Research

１．Particle-Based Cloud Modeling using the Super-Droplet Method
The super-droplet method (SDM) is a particle-based numerical algorithm that enables accurate simulation of cloud microphysics with lower computational demand (Shima et al., 2009). We are developing a highly accurate cloud model using the SDM, which will contribute to reducing the uncertainties of numerical weather prediction and climate simulation.  We are also interested in applying the methodology to other research areas, such as volcanic fumes, star formation, and spray combustion.

Precipitation from a shallow maritime cumulus (2D simulation). Click the figure to see the movie.


VR movie of a cumulus congestus, created by our collaborator Dr. Toshiki Matsushima (R-CCS).


２．Synchronization Phenomena
Self-sustained oscillators can generate macroscopic rhythms as a collective through their mutual interaction. This is referred to as synchronization phenomenon or rhythmic phenomenon. Various things are known to synchronize, ranging from mechanical ones such as metronomes, to biological ones such as heart cells, circadian rhythms, crowd clapping, firefly flashingh, and croaking of frogs.

Simulation of the collective synchronous flashing of fireflies

３．Micro-Macro Interlocked Simulation
Physical laws have a hierarchical structure. As a result, both micro and macroscopic domains may coexist in a phenomena simultaneously. Leveraging this characteristic, the Micro-Macro Interlocked Simulation methodology aims to simulate the macroscopic behavior of the phenomena of interest efficiently and accurately .

Micro-Macro Interlocked simulation of two phase singularities with attractive interaction. (a) result of the first-principle calculation. (b) and (c) result of the Micro-Macro Interlocked simulation.  The red domain indicated in (c) is simulated microscopically in (b).


４．Data Assimilation of Complex Systems
データ同化とは, 不確かさを伴うシミュレーションと観測を融合し, そこからより確かな情報を抽出する統計学的枠組みです. 社会・産業・環境に関する複雑な現象のシミュレーションにデータ同化を応用する研究を行っています.

粒子フィルタによるRosslerモデルの状態推定

Master's Thesis
・Simulation of precipitating staratocumlus over Tokyo using the super-droplet method
・Wake-induced CCN-activation behind precipitation particles
・Implementation of spontaneous breakup process to the super-droplet method
・Simulation of rain in cumulus congestus over tropical ocean using the super-droplet method
・Two-dimensional idealized simulation of orographic snow clouds in Hokuriku using the super-droplet method
・Numerical analysis of the wall adhesion and flow characteristics of micro droplets in oil mist separator
・Improvement of ice crystal growth model and the assessment of its impact to cumulonimbus simulation
・Numerical study of the impact of aerosol regeneration process on orographic clouds and precipitation using the super-droplet method
・金星の雲の共凝結成長ダイナミクスにおける分岐構造の解析
・氷水混相の雲微物理過程に関する多孔性回転楕円体モデルの性能検証と改良
・超水滴法による海洋層積雲のシミュレーションと大気境界層の乱流構造
・東シナ海における総窒素酸化物の輸送と化学的変質に関するモデル解析
・粒子の衝突併合を考慮した噴霧塗装シミュレーション
・超水滴法による海洋層積雲のシミュレーションとエアロゾル影響の評価
・スコールラインのシミュレーションに必要な空間解像度の評価
・再帰的アルゴリズムによる超水滴法の改良
・遺伝的アルゴリズムによる動物将棋の教師なし学習
・神戸市内における花粉の飛散シミュレーション
・対話型進化計算による人の好みの自動抽出
・波力発電シミュレーション
・淡路島における新しい分散電力供給システムの提案
・人の流動に着目した神戸元町商店街の活性化