中文       Go Back       Search
Associate Professor

Sep, 2008 – Jul, 2013
Ph.D. student (geodynamics)
School of Earth and Space Science, University of Science and Technology of China
Dissertation: Numerical simulations on the evolution of the thermochemical anomalies in the lowermost mantle (in Chinese)
Supervisor: Rongshan Fu

Sep, 2004 – Jul, 2008
B.S student (geophysics)
School of Geodesy and Geomatics, Wuhan University


Jan, 2021 –
associate professor
Department of Earth and Space Sciences, Southern University of Science and Technology, Shenzhen, China

Dec, 2018 –Dec, 2020
assistant professor
Department of Earth and Space Sciences, Southern University of Science and Technology, Shenzhen, China

Dec, 2016 – Dec, 2018
research fellow (geodynamics)
School of Earth Science, University of Melbourne, Australia
Supervisor: Louis Moresi

Jul, 2013 – Oct, 2016
postdoc scholar (geodynamics)
Seismological Laboratory, California Institute of Technology, USA
Supervisor: Michael Gurnis


Geodynamics aims at unifying patchy observations from different subjects with basic physical and chemical theories and providing spatially- and temporally- continuous scenarios of Earth’s evolution. To achieve a better understanding of Earth’s dynamics, I’m interested in two aspects at present. One is constraining the material properties, state and evolution history of the Earth’s mantle and lithosphere; the other is providing possible qualitative and quantitative predictions with geodynamic models and simulations. New observations will enable us to test, update, or rebuild the geodynamic models. I have worked on these projects during the last few years:

· Contrasted East Asia and South America tectonics driven by deep mantle flow

East Asia and South America have both experienced long-term subduction since at least the Jurassic, but they have had contrasting tectonic evolutions since the Late Cretaceous. East Asia was dominated by extensional tectonics with many marginal sea basins forming during the Cenozoic while South America was dominated by compressional tectonics building the Andean mountains. We suggest that the deep mantle flow revealed by seismic tomography, plate reconstructions, and geodynamic models contributed to the pan-Pacific tectonic disparity. Our geodynamic models suggest that the Atlantic Ocean opening plays an important role in promoting compression-dominated tectonics and Andean building along the South American margin by allowing fast trench-ward motion of South America. On the other hand, the long-standing downwelling flow beneath the interior of Asia since Asian assembly in the Paleozoic and the early Mesozoic helps to inhibit Atlantic-type ocean opening in Asia and restrain trench-ward motion of East Asia, promoting extension-dominated tectonics along the Asian margin. The restrained trenchward motion suggests a low probability of flat-slab subduction occurrence in East Asia during the late Mesozoic.

· Cenozoic tectonic evolution in Northeast Asia and the rapidly-aging Pacific Plate

A unified geodynamic model explaining the temporal-spatial evolution of the Cenozoic tectonic and magmatic events in Northeast Asia is still lacking. We provided a geodynamic model to fill the gap (Yang et al., 2018, EPSL). The geodynamic model explaining the first-order regional tectonic evolution suggests that (1) back-arc sea spreading is key for slab stagnation in East Asia; (2) the Japan Sea spreading was delayed during the Eocene due to the young Pacific Plate subduction; (3) the rapid aging of the Pacific Plate at the trench allowed the Japan Sea to open in the early Miocene; (4) the opening of the Japan Sea prevented further Japan Sea opening after the late Miocene; (5) the stagnant slab beneath Northeast Asia is induced by the Japan Sea spreading and thus cannot be formed earlier than 30 Ma.

· 2015 Bonin Island deep earthquake under the background of plate tectonics and slab subduction

Mantle convection models consistent with regional plate reconstructions reproduce the seismology-observed principal stress direction and slab morphology in the Izu-Bonin subduction zone (Yang et al., 2017, GRL). The isolated ~680 km deep, May 30, 2015 Mw 7.9 Bonin Islands earthquake, is a result of Pacific slab buckling in response to the slow trench retreat. Further work on the Mexican flat slab indicates that (1) the earthquake regime transitions along the subducting slab and (2) the restriction of earthquakes in the Mexican flat slab segment to depths in the range 50-80 kms are consistent with a stress regime dominated by advective viscous bending (Sandiford et al., 2018, Nature). We suggest that plate tectonics and mantle convection over millions of years strongly influence the location and focal mechanism of deep earthquakes occurring within seconds.

· Dynamic evolution of Southeast Asia since the Cenozoic

Mantle convection models combined with plate reconstructions suggest that a previously stagnant slab penetrated through the 660 km mantle discontinuity during the early Miocene. This instable process has induced a series of regional tectonic events after the early Miocene: the Sunda Trench transited from trench retreat to trench advance; southern Sundaland suffered large-scale synchronous marine inundation, lithospheric compression and basin inversion events; northern Southeast Asia witnessed anomalous post-rift basin subsidence with ultra-thick sedimentary deposits (Yang et al., 2016, GRL; Yang et al., 2016, Tectonophysics). During this project, we also developed code to rapidly calculate crustal thickness, topography, geotherm, and surface heat flux evolution after performing deforming plate reconstruction (Gurnis et al., 2017).

· High-precision oceanic residual topography agrees with mantle flow predictions at long wavelengths

Dynamic topography is one of the most important predictions made by geodynamic models. The long-standing deviations between model predictions and observations of dynamic topography make people question the reliability of geodynamic models. We demonstrate that the recovered long-wavelength dynamic topography from recent precise estimates in the oceanic realm agrees with mantle convection model predictions (Yang et al., 2017, GRL). Model predictions and oceanic observations both give ~ 1km amplitude long-wavelength dynamic topography which is positive around south Pacific and southern Africa while negative above subducted slabs. Interestingly, these observations are from the group that has been always opposing the existence of large-amplitude long-wavelength dynamic topography.

· Mantle lateral viscosity variations in global mantle flow

Although consistent with surface lithospheric tectonics, our inverted mantle viscosity has only a weak (or negative) apparent dependence on temperature at long wavelengths (Yang and Gurnis, 2016, GJI). This suggests that the cold slab’s ability to mix the mantle may be much less than previous suggestions. Our inverted lateral viscosity variations are supported by further studies (Dannberg et al., 2017, G3; Yang et al., 2017, GRL). Our understanding of mantle mixing, chemical evolution, slab behavior, etc., may be significantly changed by further global mantle (including lower mantle) convection models considering grain-size evolution and non-linear viscosity.

· Dynamics of hidden hotspot tracks beneath the continental lithosphere

The plume conduit can erode the bottom of the continental lithosphere, generating a corridor-like low seismic velocity zone downstream the plate motion direction (Yang and Leng, 2014, EPSL). The surface swell topography of this corridor is much smaller than those beneath the oceanic lithosphere, forming ‘hidden tracks’.

· The dynamics of LLSVPs in the lower mantle.

The thermochemical pile’s survival time does not monotonically vary with its viscosity. A high viscosity chemical pile (this is possible) can obstruct the horizontal flow along CMB and turn it into upwelling flow (Yang and Fu, 2014, PEPI).

· Effective elastic thickness of the continental lithosphere in China

The estimated elastic thickness is low in East China with the Cenozoic rifting basins reaching the lowest values. Elastic thickness is generally high in the Tibetan Plateau, suggesting that mid-crustal flow in Tibet might be only local features.