[1] Io's motion relative to the Jovian magnetic field generates a power of about 10 12 W, which is thought to propagate as an Alfvén wave along the magnetic field line. This power is transmitted to the electrons, which will then precipitate and generate the observed auroral phenomena from UV to radio wavelengths. A more detailed look at this hypothesis shows some difficulties: Can the Alfvén waves escape the torus or are they trapped inside? Where and how are the particles accelerated? In which direction? Is there enough power transmitted to the particles to explain the strong brightness of the auroral emissions in UV, IR, visible, and radio? In other words, can we make a global, consistent model of the Io-Jupiter interaction that matches all the observations? To answer these questions, we review the models and studies that have been proposed so far. We show that the Alfvén waves need to be filamented by a turbulent cascade process and accelerate the electrons at high latitude in order to explain the observations and to form a consistent scheme of the Io-Jupiter interaction.

The generation of nonoscillatory mirror waves is studied using a one‐dimensional periodic hybrid electromagnetic simulation. The ion dynamics are treated exactly; the electrons are approximated as a finite pressure, massless fluid. Compression of the flux tubes in the magnetosheath causes a large pressure anisotropy, and it has been proposed that this anisotropy drives a mirror instability. The mirror waves have been identified by large amplitude fluctuations of the magnetic field, anticorrelated with pressure fluctuations. The simulations are initiated in a homogeneous high beta (beta = 2.5) plasma with the ambient magnetic field at various angles to the simulation axis. It is found that ion cyclotron waves are also driven by the pressure anisotropy, in competition with the nonoscillatory mirror waves. Simulations indicate that in a pure ¹H+ plasma the much faster growing ion cyclotron waves absorb the free energy in the anisotropy to the extent that mirror waves should not be observed. Analysis of the dispersion relations of mirror waves and ion cyclotron waves in the multi‐component plasma indicates that 4He2+ and 16O6+ ions in the solar wind should stabilize the ion cyclotron waves sufficiently that the mirror waves become the dominant instability.

It is suggested that the highly structured auroral arc is caused by a current‐driven laminar electrostatic shock oblique to the geomagnetic field. Electrons are accelerated by the potential jump associated with the shock. The shock is assumed to be confined to a plane. Self‐consistent solutions to the Poisson‐Vlasov systems are calculated for the electrostatic potential. Adiabatic theory is used to calculate the ion number density in terms of the electrostatic potential and its derivatives. The electrons are assumed to be highly magnetized so they can only move parallel to the magnetic field. Solutions are exhibited for two plasma models: (1) streaming electrons and a two‐temperature distribution of ions and (2) streaming electrons and ions and thermal electrons and ions. In the latter model, solutions can be obtained for an arbitrary potential jump across the shock. The shock is identified with the linear electrostatic ion cyclotron wave, and stability of these waves is examined to determine conditions for the formation of oblique shocks. Finally, the theory is discussed in the context of the magnetosphere, and possible model shocks are exhibited and discussed in terms of auroral arc formation.

No abstract

Two‐dimensional (2‐D) hybrid simulations are carried out to study the structure of the reconnection layer in the distant magnetotail. In the simulation an initial current sheet separates the two lobes with antiparallel magnetic field components in the x direction. The current sheet normal is along the z direction. It is found that a leading bulge‐like magnetic configuration and a trailing, quasi‐steady reconnection layer are formed in a magnetic reconnection. If the duration of the reconnection is sufficiently long, the trailing reconnection layer will dominate the plasma outflow region. For the symmetric lobes with By = 0, two pairs of slow shocks are present in the quasi‐steady reconnection layer. The slow shocks are expected to be fully developed at a sufficient distance from the X line, where the separation between the two shocks is greater than a few tens of the lobe ion inertial length. The Rankine‐Hugoniot jump conditions of the slow shock are found to be better satisfied as the distance from the X line along the x axis increases. For the cases with By ≠ 0 in the two lobes, two rotational discontinuity‐like structures appear to develop in the reconnection layer. On the other hand, in the leading bulge region of a magnetic reconnection, no steady MHD discontinuities are found. Across the plasma sheet boundary layer the increase of the flow velocity appears to be much smaller than that predicted from the Rankine‐Hugoniot jump conditions for a steady discontinuity, and the increase in the ion number density is much larger. In addition, a large increase in the parallel ion temperature is found in the plasma sheet boundary layer. The 2‐D simulation results are also compared with the one‐dimensional hybrid simulations for the Riemann problem associated with the magnetotail reconnection. It is found that the 2‐D effects may lead to the presence of the non‐switch‐off slow shocks and thus the lack of coherent wave trains in slow shocks.