Major Research Outcomes

1. Thermal and mechanical interaction between continental drift and mantle convection
We performed numerical simulations of incompressible infinite Prandtl number convection in a 3-D spherical-shell with a single localized high viscosity lid like a supercontinent on the top surface to understand the possible effects of the supercontinent on plume generation. The temperature under the lid increases rapidly after the emplacement of the lid. Subsequently mantle upwellings at the bottom merge into a large scale flow and a large plume emerges under the lid. Despite the complicated 3-D thermal structure, it is dominated by a degree-one pattern controlled by the position of the HVL [Yoshida et al., 1999 (1); Honda, Yoshida, et al., 2000 (2); Yoshida, 2010a (16)]. Next, we developed a new numerical simulation model of mantle convection with deformable continents, and succeeded in realizing continental drift observed on the real Earth's surface [Yoshida, 2010b (18), 2012 (22), 2013 (26); Yoshida and Santosh, 2011b (21), 2014 (30)].
2. Numerical simulations of mantle convection with the breakup of supercontinent Pangea and continental drift
The most notable event in the process of the breakup of Pangea has been the high speed of the northward drift of the Indian subcontinent. We performed numerical simulations of 3-D spherical mantle convection and approximately reproduced the process of continental drift from the breakup of Pangea at 200 Ma to the present-day continental distribution. We found that a major factor in the northward drift of the Indian subcontinent was the large-scale cold mantle downwelling that developed spontaneously in the North Tethys Ocean. We also found that the strong lateral mantle flow caused by the high-temperature anomaly beneath Pangea, due to the thermal insulation effect, enhanced the acceleration of the Indian subcontinent during the early stage of the Pangea breakup [Yoshida, 2014 (32); Yoshida and Hamano, 2015 (35)].
3. Development of a new numerical simulation code for 3-D spherical-shell mantle convection
From 2003 to 2008, we developed a new numerical code for 3-D spherical-shell mantle convection based on the finite-difference and finite-volume method. In this code, we used a kind of the overset grid named "Yin-Yang grid" for the computational grid. The Yin-Yang grid naturally avoids the "pole problems" which are inevitable in the usual latitude-longitude grid (spherical coordinates grid). We confirmed the validity of our simulation code by careful benchmark test with other codes based on the spectral method and finite-element method that had been already developed by researchers of the world [e.g., Yoshida and Kageyama, 2004 (6); Kageyama and Yoshida, 2005 (10)].
4. Numerical simulations of mantle convection with complex rheology and plate-like surface motion
We performed a series of numerical simulations of thermal convection of Boussinesq fluid with infinite Prandtl number and with the strongly temperature- and depth-dependent viscosity in a 3-D spherical-shell to study mantle convection of single-plate terrestrial planets. The strongly temperature-dependent viscosity makes the convection under the stagnant-lid short wavelength structures. We found that in the stagnant-lid regime, numerous, cylindrical upwelling plumes are developed because of the secondary downwelling plumes arising from the bottom of lid. However, this convection pattern is inconsistent with that inferred from the geodesic observation of Venus or Mars. We found that the combination of the strongly temperature- and depth-dependent viscosity causes long-wavelength structures of convection [Yoshida and Kageyama, 2006 (12)]. In a subsequent study, we confirmed that the yield rheology is needed to realize the Earth's mantle convection with planetary-scale, sheet-like downwellings [Yoshida, 2008b (14)].
5. Estimation of the mantle viscosity structure from observed geoid anomalies and the estimation of CMB topography patterns
Instantaneous flow numerical calculations in a 3-D spherical shell are employed to investigate the effects of lateral viscosity variations (LVVs) in the lithosphere and mantle on the long-wavelength geoid anomaly. We found that when highly viscous slabs in the upper mantle are considered, the observed positive geoid anomaly over subduction zones can be accounted for only when the viscosity contrast between the reference upper mantle and the lower mantle is approximately 1000 or lower, and weak plate margins are imposed on the lithosphere [Yoshida, 2004 (7); Yoshida and Nakakuki, 2009 (13)]. Next, core-mantle boundary (CMB) topography is estimated from numerical simulations of instantaneous mantle flow in a 3-D spherical shell geometry. We found that to account for the small CMB topographic relief inferred from recent seismological results [e.g., Tanaka, 2010], lateral viscosity variations in the mantle, compositionally dense piles in the deep mantle, and a low-viscosity D'' layer are all required for the numerical models [Yoshida, 2008a (15)].
6. The origin of hotspot swells on the South Pacific region and mantle dynamics under the South Pacific seafloor
The dynamics of mantle plumes and the origin of their associated swells remain a controversial topic in geodynamics. As part of the seismic observation project on oceanic islands and the seafloor in the South Pacific (French Polynesia) region, corroborated with JAMSTEC and a French team [e.g., Suetsugu et al., 2009], we constructed a numerical model of the mantle flow beneath the French Polynesia region based on a new regional seismic tomography model. We found excellent correlations between the observed and the modeled dynamic topography and between the buoyancy fluxes obtained from our numerical model and the ones deduced from the swell morphology. These outstanding fits reveal for the first time that a direct link exists between the surface observations and mantle flows [Adam, Yoshida, et al., 2010 (17)]. In a subsequent study, we show that the superswell could be caused by the large-scale slow velocity anomalies in the lower mantle. The surface geodetic observations are explained by a model including a low-viscosity asthenosphere situated immediately beneath the lithosphere, and a lower mantle viscosity 100 times greater than the upper mantle one. Although the existence of compositional heterogeneities is often invoked to explain the dynamics of the South Pacific superplume in previous numerical and laboratory experiments, and are important to account for plume/superplumes phenomenology, we cannot definitively conclude the presence of such compositional heterogeneities from our geodynamic modeling [Adam, Yoshida, et al., 2014 (31)].
7. Behavior of the subducted oceanic crust in the mantle transition zone by 3-D numerical simulations of plate subduction
Numerical simulations of mantle convection with semi-dynamic plate subduction in 3-D geometry is performed to investigate the role of harzburgite layers in the morphology of subducting plates and the behavior of oceanic crustal layers. The results show that a buckled crustal layer is observed under the heel of the stagnant slab that begins to penetrate into the lower mantle, regardless of the magnitude of the viscosity of the harzburgite layer when the factor of viscosity increase at the boundary of the upper and lower mantle is larger than 60-100. As the harzburgite layer viscosity increases, the curvature of buckling is larger. When the viscosities of harzburgite layer and lower mantle are larger, the volumes of crustal and harzburgite materials trapped in the mantle transition zone (MTZ) are also larger, although almost all of the materials penetrate into the lower mantle. These materials are trapped in the MTZ for over tens of millions of years [e.g., Yoshida et al., 2011 (23); Yoshida, 2013 (29), 2014 (33); Yoshida and Tajima, 2013 (28); Tajima, Yoshida, Ohtani, 2015 (34)].

Lecture Slides