Kinetic simulations of astrophysical and solar plasma turbulence

HPC Project: Kinetic simulations of astrophysical and solar plasma turbulence

Project ID: pr74vi
Supercomputer cluster: LRZ/SuperMUC, Garching (Germany)
Duration of the project: 08.2016-12.2017. Extension until 06.2020.
Mcore-hours granted: 35 + 30 (extension)
Principal Investigator: Prof. Dr. Jörg Büchner
Project partner institute: Max-Planck-Institute for Plasma Physics, Garching (Germany)
Researchers: Daniel Grošelj (MPI for Plasma Physics), Patricio A. Muñoz (TU Berlin), Silvio S. Cerri (University of Pisa, Italy), Prof. Dr. Frank Jenko (MPI for Plasma Physics)
Numerical codes: fully-kinetic PIC codes OSIRIS and ACRONYM, hybrid-Vlasov code HVM

Abstract:

Addressing a key open question in space plasma physics, namely the physics of energy dissipation at the tail of the magnetohydrodynamic turbulent cascade, the project team will carry out fully kinetic and hybrid kinetic simulations using the particle-in-cell codes OSIRIS, ACRONYM and the hybrid Vlasov-Maxwell code HVM. The specific strengths of the PiC and hybrid Vlasov-Maxwell method will be combined in a single study in order to maximize scientific output and to perform a critical assessment of the two models. While it is well known that turbulence in highly collisional plasmas is dissipated by means of particle collisions, much less is known about plasma heating and turbulent dissipation in nearly collisionless plasmas, such as the solar wind. Due to the low collisionality, turbulent fluctuations in the solar wind are able to cascade to very small scales where fluid models are no longer valid and a kinetic approach has to be used instead. Several previous studies have been performed with reduced kinetic or fully kinetic models in two dimensions to reduce the computational cost. The specific circumstances under which reduced models become no longer valid are generally not well understood, and in order to reach firm conclusions a detailed comparison with fully kinetic models is necessary. Similarly, one might ask how much physical realism is lost in two-dimensional simulations compared to the computationally more costly three-dimensional equivalents. The present project aims at answering some of these compelling questions in two steps. The first part comprises a critical comparison of fully kinetic and reduced kinetic models for a set of well defined test cases, while the second part aims at state of the art, three dimensional fully kinetic simulations covering a broad range of scales. A deeper understanding of energy transfer and dissipation in plasma turbulence will help understanding observation spacecraft data as well improve our abilities for space weather prediction.

Results:

We investigated the 2D properties of the turbulence in the near-Earth, free streaming solar wind by comparing fully kinetic, hybrid-code (reduced-kinetic) with the results obtained before in the framework of a gyrokinetic plasma model [Groselj et al. 2017]. The spectral properties of the solutions obtained from the fully kinetic and gyrokinetic simulations were found to be in good agreement under typical plasma conditions. Other than as the consequences of numerous simplifying assumptions of gyrokinetic models, the fully kinetic investigations allowed to identify the  dominant type of turbulence fluctuations in two dimensions (see Figure 1). The results of the hybrid code simulations, on the other hand, were lacking of essential consequences of the kinetic electron physics, i.e. they could not reproduce the key new findings obtained by gyrokinetic and fully-kinetic models.

Our massively parallel 3D PIC-code simulations of the kinetic turbulence were among the very first targeting the end of the inertial range of the solar wind turbulence [Groselj et al. 2018].  By considering the spectral properties of the solutions and by calculating the anisotropy of the turbulence relative to the local mean magnetic field, we confirmed the predictions of the leading, so-called kinetic Alfvén wave turbulence model. This was the first study where the predictions of the kinetic Alfvén wave turbulence model were tested using only first-principle-physics assumptions without additional approximations, yielding results largely consistent with the predictions (cf. Figure 2).

Publications:

1. Grošelj, D., Cerri, S. S., Navarro, A. B., Willmott, C., Told, D., Loureiro, N. F., Jenko, F. (2017). Fully Kinetic versus Reduced-kinetic Modeling of Collisionless Plasma Turbulence. The Astrophysical Journal, 847(1), 28. https://doi.org/10.3847/1538-4357/aa894d
2. Grošelj, D., Mallet, A., Loureiro, N. F., & Jenko, F. (2018). Fully Kinetic Simulation of 3D Kinetic Alfvén Turbulence. Physical Review Letters, 120(10), 105101. https://doi.org/10.1103/PhysRevLett.120.105101