Title : Kinetic Entropy as a Diagnostic for Magnetic Reconnection
Speaker: Haoming Liang, West Virginia University
Abstract : Kinetic entropy is the entropy defined using kinetic theory for plasmas that are not necessarily in local thermodynamic equilibrium. Entropy is a natural metric of irreversible dissipation since it is conserved in ideal isolated systems and increases only when there is dissipation. It should be especially important in many heliospheric systems, where collisions can be rare so that plasmas are not in thermodynamic equilibrium. This suggests kinetic entropy can address important unsolved questions on the nature of irreversible dissipation in fundamental plasma processes such as magnetic reconnection, plasma turbulence and collisionless shocks. While entropy is often investigated in fluid and gyrokinetic systems, it is vastly underutilized in fully kinetic systems. In this work, we carry out an initial study to develop and apply the kinetic entropy diagnostic in particle-in-cell (PIC) simulations. We start with 2.5D collisionless anti-parallel reconnection. In the simulations, we calculate the commonly-used kinetic entropy written as the phase space integral of – f ln f, where f is the distribution function, and the full Boltzmann entropy related to the logarithm of the number of microstates for a specific macrostate. By decomposing kinetic entropy into a sum of velocity space and position space entropies, we show that position space entropy decreases while velocity space entropy increases during magnetic reconnection. We find that total kinetic entropy in the simulations is preserved quite well (better than three percent) and use the departure from exact conservation to quantify the effective numerical dissipation. Electrons and ions have slightly different effective collision frequencies. Finally, we use kinetic entropy to identify regions with non-Maxwellian distributions and compare the results with other approaches. Note, our work uses collisionless simulations, so we cannot yet address physical dissipation mechanisms; nonetheless, the infrastructure developed here will be useful for future studies in weakly collisional systems. It is being applied to Magnetospheric Multiscale (MMS) data.