Parallel electric fields in mixing hot and cold plasmas in the auroral downward current region: Double layers and ambipolar fields

Abstract

[1] Using a 2.5-D parallel particle-in-cell simulation of a plasma of long length parallel to an ambient magnetic field Bo, we study the processes involved in determining the distribution of an applied electric potential drop parallel to Bo. The simulated plasma consists of both hot and cold plasmas of the magnetospheric and ionospheric origins, respectively. The former plasma is at a higher positive potential with respect to the latter, and thus the simulation results are relevant to the auroral downward current regions. The parallel processing enables us to simulate a long system with the magnetic field-aligned dimension Lz ∼ 8192 λdo, where λdo is the plasma Debye length. We find that when the initially empty simulation box accumulates sufficient plasma supplied from hot plasma from the top and cold plasma from the bottom, a density cavity forms at the interface between the hot and cold plasmas. A part of the applied potential drop occurs in the cavity as a double layer (DL), while the rest of it as ambipolar fields supported by the density gradient in the hot plasma density on the high-potential side (HPS) of the DL. The DL propagates upward. The HPS of the DL is rich in large-amplitude electron holes. At later times in the evolution of plasma and fields as the DL reaches the top boundary, we find that a major part of the applied potential is distributed over long distances giving only a weak ambipolar type of parallel electric fields. Then again the distributed potential evolves into localized potential drops like in a stack of multiple double layers. The double layers and associated cavities emerge from low-frequency and long-wavelength oscillations in the presence of very hot ions. Parallel currents in the plasma seem to be the only source of free energy for driving the oscillations. We report the evolution of the electron velocity distribution functions as the potential distribution evolves.