Mukautuva hilantarkennus Vlasiator-plasmasimulaattorissa

TL;DR

This paper explores adaptive mesh refinement in Vlasiator to optimize space plasma simulations by dynamically adjusting resolution based on plasma behavior, reducing computational load while maintaining accuracy.

Contribution

It introduces a novel adaptive mesh refinement method based on simulation data, enhancing efficiency in Vlasiator's plasma modeling.

Findings

Adaptive refinement produces similar results to static methods.

Refinement based on variable gradients effectively identifies regions needing higher resolution.

Results indicate potential for runtime adaptive refinement to further improve efficiency.

Abstract

Simulating space plasma in a global scale is computationally demanding. Modeling different regions with different resolution can save computational resources without compromising too much on simulation accuracy. This thesis examines adaptive mesh refinement as a method of optimizing simulation in Vlasiator. The thesis examines behavior of plasma and different characteristic scales that need to be factored in simulation. Kinetic models using statistical methods and fluid methods are examined. Both have their advantages, and Vlasiator uses a combination of these methods. Modeling electrons kinetically requires a resolution orders of magnitude greater than ions, so ions are modeled kinetically and electrons as a fluid. Targeted refinement used in Vlasiator is introduced as a method to save memory and computation. Due to the structure of the magnetosphere, the required resolution isn't uniform in the simulation region. Regions where a higher resolution is required can be refined such that the refined area is given as simulation parameters. Adaptive mesh refinement based on the simulation data is introduced as an evolution of this method. Refinement is done based on dimensionless gradients of different variables; regions where the examined variables change rapidly are refined, and conversely regions with little change are coarsened. Test results show adaptive refinement to be a promising way to improve Vlasiator. Refinement parameters produce results similar to the static method, while giving somewhat different refinement regions. A further goal is runtime refinement.