Rice convection model simulations of the centrifugal interchange instability in the magnetospheres of Jupiter and Saturn

Radial plasma transport in Jupiter's and Saturn's magnetospheres is driven by the centrifugal interchange instability. Simulations with the Rice Convection Model (RCM) of the injection-dispersion phenomenon in Saturn's magnetosphere produce interchange convection cells stretching outward from the outer edge of the plasma torus centered near the orbit of Enceladus. We first use an idealized cold plasma torus near L = 4 (the orbit of Enceladus) as an initial distribution. Small perturbations in the torus region grow into interchange convection cells with alternating sectors of inflow and outflow. We show that the velocity-dependent Coriolis effects can be included by an effective Hall conductance, which is first implemented here using the grid based RCM. The simulation shows the following effects introduced by the Coriolis force: (1) bending of the convection cells in the retrograde direction, (2) slowing of their growth, and (3) broadening the out-moving fingers. Our simulation results support and distinguish the predictions made by Vasyliunas [GRL, 21, 401, 1994] and Pontius [GRL, 24, 2961, 1997]. Then we address two curious properties revealed by Cassini observations. (1) Several magnetospheric properties, including electron density, are modulated by the SKR longitude of the spacecraft [e.g., Gurnett et al., Science, 316 , 442, 2007]. A two-cell corotating convection model has been invoked by Gurnett et al. and by Goldreich and Farmer [JGR, 112, A05225, 2007] to explain this modulation. We have simulated this system by imposing an initially asymmetric plasma distribution in the Enceladus torus region, and find that the initial distribution must be at least ∼40% asymmetric in order to give a factor ∼2 asymmetry at 5 Saturn radii as reported. (2) The inflow sectors ("injection events") are much narrower in longitude than the interspersed outflow sectors. There is no obvious explanation for this in linear instability theory, nor in previous RCM simulations without an active plasma source. Our inclusion of an extended active continuous plasma source, combined with the Coriolis effects, provides a possible explanation. Finally, the initial results of inclusion of the pickup currents are reported