Enabling Optimizations for Global Magnetospheric Kinetic Simulations with Reduced Kinetic Spectral Models

Efficient coupling of the microscopic physics into the macroscopic system-scale dynamics (called "fluid-kinetic" coupling) is probably the most important and unresolved problem of computational plasma physics. It impacts most of plasma physics areas including space physics and fusion systems. Majority of conventional simulation tools capable of describing large scale dynamics are usually limited to simplified fluid/magnetohydrodynamics description of the plasma, because of the large spatial and temporal scale separation typical of plasmas. Yet, fluid models lack the microscopic physics, which is known to be important in many applications (e.g., reconnection, shock physics, etc.) A way forward is to build models that combine kinetic and fluid description in one consistent framework. The development of such methods can bridge the scale gap to successfully handle coupling of large-scale dynamics and microscopic processes.

In this presentation, we describe a novel simulation method, where the kinetic equation is solved using a spectral expansion of the plasma distribution function. The low-order terms in the expansion capture the large-scale dynamics of the system, while higher-order terms add microscopic physics incrementally, similar to classical fluid-moment expansion. Such a method is ideally suited for problems involving fluid-kinetic coupling, since the number of expansion terms could be adapted in space and time. Furthermore, the spectral basis itself adaptively changes in space and time, adjusting to plasma mean flow and temperature, thus making the representation of the particle distribution function very efficient. We show that our reduced kinetic model with just a ~(4-6)^3 velocity-space moments agrees well with results from fully-kinetic simulations on some examples. In addition to describing the method, we will present several examples illustrating its application to various problems, including solar wind-magnetosphere interaction.