Japan Geoscience Union Meeting 2023

Presentation information

[J] Oral

S (Solid Earth Sciences ) » S-CG Complex & General

[S-CG56] Dynamics in mobile belts

Fri. May 26, 2023 9:00 AM - 10:30 AM 302 (International Conference Hall, Makuhari Messe)

convener:Yukitoshi Fukahata(Disaster Prevention Research Institute, Kyoto University), Hikaru Iwamori(Earthquake Research Institute, The University of Tokyo), Kiyokazu Oohashi(Graduate School of Sciences and Technology for Innovation, Yamaguchi University), Chairperson:Hikaru Iwamori(Earthquake Research Institute, The University of Tokyo), Takato Takemura(Nihon University)

9:00 AM - 9:15 AM

[SCG56-01] Reconstruction of mantle thermal convection using adjoint-based data assimilation

★Invited Papers

*Atsushi Nakao1, Tatsu Kuwatani1, Shin-ichi Ito2, Hiromichi Nagao2 (1.Research Institute for Marine Geodynamics, Japan Agency for Marine-Earth Science and Technology , 2.Earthquake Research Institute, The University of Tokyo)

Keywords:Adjoint method, Inverse problem, Laminar flow, Mantle convection, Fluid mechanics

Active tectonic zones, including the Japan Island Arc, are characterized by partial melting and complex deformation due to material and heat supply and stress concentration. Quantifying active tectonic zones is therefore key to understanding the volcanism and seismicity. However, we have little direct information about active tectonic zones: geophysical observations, including seismic and electromagnetic tomography, provide only the current crust and mantle state and no time series, while geological observations provide only surface information and no information about the Earth's deep interior. Therefore, it is necessary to resolve the unknown parameters of active tectonic zones, such as deep stress, flow, and deformation processes, by combining observations and geodynamic theory. We then developed an adjoint-based data assimilation code for plate motion and mantle convection. The algorithm iteratively solves the forward and adjoint equations of mass, momentum, and energy conservation to minimize the error between observed and modeled values of plate velocities and mantle temperature. We will present the results of data assimilation on a simple 2-D Cartesian coordinate using synthetic data, and show how past mantle thermal convection is reconstructed as a function of the assimilated data and flow structures.