Topic 2: Continental rifting to seafloor spreading via 3D numerical modeling.

Continental rifting to seafloor spreading is a continuous process. Early rifting history influences the following spreading process, but rifting inheritance in spreading remains enigmatic. Using 3D thermo-mechanical coupled viscso-plastic numerical models, we investigate the complete extension process and the inheritance of continental rifting in oceanic spreading. Our modeling results show that the initial continental lithospheric coupling/decoupling affects oceanic spreading in two manners: (1) coupled model (a strong lower crust mechanically couples upper crust and upper mantle lithosphere) generates large lithospheric shear zones and fast rifting, which promotes symmetric oceanic accretion and leads to a relatively straight oceanic ridge, while (2) decoupled model (a weak ductile lower crust mechanically decouples upper crust and upper mantle lithosphere) generates separate crustal and mantle shear zones and favors asymmetric oceanic accretion involving development of active detachment faults with 3D features. (See more results in Liao & Gerya, 2015, GR).

Fig. 1-1 Possible reasons for the presence of a weak layer in cratonic lithosphere. A mid-lithospheric boundary (~ 100 km depth) is widely detected in cratonic regions by seismic studies, although the exact nature of this interface is complex. This figure summarizes the possible origins of a weak layer (red lines) proposed by different authors. See my paper for detailed explanations (Liao and Gerya, 2014).

Fig. 1-2 Comparison of modeling results between homogenous model and layered model (including a weak layer). If a weak layer is widely present in cratonic lithosphere, incorporating such a weak layer in numerical models is crucial to investigate craton evolution.

Fig. 2-2 Curved detachment fault initiation, development and termination.

Fig. 2-1 Three dimensional modeling evolution from continental rifting to seafloor spreading.

Topic 1: Craton extension via 2D numerical modeling.

Lithospheric mantle stratification is a common feature in cratonic areas which has been demonstrated by geophysical and geochemical studies. The influence of lithospheric mantle stratification during craton evolution remains poorly understood. We use a 2D thermo-mechanical coupled numerical model to study the influence of stratified lithospheric mantle on craton extension. A rheologically weak layer representing hydrated and/or metasomatized composition is implemented in the lithospheric mantle. Our results show that the weak mantle layer changes the dynamics of lithospheric extension by enhancing the deformation of the overlying mantle and crust and inhibiting deformation of the underlying mantle. (See more results in Liao. et al. 2013, GRL; Liao and Gerya, 2014, Tectonophysics).

Topic 3 (extending topic 1): Influence of wet olivine flow law on lithospheric mantle dynamics: Implication for craton deformation

Although most cratons observed today are stable (due to their intrinsic physical properties), there are some exceptions, such as the North China craton, North Atlantic craton, and North American craton, which have experienced dramatic deformation. The reasons why cratons undergo destruction are enigmatic. Rheological weakening due to hydration (where water is released from subduction slabs) can be a possible reason, such as the North China craton and North American craton, which are neighbored by the west and east Pacific subduction, respectively. In this study, we investigate the influence of wet mantle flow laws from different authors on lithosphere dynamics. Water content has been systematically tested. (Liao, J., Q. Wang, T. Gerya, M. Ballmer, J. Geophys. Res., In press)   

Fig. 3-1 (a) Rheological weakening of cratonic lithosphere due to hydration. Several possible ways of water transportation from a subduction slab. (b) Viscosity reduce influenced by water content (flow laws are from Hirth and Kohlstedt, 2003).

Topic 4 Subduction (In collaboration with J. Sheng): Dehydration of deep subducting slab: Implication for intra-plate magmatism

Origin of the intra-plate volcanism/magmatism (with a distance > 500 km from the trench, such as the Changbai Mountain in the Northeast China) remains a hot topic. Among several possible hypotheses proposed previously, occurrence of partial melting due to dehydration of the deep subducting slab becomes the favorite one. We will employ 2D thermo-mechanical coupled numerical code (I2ELVIS) , to investigate the whole continuous dynamic process, from oceanic plate subduction to magma generation, including the dynamic process of water. The focus of this study is how water preserved in the slab, transported during subduction, released from the deep slab, and triggered partial melting in the mantle. We will systematically quantify the influence of the following important parameters: (1) age of the subducting slab, (2) velocity of the subducting slab, and (3) intensity of serpentinization. Establish the regime s of parameters on dehydration and magmatism is the last goal of this work. Based on this study, we hope to provide quantitative insight of magmatism due to dehydration of deep subducting oceanic slab, and shed light on the origins of intra-plate volcanism/magmatism. (See more results in Sheng J., Liao J., and Gerya T., 2016, JAES)

Fig. 4-1 Modeling results show the dynamic processes of water transportation in the hydrated slab, slab propagation, and slab dehydration. Partial melting is generated beneath continental lithosphere.  

(PhD work) From continental rifting to seafloor spreading: Insights from 2D and 3D numerical modeling

Fig. 3-2 Model results with both dislocation and diffusion creeps of wet olivine. Lithospheric dripping occurs in the models with wet olivine rheology, and high water content promotes fast deformation.

Topic 5 Long transform fault formation through oblique continental rifting and breakup: Insights from 3D geodynamic modeling (Supervising the master student Noel Ammann at ETH Zurich)

The length of oceanic transform faults is highly variable in nature and can range from zero-offset fracture zones to longer than thousand kilometers. The initiation and evolution of relatively short (tens of km) transform faults was successfully reproduced by previous thermomechanical models. However, the origin of long (hundreds of km) structures remains more enigmatic and requires further numerical modeling effort. In this study, we use a 3D thermomechanical model of oblique continental rifting and breakup followed by oceanic spreading to study the formation of long transform faults. Our starting model geometry consists of an initial oblique weak zone (mobile belt) located in the mantle lithosphere. The obliquity of the weak zone, extension rate, mantle potential temperature, lower crust rheology, and lateral boundary conditions are the main parameters examined by our experiments. By varying these model parameters, mature oceanic spreading with long transform faults was obtained. We track the initiation and evolution of long transform faults from the initiation of oblique rifting to the point where steady-state seafloor spreading is observed. (Ammann N., Liao J.*, Gerya T.*, Ball P., 2017, Tectonophysics)

Fig. 5-1 Long transform fault formation through an initial oblique continental rifting. The length of the transformation fault is longer than 200 km (which is the threshold of the long transform faults). 

Fig. 5-2 Comparison between natural observations and model results in terms of margin geometry and long transform faults/fractures. (a) Observations showing the correlation of passive margins and long transform fault/fractures: margin offset corresponds to long fracture zones. (b) Model simulations showing the development of margin offset and long transform and fracture zones. LTF: long transform fault.

Fig. 3-3 A test model without boundary extension. Convective instability occurred in the base of the lithospheric mantle.

Fig. 3-4 Semi-analytical model results. Semi-analytical plot of the onset of convective instability. Numerical models are imposed in the plots.