction, the partial melting and magmatism migrate towards the inner land of the overriding plate, which will be reversed and backward to the trench during the transition from flat to steep subduction. On the other hand, the deep hydration in the BMW regime is strongly dependent on the sub-crustal serpentinite layer in the subducting slab, whereas the oceanic crust cannot carry water to the transition zone. The modification of the overriding lithosphere in the BMW regime occurs in a larger region of >1000 km from trench, which is however generally slower and weaker. The modification and destruction of North China Craton is more likely to be controlled by the flat subduction of paleo-Pacific plate in the late Jurassic to early Cretaceous, which may be accompanied by the effects of deep water cycling in the BMW regime.

The collision-induced subduction transference is a composite dynamic process including both the terrane collision/accretion and the subduction initiation (SI) at the neighboring passive margin. This process occurred repeatedly during the evolution of Tethyan systems, with multiple ribbon-like continents or micro-continents drifting from Gondwana in the southern hemisphere and accreting to the Eurasian continent since Paleozoic. In the previous numerical studies, the dynamics of terrane collision and induced SI are investigated individually, which however need to be integrated to study the controlling factors and time scales of collision-induced subduction transference. Systematic numerical models are conducted with variable properties of converging plates and different boundary conditions. The model results indicate that the forced convergence, rather than pure free subduction, is required to trigger and sustain the SI at the neighboring passive margin after terrane collision. In addition, a weak passive margin can significantly promote the occurrence of SI, by decreasing the required boundary force to reasonable value of plate tectonics. The lengths of subducted oceanic slab and accreting terrane play secondary roles in the occurrence of SI after collision. Under the favorable conditions of collision-induced subduction transference, the time required for SI after collision is generally short within 10 Myrs, which is consistent with the general geological records of Cimmerian collision and the following Neo-Tethyan SI. In contrast, the stable Indian passive margin and absence of SI in the present Indian Ocean may due to the low convergent force and/or the lack of proper weak zones, which remains an open question.

In this study, we for the first time applied a joint geodynamic-geophysical inversion (JGGI) approach to oceanic plateau subduction models, and compared the subduction style and corresponding topography and Bouguer gravity of two representative subduction scenarios with passive or active collision. We showed that the case of passive collision of the Ontong Java Plateau (OJP) crust better explains the topography, gravity, and seismic data than the active collision scenario. This implies that the OJP did not control the regional dynamics during the collisional process. We conclude that previous studies may have overestimated the role of the OJP in triggering subduction initiation, subduction polarity reversal, and even Pacific Plate rotation.

It is generally accepted that the subduction polarity reversal (SPR) results from the strong collision of two plates. Yet, the SPR of the Solomon Back-arc Basin is started in the “soft docking” stage, and the mechanism by which the “soft docking” induced subduction initiation (SI) remains elusive. We find that the mass depletion of the plateau influences the evolution of the subduction patterns during SI. And the island arc rheological strength affects the development of the shear zone between an island arc and back-arc basin which favors SI. What’s more, with the increase of the rheological strength difference, the SI is more easily to occur, and the contribution of the plateau collision to SI weakens. Hence, by combining the available geological evidence, we suggest that the Solomon Island Arc rheological strength and the Ontong-Java plateau collision jointly controlled the SI during the “soft docking” stage.