2026.10.30

[日時]　2026 年 12 月 1 日 (火) 16 時 〜 17時

場所：　オンライン開催　
（接続情報は別途メールにて、参加申込・招待された方へお知らせします）

[講師]

銭谷 誠司　Dr. Seiji Zenitani (Spare Research Institute, Austrian Academy of Sciences)

[講演題目]

Recent advances in numerical algorithms for particle-in-cell (PIC) simulations

(日本語開催)

[講演概要]
　　Particle-in-cell (PIC) simulation is one of the most important research tools for understanding kinetic processes in space, astrophysical, and fusion plasmas. While the PIC method was established in the 1970s, there is a strong demand for overhauling it, as computing environments have been rapidly evolving. In this contribution, we review our recent progress in numerical algorithms for PIC simulation.
　　First, we discuss initialization methods for particle velocities. We require a Kappa distribution for heliospheric problems, a loss-cone distribution for inner-magnetospheric problems, and a Maxwell-Jüttner distribution in high-energy astrophysics, but it was not clear how to generate them. Based on fundamental random variables such as uniform, normal, and gamma variates, we have developed a series of Monte Carlo procedures for generating non-Maxwellian distributions.
　　Second, we examine particle integrators, which solve the equations of motion of all particles at every timestep. This is arguably the most important component in PIC simulations. Although the Boris solver (or the Buneman-Boris method) has long been used, higher-accuracy methods are desirable in the coming era. Using subcycling and higher-order corrections, we have developed next-generation Boris solvers. We discuss their accuracy, stability, and computational costs.
　　Finally, we propose a workaround for an MHD closure problem relevant to MHD-PIC multihierarchy simulations.

[お問合せ] 　cg-sim

• 興味をお持ちの方はどなたでもご参加いただけますが、情報セキュリティおよび貿易管理の観点から、事前申込をお願いします。お申込みは、氏名、所属、職名または学年、連絡先メールアドレスを明記の上、cg-sim メールにてお送りください。折り返し参加の可否と接続情報をお知らせします。
• 当日、参加者名簿を作成します。名簿は法に基づいて開示される場合があります。
• 各所属大学・機関において研究倫理教育を受講した上でお申し込みください。