[M-GI28] データ駆動地球惑星科学

[MGI28-P02] 熱対流復元のための粒子-セル系データ同化

*中尾 篤史^1,2、桑谷 立²、伊藤 伸一^3,4、長尾 大道^3,4 (1.秋田大学大学院理工学研究科、2.海洋研究開発機構海域地震火山部門、3.東京大学地震研究所、4.東京大学大学院情報理工学系研究科)

キーワード：アジョイント法 (４次元変分法)、マーカアンドセル法、マントル対流、流体力学、逆問題

Problems in the geosciences often involve Lagrangian particles that follow the surrounding velocity fields and record information about the system; for example, the pressure of mantle-derived rocks, the temperature of volcanic ash, and the latitude of pumices from submarine volcanoes. These Lagrangian particles can be useful for estimating unknown parameters such as their sources and the flow and thermal structures of the surrounding fluid. Conventional fluid dynamic simulations have attempted to reproduce records on Lagrangian particles based on a forward framework, but there are few approaches that use particle information to constrain the unknown parameters based on an inverse framework. To use information on Lagrangian particles to constrain unknown parameters about the surrounding fluid in an inverse manner, we have newly developed a 4D-Var (four-dimensional variational) data assimilation for thermal convection in a particle-grid coupled system. Here we consider particles advected in a thermally convecting highly viscous fluid, mimicking geochemical tracers in the Earth's mantle, and construct time series of thermal structure and flow velocity only from the tracer particle records. We will present preliminary 4D-Var results on a simple 2-D Cartesian coordinate with the passive particles using synthetic data, and show how past mantle thermal convection is reconstructed.